KGRKJGETMRETU895U-589TY5MIGM5JGB5SDFESFREWTGR54TY
Server : Apache/2.4.62
System : FreeBSD fbsdweb2.web.rcn.net 14.1-RELEASE FreeBSD 14.1-RELEASE releng/14.1-n267679-10e31f0946d8 GENERIC amd64
User : www ( 80)
PHP Version : 8.3.8
Disable Function : NONE
Directory :  /domains/highlandlabs/cqi-bin/ALFA_DATA/alfasymlink/root/usr/lib32/

Upload File :
current_dir [ Writeable ] document_root [ Writeable ]

 

Current File : /domains/highlandlabs/cqi-bin/ALFA_DATA/alfasymlink/root/usr/lib32/libpmc.so.5
ELF	4x�64 
( 444@@�%�% �% �% �%0Y0YP&P%&P%&����H�6H�6H�6H�<�6<�6<�6��R�tdP&P%&P%&����P�tdp�%p�%p�%��Q�tdtttFreeBSD�`
" =OV����������z����$,4T������������-EZf{����%�`�%B�0�%����%z�P�%L�%Wa�%����%z���%u���%�@�%V���%xi��%�;�%N?@�%�(&��%%���%4���%P
��%�]��%#���%a���%L�,�6&0��%����%%���%L�0�%W���%L]`�%9J@�%Y���%_��6`��%RB0�%�x`�%1U��%;g �%1� �%auP�%4z@�%Mmp�%�3��%;?�%R*��%1$��%r_��%��0�%j��%?V��%Y��%��0�00ӯk�P&y��(z�(z�(z
5 P��@ 0 	@��� ��G	�b@ $
 `"05:?BDIMTWZ_egx��n��Mpb�g�V�E���^"������T�Дa)ڨF���&
���[��������{�,z�RU�u�_�������$󌱐fp�����YX�*G��zW+��+0�;m�����n�+I��J�Tlx[e���N���M|�8����;�e����~���L)Dr̒�:�rN�sN|÷�Cԕ"��1u�giiEa.1_#DCR`eOZdH6WL;9>A7SXcQfPV2[Y:3hF-g=M]B+('
?!^&	
4")$5*0,I<GT/%8 \bJKU@N_init_fini__cxa_finalize_Jv_RegisterClassespmc_allocate__stack_chk_guardstrdupstrseppmc_pmu_enabledpmc_pmu_pmcallocatefreestrcasecmp__errorstrncasecmpsyscall__stack_chk_fail__mb_sb_limit_ThreadRuneLocale___tls_get_addr_CurrentRuneLocalepmc_attachpmc_capabilitiespmc_configure_logfilepmc_cpuinfopmc_detachpmc_disablepmc_enablepmc_event_names_of_classmallocpmc_flush_logfilepmc_close_logfilepmc_get_driver_statspmc_get_msrpmc_initmodfindmodstatmemcpycallocpmc_name_of_capabilitypmc_name_of_classpmc_name_of_cputypepmc_name_of_disposition_pmc_name_of_eventpmc_name_of_eventpmc_name_of_modepmc_name_of_statepmc_ncpupmc_npmcpmc_pmcinfopmc_readpmc_releasepmc_rwpmc_setpmc_startpmc_stoppmc_widthpmc_writepmc_writelogstrchrstrtolpmclog_readputsreadmemmovestrlcpymemsetpmc_pmu_event_get_by_idx__assertpmclog_feedpmclog_openpmclog_closepmc_pmu_idx_get_by_eventsysctlbynamestrcasestrpmu_events_mapprintfregcompregexecregfreepmc_pmu_sample_rate_getstrcmpgetenvabortpmc_pmu_print_counterspmc_pmu_print_counter_descpmc_pmu_print_counter_desc_longpmc_pmu_print_counter_full_Z13event_to_jsonP9pmclog_everrx_ZNSt3__112basic_stringIcNS_11char_traitsIcEENS_9allocatorIcEEE6appendEPKcj_ZdlPv_Unwind_Resumesnprintfstrlen_Znwj__cxa_allocate_exception_ZNSt12length_errorD1Ev_ZTISt12length_error__cxa_throw__cxa_free_exception_ZNSt11logic_errorC2EPKc_ZTVSt12length_error__gxx_personality_v0pmu_sys_event_tableslibcxxrt.so.1CXXABI_1.3libgcc_s.so.1GCC_3.0libc.so.7FBSD_1.0FBSD_1.3FBSD_1.6libc++.so.1libm.so.5libpmc.so.5d%&h%&p%&x%&�%&�%&�%&�%&�%&�%&�%&�%&�%&�%&�%&�%&�%&�%&�%&�%&&&&&&&&& &&(&&0&&8&&@&&H&&P&&X&&`&&h&&p&&x&&�&&�&&�&&�&&�&&�&&�&&�&&�&&�&&�&&�&&�&&�&&�&&�&&'&'&'&'& '&('&0'&8'&@'&H'&P'&X'&`'&h'&p'&x'&�'&�'&�'&�'&�'&�'&�'&�'&�'&�'&�'&�'&�'&�'&�'&�'&(&(&(&(& (&((&0(&8(&@(&H(&P(&X(&`(&h(&p(&x(&�(&�(&�(&�(&�(&�(&�(&�(&�(&�(&�(&�(&�(&�(&�(&�(&)&)&)&)& )&()&0)&8)&@)&H)&P)&X)&`)&h)&p)&x)&�)&�)&�)&�)&�)&�)&�)&�)&�)&�)&�)&�)&�)&�)&�)&�)&*&*&*&*& *&(*&0*&8*&@*&H*&P*&X*&`*&h*&p*&x*&�*&�*&�*&�*&�*&�*&�*&�*&�*&�*&�*&�*&�*&�*&�*&�*&+&+&+&+& +&(+&0+&8+&@+&H+&P+&X+&`+&h+&p+&x+&�+&�+&�+&�+&�+&�+&�+&�+&�+&�+&�+&�+&�+&�+&�+&�+&,&,&,&,& ,&(,&0,&8,&@,&H,&P,&X,&`,&h,&p,&x,&�,&�,&�,&�,&�,&�,&�,&�,&�,&�,&�,&�,&�,&�,&�,&�,&-&-&-&-& -&(-&0-&8-&@-&H-&P-&X-&`-&h-&p-&x-&�-&�-&�-&�-&�-&�-&�-&�-&�-&�-&�-&�-&�-&�-&�-&�-&.&.&.&.& .&(.&0.&8.&@.&H.&P.&X.&`.&h.&p.&x.&�.&�.&�.&�.&�.&�.&�.&�.&�.&�.&�.&�.&�.&�.&�.&�.&/&/&/&/& /&(/&0/&8/&@/&H/&P/&X/&`/&h/&p/&x/&�/&�/&�/&�/&�/&�/&�/&�/&�/&�/&�/&�/&�/&�/&�/&�/&0&0&0&0& 0&(0&00&80&@0&H0&P0&X0&`0&h0&p0&x0&�0&�0&�0&�0&�0&�0&�0&�0&�0&�0&�0&�0&�0&�0&�0&�0&1&1&1&1& 1&(1&01&81&@1&H1&P1&X1&`1&h1&p1&x1&�1&�1&�1&�1&�1&�1&�1&�1&�1&�1&�1&�1&�1&�1&�1&�1&2&2&2&2& 2&(2&02&82&@2&H2&P2&X2&`2&h2&p2&x2&�2&�2&�2&�2&�2&�2&�2&�2&�2&�2&�2&�2&�2&�2&�2&�2&3&3&3&3& 3&(3&03&83&@3&H3&P3&X3&`3&h3&p3&x3&�3&�3&�3&�3&�3&�3&�3&�3&�3&�3&�3&�3&�3&�3&�3&�3&4&4&4&4& 4&(4&04&84&@4&H4&P4&X4&`4&h4&p4&x4&�4&�4&�4&�4&�4&�4&�4&�4&�4&�4&�4&�4&�4&�4&�4&�4&5&5&5&5& 5&(5&05&85&@5&H5&P5&X5&`5&h5&p5&x5&�5&�5&�5&�5&�5&�5&�5&�5&�5&�5&�5&�5&�5&�5&�5&�5&6&6&6&6& 6&(6&06&86&@6&H6&P6&X6&`6&h6&p6&x6&�6&�6&�6&�6&�6&�6&�6&�6&�6&�6&�6&�6&�6&�6&�6&�6&7&7&7&7& 7&(7&07&87&@7&H7&P7&X7&`7&h7&p7&x7&�7&�7&�7&�7&�7&�7&�7&�7&�7&�7&�7&�7&�7&�7&�7&�7&8&8&8&8& 8&(8&08&88&@8&H8&P8&X8&`8&h8&p8&x8&�8&�8&�8&�8&�8&�8&�8&�8&�8&�8&�8&�8&�8&�8&�8&�8&9&9&9&9& 9&(9&09&89&@9&H9&P9&X9&`9&h9&p9&x9&�9&�9&�9&�9&�9&�9&�9&�9&�9&�9&�9&�9&�9&�9&�9&�9&:&:&:&:& :&(:&0:&8:&@:&H:&P:&X:&`:&h:&p:&x:&�:&�:&�:&�:&�:&�:&�:&�:&�:&�:&�:&�:&�:&�:&�:&�:&;&;&;&;& ;&(;&0;&8;&@;&H;&P;&X;&`;&h;&p;&x;&�;&�;&�;&�;&�;&�;&�;&�;&�;&�;&�;&�;&�;&�;&�;&�;&<&<&<&<& <&(<&0<&8<&@<&H<&P<&X<&`<&h<&p<&x<&�<&�<&�<&�<&�<&�<&�<&�<&�<&�<&�<&�<&�<&�<&�<&�<&=&=&=&=& =&(=&0=&8=&@=&H=&P=&X=&`=&h=&p=&x=&�=&�=&�=&�=&�=&�=&�=&�=&�=&�=&�=&�=&�=&�=&�=&�=&>&>&>&>& >&(>&0>&8>&@>&H>&P>&X>&`>&h>&p>&x>&�>&�>&�>&�>&�>&�>&�>&�>&�>&�>&�>&�>&�>&�>&�>&�>&?&?&?&?& ?&(?&0?&8?&@?&H?&P?&X?&`?&h?&p?&x?&�?&�?&�?&�?&�?&�?&�?&�?&�?&�?&�?&�?&�?&�?&�?&�?&@&@&@&@& @&(@&0@&8@&@@&H@&P@&X@&`@&h@&p@&x@&�@&�@&�@&�@&�@&�@&�@&�@&�@&�@&�@&�@&�@&�@&�@&�@&A&A&A&A& A&(A&0A&8A&@A&HA&PA&XA&`A&hA&pA&xA&�A&�A&�A&�A&�A&�A&�A&�A&�A&�A&�A&�A&�A&�A&�A&�A&B&B&B&B& B&(B&0B&8B&@B&HB&PB&XB&`B&hB&pB&xB&�B&�B&�B&�B&�B&�B&�B&�B&�B&�B&�B&�B&�B&�B&�B&�B&C&C&C&C& C&(C&0C&8C&@C&HC&PC&XC&`C&hC&pC&xC&�C&�C&�C&�C&�C&�C&�C&�C&�C&�C&�C&�C&�C&�C&�C&�C&D&D&D&D& D&(D&0D&8D&@D&HD&PD&XD&`D&hD&pD&xD&�D&�D&�D&�D&�D&�D&�D&�D&�D&�D&�D&�D&�D&�D&�D&�D&E&E&E&E& E&(E&0E&8E&@E&HE&PE&XE&`E&hE&pE&xE&�E&�E&�E&�E&�E&�E&�E&�E&�E&�E&�E&�E&�E&�E&�E&�E&F&F&F&F& F&(F&0F&8F&@F&HF&PF&XF&`F&hF&pF&xF&�F&�F&�F&�F&�F&�F&�F&�F&�F&�F&�F&�F&�F&�F&�F&�F&G&G&G&G& G&(G&0G&8G&@G&HG&PG&XG&`G&hG&pG&xG&�G&�G&�G&�G&�G&�G&�G&�G&�G&�G&�G&�G&�G&�G&�G&�G&H&H&H&H& H&(H&0H&8H&@H&HH&PH&XH&`H&hH&pH&xH&�H&�H&�H&�H&�H&�H&�H&�H&�H&�H&�H&�H&�H&�H&�H&�H&I&I&I&I& I&(I&0I&8I&@I&HI&PI&XI&`I&hI&pI&xI&�I&�I&�I&�I&�I&�I&�I&�I&�I&�I&�I&�I&�I&�I&�I&�I&J&J&J&J& J&(J&0J&8J&@J&HJ&PJ&XJ&`J&hJ&pJ&xJ&�J&�J&�J&�J&�J&�J&�J&�J&�J&�J&�J&�J&�J&�J&�J&�J&K&K&K&K& K&(K&0K&8K&@K&HK&PK&XK&`K&hK&pK&xK&�K&�K&�K&�K&�K&�K&�K&�K&�K&�K&�K&�K&�K&�K&�K&�K&L&L&L&L& L&(L&0L&8L&@L&HL&PL&XL&`L&hL&pL&xL&�L&�L&�L&�L&�L&�L&�L&�L&�L&�L&�L&�L&�L&�L&�L&�L&M&M&M&M& M&(M&0M&8M&@M&HM&PM&XM&`M&hM&pM&xM&�M&�M&�M&�M&�M&�M&�M&�M&�M&�M&�M&�M&�M&�M&�M&�M&N&N&N&N& N&(N&0N&8N&@N&HN&PN&XN&`N&hN&pN&xN&�N&�N&�N&�N&�N&�N&�N&�N&�N&�N&�N&�N&�N&�N&�N&�N&O&O&O&O& O&(O&0O&8O&@O&HO&PO&XO&`O&hO&pO&xO&�O&�O&�O&�O&�O&�O&�O&�O&�O&�O&�O&�O&�O&�O&�O&�O&P&P&P&P& P&(P&0P&8P&@P&HP&PP&XP&`P&hP&pP&xP&�P&�P&�P&�P&�P&�P&�P&�P&�P&�P&�P&�P&�P&�P&�P&�P&Q&Q&Q&Q& Q&(Q&0Q&8Q&@Q&HQ&PQ&XQ&`Q&hQ&pQ&xQ&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&�Q&R&R&R&R& R&(R&0R&8R&@R&HR&PR&XR&`R&hR&pR&xR&�R&�R&�R&�R&�R&�R&�R&�R&�R&�R&�R&�R&�R&�R&�R&�R&S&S&S&S& S&(S&0S&8S&@S&HS&PS&XS&`S&hS&pS&xS&�S&�S&�S&�S&�S&�S&�S&�S&�S&�S&�S&�S&�S&�S&�S&�S&T&T&T&T& T&(T&0T&8T&@T&HT&PT&XT&`T&hT&pT&xT&�T&�T&�T&�T&�T&�T&�T&�T&�T&�T&�T&�T&�T&�T&�T&�T&U&U&U&U& U&(U&0U&8U&@U&HU&PU&XU&`U&hU&pU&xU&�U&�U&�U&�U&�U&�U&�U&�U&�U&�U&�U&�U&�U&�U&�U&�U&V&V&V&V& V&(V&0V&8V&@V&HV&PV&XV&`V&hV&pV&xV&�V&�V&�V&�V&�V&�V&�V&�V&�V&�V&�V&�V&�V&�V&�V&�V&W&W&W&W& W&(W&0W&8W&@W&HW&PW&XW&`W&hW&pW&xW&�W&�W&�W&�W&�W&�W&�W&�W&�W&�W&�W&�W&�W&�W&�W&�W&X&X&X&X& X&(X&0X&8X&@X&HX&PX&XX&`X&hX&pX&xX&�X&�X&�X&�X&�X&�X&�X&�X&�X&�X&�X&�X&�X&�X&�X&�X&Y&Y&Y&Y& Y&(Y&0Y&8Y&@Y&HY&PY&XY&`Y&hY&pY&xY&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&�Y&Z&Z&Z&Z& Z&(Z&0Z&8Z&@Z&HZ&PZ&XZ&`Z&hZ&pZ&xZ&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&�Z&[&[&[&[& [&([&0[&8[&@[&H[&P[&X[&`[&h[&p[&x[&�[&�[&�[&�[&�[&�[&�[&�[&�[&�[&�[&�[&�[&�[&�[&�[&\&\&\&\& \&(\&0\&8\&@\&H\&P\&X\&`\&h\&p\&x\&�\&�\&�\&�\&�\&�\&�\&�\&�\&�\&�\&�\&�\&�\&�\&�\&]&]&]&]& ]&(]&0]&8]&@]&H]&P]&X]&`]&h]&p]&x]&�]&�]&�]&�]&�]&�]&�]&�]&�]&�]&�]&�]&�]&�]&�]&�]&^&^&^&^& ^&(^&0^&8^&@^&H^&P^&X^&`^&h^&p^&x^&�^&�^&�^&�^&�^&�^&�^&�^&�^&�^&�^&�^&�^&�^&�^&�^&_&_&_&_& _&(_&0_&8_&@_&H_&P_&X_&`_&h_&p_&x_&�_&�_&�_&�_&�_&�_&�_&�_&�_&�_&�_&�_&�_&�_&�_&�_&`&`&`&`& `&(`&0`&8`&@`&H`&P`&X`&``&h`&p`&x`&�`&�`&�`&�`&�`&�`&�`&�`&�`&�`&�`&�`&�`&�`&�`&�`&a&a&a&a& a&(a&0a&8a&@a&Ha&Pa&Xa&`a&ha&pa&xa&�a&�a&�a&�a&�a&�a&�a&�a&�a&�a&�a&�a&�a&�a&�a&�a&b&b&b&b& b&(b&0b&8b&@b&Hb&Pb&Xb&`b&hb&pb&xb&�b&�b&�b&�b&�b&�b&�b&�b&�b&�b&�b&�b&�b&�b&�b&�b&c&c&c&c& c&(c&0c&8c&@c&Hc&Pc&Xc&`c&hc&pc&xc&�c&�c&�c&�c&�c&�c&�c&�c&�c&�c&�c&�c&�c&�c&�c&�c&d&d&d&d& d&(d&0d&8d&@d&Hd&Pd&Xd&`d&hd&pd&xd&�d&�d&�d&�d&�d&�d&�d&�d&�d&�d&�d&�d&�d&�d&�d&�d&e&e&e&e& e&(e&0e&8e&@e&He&Pe&Xe&`e&he&pe&xe&�e&�e&�e&�e&�e&�e&�e&�e&�e&�e&�e&�e&�e&�e&�e&�e&f&f&f&f& f&(f&0f&8f&@f&Hf&Pf&Xf&`f&hf&pf&xf&�f&�f&�f&�f&�f&�f&�f&�f&�f&�f&�f&�f&�f&�f&�f&�f&g&g&g&g& g&(g&0g&8g&@g&Hg&Pg&Xg&`g&hg&pg&xg&�g&�g&�g&�g&�g&�g&�g&�g&�g&�g&�g&�g&�g&�g&�g&�g&h&h&h&h& h&(h&0h&8h&@h&Hh&Ph&Xh&`h&hh&ph&xh&�h&�h&�h&�h&�h&�h&�h&�h&�h&�h&�h&�h&�h&�h&�h&�h&i&i&i&i& i&(i&0i&8i&@i&Hi&Pi&Xi&`i&hi&pi&xi&�i&�i&�i&�i&�i&�i&�i&�i&�i&�i&�i&�i&�i&�i&�i&�i&j&j&j&j& j&(j&0j&8j&@j&Hj&Pj&Xj&`j&hj&pj&xj&�j&�j&�j&�j&�j&�j&�j&�j&�j&�j&�j&�j&�j&�j&�j&�j&k&k&k&k& k&(k&0k&8k&@k&Hk&Pk&Xk&`k&hk&pk&xk&�k&�k&�k&�k&�k&�k&�k&�k&�k&�k&�k&�k&�k&�k&�k&�k&l&l&l&l& l&(l&0l&8l&@l&Hl&Pl&Xl&`l&hl&pl&xl&�l&�l&�l&�l&�l&�l&�l&�l&�l&�l&�l&�l&�l&�l&�l&�l&m&m&m&m& m&(m&0m&8m&@m&Hm&Pm&Xm&`m&hm&pm&xm&�m&�m&�m&�m&�m&�m&�m&�m&�m&�m&�m&�m&�m&�m&�m&�m&n&n&n&n& n&(n&0n&8n&@n&Hn&Pn&Xn&`n&hn&pn&xn&�n&�n&�n&�n&�n&�n&�n&�n&�n&�n&�n&�n&�n&�n&�n&�n&o&o&o&o& o&(o&0o&8o&@o&Ho&Po&Xo&`o&ho&po&xo&�o&�o&�o&�o&�o&�o&�o&�o&�o&�o&�o&�o&�o&�o&�o&�o&p&p&p&p& p&(p&0p&8p&@p&Hp&Pp&Xp&`p&hp&pp&xp&�p&�p&�p&�p&�p&�p&�p&�p&�p&�p&�p&�p&�p&�p&�p&�p&q&q&q&q& q&(q&0q&8q&@q&Hq&Pq&Xq&`q&hq&pq&xq&�q&�q&�q&�q&�q&�q&�q&�q&�q&�q&�q&�q&�q&�q&�q&�q&r&r&r&r& r&(r&0r&8r&@r&Hr&Pr&Xr&`r&hr&pr&xr&�r&�r&�r&�r&�r&�r&�r&�r&�r&�r&�r&�r&�r&�r&�r&�r&s&s&s&s& s&(s&0s&8s&@s&Hs&Ps&Xs&`s&hs&ps&xs&�s&�s&�s&�s&�s&�s&�s&�s&�s&�s&�s&�s&�s&�s&�s&�s&t&t&t&t& t&(t&0t&8t&@t&Ht&Pt&Xt&`t&ht&pt&xt&�t&�t&�t&�t&�t&�t&�t&�t&�t&�t&�t&�t&�t&�t&�t&�t&u&u&u&u& u&(u&0u&8u&@u&Hu&Pu&Xu&`u&hu&pu&xu&�u&�u&�u&�u&�u&�u&�u&�u&�u&�u&�u&�u&�u&�u&�u&�u&v&v&v&v& v&(v&0v&8v&@v&Hv&Pv&Xv&`v&hv&pv&xv&�v&�v&�v&�v&�v&�v&�v&�v&�v&�v&�v&�v&�v&�v&�v&�v&w&w&w&w& w&(w&0w&8w&@w&Hw&Pw&Xw&`w&hw&pw&xw&�w&�w&�w&�w&�w&�w&�w&�w&�w&�w&�w&�w&�w&�w&�w&�w&x&x&x&x& x&(x&0x&8x&@x&Hx&Px&Xx&`x&hx&px&xx&�x&�x&�x&�x&�x&�x&�x&�x&�x&�x&�x&�x&�x&�x&�x&�x&y&y&y&y& y&(y&0y&8y&@y&Hy&Py&Xy&`y&hy&py&xy&�y&�y&�y&�y&�y&�y&�y&�y&�y&�y&�y&�y&�y&�y&�y&�y&z&z&z&z& z&(z&0z&8z&@z&Hz&Pz&Xz&`z&hz&pz&xz&�z&�z&�z&�z&�z&�z&�z&�z&�z&�z&�z&�z&�z&�z&�z&�z&{&{&{&{& {&({&0{&8{&@{&H{&P{&X{&`{&h{&p{&x{&�{&�{&�{&�{&�{&�{&�{&�{&�{&�{&�{&�{&�{&�{&�{&�{&|&|&|&|& |&(|&0|&8|&@|&H|&P|&X|&`|&h|&p|&x|&�|&�|&�|&�|&�|&�|&�|&�|&�|&�|&�|&�|&�|&�|&�|&�|&}&}&}&}& }&(}&0}&8}&@}&H}&P}&X}&`}&h}&p}&x}&�}&�}&�}&�}&�}&�}&�}&�}&�}&�}&�}&�}&�}&�}&�}&�}&~&~&~&~& ~&(~&0~&8~&@~&H~&P~&X~&`~&h~&p~&x~&�~&�~&�~&�~&�~&�~&�~&�~&�~&�~&�~&�~&�~&�~&�~&�~&&&&& &(&0&8&@&H&P&X&`&h&p&x&�&�&�&�&�&�&�&�&�&�&�&�&�&�&�&�&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȁ&Ѐ&؀&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȁ&Ё&؁&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȃ&Ђ&؂&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȃ&Ѓ&؃&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȅ&Є&؄&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȅ&Ѕ&؅&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȇ&І&؆&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȇ&Ї&؇&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȉ&Ј&؈&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȉ&Љ&؉&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȋ&Њ&؊&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȋ&Ћ&؋&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȍ&Ќ&،&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȍ&Ѝ&؍&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȏ&Ў&؎&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȏ&Џ&؏&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȑ&А&ؐ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȑ&Б&ؑ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȓ&В&ؒ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȓ&Г&ؓ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȕ&Д&ؔ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȕ&Е&ؕ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȗ&Ж&ؖ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȗ&З&ؗ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ș&И&ؘ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ș&Й&ؙ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ț&К&ؚ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ț&Л&؛&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȝ&М&؜&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȝ&Н&؝&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȟ&О&؞&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȟ&П&؟&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ƞ&Р&ؠ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȡ&С&ء&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȣ&Т&آ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȣ&У&أ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȥ&Ф&ؤ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȥ&Х&إ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȧ&Ц&ئ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȧ&Ч&ا&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȩ&Ш&ب&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȩ&Щ&ة&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȫ&Ъ&ت&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȫ&Ы&ث&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȭ&Ь&ج&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȭ&Э&ح&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȯ&Ю&خ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȯ&Я&د&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȱ&а&ذ&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȱ&б&ر&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&Ȳ&в&ز&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȳ&г&س&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȴ&д&ش&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&ȵ&е&ص&�&�&�&��&�&�&�&�& �&(�&0�&8�&@�&H�&P�&X�&`�&h�&p�&x�&��&��&��&��&��&��&��&��&��&��&��&Ķ&ȶ&̶&ж&Զ&ض&ܶ&�&�&�&�&��&�&�&�&�&$�&,�&4�&<�&D�&L�&T�&\�&d�&l�&t�&|�&��&��&��&��&��&��&��&��&ķ&̷&Է&ܷ&�&�&�&��&�&�&�&�&$�&,�&4�&<�&D�&L�&T�&\�&d�&l�&t�&|�&��&��&��&��&��&��&��&��&ĸ&ȸ&̸&и&Ը&ظ&ܸ&�&�&�&�&�&�&��&��&�&�&�&$�&0�&<�&T�&`�&l�&x�&��&��&��&��&��&̹&�&�&��&�&�&,�&8�&P�&\�&h�&��&��&��&��&��&Ⱥ&Ժ&�&��&�&�&(�&@�&L�&X�&d�&|�&��&��&��&��&Ļ&л&�&�&�&�&$�&0�&H�&T�&`�&l�&��&��&��&��&��&��&̼&�&�&��&�& �&,�&8�&D�&\�&`�&d�&h�&l�&p�&t�&x�&|�&��&��&��&��&��&��&��&��&��&��&��&��&��&��&��&��&��&Ľ&Ƚ&̽&н&Խ&ؽ&ܽ&�&�&�&�&��&��&�&�&�&�&�&�&�&�& �&$�&(�&,�&0�&4�&8�&<�&@�&D�&H�&L�&P�&T�&X�&\�&`�&d�&h�&l�&p�&t�&x�&|�&��&��&��&��&��&��&��&��&��&��&��&��&��&ľ&Ⱦ&̾&о&Ծ&ؾ&ܾ&�&�&�&�&�&�&��&��&�&�&�&�&�&�&�&�& �&$�&(�&,�&0�&4�&8�&<�&@�&P�&T�&X�&l�&p�&t�&��&��&��&��&��&��&ȿ&̿&п&�&�&�&�&�&�& �&$�&(�&@�&D�&H�&\�&`�&d�&l�&|�&��&��&��&��&��&��&��&��&�&�&�&�&��&��&�&�&�&0�&4�&L�&P�&T�&l�&p�&��&��&��&��&��&�&�&�&�&�&�&�&�& �&$�&<�&@�&D�&\�&`�&x�&|�&��&��&��&��&��&��&��&��&��&��&��&��&�&�&,�&0�&4�&L�&P�&h�&l�&p�&��&��&��&��&��&��&��&��&��&��&�&�&�&�& �&$�&<�&@�&D�&X�&\�&`�&x�&|�&��&��&��&��&��&��&��&��&��&��&�&�&�&,�&0�&H�&L�&P�&h�&l�&��&��&��&��&��&��&��&��&��&��&��&�&�&�& �&8�&<�&@�&X�&\�&t�&x�&|�&��&��&��&��&��&��&��&��&��&��&�&�&(�&,�&0�&H�&L�&P�&d�&h�&l�&��&��&��&��&��&��&��&��&��&��&��&��&��&�&�&�&�& �&8�&<�&@�&T�&X�&\�&t�&x�&��&��&��&��&��&��&��&��&��&��&��&�&�&�&(�&,�&D�&H�&L�&d�&h�&��&��&��&��&��&��&��&��&��&��&��&��&�&�&�&4�&8�&<�&T�&X�&p�&t�&x�&��&��&��&��&��&��&��&��&��&��&�&�&$�&(�&,�&D�&H�&`�&d�&h�&��&��&��&��&��&��&��&��&��&��&��&��&��&�&�&�&4�&8�&P�&T�&X�&p�&t�&��&��&��&��&��&��&��&��&��&��&�&�&�&$�&(�&@�&D�&H�&`�&d�&|�&��&��&��&��&��&��&��&��&��&��&��&��&�&�&0�&4�&8�&P�&T�&l�&p�&t�&��&��&��&��&��&��&��&��&��&��&�&�& �&$�&(�&@�&D�&\�&`�&d�&|�&��&��&��&��&��&��&��&��&��&��&��&�&�&�&0�&4�&L�&P�&T�&l�&p�&��&��&��&��&��&��&��&��&��&��&��&��&�&�& �&$�&(�&P�&X�&\�&`�&d�&��&��&��&��&��&��&��&��&��&�&�&�&�&�&@�&H�&L�&P�&T�&|�&��&��&��&��&��&��&��&��&��&��&��&�&�&0�&8�&<�&@�&D�&l�&t�&x�&|�&��&��&��&��&��&��&��&��&��& �&(�&,�&0�&4�&\�&d�&h�&l�&p�&��&��&��&��&��&��&��&��&�&�&�& �&$�&L�&T�&X�&\�&`�&��&��&��&��&��&��&��&��&��&��&�&�&�&�&<�&D�&H�&L�&x�&��&��&��&��&��&��&��&��&��&��&�&�&,�&4�&8�&<�&@�&h�&p�&t�&x�&|�&��&��&��&��&��&��&��&��&��&��&�&$�&(�&,�&0�&X�&`�&d�&h�&l�&��&��&��&��&��&��&��&��&��&��&�&�&�&�& �&H�&P�&T�&X�&\�&��&��&��&��&��&��&��&��&��&��&��&�&�&�&�&8�&@�&D�&H�&L�&t�&|�&��&��&��&��&��&��&��&��&��&��&��&(�&0�&4�&8�&d�&l�&p�&t�&x�&��&��&��&��&��&��&��&��&��&��&�& �&$�&(�&,�&T�&\�&`�&d�&h�&��&��&��&��&��&��&��&��&��&�&�&�&�&�&D�&L�&P�&T�&X�&��&��&��&��&��&��&��&��&��&��&��&�&�&�&�&4�&<�&@�&D�&H�&p�&x�&|�&��&��&��&��&��&��&��&��&��&��&��&��&$�&,�&0�&4�&8�&`�&h�&l�&p�&t�&��&��&��&��&��&��&��&��&��&��&�&�& �&$�&(�&P�&X�&\�&`�&d�&��&��&��&��&��&��&��&��&��&��&�&�&�&�&�&@�&H�&L�&P�&|�&��&��&��&��&��&��&��&��&��&��&�&�&�&0�&8�&<�&@�&l�&t�&x�&|�&��&��&��&��&��&��&��&��& �&(�&,�&0�&\�&d�&h�&l�&��&��&��&��&��&��&��&��&�&�&�& �&L�&T�&X�&\�&��&��&��&��&��&��&��&��&��&��&�&�&�&�&�&<�&D�&H�&L�&P�&x�&��&��&��&��&��&��&��&��&��&��&��&�&,�&4�&8�&<�&@�&h�&p�&t�&x�&|�&��&��&��&��&��&��&��&��&��&�&$�&(�&,�&0�&X�&`�&d�&h�&l�&��&��&��&��&��&��&��&��&��&��&�&�&�&�& �&H�&P�&T�&X�&\�&��&��&��&��&��&��&��&��&��&��&��&�&�&�&�&8�&@�&D�&H�&L�&t�&|�&��&��&��&��&��&��&��&��&��&��&��&��&�&(�&0�&4�&8�&<�&d�&l�&p�&t�&x�&��&��&��&��&��&��&��&��&��&��&�& �&$�&(�&,�&T�&\�&`�&d�&h�&��&��&��&��&��&��&��&��&��&�&�&�&�&�&D�&L�&P�&T�&X�&��&��&��&��&��&��&��&��&��&��&�&�&�&4�&<�&@�&D�&p�&x�&|�&��&��&��&��&��&��&��&��&��&��&��&��&$�&,�&0�&4�&8�&`�&h�&l�&p�&t�&��&��&��&��&��&��&��&��&��&��&�&�& �&$�&(�&P�&X�&\�&`�&d�&��&��&��&��&��&��&��&��&��&��&�&�&�&�&�&@�&H�&L�&P�&T�&|�&��&��&��&��&��&��&��&��&��&��&��&�&�&�&0�&8�&<�&@�&D�&l�&t�&x�&|�&��&��&��&��&��&��&��&��&��&��&��& �&(�&,�&0�&4�&\�&d�&h�&l�&p�&��&��&��&��&��&��&��&��&��&��&�&�&�& �&$�&L�&T�&X�&\�&`�&��&��&��&��&��&��&��&��&��&�&�&�&�&�&<�&D�&H�&L�&P�&x�&��&��&��&��&��&��&��&��&��&��&��&�&�&,�&4�&8�&<�&@�&h�&p�&t�&x�&|�&��&��&��&��&��&�&�&�&�&�&�&$�&(�&,�&0�&X�&`�&d�&h�&l�&��&��&��&��&��&�&�&�&�&�&�&�&�&�& �&H�&P�&T�&X�&\�&��&��&��&��&��&��&�&�&�&�&��&�&�&�&�&8�&@�&D�&H�&L�&t�&|�&��&��&��&��&��&��&��&�&�&�&��&��&�&(�&0�&4�&8�&<�&d�&l�&p�&t�&x�&��&��&��&��&��&�&�&�&�&�&�& �&$�&(�&,�&T�&\�&`�&d�&h�&��&��&��&��&��&�&�&�&�&�&�&�&�&�&�&D�&L�&P�&T�&X�&��&��&��&��&��&��&�&�&�&�&��&�&�&�&�&4�&<�&@�&D�&H�&p�&x�&|�&��&��&��&��&��&��&��&�&�&�&��&��&$�&,�&0�&4�&8�&`�&h�&l�&p�&t�&��&��&��&��&��&�&�&�&�&�&�&�& �&$�&(�&P�&X�&\�&`�&��&��&��&��&��&�&�&�&�&�&�&�&�&�&�&@�&H�&L�&P�&T�&|�&��&��&��&��&��&��&�&�&�&�&��&''0'8'<'@'D'l't'x'|'�'�'�'�'�'�'�'�'�' '(','0'4'\'d'h'l'�'�'�'�'�'�'�'�'�'''' '$'L'T'X'\'�'�'�'�'�'�'�'�'�'''''<'D'H'L'P'x'�'�'�'�'�'�'�'�'�'�'�''','4'8'<'h'p't'x'|'�'�'�'�'�'�'�'�'�''$'(','0'X'`'d'h'�'�'�'�'�'�'�'�'�''''' 'H'P'T'X'\'�'�'�'�'�'�'�'�'�'�'�'''''8'@'D'H'L't'|'�'�'�'�'�'�'�'�'�'�''('0'4'8'<'d'l'p't'x'�'�'�'�'�'�'�'�'�'	' 	'$	'(	'T	'\	'`	'd	'�	'�	'�	'�	'�	'�	'�	'�	'�	'
'
'
'
'
'D
'L
'P
'T
'X
'�
'�
'�
'�
'�
'�
'�
'�
'�
''''4'<'@'D'H'p'x'|'�'�'�'�'�'�'�'�'�'�'�'�'$','0'4'8'`'h'l'p't'�'�'�'�'�'�'�'�'�'
'
' 
'$
'(
'P
'X
'\
'`
'd
'�
'�
'�
'�
'�
'�
'�
'�
'�
'�
''''''@'H'L'P'T'|'�'�'�'�'�'�'�'�'�'�'�'''0'8'<'@'l't'x'|'�'�'�'�'�'�'�'�' '(','0'\'d'h'l'�'�'�'�'�'�'�'�'''' '$'L'T'X'\'`'�'�'�'�'�'�'�'�'�''''''<'D'H'L'x'�'�'�'�'�'�'�'�'�'�'�''','4'8'<'h'p't'x'|'�'�'�'�'�'�'�'�'�''$'(','X'`'d'h'l'�'�'�'�'�'�'�'�'�'''''H'P'T'X'\'�'�'�'�'�'�'�'�'�'�''''8'@'D'H'L't'|'�'�'�'�'�'�'�'�'�'�'�'�''('0'4'8'<'d'l'p't'|'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'' '$'('0'4'8'T'\'`'d'l'p't'�'�'�'�'�'�'�'�'�'�'�'�'�'�''''' '$'('D'L'P'T'\'`'d'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'''''''4'<'@'D'L'P'T'p'x'|'�'�'�'�'�'�'�'�'�'�'�'�'�'�''''$','0'4'<'@'D'`'h'l'p'x'|'�'�'�'�'�'�'�'�'�'�'�'�'�'�''' '$','4'P'X'\'`'h'p'�'�'�'�'�'�'�'�'�'�'�'�''''''$'@'H'L'P'X'`'|'�'�'�'�'�'�'�'�'�'�'�'�'�'�''''''0'8'<'@'H'L'P'l't'x'|'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�''' '(','0'8'@'\'d'h'l't'|'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�'�' ' ' ' '  '( '0 'L 'T 'X '\ 'd 'l '� '� '� '� '� '� '� '� '� '� '� '� '!'!'!'!'!' !'(!',!'<!'D!'H!'L!'T!'\!'d!'h!'x!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'�!'"'"'"'"'"',"'4"'8"'<"'D"'L"'T"'X"'h"'p"'t"'x"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'�"'#'$#'(#',#'X#'`#'d#'h#'�#'�#'�#'�#'�#'�#'�#'�#'$'$'$'$' $'H$'P$'T$'X$'\$'�$'�$'�$'�$'�$'�$'�$'�$'�$'�$'�$'%'%'%'%'8%'@%'D%'H%'t%'|%'�%'�%'�%'�%'�%'�%'�%'�%'�%'�%'(&'0&'4&'8&'<&'d&'l&'p&'t&'x&'�&'�&'�&'�&'�&'�&'�&'�&'�&'�&''' ''$''('',''T''\''`''d''h''�''�''�''�''�''�''�''�''�''('('('('D('L('P('T('�('�('�('�('�('�('�('�('�('�(')')')')'4)'<)'@)'D)'H)'p)'x)'|)'�)'�)'�)'�)'�)'�)'�)'�)'�)'�)'$*',*'0*'4*'`*'h*'l*'p*'�*'�*'�*'�*'�*'�*'�*'�*'+'+' +'$+'P+'X+'\+'`+'d+'�+'�+'�+'�+'�+',',',',,'0,'4,'L,'P,'T,'h,'l,'p,'�,'�,'�,'�,'�,'�,'�,'�,'�,'�,'�,'�,'-'-'-'-' -'$-',-'<-'@-'D-'X-'\-'`-'x-'|-'�-'�-'�-'�-'�-'�-'�-'�-'�-'�-'�-'�-'.'.'.',.'0.'H.'L.'P.'h.'l.'�.'�.'�.'�.'�.'�.'�.'�.'�.'�.'�.'/'/'/' /'8/'</'@/'X/'\/'t/'x/'|/'�/'�/'�/'�/'�/'�/'�/'�/'�/'�/'�/'0'0'(0',0'00'H0'L0'd0'h0'l0'�0'�0'�0'�0'�0'�0'�0'�0'�0'�0'�0'�0'1'1'1'1' 1'81'<1'T1'X1'\1't1'x1'�1'�1'�1'�1'�1'�1'�1'�1'�1'�1'2'2'2'(2',2'D2'H2'L2'd2'h2'�2'�2'�2'�2'�2'�2'�2'�2'�2'�2'�2'�2'3'3'3'43'83'<3'T3'X3'p3't3'x3'�3'�3'�3'�3'�3'�3'�3'�3'�3'�3'4'4'4'$4'(4',4'D4'H4'L4'`4'd4'h4'�4'�4'�4'�4'�4'�4'�4'�4'�4'�4'�4'�4'�4'�4'5'5'5'5'45'85'P5'T5'X5'p5't5'�5'�5'�5'�5'�5'�5'�5'�5'�5'�5'�5'6'6'6'$6'(6'@6'D6'H6'`6'd6'|6'�6'�6'�6'�6'�6'�6'�6'�6'�6'�6'�6'�6'7'7'07'47'87'P7'T7'l7'p7't7'�7'�7'�7'�7'�7'�7'�7'�7'�7'�7'8'8' 8'$8'(8'@8'D8'\8'`8'd8'|8'�8'�8'�8'�8'�8'�8'�8'�8'�8'�8'�8'�8'9'9'9'09'49'L9'P9'T9'l9'p9'�9'�9'�9'�9'�9'�9'�9'�9'�9'�9':':':' :'$:'<:'@:'D:'\:'`:'x:'|:'�:'�:'�:'�:'�:'�:'�:'�:'�:'�:'�:';';',;'0;'4;'L;'P;'h;'l;'p;'�;'�;'�;'�;'�;'�;'�;'�;'�;'�;'<'<'<' <'$<'<<'@<'X<'\<'`<'x<'|<'�<'�<'�<'�<'�<'�<'�<'�<'�<'�<'='='=' ='(=',='0='4='\='d='h='l='�='�='�='�='�='�='�='�='�='�='>'>'>' >'L>'T>'X>'\>'`>'�>'�>'�>'�>'�>'�>'�>'�>'�>'�>'?'?'?'?'?'<?'D?'H?'L?'P?'x?'�?'�?'�?'�?'�?'�?'�?'�?'�?'�?'�?'@'@',@'4@'8@'<@'h@'p@'t@'x@'�@'�@'�@'�@'�@'�@'�@'�@'�@'�@'A'$A'(A',A'XA'`A'dA'hA'�A'�A'�A'�A'�A'�A'�A'�A'�A'�A'B'B'B'B' B'HB'PB'TB'XB'\B'�B'�B'�B'�B'�B'�B'�B'�B'�B'C'C'C'8C'@C'DC'HC'tC'|C'�C'�C'�C'�C'�C'�C'�C'�C'�C'�C'�C'�C'D'(D'0D'4D'8D'<D'dD'lD'pD'tD'xD'�D'�D'�D'�D'�D'�D'�D'�D'�D'�D'E' E'$E'(E',E'TE'\E'`E'dE'hE'�E'�E'�E'�E'�E'�E'�E'�E'�E'�E'F'F'F'F'F'DF'LF'PF'TF'XF'�F'�F'�F'�F'�F'�F'�F'�F'�F'�F'�F'G'G'G'G'4G'<G'@G'DG'HG'pG'xG'|G'�G'�G'�G'�G'�G'�G'�G'�G'�G'�G'�G'�G'$H',H'0H'4H'8H'`H'hH'lH'pH'tH'�H'�H'�H'�H'�H'�H'�H'�H'�H'I'I' I'$I'(I'PI'XI'\I'`I'dI'�I'�I'�I'�I'�I'�I'�I'�I'�I'�I'J'J'J'J'J'@J'HJ'LJ'PJ'TJ'|J'�J'�J'�J'�J'�J'�J'�J'�J'�J'�J'�J'K'K'K'0K'8K'<K'@K'DK'lK'tK'xK'|K'�K'�K'�K'�K'�K'�K'�K'�K'�K'�K'�K' L'(L',L'0L'4L'\L'dL'hL'lL'pL'�L'�L'�L'�L'�L'�L'�L'�L'�L'�L'M'M'M' M'$M'LM'TM'XM'\M'�M'�M'�M'�M'�M'�M'�M'�M'�M'N'N'N'N'N'<N'DN'HN'LN'PN'xN'�N'�N'�N'�N'�N'�N'�N'�N'�N'�N'�N'�N'O'O',O'4O'8O'<O'@O'hO'pO'tO'xO'|O'�O'�O'�O'�O'�O'�O'�O'�O'�O'�O'P'$P'(P',P'0P'XP'`P'dP'hP'lP'�P'�P'�P'�P'�P'�P'�P'�P'�P'�P'Q'Q'Q'Q' Q'HQ'PQ'TQ'XQ'\Q'�Q'�Q'�Q'�Q'�Q'�Q'�Q'�Q'�Q'�Q'�Q'R'R'R'R'8R'@R'DR'HR'LR'tR'|R'�R'�R'�R'�R'�R'�R'�R'�R'�R'�R'�R'�R'S'(S'0S'4S'8S'<S'dS'lS'pS'tS'xS'�S'�S'�S'�S'�S'�S'�S'�S'�S'�S'T' T'$T'(T',T'TT'\T'`T'dT'hT'�T'�T'�T'�T'�T'�T'�T'�T'�T'�T'U'U'U'U'U'DU'LU'PU'TU'XU'�U'�U'�U'�U'�U'�U'�U'�U'�U'�U'�U'V'V'V'V'4V'<V'@V'DV'HV'pV'xV'|V'�V'�V'�V'�V'�V'�V'�V'�V'�V'�V'�V'�V'$W',W'0W'4W'8W'`W'hW'lW'pW'tW'�W'�W'�W'�W'�W'�W'�W'�W'�W'�W'X'X' X'$X'(X'PX'XX'\X'`X'dX'�X'�X'�X'�X'�X'�X'�X'�X'�X'�X'Y'Y'Y'Y'Y'@Y'HY'LY'PY'TY'|Y'�Y'�Y'�Y'�Y'�Y'�Y'�Y'�Y'�Y'�Y'�Y'Z'Z'Z'0Z'8Z'<Z'@Z'DZ'lZ'tZ'xZ'|Z'�Z'�Z'�Z'�Z'�Z'�Z'�Z'�Z'�Z'�Z'�Z' ['([',['0['4['\['d['h['l['p['�['�['�['�['�['�['�['�['�['�['\'\'\' \'$\'L\'T\'X\'\\'`\'�\'�\'�\'�\'�\'�\'�\'�\'�\'�\']']']']']'<]'D]'H]'L]'P]'x]'�]'�]'�]'�]'�]'�]'�]'�]'�]'�]'�]'�]'^'^',^'4^'8^'<^'@^'h^'p^'t^'x^'|^'�^'�^'�^'�^'�^'�^'�^'�^'�^'�^'_'$_'(_',_'0_'X_'`_'d_'h_'l_'�_'�_'�_'�_'�_'�_'�_'�_'�_'�_'`'`'`'`' `'H`'P`'T`'X`'\`'�`'�`'�`'�`'�`'�`'�`'�`'�`'�`'�`'a'a'a'a'8a'@a'Da'Ha'La'ta'|a'�a'�a'�a'�a'�a'�a'�a'�a'�a'�a'�a'�a'b'(b'0b'4b'8b'<b'db'lb'pb'tb'xb'�b'�b'�b'�b'�b'�b'�b'�b'�b'�b'c' c'$c'(c',c'Tc'\c'`c'dc'hc'�c'�c'�c'�c'�c'�c'�c'�c'�c'�c'd'd'd'd'd'Dd'Ld'Pd'Td'Xd'�d'�d'�d'�d'�d'�d'�d'�d'�d'�d'�d'e'e'e'e'4e'<e'@e'De'He'pe'xe'|e'�e'�e'�e'�e'�e'�e'�e'�e'�e'�e'�e'�e'$f',f'0f'4f'8f'`f'hf'lf'pf'tf'�f'�f'�f'�f'�f'�f'�f'�f'�f'�f'g'g' g'$g'(g'Pg'Xg'\g'`g'dg'�g'�g'�g'�g'�g'�g'�g'�g'�g'�g'h'h'h'h'h'@h'Hh'Lh'Ph'Th'|h'�h'�h'�h'�h'�h'�h'�h'�h'�h'�h'�h'i'i'i'0i'8i'<i'@i'Di'li'ti'xi'|i'�i'�i'�i'�i'�i'�i'�i'�i'�i'�i'�i' j'(j',j'0j'4j'\j'dj'hj'lj'pj'�j'�j'�j'�j'�j'�j'�j'�j'�j'�j'k'k'k' k'$k'Lk'Tk'Xk'\k'`k'�k'�k'�k'�k'�k'�k'�k'�k'�k'�k'l'l'l'l'l'<l'Dl'Hl'Ll'Pl'xl'�l'�l'�l'�l'�l'�l'�l'�l'�l'�l'�l'�l'm'm',m'4m'8m'<m'@m'hm'pm'tm'xm'|m'�m'�m'�m'�m'�m'�m'�m'�m'�m'�m'n'$n'(n',n'0n'Xn'`n'dn'hn'ln'�n'�n'�n'�n'�n'�n'�n'�n'�n'�n'o'o'o'o' o'Ho'Po'To'Xo'\o'�o'�o'�o'�o'�o'�o'�o'�o'�o'�o'�o'p'p'p'p'8p'@p'Dp'Hp'Lp'tp'|p'�p'�p'�p'�p'�p'�p'�p'�p'�p'�p'�p'�p'q'(q'0q'4q'8q'<q'dq'lq'pq'tq'xq'�q'�q'�q'�q'�q'�q'�q'�q'�q'�q'r' r'$r'(r',r'Tr'\r'`r'dr'hr'�r'�r'�r'�r'�r'�r'�r'�r'�r'�r's's's's's'Ds'Ls'Ps'Ts'Xs'�s'�s'�s'�s'�s'�s'�s'�s'�s'�s'�s't't't't'4t'<t'@t'Dt'Ht'pt'xt'|t'�t'�t'�t'�t'�t'�t'�t'�t'�t'�t'�t'�t'$u',u'0u'4u'8u'`u'hu'lu'pu'tu'�u'�u'�u'�u'�u'�u'�u'�u'�u'�u'v'v' v'$v'(v'Pv'Xv'\v'`v'dv'�v'�v'�v'�v'�v'�v'�v'�v'�v'�v'w'w'w'w'w'@w'Hw'Lw'Pw'Tw'|w'�w'�w'�w'�w'�w'�w'�w'�w'�w'�w'�w'x'x'x'0x'8x'<x'@x'Dx'lx'tx'xx'|x'�x'�x'�x'�x'�x'�x'�x'�x'�x'�x'�x' y'(y',y'0y'4y'\y'dy'hy'ly'py'�y'�y'�y'�y'�y'�y'�y'�y'�y'�y'z'z'z' z'$z'Lz'Tz'Xz'\z'`z'�z'�z'�z'�z'�z'�z'�z'�z'�z'�z'{'{'{'{'{'<{'D{'H{'L{'P{'x{'�{'�{'�{'�{'�{'�{'�{'�{'�{'�{'�{'�{'|'|',|'4|'8|'<|'@|'h|'p|'t|'x|'||'�|'�|'�|'�|'�|'�|'�|'�|'�|'�|'}'$}'(}',}'0}'X}'`}'d}'h}'l}'�}'�}'�}'�}'�}'�}'�}'�}'~'~'~'~'H~'P~'T~'X~'�~'�~'�~'�~'�~'�~'�~'�~'�~''''8'@'D'H't'|'�'�'�'�'�'�'�'�'�'�'�'(�'0�'4�'8�'<�'d�'l�'p�'t�'x�'��'��'��'��'��'܀'�'�'�'�'�' �'$�'(�'T�'\�'`�'d�'��'��'��'��'��'́'ԁ'؁'܁'�'�'�'�'�'�'D�'L�'P�'T�'��'��'��'��'��'��'Ă'Ȃ'̂'Ђ'��'�'�'�'�'4�'<�'@�'D�'H�'p�'x�'|�'��'��'��'��'��'��'��'�'�'�'��'��'$�',�'0�'4�'8�'`�'h�'l�'p�'t�'��'��'��'��'��'؄'�'�'�'�'�'�' �'$�'(�'P�'X�'\�'`�'d�'��'��'��'��'��'ȅ'Ѕ'ԅ'؅'܅'�'�'�'�'�'@�'H�'L�'P�'T�'|�'��'��'��'��'��'��'Ć'Ȇ'̆'�'��'�'�'0�'8�'<�'@�'D�'l�'t�'x�'|�'��'��'��'��'��'��'�'�'�'�' �'(�',�'0�'4�'\�'d�'h�'l�'��'��'��'��'Ԉ'܈'�'�'�'�'�'�' �'$�'L�'T�'X�'\�'`�'��'��'��'��'��'ĉ'̉'Љ'ԉ'؉'�'�'�'�'�'<�'D�'H�'L�'P�'x�'��'��'��'��'��'��'��'Ċ'Ȋ'�'��'��'�'�',�'4�'8�'<�'@�'h�'p�'t�'x�'|�'��'��'��'��'��'�'�'�'�'�'�'$�'(�',�'0�'X�'`�'d�'h�'l�'��'��'��'��'��'Ќ'،'܌'�'�'�'�'�'�' �'H�'P�'T�'X�'\�'��'��'��'��'��'��'ȍ'̍'Ѝ'ԍ'��'�'�'�'�'8�'@�'D�'H�'L�'t�'|�'��'��'��'��'��'��'��'Ď'�'�'��'��'�'(�'0�'4�'8�'<�'d�'l�'p�'t�'x�'��'��'��'��'��'܏'�'�'�'�'�' �'$�'(�',�'T�'\�'`�'d�'h�'��'��'��'��'��'̐'Ԑ'ؐ'ܐ'�'�'�'�'�'�'D�'L�'P�'T�'X�'��'��'��'��'��'��'đ'ȑ'̑'Б'��'�'�'�'�'4�'<�'@�'D�'H�'p�'x�'|�'��'��'��'��'��'��'��'�'�'�'��'��'$�',�'0�'4�'8�'`�'h�'l�'p�'t�'��'��'��'��'��'ؓ'�'�'�'�'�'�' �'$�'(�'P�'X�'\�'`�'d�'��'��'��'��'��'Ȕ'Д'Ԕ'ؔ'ܔ'�'�'�'�'�'@�'H�'L�'P�'T�'|�'��'��'��'��'��'��'ĕ'ȕ'̕'�'��'�'�'�'0�'8�'<�'@�'D�'l�'t�'x�'|�'��'��'��'��'��'��'�'�'�'�'��' �'(�',�'0�'4�'\�'d�'h�'l�'p�'��'��'��'��'��'ԗ'ܗ'�'�'�'�'�'�' �'$�'L�'T�'X�'\�'`�'��'��'��'��'��'Ę'̘'И'Ԙ'ؘ'�'�'�'�'�'<�'D�'H�'L�'P�'x�'��'��'��'��'��'��'��'ę'ș'�'��'��'�'�',�'4�'8�'<�'@�'h�'p�'t�'x�'|�'��'��'��'��'��'�'�'�'�'�'�'$�'(�',�'0�'X�'`�'d�'h�'l�'��'��'��'��'��'Л'؛'ܛ'�'�'�'�'�'�' �'H�'P�'T�'X�'\�'��'��'��'��'��'��'Ȝ'̜'М'Ԝ'��'�'�'�'�'8�'@�'D�'H�'L�'t�'|�'��'��'��'��'��'��'��'ĝ'�'�'��'��'�'(�'0�'4�'8�'<�'d�'l�'p�'t�'x�'��'��'��'��'��'ܞ'�'�'�'�'�' �'$�'(�',�'T�'\�'`�'d�'h�'��'��'��'��'��'̟'ԟ'؟'ܟ'�'�'�'�'�'�'D�'L�'P�'T�'X�'��'��'��'��'��'��'Ġ'Ƞ'̠'Р'��'�'�'�'�'4�'<�'@�'D�'H�'p�'x�'|�'��'��'��'��'��'��'��'�'�'�'��'��'$�',�'0�'4�'8�'`�'h�'l�'p�'t�'��'��'��'��'��'آ'�'�'�'�'�'�' �'$�'(�'P�'X�'\�'`�'d�'��'��'��'��'��'ȣ'У'ԣ'أ'ܣ'�'�'�'�'�'@�'H�'L�'P�'T�'|�'��'��'��'��'��'��'Ĥ'Ȥ'̤'�'��'�'�'�'0�'8�'<�'@�'D�'l�'t�'x�'|�'��'��'��'��'��'��'�'�'�'�'��' �'(�',�'0�'4�'\�'d�'h�'l�'p�'��'��'��'��'��'Ԧ'ܦ'�'�'�'�'�'�' �'$�'L�'T�'X�'\�'`�'��'��'��'��'��'ħ'̧'Ч'ԧ'ا'�'�'�'�'�'<�'D�'H�'L�'P�'x�'��'��'��'��'��'��'��'Ĩ'Ȩ'�'��'��'�'�',�'4�'8�'<�'@�'h�'p�'t�'x�'|�'��'��'��'��'��'�'�'�'�'�'�'$�'(�',�'0�'X�'`�'d�'h�'l�'��'��'��'��'��'Ъ'ت'ܪ'�'�'�'�'�'�' �'H�'P�'T�'X�'\�'��'��'��'��'��'��'ȫ'̫'Ы'ԫ'��'�'�'�'�'8�'@�'D�'H�'L�'t�'|�'��'��'��'��'��'��'��'Ĭ'�'�'��'��'�'(�'0�'4�'8�'<�'d�'l�'p�'t�'x�'��'��'��'��'��'ܭ'�'�'�'�'�' �'$�'(�',�'T�'\�'`�'d�'h�'��'��'��'��'��'̮'Ԯ'خ'ܮ'�'�'�'�'�'�'D�'L�'P�'T�'X�'��'��'��'��'��'��'į'ȯ'̯'Я'��'�'�'�'�'4�'<�'@�'D�'H�'p�'x�'|�'��'��'��'��'��'��'��'�'�'�'��'��'$�',�'0�'4�'8�'`�'h�'l�'p�'t�'��'��'��'��'��'ر'�'�'�'�'�'�' �'$�'(�'P�'X�'\�'`�'d�'��'��'��'��'��'Ȳ'в'Բ'ز'ܲ'�'�'�'�'�'@�'H�'L�'P�'T�'|�'��'��'��'��'��'��'ij'ȳ'̳'�'��'�'�'�'0�'8�'<�'@�'D�'l�'t�'x�'|�'��'��'��'��'��'��'�'�'�'�'��' �'(�',�'0�'4�'\�'d�'h�'l�'p�'��'��'��'��'��'Ե'ܵ'�'�'�'�'�'�' �'$�'L�'T�'X�'\�'`�'��'��'��'��'��'Ķ'̶'ж'Զ'ض'�'�'�'�'�'<�'D�'H�'L�'P�'x�'��'��'��'��'��'��'��'ķ'ȷ'�'��'��'�'�',�'4�'8�'<�'@�'h�'p�'t�'x�'|�'��'��'��'��'��'�'�'�'�'�'�'$�'(�',�'0�'X�'`�'d�'h�'l�'��'��'��'��'��'й'ع'ܹ'�'�'�'�'�'�' �'H�'P�'T�'X�'\�'��'��'��'��'��'��'Ⱥ'̺'к'Ժ'��'�'�'�'�'8�'@�'D�'H�'L�'t�'|�'��'��'��'��'��'��'��'Ļ'�'�'��'��'�'(�'0�'4�'8�'<�'d�'l�'p�'t�'x�'��'��'��'��'��'ܼ'�'�'�'�'�' �'$�'(�',�'T�'\�'`�'d�'h�'��'��'��'��'��'̽'Խ'ؽ'ܽ'�'�'�'�'�'�'D�'L�'P�'T�'��'��'��'��'��'��'ľ'Ⱦ'̾'о'��'�'�'�'�'4�'<�'@�'D�'p�'x�'|�'��'��'��'��'��'��'��'�'�'�'��'��'$�',�'0�'4�'8�'`�'h�'l�'p�'t�'��'��'��'��'��'�'�'�'�'�'�'�' �'$�'(�'P�'X�'\�'`�'d�'��'��'��'��'��'�'�'�'�'�'�'�'�'�'�'@�'H�'L�'P�'T�'|�'��'��'��'��'��'��'��'��'��'��'��'�'�'�'0�'8�'<�'@�'D�'l�'t�'x�'|�'��'��'��'��'��'��'��'��'��'��'��' �'(�',�'0�'4�'\�'d�'h�'l�'p�'��'��'��'��'��'��'��'��'��'��'�'�'�' �'$�'L�'T�'X�'\�'`�'��'��'��'��'��'��'��'��'��'��'�'�'�'�'�'<�'D�'H�'L�'P�'x�'��'��'��'��'��'��'��'��'��'��'��'��'�'�',�'4�'8�'<�'@�'h�'p�'t�'x�'|�'��'��'��'��'��'��'��'��'��'��'�'$�'(�',�'0�'X�'`�'d�'h�'l�'��'��'��'��'��'��'��'��'��'�'�'�'�' �'H�'P�'T�'X�'\�'��'��'��'��'��'��'��'��'��'��'��'�'�'�'�'8�'@�'D�'H�'L�'t�'|�'��'��'��'��'��'��'��'��'��'��'��'(�'0�'4�'8�'d�'l�'p�'t�'x�'��'��'��'��'��'��'��'��'��'�' �'$�'(�',�'T�'\�'`�'d�'��'��'��'��'��'��'��'��'��'�'�'�'�'D�'L�'P�'T�'X�'��'��'��'��'��'��'��'��'��'��'�'�'�'�'4�'<�'@�'D�'p�'x�'|�'��'��'��'��'��'��'��'��'��'��'��'$�',�'0�'4�'`�'h�'l�'p�'t�'��'��'��'��'��'��'��'��'��'�'�' �'$�'P�'X�'\�'`�'d�'��'��'��'��'��'��'��'��'��'��'�'�'�'�'�'@�'H�'L�'P�'T�'|�'��'��'��'��'��'��'��'��'��'��'��'�'�'0�'8�'<�'@�'l�'t�'x�'|�'��'��'��'��'��'��'��'��'��'��'��' �'(�',�'0�'4�'\�'d�'h�'l�'��'��'��'��'��'��'��'��'�'�'�' �'L�'T�'X�'\�'`�'��'��'��'��'��'��'��'��'��'��'�'�'�'�'<�'D�'H�'L�'x�'��'��'��'��'��'��'��'��'��'��'��'�'�',�'4�'8�'<�'@�'h�'p�'t�'x�'|�'��'��'��'��'��'��'��'��'��'��'�'$�'(�',�'X�'`�'d�'h�'l�'��'��'��'��'��'��'��'��'��'��'�'�'�'�' �'H�'P�'T�'X�'\�'��'��'��'��'��'��'��'��'��'��'��'�'�'�'�'8�'@�'D�'H�'L�'t�'|�'��'��'��'��'��'��'��'��'��'��'(�'0�'4�'8�'d�'l�'p�'t�'��'��'��'��'��'��'��'��'�' �'$�'(�'T�'\�'`�'d�'��'��'��'��'��'��'��'��'��'��'�'�'�'�'�'D�'L�'P�'T�'��'��'��'��'��'��'��'��'��'��'�'�'�'�'4�'<�'@�'D�'p�'x�'|�'��'��'��'��'��'��'��'��'��'��'��'$�',�'0�'4�'`�'h�'l�'p�'t�'��'��'��'��'��'��'��'��'��'�'�' �'$�'P�'X�'\�'`�'d�'��'��'��'��'��'��'��'��'��'�'�'�'�'�'@�'H�'L�'P�'|�'��'��'��'��'��'��'��'��'��'��'�'�'�'0�'8�'<�'@�'D�'l�'t�'x�'|�'��'��'��'��'��'��'��'��'��'��'��'��' �'(�',�'0�'4�'8�'\�'d�'h�'l�'p�'t�'��'��'��'��'��'��'��'��'��'��'��'��'�'�'�' �'$�'(�'L�'T�'X�'\�'`�'d�'��'��'��'��'��'��'��'��'��'��'��'��'�'�'�'�'�'�'<�'D�'H�'L�'P�'T�'x�'��'��'��'��'��'��'��'��'��'��'��'��'��'��'�'�'�',�'4�'8�'<�'@�'D�'h�'p�'t�'x�'|�'��'��'��'��'��'��'��'��'��'��'��'��'��'�'$�'(�',�'0�'4�'X�'`�'d�'h�'l�'p�'��'��'��'��'��'��'��'��'��'��'��'��'�'�'�'�' �'$�'H�'P�'T�'X�'\�'��'��'��'��'��'��'��'��'��'�'�'�'8�'@�'D�'H�'t�'|�'��'��'��'��'��'��'��'��'��'��'��'��'�'(�'0�'4�'8�'<�'d�'l�'p�'t�'x�'��'��'��'��'��'��'��'��'�' �'$�'(�'T�'\�'`�'d�'��'��'��'��'��'��'��'��'��'��'�'�'�'�'�'D�'L�'P�'T�'X�'��'��'��'��'��'��'��'��'��'��'��'�'�'�'�'4�'<�'@�'D�'p�'x�'|�'��'��'��'��'��'��'��'��'��'$�',�'0�'4�'8�'`�'h�'l�'p�'t�'��'��'��'��'��'��'��'��'��'��'�'�' �'$�'P�'X�'\�'`�'��'��'��'��'��'��'��'��'�'�'�'�'@�'H�'L�'P�'|�'��'��'��'��'��'��'��'��'��'��'�'�'�'x�'|�'��'��'��'��'��'��'��'��'��'��'��'��'��'�'�'�',�'0�'4�'H�'L�'P�'h�'l�'p�'��'��'��'��'��'��'��'��'��'��'��'��'��'��'�'�'�' �'$�'8�'<�'@�'X�'\�'t�'x�'|�'��'��'��'��'��'��'��'��'��'��'�'�'(�',�'0�'H�'L�'d�'h�'l�'��'��'��'��'��'��'��'��'��'��'��'�'�'�'�' �'8�'<�'T�'X�'\�'t�'x�'��'��'��'��'��'��'��'��'��'��'�'�'�'(�',�'0�'D�'H�'L�'d�'h�'l�'��'��'��'��'��'��'��'�'�'�'��'��'�'�'�'4�'8�'<�'T�'X�'p�'t�'x�'��'��'��'��'��'�'�'�'�'�'�'�'$�'(�',�'D�'H�'`�'d�'h�'��'��'��'��'��'��'��'�'�'�'��'��'�'�'�'4�'8�'P�'T�'X�'p�'t�'x�'��'��'��'��'��'��'�'�'�'�'�'�'�'�'�'$�'(�',�'@�'D�'H�'`�'d�'h�'|�'��'��'��'��'��'��'��'�'�'�'��'��'�'�'�'0�'4�'8�'P�'T�'l�'p�'t�'��'��'��'��'��'�'�'�'�'�'�'�' �'$�'(�'@�'D�'\�'`�'d�'|�'��'��'��'��'��'��'�'�'�'�'��'�'�'�'0�'4�'L�'P�'T�'l�'p�'��'��'��'��'��'�'�'�'�'�'�'�'�'�' �'$�'<�'@�'D�'\�'`�'x�'|�'��'��'��'��'��'��'�'�'�'�'��'�'�',�'0�'4�'L�'P�'h�'l�'p�'��'��'��'��'��'�'�'�'�'�'�'�'�' �'$�'<�'@�'X�'\�'`�'x�'|�'��'��'��'��'��'�'�'�'�'�'(((,(0(H(L(P(h(l(�(�(�(�(�(�(�(�(�(�(�(((( (8(<(@(X(\(t(x(|(�(�(�(�(�(�(�(�(�(�(((((<(D(H(L(P(x(�(�(�(�(�(�(�(�(�(�(�(((,(4(8(<(@(h(p(t(x(|(�(�(�(�(�(�(�(�(�(�(($(((,(X(`(d(h(l(�(�(�(�(�(�(�(�(�(((((H(P(T(X(\(�(�(�(�(�(�(�(�(�(�((((8(@(D(H(L(t(|(�(�(�(�(�(�(�(�(�(�(�(�((((0(4(8(d(l(p(t(�(�(�(�(�(�(�(�(( ($(((,(T(\(`(d(h(�(�(�(�(�(�(�(�(�(�(	(	(	(	(	(D	(L	(P	(T	(X	(�	(�	(�	(�	(�	(�	(�	(�	(�	(�	(�	(
(
(
(
(4
(<
(@
(D
(H
(p
(x
(|
(�
(�
(�
(�
(�
(�
(�
(�
(�
(�
(�
(�
($(,(0(4(8(`(h(l(p(t(�(�(�(�(�(�(�(�(�((( ($(P(X(\(`(�(�(�(�(�(�(�(�(�(�(
(
(
(
(
(@
(H
(L
(P
(T
(|
(�
(�
(�
(�
(�
(�
(�
(�
(�
(�
(�
(((0(8(<(@(D(l(t(x(|(�(�(�(�(�(�(�(�(�(�(�( (((,(0(4(\(d(h(l(p(�(�(�(�(�(�(�(�(�(�(((( ($(L(T(X(\(`(�(�(�(�(�(�(�(�(�(�((((((<(D(H(L(P(x(�(�(�(�(�(�(�(�(�(�(�(�(((,(4(8(<(@(h(p(t(x(|(�(�(�(�(�(�(�(�(�(($(((,(X(`(d(h(l(�(�(�(�(�(�(�(�(�(�((((( (H(P(T(X(\(�(�(�(�(�(�(�(�(�(�(�(((((8(@(D(H(L(t(|(�(�(�(�(�(�(�(�(�(�(�(�((((0(4(8(<(d(l(p(t(�(�(�(�(�(�(�(�(( ($(((T(\(`(d(�(�(�(�(�(�(�(�(((((D(L(P(T(�(�(�(�(�(�(�(�(�(�(((((4(<(@(D(H(p(x(|(�(�(�(�(�(�(�(�(�(�(�($(,(0(4(`(h(l(p(t(�(�(�(�(�(�(�(�(�(�((( ($(P(X(\(`(d(�(�(�(�(�(�(�(�(�(�((((((@(H(L(P(T(|(�(�(�(�(�(�(�(�(�(�(�((((0(8(<(@(D(l(t(x(|(�(�(�(�(�(�(�(�(�(�(�( (((,(0(4(\(d(h(l(p(�(�(�(�(�(�(�(�(�(�(((( ($(L(T(X(\(`(�(�(�(�(�(�(�(�(�( ( ( ( ( (< (D (H (L (P (x (� (� (� (� (� (� (� (� (� (� (� (!(!(,!(4!(8!(<!(h!(p!(t!(x!(�!(�!(�!(�!(�!(�!(�!(�!(�!(�!("($"(("(,"(0"(X"(`"(d"(h"(l"(�"(�"(�"(�"(�"(�"(�"(�"(�"(�"(#(#(#(#( #(H#(P#(T#(X#(\#(�#(�#(�#(�#(�#(�#(�#(�#(�#(�#(�#($($($($(8$(@$(D$(H$(L$(t$(|$(�$(�$(�$(�$(�$(�$(�$(�$(�$(�$(�$(�$(%((%(0%(4%(8%(<%(d%(l%(p%(t%(x%(�%(�%(�%(�%(�%(�%(�%(�%(�%(�%(&( &($&((&(,&(T&(\&(`&(d&(h&(�&(�&(�&(�&(�&(�&(�&(�&(�&(�&('('('('('(D'(L'(P'(T'(X'(�'(�'(�'(�'(�'(�'(�'(�'(�'(�'(�'(((((((((4((<((@((D((H((p((x((|((�((�((�((�((�((�((�((�((�((�((�((�(($)(,)(0)(4)(8)(`)(h)(l)(p)(t)(�)(�)(�)(�)(�)(�)(�)(�)(�)(�)(*(*( *($*((*(P*(X*(\*(`*(d*(�*(�*(�*(�*(�*(�*(�*(�*(�*(�*(+(+(+(+(+(@+(H+(L+(P+(T+(|+(�+(�+(�+(�+(�+(�+(�+(�+(�+(�+(�+(,(,(,(0,(8,(<,(@,(D,(l,(t,(x,(|,(�,(�,(�,(�,(�,(�,(�,(�,(�,(�,( -((-(,-(0-(4-(\-(d-(h-(l-(�-(�-(�-(�-(�-(�-(�-(�-(�-(�-(.(.(.( .($.(L.(T.(X.(\.(`.(�.(�.(�.(�.(�.(�.(�.(�.(�.(�.(/(/(/(/(/(</(D/(H/(L/(P/(x/(�/(�/(�/(�/(�/(�/(�/(�/(�/(�/(�/(�/(0(0(,0(40(80(<0(@0(h0(p0(t0(x0(|0(�0(�0(�0(�0(�0(�0(�0(�0(�0(�0(1($1((1(,1(01(X1(`1(d1(h1(l1(�1(�1(�1(�1(�1(�1(�1(�1(�1(�1(2(2(2(2( 2(H2(P2(T2(X2(\2(�2(�2(�2(�2(�2(�2(�2(�2(�2(�2(�2(3(3(3(3(83(@3(D3(H3(L3(t3(|3(�3(�3(�3(�3(�3(�3(�3(�3(�3(�3(�3(�3(4((4(04(44(84(<4(d4(l4(p4(t4(x4(�4(�4(�4(�4(�4(�4(�4(�4(�4(�4(5( 5($5((5(,5(T5(\5(`5(d5(h5(�5(�5(�5(�5(�5(�5(�5(�5(�5(�5(6(6(6(6(6(D6(L6(P6(T6(X6(�6(�6(�6(�6(�6(�6(�6(�6(�6(�6(�6(7(7(7(47(<7(@7(D7(H7(p7(x7(|7(�7(�7(�7(�7(�7(�7(�7(�7(�7(�7(�7(�7($8(,8(08(48(88(`8(h8(l8(p8(t8(�8(�8(�8(�8(�8(�8(�8(�8(�8(9(9( 9($9(P9(X9(\9(`9(�9(�9(�9(�9(�9(�9(�9(�9(�9(�9(:(:(:(:(@:(H:(L:(P:(T:(|:(�:(�:(�:(�:(�:(�:(�:(�:(�:(�:(;(;(0;(8;(<;(@;(l;(t;(x;(|;(�;(�;(�;(�;(�;(�;(�;(�;(�;(�;( <((<(,<(0<(4<(\<(d<(h<(l<(�<(�<(�<(�<(�<(�<(�<(�<(�<(=(=(=( =($=(L=(T=(X=(\=(�=(�=(�=(�=(�=(�=(�=(�=(�=(�=(>(>(>(>(<>(D>(H>(L>(x>(�>(�>(�>(�>(�>(�>(�>(�>(�>(�>(�>(�>(?(?(,?(4?(8?(<?(@?(h?(p?(t?(x?(|?(�?(�?(�?(�?(�?(�?(�?(�?(�?(�?(@($@((@(,@(X@(`@(d@(h@(�@(�@(�@(�@(�@(�@(�@(�@(�@(�@(A(A(A(A( A(HA(PA(TA(XA(\A(�A(�A(�A(�A(�A(�A(�A(�A(�A(B(B(B(8B(@B(DB(HB(tB(|B(�B(�B(�B(�B(�B(�B(�B(�B(�B(�B(�B(�B(C((C(0C(4C(8C(dC(lC(pC(tC(�C(�C(�C(�C(�C(�C(�C(�C(�C(D( D($D((D(,D(TD(\D(`D(dD(hD(�D(�D(�D(�D(�D(�D(�D(�D(�D(�D(E(E(E(E(E(DE(LE(PE(TE(�E(�E(�E(�E(�E(�E(�E(�E(�E(�E(�E(F(F(F(F(4F(<F(@F(DF(HF(pF(xF(|F(�F(�F(�F(�F(�F(�F(�F(�F(�F(�F(�F(�F($G(,G(0G(4G(8G(`G(hG(lG(pG(tG(�G(�G(�G(�G(�G(�G(�G(�G(H(H( H($H(PH(XH(\H(`H(�H(�H(�H(�H(�H(�H(�H(�H(I(I(I(I(@I(HI(LI(PI(|I(�I(�I(�I(�I(�I(�I(�I(�I(�I(�I(J(J(J(0J(8J(<J(@J(DJ(lJ(tJ(xJ(|J(�J(�J(�J(�J(�J(�J(�J(�J(�J( K((K(,K(0K(4K(\K(dK(hK(lK(�K(�K(�K(�K(�K(�K(�K(�K(�K(L(L(L( L($L(LL(TL(XL(\L(�L(�L(�L(�L(�L(�L(�L(�L(�L(M(M(M(M(M(<M(DM(HM(LM(xM(�M(�M(�M(�M(�M(�M(�M(�M(�M(�M(�M(N(N(,N(4N(8N(<N(@N(hN(pN(tN(xN(�N(�N(�N(�N(�N(�N(�N(�N(�N(O($O((O(,O(0O(XO(`O(dO(hO(lO(�O(�O(�O(�O(�O(�O(�O(�O(�O(�O(P(P(P(P($P(,P(HP(PP(TP(XP(`P(dP(hP(�P(�P(�P(�P(�P(�P(�P(�P(�P(�P(�P(�P(�P(�P(�P(Q(Q(Q(Q(Q(Q(8Q(@Q(DQ(HQ(PQ(TQ(XQ(tQ(|Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(�Q(R(R(R((R(0R(4R(8R(@R(DR(HR(dR(lR(pR(tR(|R(�R(�R(�R(�R(�R(�R(�R(�R(�R(�R(�R(�R(�R(�R(�R(�R(S( S($S((S(0S(8S(TS(\S(`S(dS(lS(pS(tS(�S(�S(�S(�S(�S(�S(�S(�S(�S(�S(�S(�S(�S(�S(T(T(T(T( T($T((T(DT(LT(PT(TT(\T(`T(dT(�T(�T(�T(�T(�T(�T(�T(�T(�T(�T(�T(�T(�T(U(U(U(U(U(4U(<U(@U(DU(LU(TU(pU(xU(|U(�U(�U(�U(�U(�U(�U(�U(�U(�U(�U(�U(�U(�U(V(V($V(,V(0V(4V(<V(DV(`V(hV(lV(pV(xV(|V(�V(�V(�V(�V(�V(�V(�V(�V(�V(�V(�V(�V(�V(�V(�V(W(W( W($W(,W(0W(4W(PW(XW(\W(`W(hW(pW(�W(�W(�W(�W(�W(�W(�W(�W(�W(�W(�W(�W(�W(�W(X(X(X(X(X( X($X(@X(HX(LX(PX(XX(\X(`X(|X(�X(�X(�X(�X(�X(�X(�X(�X(�X(�X(�X(�X(�X(�X(�X(Y(Y(Y(Y(Y( Y(0Y(8Y(<Y(@Y(HY(PY(XY(\Y(lY(tY(xY(|Y(�Y(�Y(�Y(�Y(�Y(�Y(�Y(�Y(�Y(�Y(�Y(�Y(�Y(Z( Z((Z(,Z(0Z(8Z(@Z(\Z(dZ(hZ(lZ(tZ(|Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z(�Z([([([([( [(([(0[(8[(<[(L[(T[(X[(\[(d[(l[(t[(x[(�[(�[(�[(�[(�[(�[(�[(�[(�[(�[(�[(�[(�[(�[(�[(�[(\(\(\(\(\( \((\(,\(<\(D\(H\(L\(P\(x\(�\(�\(�\(�\(�\(�\(�\(�\(�\(�\(](,](4](8](<](h](p](t](x](|](�](�](�](�](�](�](�](�](�](�](^($^((^(,^(0^(X^(`^(d^(h^(l^(�^(�^(�^(�^(�^(�^(�^(�^(_(_(_(_(H_(P_(T_(X_(�_(�_(�_(�_(�_(�_(�_(�_(�_(�_(�_(`(`(`(`(8`(@`(D`(H`(L`(t`(|`(�`(�`(�`(�`(�`(�`(�`(�`(�`(�`(�`(�`(a((a(0a(4a(8a(da(la(pa(ta(�a(�a(�a(�a(�a(�a(�a(�a(b( b($b((b(,b(Tb(\b(`b(db(hb(�b(�b(�b(�b(�b(�b(�b(�b(�b(�b(c(c(c(c(Dc(Lc(Pc(Tc(�c(�c(�c(�c(�c(�c(�c(�c(�c(d(d(d(4d(<d(@d(Dd(pd(xd(|d(�d(�d(�d(�d(�d(�d(�d(�d(�d(�d(�d(`e(he(le(pe(�e(�e(�e(�e(�e(�e(�e(�e(f(f( f($f(Pf(Xf(\f(`f(�f(�f(�f(�f(�f(�f(�f(�f(g(g(g(g(@g(Hg(Lg(Pg(|g(�g(�g(�g(�g(�g(�g(�g(�g(�g(h(h(0h(8h(<h(@h(lh(th(xh(|h(�h(�h(�h(�h(�h(�h(�h(�h( i((i(,i(0i(\i(di(hi(li(�i(�i(�i(�i(�i(�i(�i(�i(j(j(j( j(Lj(Tj(Xj(\j(�j(�j(�j(�j(�j(�j(�j(�j(k(k(k(k(<k(Dk(Hk(Lk(xk(�k(�k(�k(�k(�k(�k(�k(�k(�k(�k(l(,l(4l(8l(<l(hl(pl(tl(xl(�l(�l(�l(�l(�l(�l(�l(�l(m($m((m(,m(Xm(`m(dm(hm(�m(�m(�m(�m(�m(�m(�m(�m(n(n(n(n(Hn(Pn(Tn(Xn(�n(�n(�n(�n(�n(�n(�n(�n(�n(o(o(o(8o(@o(Do(Ho(to(|o(�o(�o(�o(�o(�o(�o(�o(�o(�o(�o((p(0p(4p(8p(dp(lp(pp(tp(�p(�p(�p(�p(�p(�p(�p(�p(q( q($q((q(Tq(\q(`q(dq(�q(�q(�q(�q(�q(�q(�q(�q(r(r(r(r(Dr(Lr(Pr(Tr(�r(�r(�r(�r(�r(�r(�r(�r(�r(s(s(s(4s(<s(@s(Ds(ps(xs(|s(�s(�s(�s(�s(�s(�s(�s(�s(�s($t(,t(0t(4t(`t(ht(lt(pt(�t(�t(�t(�t(�t(�t(�t(�t(u(u( u($u(Pu(Xu(\u(`u(�u(�u(�u(�u(�u(�u(�u(�u(v(v(v(v(@v(Hv(Lv(Pv(|v(�v(�v(�v(�v(�v(�v(�v(�v(�v(w(w(0w(8w(<w(@w(lw(tw(xw(|w(�w(�w(�w(�w(�w(�w(�w(�w( x((x(,x(0x(\x(dx(hx(lx(�x(�x(�x(�x(�x(�x(�x(�x(y(y(y( y(Ly(Ty(Xy(\y(�y(�y(�y(�y(�y(�y(�y(�y(z(z(z(z(<z(Dz(Hz(Lz(xz(�z(�z(�z(�z(�z(�z(�z(�z(�z(�z({(,{(4{(8{(<{(h{(p{(t{(x{(�{(�{(�{(�{(�{(�{(�{(�{(|($|((|(,|(X|(`|(d|(h|(�|(�|(�|(�|(�|(�|(�|(�|(}(}(}(}(H}(P}(T}(X}(�}(�}(�}(�}(�}(�}(�}(�}(�}(~(~(~(8~(@~(D~(H~(t~(|~(�~(�~(�~(�~(�~(�~(�~(�~(�~(�~(((0(4(8(d(l(p(t(�(�(�(�(�(�(�(�(�( �($�((�(T�(\�(`�(d�(��(��(��(��(̀(Ԁ(؀(܀(�(�(�(�(D�(L�(P�(T�(��(��(��(��(��(ā(ȁ(́(��(�(�(�(4�(<�(@�(D�(p�(x�(|�(��(��(��(��(��(�(�(�(��($�(,�(0�(4�(`�(h�(l�(p�(��(��(��(��(؃(�(�(�(�(�( �($�(P�(X�(\�(`�(��(��(��(��(Ȅ(Є(Ԅ(؄(�(�(�(�(@�(H�(L�(P�(|�(��(��(��(��(��(ą(ȅ(�(��(�(�(0�(8�(<�(@�(l�(t�(x�(|�(��(��(��(��(�(�(�(�( �((�(,�(0�(\�(d�(h�(l�(��(��(��(��(ԇ(܇(�(�(�(�(�( �(L�(T�(X�(\�(��(��(��(��(Ĉ(̈(Ј(Ԉ(�(�(�(�(<�(D�(H�(L�(x�(��(��(��(��(��(��(ĉ(�(��(��(�(,�(4�(8�(<�(h�(p�(t�(x�(��(��(��(��(�(�(�(�(�($�((�(,�(X�(`�(d�(h�(��(��(��(��(Ћ(؋(܋(�(�(�(�(�(H�(P�(T�(X�(��(��(��(��(��(Ȍ(̌(Ќ(��(�(�(�(8�(@�(D�(H�(t�(|�(��(��(��(��(��(��(�(�(��(��((�(0�(4�(8�(d�(l�(p�(t�(��(��(��(��(܎(�(�(�(�( �($�((�(T�(\�(`�(d�(��(��(��(��(̏(ԏ(؏(܏(�(�(�(�(D�(L�(P�(T�(��(��(��(��(��(Đ(Ȑ(̐(��(�(�(�(4�(<�(@�(D�(p�(x�(|�(��(��(��(��(��(�(�(�(��($�(,�(0�(4�(`�(h�(l�(p�(��(��(��(��(ؒ(�(�(�(�(�( �($�(P�(X�(\�(`�(��(��(��(��(ȓ(Г(ԓ(ؓ(�(�(�(�(@�(H�(L�(P�(|�(��(��(��(��(��(Ĕ(Ȕ(�(��(�(�(0�(8�(<�(@�(l�(t�(x�(|�(��(��(��(��(�(�(�(�( �((�(,�(0�(\�(d�(h�(l�(��(��(��(��(Ԗ(ܖ(�(�(�(�(�( �(L�(T�(X�(\�(��(��(��(��(ė(̗(З(ԗ(�(�(�(�(<�(D�(H�(L�(x�(��(��(��(��(��(��(Ę(�(��(��(�(,�(4�(8�(<�(h�(p�(t�(x�(��(��(��(��(�(�(�(�(�($�((�(,�(X�(`�(d�(h�(��(��(��(��(К(ؚ(ܚ(�(�(�(�(�(H�(P�(T�(X�(��(��(��(��(��(ț(̛(Л(��(�(�(�(8�(@�(D�(H�(t�(|�(��(��(��(��(��(��(�(�(��(��((�(0�(4�(8�(d�(l�(p�(t�(��(��(��(��(ܝ(�(�(�(�( �($�((�(T�(\�(`�(d�(��(��(��(��(̞(Ԟ(؞(ܞ(�(�(�(�(D�(L�(P�(T�(��(��(��(��(��(ğ(ȟ(̟(��(�(�(�(4�(<�(@�(D�(p�(x�(|�(��(��(��(��(��(�(�(�(��($�(,�(0�(4�(`�(h�(l�(p�(��(��(��(��(ء(�(�(�(�(�( �($�(P�(X�(\�(`�(��(��(��(��(Ȣ(Т(Ԣ(آ(�(�(�(�(@�(H�(L�(P�(|�(��(��(��(��(��(ģ(ȣ(�(��(�(�(0�(8�(<�(@�(l�(t�(x�(|�(�(�(�(�(��( �((�(,�(0�(4�(\�(d�(h�(l�(p�(��(��(��(��(��(ԥ(ܥ(�(�(�(�(�(�( �($�(L�(T�(X�(\�(`�(��(��(��(��(��(Ħ(̦(Ц(Ԧ(ئ(�(�(�(�(�(<�(D�(H�(L�(P�(x�(��(��(��(��(��(��(��(ħ(ȧ(�(��(��(�(�(,�(4�(8�(<�(@�(h�(p�(t�(x�(|�(��(��(��(��(��(�(�(�(�(�(�($�((�(,�(0�(X�(`�(d�(h�(l�(��(��(��(��(Щ(ة(ܩ(�(�(�(�(�(�( �(H�(P�(T�(X�(\�(��(��(��(��(��(��(Ȫ(̪(Ъ(Ԫ(��(�(�(�(�(8�(@�(D�(H�(L�(t�(|�(��(��(��(��(��(��(��(ī(�(�(��(��(�((�(0�(4�(8�(<�(d�(l�(p�(t�(x�(��(��(��(��(��(ܬ(�(�(�(�(�( �($�((�(,�(T�(\�(`�(d�(h�(��(��(��(��(��(̭(ԭ(ح(ܭ(�(�(�(�(�(�(D�(L�(P�(T�(X�(��(��(��(��(��(��(Į(Ȯ(̮(Ю(��(�(�(�(�(4�(<�(@�(D�(H�(p�(x�(|�(��(��(��(��(��(��(��(�(�(�(��(��($�(,�(0�(4�(8�(`�(h�(l�(p�(t�(��(��(��(��(��(ذ(�(�(�(�(�(�( �($�((�(P�(X�(\�(`�(d�(��(��(��(��(��(ȱ(б(Ա(ر(ܱ(�(�(�(�(�(@�(H�(L�(P�(T�(|�(��(��(��(��(��(��(IJ(Ȳ(̲(�(��(�(�(�(0�(8�(<�(@�(D�(l�(t�(x�(|�(��(��(��(��(��(��(�(�(�(�(��( �((�(,�(0�(4�(\�(d�(h�(l�(p�(��(��(��(��(��(Դ(ܴ(�(�(�(�(�(�( �($�(L�(T�(X�(\�(`�(��(��(��(��(��(ĵ(̵(е(Ե(ص(�(�(�(�(�(<�(D�(H�(L�(P�(x�(��(��(��(��(��(��(��(Ķ(ȶ(�(��(��(�(�(,�(4�(8�(<�(@�(h�(p�(t�(x�(|�(��(��(��(��(��(�(�(�(�(�(�($�((�(,�(0�(X�(`�(d�(h�(l�(��(��(��(��(��(и(ظ(ܸ(�(�(�(�(�(�( �(H�(P�(T�(X�(\�(��(��(��(��(��(��(ȹ(̹(й(Թ(��(�(�(�(�(8�(@�(D�(H�(L�(t�(|�(��(��(��(��(��(��(��(ĺ(�(�(��(��(�((�(0�(4�(8�(<�(d�(l�(p�(t�(x�(��(��(��(��(��(ܻ(�(�(�(�(�( �($�((�(,�(T�(\�(`�(d�(h�(��(��(��(��(��(̼(Լ(ؼ(ܼ(�(�(�(�(�(�(D�(L�(P�(T�(X�(��(��(��(��(��(��(Ľ(Ƚ(̽(н(��(�(�(�(�(4�(<�(@�(D�(H�(p�(x�(|�(��(��(��(��(��(��(��(�(�(�(��(��($�(,�(0�(4�(8�(`�(h�(l�(p�(t�(��(��(��(��(��(ؿ(�(�(�(�(�(�( �($�((�(P�(X�(\�(`�(d�(��(��(��(��(��(�(�(�(�(�(�(�(�(�(�(@�(H�(L�(P�(T�(|�(��(��(��(��(��(��(�(�(�(�(��(�(�(�(0�(8�(<�(@�(D�(l�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��(��( �((�(,�(0�(4�(\�(d�(h�(l�(p�(��(��(��(��(��(��(��(��(��(��(�(�(�( �($�(L�(T�(X�(\�(`�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(<�(D�(H�(L�(P�(x�(��(��(��(��(��(��(��(��(��(��(��(��(�(�(,�(4�(8�(<�(@�(h�(p�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��(�($�((�(,�(0�(X�(`�(d�(h�(l�(��(��(��(��(��(��(��(��(��(��(�(�(�(�( �(H�(P�(T�(X�(\�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(�(8�(@�(D�(H�(L�(t�(|�(��(��(��(��(��(��(��(��(��(��(��(��(�((�(0�(4�(8�(<�(d�(l�(p�(t�(x�(��(��(��(��(��(��(��(��(��(��(�( �($�((�(,�(T�(\�(`�(d�(h�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(D�(L�(P�(T�(X�(��(��(��(��(��(��(�(�(�(�(4�(<�(@�(D�(H�(p�(x�(|�(��(��(��(��(��(��(��(��(��(��(��(��($�(,�(0�(4�(8�(`�(h�(l�(p�(t�(��(��(��(��(��(��(��(��(��(��(�(�( �($�((�(P�(X�(\�(`�(d�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(@�(H�(L�(P�(T�(|�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(0�(8�(<�(@�(D�(l�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��( �((�(,�(0�(4�(\�(d�(h�(l�(p�(��(��(��(��(��(��(��(��(��(��(�(�(�( �($�(L�(T�(X�(\�(`�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(<�(D�(H�(L�(P�(x�(��(��(��(��(��(��(��(��(��(��(��(��(�(�(,�(4�(8�(<�(@�(h�(p�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��(�($�((�(,�(0�(X�(`�(d�(h�(l�(��(��(��(��(��(��(��(��(��(��(�(�(�(�( �(H�(P�(T�(X�(\�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(�(8�(@�(D�(H�(L�(t�(|�(��(��(��(��(��(��(��(��(��(��(��(��(�((�(0�(4�(8�(<�(d�(l�(p�(t�(x�(��(��(��(��(��(��(��(��(��(��(�( �($�((�(,�(T�(\�(`�(d�(h�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(D�(L�(P�(T�(X�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(�(4�(<�(@�(D�(H�(p�(x�(|�(��(��(��(��(��(��(��(��(��(��(��(��($�(,�(0�(4�(8�(`�(h�(l�(p�(t�(��(��(��(��(��(��(��(��(��(��(�(�( �($�((�(P�(X�(\�(`�(d�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(@�(H�(L�(P�(T�(|�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(0�(8�(<�(@�(D�(l�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��(��( �((�(,�(0�(4�(\�(d�(h�(l�(p�(��(��(��(��(��(��(��(��(��(��(�(�(�( �($�(L�(T�(X�(\�(`�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(<�(D�(H�(L�(P�(x�(��(��(��(��(��(��(��(��(��(��(��(��(�(�(,�(4�(8�(<�(@�(h�(p�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��(�($�((�(,�(0�(X�(`�(d�(h�(l�(��(��(��(��(��(��(��(��(��(��(�(�(�(�( �(H�(P�(T�(X�(\�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(�(8�(@�(D�(H�(L�(t�(|�(��(��(��(��(��(��(��(��(��(��(��(��(�((�(0�(4�(8�(<�(d�(l�(p�(t�(x�(��(��(��(��(��(��(��(��(��(��(�( �($�((�(,�(T�(\�(`�(d�(h�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(D�(L�(P�(T�(X�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(�(4�(<�(@�(D�(H�(p�(x�(|�(��(��(��(��(��(��(��(��(��(��(��(��($�(,�(0�(4�(8�(`�(h�(l�(p�(t�(��(��(��(��(��(��(��(��(��(��(�(�( �($�((�(P�(X�(\�(`�(d�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(@�(H�(L�(P�(T�(|�(��(��(��(��(��(��(��(��(��(��(��(�(�(�(0�(8�(<�(@�(D�(l�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��(��( �((�(,�(0�(4�(\�(d�(h�(l�(p�(��(��(��(��(��(��(��(��(��(��(�(�(�( �($�(L�(T�(X�(\�(`�(��(��(��(��(��(��(��(��(��(��(�(�(�(�(�(<�(D�(H�(L�(P�(x�(��(��(��(��(��(��(��(��(��(��(��(��(�(�(,�(4�(8�(<�(@�(h�(p�(t�(x�(|�(��(��(��(��(��(��(��(��(��(��(�($�((�(,�(0�(X�(`�(d�(h�(l�(��(��(��(��(��(��(��(��(��(��(�(�(�(�( �(H�(P�(T�(X�(\�(��(��(��(��(��(��(�(�(�(�(��(�(�(�(�(8�(@�(D�(H�(L�(t�(|�(��(��(��(��(��(��(��(�((�(0�(4�(8�(<�(d�(l�(p�(t�(��(��(��(��(��(�(�(�(�(�( �($�((�(T�(\�(`�(d�(��(��(��(��(��(�(�(�(�(�(�(�(�(�(�(D�(L�(P�(T�(X�(��(��(��(��(��(��(�(�(�(�(��(�(�(�(�(4�(<�(@�(D�(H�(p�(x�(|�(��(��(��(��(��(��(��(�(�(�(��(��($�(,�(0�(4�(8�(`�(h�(l�(p�(t�(��(��(��(��(��(�(�(�(�(�(�(�( �($�((�(P�(X�(\�(`�(d�(��(��(��(��(��(�(�(�(�(�(�(�(�(�(�(@�(H�(L�(P�(T�(|�(��(��(��(��(��(��(�(�(�(�(��(�(�(�(0�(8�(<�(@�(D�(l�(t�(x�(|�(��(��(��(��(��(��(�(�(�(�(��( �((�(,�(0�(4�(\�(d�(h�(l�(p�(��(��(��(��(��(�(�(�(�(�())) )$)L)T)X)\)`)�)�)�)�)�)�)�)�)))))<)D)H)L)x)�)�)�)�)�)�)�)�)�)�)�)),)4)8)<)@)h)p)t)x)�)�)�)�)�)�)�)�)�)�))$)(),)0)X)`)d)h)�)�)�)�)�)�)�)�)))))H)P)T)X)�)�)�)�)�)�)�)�)�)))))8)@)D)H)L)t)|)�)�)�)�)�)�)�)�)�)�)�)�)()0)4)8)<)d)l)p)t)�)�)�)�)�)�)�)�)) )$)(),)T)\)`)d)h)�)�)�)�)�)�)�)�)�)))))D)L)P)T)X)�)�)�)�)�)�)�)�)�)�)�)	)	)	)	)4	)<	)@	)D	)H	)p	)x	)|	)�	)�	)�	)�	)�	)�	)�	)�	)�	)�	)�	)�	)$
),
)0
)4
)8
)`
)h
)l
)p
)t
)�
)�
)�
)�
)�
)�
)�
)�
)�
)�
))) )$)()P)X)\)`)d)�)�)�)�)�)�)�)�)�)�))))))@)H)L)P)T)|)�)�)�)�)�)�)�)�)�)�)�)
)
)0
)8
)<
)@
)D
)l
)t
)x
)|
)�
)�
)�
)�
)�
)�
)�
)�
)�
)�
)�
) )(),)0)4)\)d)h)l)p)�)�)�)�)�)�)�)�)�)�)))) )L)T)X)\)�)�)�)�)�)�)�)�)�)))))<)D)H)L)x)�)�)�)�)�)�)�)�)�)�)�))),)4)8)<)@)h)p)t)x)|)�)�)�)�)�)�)�)�)�))$)(),)0)X)`)d)h)l)�)�)�)�)�)�)�)�)�))))) )H)P)T)X)\)�)�)�)�)�)�)�)�)�)))))8)@)D)H)L)t)|)�)�)�)�)�)�)�)�)�)�)�))()0)4)8)<)d)l)p)t)�)�)�)�)�)�)�)�)) )$)()`)d)h)|)�)�)�)�)�)�)�)�)�)�)�)�)�)�))))0)4)8)P)T)X)l)p)t)|)�)�)�)�)�)�)�)�)�)�)�)�)))) )$)()@)D)\)`)d)|)�)�)�)�)�)�)�)�)�)�)�))))0)4)L)P)T)l)p)�)�)�)�)�)�)�)�)�)�)))) )$)<)@)D)\)`)x)|)�)�)�)�)�)�)�)�)�)�)�))),)0)4)L)P)h)l)p)�)�)�)�)�)�)�)�)�)�)))) )$)<)@)X)\)`)x)|)�)�)�)�)�)�)�)�)�)�)�)))),)0)H)L)P)h)l)�)�)�)�)�)�)�)�)�)�)�)))) )8)<)@)X)\)t)x)|)�)�)�)�)�)�)�)�)�)�)�)))(),)0)H)L)d)h)l)�)�)�)�)�)�)�)�)�)�)�) ) ) )  )8 )< )T )X )\ )t )x )� )� )� )� )� )� )� )� )� )� )!)!)!)(!),!)D!)H!)L!)d!)h!)�!)�!)�!)�!)�!)�!)�!)�!)�!)�!)�!)�!)")")")4")8")<")T")X")p")t")x")�")�")�")�")�")�")�")�")�")�")#)#)$#)(#),#)8#)@#)D#)H#)t#)|#)�#)�#)�#)�#)�#)�#)�#)�#)�#)�#)($)0$)4$)8$)d$)l$)p$)t$)x$)�$)�$)�$)�$)�$)�$)�$)�$)%) %)$%)(%),%)T%)\%)`%)d%)�%)�%)�%)�%)�%)�%)�%)�%)&)&)&)&)D&)L&)P&)T&)�&)�&)�&)�&)�&)�&)�&)�&)�&)')')')4')<')@')D')H')p')x')|')�')�')�')�')�')�')�')�')�')�')�')�')$(),()0()4()8()`()h()l()p()t()�()�()�()�()�()�()�()�()�()�())))) ))$))())P))X))\))`))d))�))�))�))�))�))�))�))�))�))�))*)*)*)*)*)@*)H*)L*)P*)T*)|*)�*)�*)�*)�*)�*)�*)�*)�*)�*)�*)�*)+)+)+)0+)8+)<+)@+)D+)l+)t+)x+)|+)�+)�+)�+)�+)�+)�+)�+)�+)�+)�+)�+) ,)(,),,)0,)4,)\,)d,)h,)l,)p,)�,)�,)�,)�,)�,)�,)�,)�,)�,)-)-)-) -)L-)T-)X-)\-)�-)�-)�-)�-)�-)�-)�-)�-)�-).).).).)<.)D.)H.)L.)x.)�.)�.)�.)�.)�.)�.)�.)�.)�.)�.)/),/)4/)8/)</)h/)p/)t/)x/)�/)�/)�/)�/)�/)�/)�/)�/)0)$0)(0),0)X0)`0)d0)h0)�0)�0)�0)�0)�0)�0)�0)�0)1)1)1)1)H1)P1)T1)X1)\1)�1)�1)�1)�1)�1)�1)�1)�1)�1)�1)2)2)2)2)82)@2)D2)H2)t2)|2)�2)�2)�2)�2)�2)�2)�2)�2)�2)�2)�2)(3)03)43)83)d3)l3)p3)t3)�3)�3)�3)�3)�3)�3)�3)�3)4) 4)$4)(4)T4)\4)`4)d4)�4)�4)�4)�4)�4)�4)�4)�4)5)5)5)5)D5)L5)P5)T5)�5)�5)�5)�5)�5)�5)�5)�5)�5)�5)6)6)6)6)46)<6)@6)D6)H6)p6)x6)|6)�6)�6)�6)�6)�6)�6)�6)�6)�6)�6)$7),7)07)47)87)`7)h7)l7)p7)t7)�7)�7)�7)�7)�7)�7)�7)�7)�7)8)8) 8)$8)P8)X8)\8)`8)�8)�8)�8)�8)�8)�8)�8)�8)9)9)9)9)@9)H9)L9)P9)|9)�9)�9)�9)�9)�9)�9)�9)�9)�9)�9)�9):):)0:)8:)<:)@:)D:)l:)t:)x:)|:)�:)�:)�:)�:)�:)�:)�:)�:)�:)�:) ;)(;),;)0;)\;)d;)h;)l;)p;)�;)�;)�;)�;)�;)�;)�;)�;)�;)�;)<)<)<) <)$<)L<)T<)X<)\<)�<)�<)�<)�<)�<)�<)�<)�<)�<)=)=)=)=)=)<=)D=)H=)L=)P=)x=)�=)�=)�=)�=)�=)�=)�=)�=)�=)�=)�=)�=)>)>),>)4>)8>)<>)@>)h>)p>)t>)x>)|>)�>)�>)�>)�>)�>)�>)�>)�>)�>)�>)?)$?)(?),?)0?)X?)`?)d?)h?)l?)�?)�?)�?)�?)�?)�?)�?)�?)�?)�?)@)@)@)@) @)H@)P@)T@)X@)\@)�@)�@)�@)�@)�@)�@)�@)�@)�@)�@)�@)A)A)A)A)8A)@A)DA)HA)LA)tA)|A)�A)�A)�A)�A)�A)�A)�A)�A)�A)�A)�A)(B)0B)4B)8B)<B)dB)lB)pB)tB)�B)�B)�B)�B)�B)�B)�B)�B)C) C)$C)(C),C)TC)\C)`C)dC)hC)�C)�C)�C)�C)�C)�C)�C)�C)�C)�C)D)D)D)D)D)DD)LD)PD)TD)�D)�D)�D)�D)�D)�D)�D)�D)�D)�D)E)E)E)E)4E)<E)@E)DE)HE)pE)xE)|E)�E)�E)�E)�E)�E)�E)�E)�E)�E)$F),F)0F)4F)`F)hF)lF)pF)�F)�F)�F)�F)�F)�F)�F)�F)G)G) G)$G)PG)XG)\G)`G)�G)�G)�G)�G)�G)�G)�G)�G)�G)H)H)H)H)@H)HH)LH)PH)|H)�H)�H)�H)�H)�H)�H)�H)�H)�H)I)I)I)0I)8I)<I)@I)DI)lI)tI)xI)|I)�I)�I)�I)�I)�I)�I)�I)�I)�I) J)(J),J)0J)4J)\J)dJ)hJ)lJ)pJ)�J)�J)�J)�J)�J)�J)�J)�J)�J)�J)K)K)K) K)$K)LK)TK)XK)\K)�K)�K)�K)�K)�K)�K)�K)�K)�K)L)L)L)L)L)<L)DL)HL)LL)PL)xL)�L)�L)�L)�L)�L)�L)�L)�L)�L)�L)�L)�L)M)M),M)4M)8M)<M)@M)hM)pM)tM)xM)|M)�M)�M)�M)�M)�M)�M)�M)�M)�M)N)$N)(N),N)XN)`N)dN)hN)lN)�N)�N)�N)�N)�N)�N)�N)�N)�N)�N)O)O)O)O)HO)PO)TO)XO)\O)�O)�O)�O)�O)�O)�O)�O)�O)�O)�O)P)P)P)8P)@P)DP)HP)tP)|P)�P)�P)�P)�P)�P)�P)�P)�P)�P)�P)�P)�P)�P)�P)�P)�P)Q)Q)Q)(Q)0Q)4Q)8Q)<Q)@Q)HQ)dQ)lQ)pQ)tQ)xQ)|Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)�Q)R) R)$R)(R),R)0R)8R)TR)\R)`R)dR)hR)lR)tR)�R)�R)�R)�R)�R)�R)�R)�R)�R)�R)�R)�R)�R)�R)S)S)S)S)S) S)(S)DS)LS)PS)TS)XS)\S)dS)�S)�S)�S)�S)�S)�S)�S)�S)�S)�S)�S)�S)�S)�S)�S)T)T)T)T)T)T)4T)<T)@T)DT)HT)LT)TT)pT)xT)|T)�T)�T)�T)�T)�T)�T)�T)�T)�T)�T)�T)�T)�T)�T)�T)�T)U)U)$U),U)0U)4U)8U)<U)DU)`U)hU)lU)pU)tU)xU)�U)�U)�U)�U)�U)�U)�U)�U)�U)�U)�U)�U)�U)�U)�U)V)V) V)$V)(V),V)4V)PV)XV)\V)`V)dV)hV)pV)�V)�V)�V)�V)�V)�V)�V)�V)�V)�V)�V)�V)�V)�V)W)W)W)W)W)$W)@W)HW)LW)PW)TW)XW)`W)|W)�W)�W)�W)�W)�W)�W)�W)�W)�W)�W)�W)�W)�W)�W)�W)X)X)X)X)0X)8X)<X)@X)DX)HX)lX)tX)xX)|X)�X)�X)�X)�X)�X)�X)�X)�X)�X)�X)�X)�X)�X)�X) Y)(Y),Y)0Y)4Y)8Y)\Y)dY)hY)lY)pY)tY)�Y)�Y)�Y)�Y)�Y)�Y)�Y)�Y)�Y)�Y)�Y)�Y)Z)Z)Z) Z)$Z)(Z)LZ)TZ)XZ)\Z)`Z)dZ)�Z)�Z)�Z)�Z)�Z)�Z)�Z)�Z)�Z)�Z)�Z)�Z)[)[)[)[)[)[)<[)D[)H[)L[)P[)T[)x[)�[)�[)�[)�[)�[)�[)�[)�[)�[)�[)�[)�[)�[)�[)\)\)\),\)4\)8\)<\)@\)D\)h\)p\)t\)x\)|\)�\)�\)�\)�\)�\)�\)�\)�\)�\)�\)�\)�\)�\)])$])(]),])0])4])X])`])d])h])l])p])�])�])�])�])�])�])�])�])�])�])�])�])^)^)^)^) ^)$^)H^)P^)T^)X^)\^)`^)�^)�^)�^)�^)�^)�^)�^)�^)�^)�^)�^)�^)�^)_)_)_)_)_)8_)@_)D_)H_)L_)t_)|_)�_)�_)�_)�_)�_)�_)�_)�_)�_)�_)�_)�_)`)(`)0`)4`)8`)<`)d`)l`)p`)t`)x`)�`)�`)�`)�`)�`)�`)�`)�`)�`)a) a)$a)(a),a)Ta)\a)`a)da)ha)�a)�a)�a)�a)�a)�a)�a)�a)�a)�a)b)b)b)b)b)Db)Lb)Pb)Tb)Xb)�b)�b)�b)�b)�b)�b)�b)�b)�b)�b)�b)c)c)c)4c)<c)@c)Dc)Hc)pc)xc)|c)�c)�c)�c)�c)�c)�c)�c)�c)�c)�c)�c)$d),d)0d)4d)8d)`d)hd)ld)pd)td)�d)�d)�d)�d)�d)�d)�d)�d)�d)�d)e)e) e)$e)(e)Pe)Xe)\e)`e)de)�e)�e)�e)�e)�e)�e)�e)�e)�e)f)f)f)f)f)@f)Hf)Lf)Pf)Tf)|f)�f)�f)�f)�f)�f)�f)�f)�f)�f)�f)�f)g)g)g)0g)8g)<g)@g)Dg)lg)tg)xg)|g)�g)�g)�g)�g)�g)�g)�g)�g) h)(h),h)0h)\h)dh)hh)lh)�h)�h)�h)�h)�h)�h)�h)�h)i)i)i) i)Li)Ti)Xi)\i)`i)�i)�i)�i)�i)�i)�i)�i)�i)�i)�i)j)j)j)j)j)<j)Dj)Hj)Lj)Pj)xj)�j)�j)�j)�j)�j)�j)�j)k)k),k)4k)8k)<k)hk)pk)tk)xk)|k)�k)�k)�k)�k)�k)�k)�k)�k)l)$l)(l),l)Xl)`l)dl)hl)ll)�l)�l)�l)�l)�l)�l)�l)�l)�l)�l)m)m)m)m) m)Hm)Pm)Tm)Xm)\m)�m)�m)�m)�m)�m)�m)�m)�m)�m)�m)�m)n)n)n)n)8n)@n)Dn)Hn)Ln)tn)|n)�n)�n)�n)�n)�n)�n)�n)�n)�n)�n)�n)�n)o)(o)0o)4o)8o)<o)do)lo)po)to)xo)�o)�o)�o)�o)�o)�o)�o)�o)�o)�o)p) p)$p)(p),p)Tp)\p)`p)dp)hp)�p)�p)�p)�p)�p)�p)�p)�p)�p)�p)q)q)q)q)q)Dq)Lq)Pq)Tq)Xq)�q)�q)�q)�q)�q)�q)�q)�q)�q)�q)�q)r)r)r)r)4r)<r)@r)Dr)Hr)pr)xr)|r)�r)�r)�r)�r)�r)�r)�r)�r)�r)�r)�r)�r)$s),s)0s)4s)8s)`s)hs)ls)ps)ts)�s)�s)�s)�s)�s)�s)�s)�s)�s)�s)t)t) t)$t)(t)Pt)Xt)\t)`t)�t)�t)�t)�t)�t)�t)�t)�t)u)u)u)u)@u)Hu)Lu)Pu)Tu)|u)�u)�u)�u)�u)�u)�u)�u)�u)�u)v)v)0v)8v)<v)@v)lv)tv)xv)|v)�v)�v)�v)�v)�v)�v)�v)�v)�v) w)(w),w)0w)4w)\w)dw)hw)lw)pw)�w)�w)�w)�w)�w)�w)�w)�w)�w)x)x)x) x)Lx)Tx)Xx)\x)�x)�x)�x)�x)�x)�x)�x)�x)y)y)y)y)<y)Dy)Hy)Ly)xy)�y)�y)�y)�y)�y)�y)�y)�y)�y)�y)�y)�y)z)z),z)4z)8z)<z)@z)hz)pz)tz)xz)�z)�z)�z)�z)�z)�z)�z)�z)�z){)${)({),{)X{)`{)d{)h{)�{)�{)�{)�{)�{)�{)�{)�{)�{)�{)|)|)|)|)H|)P|)T|)X|)\|)�|)�|)�|)�|)�|)�|)�|)�|)�|)�|)})})})})8})@})D})H})L})t})|})�})�})�})�})�})�})�})�})�})�})�})�})~)(~)0~)4~)8~)<~)d~)l~)p~)t~)x~)�~)�~)�~)�~)�~)�~)�~)�~)�~)�~)) )$)(),)T)\)`)d)h)�)�)�)�)�)�)�)�)�)�)�)�)�)�)�)D�)L�)P�)T�)X�)��)��)��)��)��)��)Ā)Ȁ)̀)Ѐ)��)�)�)�)�)4�)<�)@�)D�)H�)p�)x�)|�)��)��)��)��)��)��)�)�)�)��)��)$�),�)0�)4�)8�)`�)h�)l�)p�)t�)��)��)��)��)��)؂)�)�)�)�)�)�) �)$�)(�)P�)X�)\�)`�)d�)��)��)��)��)ȃ)Ѓ)ԃ)؃)�)�)�)�)�)@�)H�)L�)P�)|�)��)��)��)��)��)Ą)Ȅ)�)��)�)�)0�)8�)<�)@�)D�)l�)t�)x�)|�)��)��)��)��)��)��)�)�)�)�)��) �)(�),�)0�)4�)\�)d�)h�)l�)��)��)��)��)��)Ԇ)܆)�)�)�)�)�)�) �)$�)L�)T�)X�)\�)��)��)��)��)��)ć)̇)Ї)ԇ)؇)�)�)�)�)<�)D�)H�)L�)x�)��)��)��)��)��)��)��)Ĉ)Ȉ)�)��)��)�),�)4�)8�)<�)@�)h�)p�)t�)x�)|�)��)��)��)��)��)�)�)�)�)�)$�)(�),�)X�)`�)d�)h�)��)��)��)��)܊)�)�)��)��)�)�)�) �)4�)8�)<�)T�)X�)\�)p�)t�)x�)��)��)��)��)��)��)̋)Ћ)ԋ)�)�)�)��)�)�)�)$�)(�),�)D�)H�)L�)`�)d�)h�)��)��)��)��)��)��)��)��)،)܌)�)��)��)�)�)�)4�)8�)P�)T�)X�)p�)t�)��)��)��)��)��)ȍ)̍)Ѝ)�)�)�)�)�)$�)(�)@�)D�)H�)`�)d�)|�)��)��)��)��)��)��)��)؎)܎)�)��)��)�)�)0�)4�)8�)P�)T�)l�)p�)t�)��)��)��)��)��)ȏ)̏)�)�)�)�)�) �)$�)(�)@�)D�)\�)`�)d�)|�)��)��)��)��)��)��)Ԑ)ؐ)ܐ)�)��)�)�)�)0�)4�)L�)P�)T�)l�)p�)t�)��)��)��)��)��)đ)ȑ)̑)�)�)�)�)�) �)$�)<�)@�)D�)\�)`�)x�)|�)��)��)��)��)��)��)Ԓ)ؒ)�)�)��)�)�),�)0�)4�)L�)P�)h�)l�)p�)x�)��)��)��)��)��)ē)ȓ)�)�)�)�)�)�) �)$�)<�)@�)X�)\�)`�)x�)|�)��)��)��)��)��)Д)Ԕ)ؔ)�)�)�)�)�),�)0�)H�)L�)P�)h�)l�)��)��)��)��)��)��)ĕ)ȕ)�)�)��)�)�)�) �)8�)<�)@�)X�)\�)t�)x�)|�)��)��)��)��)��)Ж)Ԗ)�)�)�)�)�)(�),�)0�)H�)L�)d�)h�)l�)��)��)��)��)��)��)ė)ܗ)�)�)�)��)��)�),�)4�)8�)<�)h�)p�)t�)x�)��)��)��)��)�)�)�)�)�)$�)(�),�)0�)X�)`�)d�)h�)��)��)��)��)Й)ؙ)ܙ)�)�)�)�)�)�)H�)P�)T�)X�)��)��)��)��)��)Ț)̚)К)��)�)�)�)8�)@�)D�)H�)t�)|�)��)��)��)��)��)��)�)�)��)��)�)(�)0�)4�)8�)<�)d�)l�)p�)t�)x�)��)��)��)��)��)ܜ)�)�)�)�)�) �)$�)(�),�)T�)\�)`�)d�)h�)��)��)��)��)��)̝)ԝ)؝)ܝ)�)�)�)�)�)�)D�)L�)P�)T�)X�)��)��)��)��)��)��)Ğ)Ȟ)̞)О)��)�)�)�)�)4�)<�)@�)D�)H�)p�)x�)|�)��)��)��)��)��)��)��)�)�)�)��)��)$�),�)0�)4�)8�)`�)h�)l�)p�)t�)��)��)��)��)��)ؠ)�)�)�)�)�)�) �)$�)(�)P�)X�)\�)`�)d�)��)��)��)��)��)ȡ)С)ԡ)ء)ܡ)�)�)�)�)�)@�)H�)L�)P�)T�)|�)��)��)��)��)��)��)Ģ)Ȣ)̢)�)��)�)�)0�)8�)<�)@�)D�)l�)t�)x�)|�)��)��)��)��)�)�)�)�) �)(�),�)0�)4�)\�)d�)h�)l�)��)��)��)��)Ԥ)ܤ)�)�)�)�)�) �)L�)T�)X�)\�)��)��)��)��)ĥ)̥)Х)ԥ)�)�)�)�)<�)D�)H�)L�)x�)��)��)��)��)��)��)Ħ)�)��)��)�),�)4�)8�)<�)h�)p�)t�)x�)��)��)��)��)��)�)�)�)�)�)$�)(�),�)0�)X�)`�)d�)h�)l�)��)��)��)��)Ш)ب)ܨ)�)�)�)�)�)�)H�)P�)T�)X�)��)��)��)��)��)ȩ)̩)Щ)��)�)�)�)8�)@�)D�)H�)t�)|�)��)��)��)��)��)��)�)�)��)��)(�)0�)4�)8�)d�)l�)p�)t�)��)��)��)��)ܫ)�)�)�)�)�) �)$�)(�)T�)\�)`�)d�)h�)��)��)��)��)��)̬)Ԭ)ج)ܬ)�)�)�)�)D�)L�)P�)T�)X�)��)��)��)��)��)��)ĭ)ȭ)̭)Э)��)�)�)�)�)4�)<�)@�)D�)p�)x�)|�)��)��)��)��)��)�)�)�)��)$�),�)0�)4�)`�)h�)l�)p�)��)��)��)��)د)�)�)�)�)�)�) �)$�)(�)P�)X�)\�)`�)��)��)��)��)��)Ȱ)а)԰)ذ)�)�)�)�)�)@�)H�)L�)P�)T�)|�)��)��)��)��)��)ı)ȱ)̱)�)��)�)�)�)0�)8�)<�)@�)D�)l�)t�)x�)|�)��)��)��)��)��)�)�)�)�)��) �)(�),�)0�)\�)d�)h�)l�)p�)��)��)��)��)��)Գ)ܳ)�)�)�)�)�)�) �)$�)L�)T�)X�)\�)`�)��)��)��)��)��)Ĵ)̴)д)Դ)ش)�)�)�)�)�)<�)D�)H�)L�)P�)x�)��)��)��)��)��)��)��)ĵ)ȵ)�)��)��)�)�),�)4�)8�)<�)@�)h�)p�)t�)x�)|�)��)��)��)��)��)�)�)�)�)�)�)$�)(�),�)0�)X�)`�)d�)h�)l�)��)��)��)��)��)з)ط)ܷ)�)�)�)�)�)�)H�)P�)T�)X�)��)��)��)��)��)��)ȸ)̸)и)��)�)�)�)8�)@�)D�)H�)t�)|�)��)��)��)��)��)��)��)Ĺ)�)�)��)��)�)(�)0�)4�)8�)<�)d�)l�)p�)t�)x�)��)��)��)��)ܺ)�)�)�)�) �)$�)(�),�)T�)\�)`�)d�)h�)��)��)��)��)��)̻)Ի)ػ)ܻ)�)�)�)�)D�)L�)P�)T�)��)��)��)��)��)ļ)ȼ)̼)��)�)�)�)4�)<�)@�)D�)p�)x�)|�)��)��)��)��)��)�)�)�)��)��)$�),�)0�)4�)`�)h�)l�)p�)��)��)��)��)ؾ)�)�)�)�)�) �)$�)P�)X�)\�)`�)d�)��)��)��)��)��)ȿ)п)Կ)ؿ)ܿ)�)�)�)�)@�)H�)L�)P�)|�)��)��)��)��)��)��)�)�)�)�)��)�)�)�)0�)8�)<�)@�)D�)l�)t�)x�)|�)��)��)��)��)��)�)�)�)�)��) �)(�),�)0�)\�)d�)h�)l�)p�)��)��)��)��)��)��)��)��)��)��)�)�)�) �)$�)L�)T�)X�)\�)`�)��)��)��)��)��)��)��)��)��)��)�)�)�)�)<�)D�)H�)L�)P�)x�)��)��)��)��)��)��)��)��)��)��)��)�)�),�)4�)8�)<�)@�)h�)p�)t�)x�)��)��)��)��)��)��)��)��)��)�)$�)(�),�)0�)X�)`�)d�)h�)��)��)��)��)��)��)��)��)��)��)�)�)�)�)$�)(�),�)H�)P�)T�)X�)`�)d�)h�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�)�)8�)@�)D�)H�)P�)T�)X�)t�)|�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)(�)0�)4�)8�)@�)D�)H�)d�)l�)p�)t�)|�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�) �)$�)(�)0�)4�)8�)T�)\�)`�)d�)l�)p�)t�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�) �)$�)(�)D�)L�)P�)T�)\�)d�)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�)4�)<�)@�)D�)L�)T�)p�)x�)|�)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)$�),�)0�)4�)<�)@�)D�)`�)h�)l�)p�)x�)|�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�) �)$�),�)4�)P�)X�)\�)`�)h�)l�)p�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�) �)$�)@�)H�)L�)P�)X�)`�)|�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�) �)0�)8�)<�)@�)H�)P�)l�)t�)x�)|�)��)��)��)��)��)��)��)��)��)��)��)��)��)�) �)(�),�)0�)8�)@�)H�)L�)\�)d�)h�)l�)t�)|�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�) �)(�)0�)8�)<�)L�)T�)X�)\�)d�)l�)t�)x�)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�)<�)D�)H�)L�)P�)x�)��)��)��)��)��)��)��)��)��)��)��)��)�)�),�)4�)8�)<�)@�)h�)p�)t�)x�)��)��)��)��)��)��)��)��)��)��)�)$�)(�),�)0�)X�)`�)d�)h�)l�)��)��)��)��)��)��)��)��)��)��)�)�)�)�) �)H�)P�)T�)X�)\�)��)��)��)��)��)��)��)��)��)��)�)�)�)�)8�)@�)D�)H�)L�)t�)|�)��)��)��)��)��)��)��)��)��)��)��)�)(�)0�)4�)8�)<�)d�)l�)p�)t�)x�)��)��)��)��)��)��)��)��)��)��)�) �)$�)(�),�)T�)\�)`�)d�)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�)D�)L�)P�)T�)X�)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�)4�)<�)@�)D�)p�)x�)|�)��)��)��)��)��)��)��)��)��)$�),�)0�)4�)`�)h�)l�)p�)��)��)��)��)��)��)��)��)�)�) �)$�)(�)P�)X�)\�)`�)d�)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�)@�)H�)L�)P�)T�)��)��)��)��)��)��)��)�)�)�)0�)8�)<�)@�)D�)l�)t�)x�)|�)��)��)��)��)��)��)��)��)��) �)(�),�)0�)4�)\�)d�)h�)l�)p�)��)��)��)��)��)��)��)��)��)��)�)�)�) �)$�)L�)T�)X�)\�)`�)��)��)��)��)��)��)��)��)��)��)�)�)�)�)�)<�)D�)H�)L�)P�)x�)��)��)��)��)��)��)��)��)��)��)��)��)�)�),�)4�)8�)<�)@�)h�)p�)t�)x�)|�)��)��)��)��)��)��)��)��)��)��)�)$�)(�),�)0�)X�)`�)d�)h�)l�)��)��)��)��)��)��)��)��)��)��)�)�)�)�) �)H�)P�)T�)X�)\�)��)��)��)��)��)��)��)��)��)��)��)�)�)�)�)8�)@�)D�)H�)L�)t�)|�)��)��)��)��)��)��)��)��)��)��)��)��)�)(�)0�)4�)8�)<�)d�)l�)p�)t�)x�)��)��)��)��)��)��)��)��)��)��)�) �)$�)(�),�)T�)\�)`�)d�)h�)��)��)��)��)��)��)��)��)��)�)�)�)�)D�)L�)P�)T�)��)��)��)��)��)��)��)��)��)��)�)�)�)4�)<�)@�)D�)p�)x�)|�)��)��)��)��)��)��)��)��)��)$�),�)0�)4�)`�)h�)l�)p�)��)��)��)��)��)��)��)��)�)�) �)$�)P�)X�)\�)`�)��)��)��)��)��)��)��)��)�)�)�)�)@�)H�)L�)P�)T�)|�)��)��)��)��)��)��)��)��)��)��)��)�)�)�)0�)8�)<�)@�)D�)l�)t�)x�)|�)��)��)��)��)��)��)��)��)��)��)��) �)(�),�)0�)4�)\�)d�)h�)l�)p�)��)��)��)��)��)��)��)��)��)��)�)�)�) �)$�)L�)T�)X�)\�)`�)��)��)��)��)��)��)��)��)�)�)�)�)<�)D�)H�)L�)x�)��)��)��)��)��)��)��)��)��)��)�),�)4�)8�)<�)h�)p�)t�)x�)��)��)��)��)��)��)��)��)�)$�)(�),�)X�)`�)d�)h�)��)��)��)��)��)��)��)��)�)�)�)�)H�)P�)T�)X�)��)��)��)��)��)��)��)��)��)�)�)�)8�)@�)D�)H�)t�)|�)��)��)��)��)��)��)�)�)��)��)(�)0�)4�)8�)d�)l�)p�)t�)��)��)��)��)�)�)�)�)�) �)$�)(�),�)T�)\�)`�)d�)h�)��)��)��)��)��)�)�)�)�)�)�)�)�)�)�)D�)L�)P�)T�)X�)��)��)��)��)��)��)�)�)�)�)��)�)�)�)�)4�)<�)@�)D�)H�)p�)x�)|�)��)��)��)��)��)�)�)�)��)��)$�),�)0�)4�)`�)h�)l�)p�)��)��)��)��)��)�)�)�)�)�)�)�) �)$�)(�)P�)X�)\�)`�)d�)��)��)��)��)��)�)�)�)�)�)�)�)�)�)�)@�)H�)L�)P�)T�)|�)��)��)��)��)��)��)�)�)�)�)��)�)�)�)0�)8�)<�)@�)D�)l�)t�)x�)|�)��)��)��)��)��)��)�)�)�)�)��) �)(�),�)0�)4�)\�)d�)h�)l�)p�)��)��)��)��)��)�)�)�)�)�)�)�)�) �)$�)L�)T�)X�)\�)`�)��)��)��)��)��)�)�)�)�)�)*****<*D*H*L*P*x*�*�*�*�*�*�*�*�*�*�*�*�**,*4*8*<*h*p*t*x*�*�*�*�*�*�*�*�*(*,*0*D*H*L*d*h*l*�*�*�*�*�*�*�*�*�*�*�*�*�*�**** *4*8*<*D*T*X*\*p*t*x*�*�*�*�*�*�*�*�*�*�*�*�***$*(*,*D*H*`*d*h*�*�*�*�*�*�*�*�*�*�*�*�****4*8*P*T*X*p*t*�*�*�*�*�*�*�*�*�*�*�****$*(*@*D*H*`*d*|*�*�*�*�*�*�*�*�*�*�*�*�***0*4*8*P*T*l*p*t*�*�*�*�*�*�*�*�*�*�*** *$*(*@*D*\*`*d*|*�*�*�*�*�*�*�*�*�*�*�*�*	*	*	*0	*4	*L	*P	*T	*l	*p	*�	*�	*�	*�	*�	*�	*�	*�	*�	*�	*�	*
*
*
* 
*$
*<
*@
*D
*\
*`
*x
*|
*�
*�
*�
*�
*�
*�
*�
*�
*�
*�
*�
***,*0*4*L*P*h*l*p*�*�*�*�*�*�*�*�*�*�*�**** *$*<*@*X*\*`*x*|*�*�*�*�*�*�*�*�*�*�*
*
*
*,
*0
*H
*L
*P
*h
*l
*�
*�
*�
*�
*�
*�
*�
*�
*�
*�
*�
**** *8*<*@*X*\*t*x*|*�*�*�*�*�*�*�*�*�*�***(*,*0*H*L*d*h*l*�*�*�*�*�*�*�*�*�*�*�**** *,*4*8*<*h*p*t*x*|*�*�*�*�*�*�*�*�*�*�**$*(*,*0*X*`*d*h*l*�*�*�*�*�*�*�*�*�*�***** *H*P*T*X*�*�*�*�*�*�*�*�*�*�*�****8*@*D*H*t*|*�*�*�*�*�*�*�*�*�*�*(*0*4*8*d*l*p*t*�*�*�*�*�*�*�*�*�*�** *$*(*,*T*\*`*d*h*�*�*�*�*�*�*�*�*�*�******D*L*P*T*X*�*�*�*�*�*�*�*�*�*�*�*****4*<*@*D*p*x*|*�*�*�*�*�*�*�*�*�*�*�*�*$*,*0*4*8*`*h*l*p*t*�*�*�*�*�*�*�*�*�*�*** *$*(*P*X*\*`*�*�*�*�*�*�*�*�*�*****@*H*L*P*|*�*�*�*�*�*�*�*�*�****0*8*<*@*D*l*t*x*|*�*�*�*�*�*�*�*�*�*�*�* *(*,*0*4*\*d*h*l*p*�*�*�*�*�*�*�*�*�*�**** *$*L*T*X*\*�*�*�*�*�*�*�*�*****<*D*H*L*P*x*�*�*�*�*�*�*�*�*�*�**,*4*8*<*@*h*p*t*x*�*�*�*�*�*�*�*�*�* *$ *( *, *0 *X *` *d *h *l *� *� *� *� *� *� *� *� *� *!*!*!*!*H!*P!*T!*X!*\!*�!*�!*�!*�!*�!*�!*�!*�!*�!*�!*"*"*"*"*8"*@"*D"*H"*t"*|"*�"*�"*�"*�"*�"*�"*�"*�"*�"*�"*�"*#*(#*0#*4#*8#*d#*l#*p#*t#*�#*�#*�#*�#*�#*�#*�#*�#*�#*$* $*$$*($*T$*\$*`$*d$*�$*�$*�$*�$*�$*�$*�$*�$*�$*�$*%*%*%*%*D%*L%*P%*T%*�%*�%*�%*�%*�%*�%*�%*�%*�%*�%*&*&*&*&*4&*<&*@&*D&*H&*p&*x&*|&*�&*�&*�&*�&*�&*�&*�&*�&*�&*�&*�&*�&*$'*,'*0'*4'*8'*`'*h'*l'*p'*�'*�'*�'*�'*�'*�'*�'*�'*�'*(*(* (*$(*((*P(*X(*\(*`(*�(*�(*�(*�(*�(*�(*�(*�(*)*)*)*)*)*@)*H)*L)*P)*|)*�)*�)*�)*�)*�)*�)*�)*�)*�)*�)*******0**8**<**@**D**l**t**x**|**�**�**�**�**�**�**�**�**�**�** +*(+*,+*0+*4+*\+*d+*h+*l+*p+*�+*�+*�+*�+*�+*�+*�+*�+*�+*,*,*,* ,*$,*L,*T,*X,*\,*`,*�,*�,*�,*�,*�,*�,*�,*�,*�,*�,*-*-*-*-*-*<-*D-*H-*L-*P-*x-*�-*�-*�-*�-*�-*�-*�-*�-*�-*�-*�-*�-*.*.*,.*4.*8.*<.*@.*h.*p.*t.*x.*|.*�.*�.*�.*�.*�.*�.*�.*�.*�.*�.*/*$/*(/*,/*0/*X/*`/*d/*h/*l/*�/*�/*�/*�/*�/*�/*�/*�/*�/*0*0*0*0* 0*H0*P0*T0*X0*\0*�0*�0*�0*�0*�0*�0*�0*�0*�0*�0*�0*1*1*1*1*81*@1*D1*H1*L1*t1*|1*�1*�1*�1*�1*�1*�1*�1*�1*�1*�1*�1*(2*02*42*82*d2*l2*p2*t2*�2*�2*�2*�2*�2*�2*�2*�2*�2*�2*�2*3* 3*$3*(3*,3*03*83*T3*\3*`3*d3*h3*l3*t3*�3*�3*�3*�3*�3*�3*�3*�3*�3*�3*�3*�3*�3*�3*4*4*4*4*4* 4*(4*D4*L4*P4*T4*X4*\4*d4*�4*�4*�4*�4*�4*�4*�4*�4*�4*�4*�4*�4*�4*�4*�4*5*5*5*5*5*5*45*<5*@5*D5*H5*L5*T5*p5*x5*|5*�5*�5*�5*�5*�5*�5*�5*�5*�5*�5*�5*�5*�5*�5*�5*�5*6*6*$6*,6*06*46*86*<6*D6*`6*h6*l6*p6*t6*x6*�6*�6*�6*�6*�6*�6*�6*�6*�6*�6*�6*�6*�6*�6*�6*7*7* 7*$7*(7*,7*47*P7*X7*\7*`7*d7*h7*p7*�7*�7*�7*�7*�7*�7*�7*�7*�7*�7*�7*�7*�7*�7*8*8*8*8*8*8*$8*@8*H8*L8*P8*T8*X8*`8*|8*�8*�8*�8*�8*�8*�8*�8*�8*�8*�8*�8*�8*�8*�8*�8*9*9*9*9*9*09*89*<9*@9*D9*H9*P9*l9*t9*x9*|9*�9*�9*�9*�9*�9*�9*�9*�9*�9*�9*�9*�9*�9*�9*�9*�9*:* :*(:*,:*0:*4:*8:*@:*\:*d:*h:*l:*p:*t:*|:*�:*�:*�:*�:*�:*�:*�:*�:*�:*�:*�:*�:*�:*;*;*;* ;*$;*(;*L;*T;*X;*\;*`;*d;*�;*�;*�;*�;*�;*�;*�;*�;*�;*�;*�;*�;*<*<*<*<*<*<*<<*D<*H<*L<*P<*T<*x<*�<*�<*�<*�<*�<*�<*�<*�<*�<*�<*�<*�<*�<*�<*=*=*=*,=*4=*8=*<=*@=*D=*h=*p=*t=*x=*|=*�=*�=*�=*�=*�=*�=*�=*�=*�=*�=*�=*�=*�=*>*$>*(>*,>*0>*4>*X>*`>*d>*h>*l>*p>*�>*�>*�>*�>*�>*�>*�>*�>*�>*�>*�>*�>*?*?*?*?* ?*$?*H?*P?*T?*X?*\?*`?*�?*�?*�?*�?*�?*�?*�?*�?*�?*�?*�?*�?*�?*@*@*@*@*@*8@*@@*D@*H@*L@*P@*t@*|@*�@*�@*�@*�@*�@*�@*�@*�@*�@*�@*�@*�@*�@*�@*(A*0A*4A*8A*<A*dA*lA*pA*tA*xA*�A*�A*�A*�A*�A*�A*�A*�A*�A*�A*B* B*$B*(B*,B*TB*\B*`B*dB*hB*�B*�B*�B*�B*�B*�B*�B*�B*�B*�B*C*C*C*C*DC*LC*PC*TC*XC*�C*�C*�C*�C*�C*�C*�C*�C*�C*�C*�C*D*D*D*D*4D*<D*@D*DD*HD*pD*xD*|D*�D*�D*�D*�D*�D*�D*�D*�D*�D*�D*�D*�D*`E*hE*lE*pE*tE*�E*�E*�E*�E*�E*�E*�E*�E*�E*�E*F*F* F*$F*PF*XF*\F*`F*dF*�F*�F*�F*�F*�F*�F*�F*�F*�F*G*G*G*G*G*@G*HG*LG*PG*TG*|G*�G*�G*�G*�G*�G*�G*�G*�G*�G*�G*�G*H*H*H*0H*8H*<H*@H*DH*lH*tH*xH*|H*�H*�H*�H*�H*�H*�H*�H*�H*�H*�H*�H* I*(I*,I*0I*4I*\I*dI*hI*lI*pI*�I*�I*�I*�I*�I*�I*�I*�I*�I*�I*J*J*J* J*$J*LJ*TJ*XJ*\J*`J*�J*�J*�J*�J*�J*�J*�J*�J*�J*�J*K*K*K*K*K*<K*DK*HK*LK*PK*xK*�K*�K*�K*�K*�K*�K*�K*�K*�K*�K*�K*�K*L*L*,L*4L*8L*<L*@L*hL*pL*tL*xL*|L*�L*�L*�L*�L*�L*�L*�L*�L*�L*�L*M*$M*(M*,M*0M*XM*`M*dM*hM*lM*�M*�M*�M*�M*�M*�M*�M*�M*�M*�M*N*N*N*N* N*HN*PN*TN*XN*\N*�N*�N*�N*�N*�N*�N*�N*�N*�N*�N*�N*O*O*O*O*8O*@O*DO*HO*LO*tO*|O*�O*�O*�O*�O*�O*�O*�O*�O*�O*�O*(P*0P*4P*8P*dP*lP*pP*tP*�P*�P*�P*�P*�P*�P*�P*�P*Q* Q*$Q*(Q*TQ*\Q*`Q*dQ*�Q*�Q*�Q*�Q*�Q*�Q*�Q*�Q*R*R*R*R*DR*LR*PR*TR*�R*�R*�R*�R*�R*�R*�R*�R*�R*S*S*S*4S*<S*@S*DS*pS*xS*|S*�S*�S*�S*�S*�S*�S*�S*�S*�S*$T*,T*0T*4T*`T*hT*lT*pT*�T*�T*�T*�T*�T*�T*�T*�T*�T*�T*U*U* U*$U*(U*PU*XU*\U*`U*dU*�U*�U*�U*�U*�U*�U*�U*�U*�U*�U*V*V*V*V*V*@V*HV*LV*PV*TV*|V*�V*�V*�V*�V*�V*�V*�V*�V*�V*�V*�V*W*W*W*0W*8W*<W*@W*DW*lW*tW*xW*|W*�W*�W*�W*�W*�W*�W*�W*�W*�W*�W* X*(X*,X*0X*\X*dX*hX*lX*�X*�X*�X*�X*�X*�X*�X*�X*Y*Y*Y* Y*LY*TY*XY*\Y*�Y*�Y*�Y*�Y*�Y*�Y*�Y*�Y*Z*Z*Z*Z*<Z*DZ*HZ*LZ*xZ*�Z*�Z*�Z*�Z*�Z*�Z*�Z*�Z*�Z*�Z*[*,[*4[*8[*<[*h[*p[*t[*x[*�[*�[*�[*�[*�[*�[*�[*�[*\*$\*(\*,\*X\*`\*d\*h\*�\*�\*�\*�\*�\*�\*�\*�\*]*]*]*]*H]*P]*T]*X]*�]*�]*�]*�]*�]*�]*�]*�]*�]*^*^*^*8^*@^*D^*H^*t^*|^*�^*�^*�^*�^*�^*�^*�^*�^*�^*�^*(_*0_*4_*8_*d_*l_*p_*t_*�_*�_*�_*�_*�_*�_*�_*�_*`* `*$`*(`*T`*\`*``*d`*�`*�`*�`*�`*�`*�`*�`*�`*a*a*a*a*a*Da*La*Pa*Ta*Xa*�a*�a*�a*�a*�a*�a*�a*�a*�a*�a*�a*b*b*b*b*4b*<b*@b*Db*Hb*pb*xb*|b*�b*�b*�b*�b*�b*�b*�b*�b*�b*�b*�b*�b*$c*,c*0c*4c*8c*`c*hc*lc*pc*�c*�c*�c*�c*�c*�c*�c*�c*�c*d*d* d*$d*Pd*Xd*\d*`d*�d*�d*�d*�d*�d*�d*�d*�d*�d*�d*e*e*e*e*e*@e*He*Le*Pe*Te*|e*�e*�e*�e*�e*�e*�e*�e*�e*�e*�e*�e*f*f*f*0f*8f*<f*@f*Df*lf*tf*xf*|f*�f*�f*�f*�f*�f*�f*�f*�f*�f*�f*�f* g*(g*,g*0g*4g*\g*dg*hg*lg*pg*�g*�g*�g*�g*�g*�g*�g*�g*�g*�g*h*h*h* h*$h*Lh*Th*Xh*\h*`h*�h*�h*�h*�h*�h*�h*�h*�h*�h*�h*i*i*i*i*i*<i*Di*Hi*Li*Pi*xi*�i*�i*�i*�i*�i*�i*�i*�i*�i*�i*�i*�i*j*j*,j*4j*8j*<j*@j*hj*pj*tj*xj*|j*�j*�j*�j*�j*�j*�j*�j*�j*�j*k*$k*(k*,k*Xk*`k*dk*hk*�k*�k*�k*�k*�k*�k*�k*�k*l*l* l*4l*8l*<l*Tl*Xl*\l*pl*tl*xl*�l*�l*�l*�l*�l*�l*�l*�l*�l*�l*�l*�l*m*m*m*$m*(m*,m*4m*Dm*Hm*Lm*`m*dm*hm*�m*�m*�m*�m*�m*�m*�m*�m*�m*�m*�m*�m*�m*�m*n*n*n*4n*8n*Pn*Tn*Xn*pn*tn*�n*�n*�n*�n*�n*�n*�n*�n*�n*�n*o*o*o*$o*(o*@o*Do*Ho*`o*do*|o*�o*�o*�o*�o*�o*�o*�o*�o*�o*�o*�o*�o*�o*p*p*0p*4p*8p*Pp*Tp*lp*pp*tp*�p*�p*�p*�p*�p*�p*�p*�p*�p*�p*q*q* q*$q*(q*@q*Dq*\q*`q*dq*|q*�q*�q*�q*�q*�q*�q*�q*�q*�q*�q*�q*r*r*r*0r*4r*Lr*Pr*Tr*lr*pr*�r*�r*�r*�r*�r*�r*�r*�r*�r*�r*�r*s*s*s* s*$s*<s*@s*Ds*\s*`s*xs*|s*�s*�s*�s*�s*�s*�s*�s*�s*�s*�s*�s*�s*t*t*,t*0t*4t*Lt*Pt*ht*lt*pt*�t*�t*�t*�t*�t*�t*�t*�t*�t*�t*u*u*u* u*$u*<u*@u*Xu*\u*`u*xu*|u*�u*�u*�u*�u*�u*�u*�u*�u*�u*�u*�u*v*v*v*,v*0v*Hv*Lv*Pv*hv*lv*�v*�v*�v*�v*�v*�v*�v*�v*�v*�v*�v*w*w*w* w*8w*<w*@w*Xw*\w*tw*xw*|w*�w*�w*�w*�w*�w*�w*�w*�w*�w*�w*x*x*(x*,x*0x*Hx*Lx*dx*hx*lx*�x*�x*�x*�x*�x*�x*�x*�x*�x*�x*�x*y*y*y* y*8y*<y*Ty*Xy*\y*ty*xy*�y*�y*�y*�y*�y*�y*�y*�y*�y*�y*�y*�y*z*$z*(z*,z*0z*Xz*`z*dz*hz*lz*�z*�z*�z*�z*�z*�z*�z*�z*�z*�z*{*{*{*{* {*H{*P{*T{*X{*\{*�{*�{*�{*�{*�{*�{*�{*�{*�{*�{*�{*|*|*|*8|*@|*D|*H|*L|*t|*||*�|*�|*�|*�|*�|*�|*�|*�|*�|*�|*�|*(}*0}*4}*8}*d}*l}*p}*t}*�}*�}*�}*�}*�}*�}*�}*�}*~* ~*$~*(~*T~*\~*`~*d~*�~*�~*�~*�~*�~*�~*�~*�~*****D*L*P*T*�*�*�*�*�*�*�*�*�*�*�*�*4�*<�*@�*D�*p�*x�*|�*��*��*��*��*��*�*�*�*��*$�*,�*0�*4�*`�*h�*l�*p�*��*��*��*��*؁*�*�*�*�*�* �*$�*P�*X�*\�*`�*��*��*��*��*Ȃ*Ђ*Ԃ*؂*�*�*�*�*@�*H�*L�*P�*|�*��*��*��*��*��*��*ă*ȃ*̃*�*��*�*�*�*0�*8�*<�*@�*D�*l�*t�*x�*|�*��*��*��*��*��*��*�*�*�*�*��* �*(�*,�*0�*4�*\�*d�*h�*l�*p�*��*��*��*��*��*ԅ*܅*�*�*�*�*�*�* �*L�*T�*X�*\�*`�*��*��*��*��*��*Ć*̆*І*Ԇ*؆*�*�*�*�*�*<�*D�*H�*L�*P�*x�*��*��*��*��*��*��*��*ć*ȇ*�*��*��*�*�*,�*4�*8�*<�*h�*p�*t�*x�*��*��*��*��*��*�*�*�*�*�*$�*(�*,�*X�*`�*d�*h�*��*��*��*��*Љ*؉*܉*�*�*�*�*�*�* �*H�*P�*T�*X�*\�*��*��*��*��*��*��*Ȋ*̊*Њ*Ԋ*��*�*�*�*�*8�*@�*D�*H�*L�*t�*|�*��*��*��*��*��*��*��*ċ*�*�*��*��*�*(�*0�*4�*8�*d�*l�*p�*t�*��*��*��*��*܌*�*�*�*�* �*$�*(�*,�*T�*\�*`�*d�*��*��*��*��*̍*ԍ*؍*܍*�*�*�*�*�*D�*L�*P�*T�*��*��*��*��*��*Ď*Ȏ*̎*Ў*��*�*�*�*�*4�*<�*@�*D�*H�*p�*x�*|�*��*��*��*��*��*��*�*�*�*��*$�*,�*0�*4�*8�*`�*h�*l�*p�*t�*��*��*��*��*ؐ*�*�*�*�*�*�* �*$�*P�*X�*\�*`�*d�*��*��*��*��*ȑ*Б*ԑ*ؑ*ܑ*�*�*�*�*@�*H�*L�*P�*|�*��*��*��*��*��*Ē*Ȓ*̒*�*��*�*�*0�*8�*<�*@�*l�*t�*x�*|�*��*��*��*��*��*��*�*�*�*�* �*(�*,�*0�*\�*d�*h�*l�*��*��*��*��*��*Ԕ*ܔ*�*�*�*�*�*�* �*$�*L�*T�*X�*\�*`�*��*��*��*��*��*ĕ*̕*Е*ԕ*ؕ*�*�*�*�*�*<�*D�*H�*L�*x�*��*��*��*��*��*��*Ė*Ȗ*�*��*��*�*�*,�*4�*8�*<�*h�*p�*t�*x�*��*��*��*��*�*�*�*�*�*�*$�*(�*,�*X�*`�*d�*h�*��*��*��*��*��*И*ؘ*ܘ*�*�*�*�*�*�* �*H�*P�*T�*X�*��*��*��*��*��*��*ș*̙*Й*ԙ*��*�*�*�*�*8�*@�*D�*H�*L�*t�*|�*��*��*��*��*��*��*��*�*�*��*��*�*(�*0�*4�*8�*<�*d�*l�*p�*t�*x�*��*��*��*��*��*ܛ*�*�*�*�*�* �*$�*(�*,�*T�*\�*`�*d�*h�*��*��*��*��*��*̜*Ԝ*؜*ܜ*�*�*�*�*�*�*D�*L�*P�*T�*X�*��*��*��*��*��*��*ĝ*ȝ*̝*Н*��*�*�*�*�*4�*<�*@�*D�*H�*p�*x�*|�*��*��*��*��*��*��*�*�*�*��*��*$�*,�*0�*4�*8�*`�*h�*l�*p�*t�*��*��*��*��*��*؟*�*�*�*�*�*�* �*$�*(�*P�*X�*\�*`�*d�*��*��*��*��*Ƞ*Р*Ԡ*ؠ*�*�*�*�*@�*H�*L�*P�*|�*��*��*��*��*��*ġ*ȡ*С*ء*�*��*�*�*�*�*�*0�*8�*<�*@�*H�*L�*P�*l�*t�*x�*|�*��*��*��*��*��*��*��*��*Ģ*Ȣ*�*�*�*�*��*�*�* �*(�*,�*0�*8�*<�*@�*\�*d�*h�*l�*t�*x�*|�*��*��*��*��*��*��*��*ԣ*ܣ*�*�*�*�*�*�*�*�* �*(�*,�*0�*L�*T�*X�*\�*d�*h�*l�*��*��*��*��*��*��*��*Ĥ*̤*Ф*Ԥ*ܤ*�*�*�*�*�*�*�*�* �*<�*D�*H�*L�*T�*X�*\�*x�*��*��*��*��*��*��*��*��*��*ĥ*̥*Х*ԥ*�*��*��*�*�*�*�*,�*4�*8�*<�*D�*H�*L�*h�*p�*t�*x�*��*��*��*��*��*��*��*��*��*Ħ*�*�*�*�*��*�*�*$�*(�*,�*4�*<�*X�*`�*d�*h�*p�*x�*��*��*��*��*��*��*Ч*ا*ܧ*�*�*�*�*�*�*�*$�*(�*,�*H�*P�*T�*X�*`�*d�*h�*��*��*��*��*��*��*��*��*Ȩ*̨*Ш*ب*ܨ*�*��*�*�*�*�*�*8�*@�*D�*H�*P�*X�*`�*d�*t�*|�*��*��*��*��*��*��*��*��*��*��*ȩ*̩*Щ*�*�*��*��*�*�*�*(�*0�*4�*8�*@�*H�*d�*l�*p�*t�*|�*��*��*��*��*��*��*��*��*��*ܪ*�*�*�*�*��*�* �*$�*(�*0�*8�*@�*D�*T�*\�*`�*d�*l�*t�*|�*��*��*��*��*��*��*��*��*��*̫*ԫ*ث*ܫ*�*�*�*�*�*�* �*(�*D�*L�*P�*T�*\�*d�*l�*p�*��*��*��*��*��*��*��*��*��*Ĭ*Ȭ*̬*Ԭ*ܬ*�*�*��*�*�*�*�*�* �*$�*4�*<�*@�*D�*L�*T�*\�*`�*p�*x�*|�*��*��*��*��*��*��*��*��*��*ĭ*̭*ԭ*ح*�*�*�*��*�*�*�*�*$�*,�*0�*4�*<�*D�*L�*P�*`�*h�*l�*p�*x�*��*��*��*��*��*��*��*��*��*Į*Ȯ*خ*�*�*�*�*��*�*�*�*�* �*$�*,�*4�*<�*@�*P�*X�*\�*`�*h�*p�*x�*|�*��*��*��*��*��*��*��*��*ȯ*Я*ԯ*د*�*�*�*�*�*�*�*�*�*$�*,�*0�*@�*H�*L�*P�*X�*`�*h�*l�*|�*��*��*��*��*��*��*��*��*��*İ*Ȱ*а*ذ*�*�*�*��*�*�*�*�*�* �*0�*8�*<�*@�*H�*P�*X�*\�*l�*t�*x�*|�*��*��*��*��*��*��*��*��*��*ȱ*б*Ա*�*�*�*�*��*�*�*�* �*(�*,�*0�*\�*d�*h�*l�*��*��*��*��*Բ*ܲ*�*�*�*�*�*�* �*$�*L�*T�*X�*\�*`�*��*��*��*��*��*ij*̳*г*Գ*س*�*�*�*�*�*<�*D�*H�*L�*P�*x�*��*��*��*��*��*��*��*Ĵ*�*��*��*�*�*,�*4�*8�*<�*@�*h�*p�*t�*x�*|�*��*��*��*��*��*�*�*�*�*�*�*$�*(�*,�*0�*X�*`�*d�*h�*l�*��*��*��*��*��*�*�*�*�*H�*P�*T�*X�*��*��*��*��*��*ȷ*̷*з*Է*��*�*�*�*8�*@�*D�*H�*t�*|�*��*��*��*��*��*��*�*�*��*��*(�*0�*4�*8�*d�*l�*p�*t�*��*��*��*��*ܹ*�*�*�*�* �*$�*(�*T�*\�*`�*d�*h�*��*��*��*��*̺*Ժ*غ*ܺ*�*�*�*�*D�*L�*P�*T�*��*��*��*��*��*Ļ*Ȼ*̻*��*�*�*�*4�*<�*@�*D�*p�*x�*|�*��*��*��*��*��*�*�*�*��*$�*,�*0�*4�*`�*h�*l�*p�*��*��*��*��*ؽ*�*�*�*�*�* �*$�*P�*X�*\�*`�*��*��*��*��*Ⱦ*о*Ծ*ؾ*�*�*�*�*@�*H�*L�*P�*|�*��*��*��*��*��*Ŀ*ȿ*�*��*�*�*0�*8�*<�*@�*l�*t�*x�*|�*��*��*��*��*�*�*�*�* �*(�*,�*0�*\�*d�*h�*l�*��*��*��*��*�*�*�*�*�*�*�* �*L�*T�*X�*\�*��*��*��*��*��*��*��*��*��*��*�*�*�*�*<�*D�*H�*L�*x�*��*��*��*��*��*��*��*��*��*��*�*,�*4�*8�*<�*h�*p�*t�*x�*��*��*��*��*��*��*��*��*��*�*$�*(�*,�*0�*X�*`�*d�*h�*l�*��*��*��*��*��*��*��*��*��*�*�*�*�* �*H�*P�*T�*X�*��*��*��*��*��*��*��*��*��*�*�*�*8�*@�*D�*H�*t�*|�*��*��*��*��*��*��*��*��*��*��*(�*0�*4�*8�*d�*l�*p�*t�*��*��*��*��*��*��*��*��*�* �*$�*(�*T�*\�*`�*d�*��*��*��*��*��*��*��*��*�*�*�*�*D�*L�*P�*T�*��*��*��*��*��*��*��*��*��*�*�*�*4�*<�*@�*D�*p�*x�*|�*��*��*��*��*��*��*��*��*��*$�*,�*0�*4�*`�*h�*l�*p�*��*��*��*��*��*��*��*��*�*�* �*$�*P�*X�*\�*`�*��*��*��*��*��*��*��*��*�*�*�*�*@�*H�*L�*P�*|�*��*��*��*��*��*��*��*��*��*�*�*0�*8�*<�*@�*l�*t�*x�*|�*��*��*��*��*��*��*��*��* �*(�*,�*0�*\�*d�*h�*l�*��*��*��*��*��*��*��*��*�*�*�* �*L�*T�*X�*\�*��*��*��*��*��*��*��*��*�*�*�*�*<�*D�*H�*L�*x�*��*��*��*��*��*��*��*��*��*��*�*,�*4�*8�*<�*h�*p�*t�*x�*��*��*��*��*��*��*��*��*�*$�*(�*,�*X�*`�*d�*h�*��*��*��*��*��*��*��*��*�*�*�*�*H�*P�*T�*X�*��*��*��*��*��*��*��*��*��*�*�*�*8�*@�*D�*H�*t�*|�*��*��*��*��*��*��*��*��*��*��*(�*0�*4�*8�*d�*l�*p�*t�*��*��*��*��*��*��*��*��*��*�* �*$�*(�*T�*\�*`�*d�*��*��*��*��*��*��*��*��*�*�*�*�*�*D�*L�*P�*T�*��*��*��*��*��*��*��*��*��*�*�*�*4�*<�*@�*D�*p�*x�*|�*��*��*��*��*��*��*��*��*��*$�*,�*0�*4�*`�*h�*l�*p�*��*��*��*��*��*��*��*��*��*�*�* �*$�*P�*X�*\�*`�*��*��*��*��*��*��*��*��*�*�*�*�*@�*H�*L�*P�*T�*|�*��*��*��*��*��*��*��*��*��*�*�*0�*8�*<�*@�*l�*t�*x�*|�*��*��*��*��*��*��*��*�*�*�*,�*0�*4�*H�*L�*P�*h�*l�*p�*��*��*��*��*��*��*��*��*��*��*��*��*��*�*�*�*�* �*$�*8�*<�*@�*X�*\�*`�*t�*x�*|�*��*��*��*��*��*��*��*��*��*��*��*�*�*(�*,�*0�*H�*L�*d�*h�*l�*��*��*��*��*��*��*��*��*��*��*��*�*�*�* �*8�*<�*@�*T�*X�*\�*t�*x�*��*��*��*��*��*��*��*��*��*��*�*�*�*(�*,�*D�*H�*L�*d�*h�*��*��*��*��*��*��*��*��*��*��*��*��*�*�*�*4�*8�*<�*T�*X�*p�*t�*x�*��*��*��*��*��*��*��*��*��*��*�*�*$�*(�*,�*D�*H�*`�*d�*h�*��*��*��*��*��*��*��*��*��*��*��*��*�*�*�*4�*8�*P�*T�*X�*p�*t�*��*��*��*��*��*��*��*��*��*��*�*�*�*$�*(�*@�*D�*H�*`�*d�*|�*��*��*��*��*��*��*��*��*��*��*��*��*�*�*0�*4�*8�*D�*L�*P�*T�*X�*��*��*��*��*��*��*��*��*��*�*�*�*4�*<�*@�*D�*p�*x�*|�*��*��*��*��*��*��*��*��*��*$�*,�*0�*4�*`�*h�*l�*p�*��*��*��*��*��*��*��*��*�*�* �*$�*P�*X�*\�*`�*��*��*��*��*��*��*��*��*�*�*�*�*@�*H�*L�*P�*|�*��*��*��*��*��*��*��*��*��*�*�*0�*8�*<�*@�*l�*t�*x�*|�*��*��*��*��*��*��*��*��* �*(�*,�*0�*\�*d�*h�*l�*��*��*��*��*��*��*��*��*�*�*�* �*L�*T�*X�*\�*��*��*��*��*��*��*��*��*�*�*�*�*<�*D�*H�*L�*x�*��*��*��*��*��*��*��*��*��*��*�*,�*4�*8�*<�*h�*p�*t�*x�*��*��*��*��*��*��*��*��*�*$�*(�*,�*0�*X�*`�*d�*h�*��*��*��*��*��*��*��*��*�*�*�*�*H�*P�*T�*X�*��*��*��*��*��*��*��*��*��*�*�*�*8�*@�*D�*H�*t�*|�*��*��*��*��*��*��*��*��*��*��*(�*0�*4�*8�*d�*l�*p�*t�*��*��*��*��*�*�*�*�*�* �*$�*(�*T�*\�*`�*d�*��*��*��*��*�*�*�*�*�*�*�*�*D�*L�*P�*T�*��*��*��*��*��*�*�*�*��*�*�*�*4�*<�*@�*D�*p�*x�*|�*��*��*��*��*��*�*�*�*��*$�*,�*0�*4�*`�*h�*l�*p�*��*��*��*��*�*�*�*�*�*�* �*$�*(�*P�*X�*\�*`�*��*��*��*��*�*�*�*�*�*�*�*�*@�*H�*L�*P�*|�*��*��*��*��*��*�*�*�*��*�*�*0�*8�*<�*@�*D�*l�*t�*x�*|�*��*��*��*��*��*�*�*�*�*��* �*(�*,�*0�*\�*d�*h�*l�*��*��*��*��*�*�*�*�*�*�*�* �*L�*T�*X�*\�*��*��*��*��*�*�*�*�*�*�*�*�*<�*D�*H�*L�*x�*��*��*��*��*��*��*��*�*�*��*��*+,+4+8+<+h+p+t+x+�+�+�+�+�+�+�+�++$+(+,+X+`+d+h+�+�+�+�+�+�+�+�+�+++++ +H+P+T+X+�+�+�+�+�+�+�+�+�++++8+@+D+H+t+|+�+�+�+�+�+�+�+�+�+�++(+0+4+8+<+d+l+p+t+�+�+�+�+�+�+�+�+�++ +$+(+T+\+`+d+h+�+�+�+�+�+�+�+�+++++D+L+P+T+�+�+�+�+�+�+�+�+�++++4+<+@+D+p+x+|+�+�+�+�+�+�+�+�+�+$+,+0+4+`+h+l+p+�+�+�+�+�+�+�+�+	+	+ 	+$	+P	+X	+\	+`	+�	+�	+�	+�	+�	+�	+�	+�	+
+
+
+
+@
+H
+L
+P
+|
+�
+�
+�
+�
+�
+�
+�
+�
+�
+++0+8+<+@+l+t+x+|+�+�+�+�+�+�+�+�+ +(+,+0+\+d+h+l+�+�+�+�+�+�+�+�+
+
+
+ 
+L
+T
+X
+\
+�
+�
+�
+�
+�
+�
+�
+�
+�
+++++<+D+H+L+P+x+�+�+�+�+�+�+�+�+�+�+�++,+4+8+<+h+p+t+x+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+++$+(+,+4+8+<+X+`+d+h+p+t+x+�+�+�+�+�+�+�+�+�+�+�+�+�+�+++++$+(+,+H+P+T+X+`+d+h+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+++++++8+@+D+H+P+T+X+t+|+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�++++(+0+4+8+@+H+P+T+d+l+p+t+|+�+�+�+�+�+�+�+�+�+�+�+�+�++ +$+(+0+8+T+\+`+d+l+t+|+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+++++ +$+(+D+L+P+T+\+d+�+�+�+�+�+�+�+�+�+�+�+�+�++++++4+<+@+D+L+T+\+`+p+x+|+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+++$+,+0+4+<+D+`+h+l+p+x+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+++++ +$+,+4+<+@+P+X+\+`+h+p+x+|+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�++++++$+,+0+@+H+L+P+X+`+h+l+|+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�++++++ +0+8+<+@+H+P+X+\+l+t+x+|+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�++++ +(+,+0+8+@+H+L+\+d+h+l+t+|+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+++++ +(+0+8+<+L+T+X+\+d+l+t+x+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�+�++++++ +(+,+<+D+H+L+T+\+d+h+x+�+�+�+�+�+�+�+�+�+�+�++,+4+8+<+@+h+p+t+x+�+�+�+�+�+�+�+�++$+(+,+X+`+d+h+�+�+�+�+�+�+�+�+ + + + +H +P +T +X +� +� +� +� +� +� +� +� +� +� +!+!+!+t!+|!+�!+�!+�!+�!+�!+�!+�!+�!+�!+�!+�!+("+0"+4"+8"+d"+l"+p"+t"+�"+�"+�"+�"+�"+�"+�"+�"+�"+�"+#+ #+$#+(#+T#+\#+`#+d#+h#+�#+�#+�#+�#+�#+�#+�#+�#+$+$+$+$+D$+L$+P$+T$+�$+�$+�$+�$+�$+�$+�$+�$+�$+%+%+%+4%+<%+@%+D%+p%+x%+|%+�%+�%+�%+�%+�%+�%+�%+�%+�%+$&+,&+0&+4&+`&+h&+l&+p&+�&+�&+�&+�&+�&+�&+�&+�&+'+'+ '+$'+P'+X'+\'+`'+�'+�'+�'+�'+�'+�'+�'+�'+(+(+(+(+@(+H(+L(+P(+|(+�(+�(+�(+�(+�(+�(+�(+�(+�(+)+)+0)+8)+<)+@)+l)+t)+x)+|)+�)+�)+�)+�)+�)+�)+�)+�)+ *+(*+,*+0*+\*+d*+h*+l*+�*+�*+�*+�*+�*+�*+�*+�*+++++++ ++L++T++X++\++�++�++�++�++�++�++�++�++,+,+,+,+<,+D,+H,+L,+x,+�,+�,+�,+�,+�,+�,+�,+�,+�,+�,+-+,-+4-+8-+<-+h-+p-+t-+x-+�-+�-+�-+�-+�-+�-+�-+�-+.+$.+(.+,.+X.+`.+d.+h.+�.+�.+�.+�.+�.+�.+�.+�.+/+/+/+/+H/+P/+T/+X/+�/+�/+�/+�/+�/+�/+�/+�/+�/+0+0+0+80+@0+D0+H0+t0+|0+�0+�0+�0+�0+�0+�0+�0+�0+�0+�0+(1+01+41+81+d1+l1+p1+t1+�1+�1+�1+�1+�1+�1+�1+�1+2+ 2+$2+(2+T2+\2+`2+d2+�2+�2+�2+�2+�2+�2+�2+�2+3+3+3+3+D3+L3+P3+T3+�3+�3+�3+�3+�3+�3+�3+�3+�3+4+4+4+44+<4+@4+D4+p4+x4+|4+�4+�4+�4+�4+�4+�4+�4+�4+�4+$5+,5+05+45+`5+h5+l5+p5+�5+�5+�5+�5+�5+�5+�5+�5+6+6+ 6+$6+P6+X6+\6+`6+�6+�6+�6+�6+�6+�6+�6+�6+7+7+7+7+@7+H7+L7+P7+|7+�7+�7+�7+�7+�7+�7+�7+�7+�7+8+8+08+88+<8+@8+l8+t8+x8+|8+�8+�8+�8+�8+�8+�8+�8+�8+ 9+(9+,9+09+\9+d9+h9+l9+�9+�9+�9+�9+�9+�9+�9+�9+:+:+:+ :+L:+T:+X:+\:+�:+�:+�:+�:+�:+�:+�:+�:+;+;+;+;+<;+D;+H;+L;+x;+�;+�;+�;+�;+�;+�;+�;+�;+�;+�;+<+,<+4<+8<+<<+h<+p<+t<+x<+�<+�<+�<+�<+�<+�<+�<+�<+=+$=+(=+,=+X=+`=+d=+h=+�=+�=+�=+�=+�=+�=+�=+�=+>+>+>+>+H>+P>+T>+X>+�>+�>+�>+�>+�>+�>+�>+�>+�>+?+?+?+8?+@?+D?+H?+t?+|?+�?+�?+�?+�?+�?+�?+�?+�?+�?+�?+(@+0@+4@+8@+d@+l@+p@+t@+�@+�@+�@+�@+�@+�@+�@+�@+A+ A+$A+(A+TA+\A+`A+dA+�A+�A+�A+�A+�A+�A+�A+�A+B+B+B+B+DB+LB+PB+TB+�B+�B+�B+�B+�B+�B+�B+�B+�B+C+C+C+4C+<C+@C+DC+pC+xC+|C+�C+�C+�C+�C+�C+�C+�C+�C+�C+$D+,D+0D+4D+`D+hD+lD+pD+�D+�D+�D+�D+�D+�D+�D+�D+E+E+ E+$E+PE+XE+\E+`E+�E+�E+�E+�E+�E+�E+�E+�E+F+F+F+F+@F+HF+LF+PF+|F+�F+�F+�F+�F+�F+�F+�F+�F+�F+G+G+0G+8G+<G+@G+lG+tG+xG+|G+�G+�G+�G+�G+�G+�G+�G+�G+ H+(H+,H+0H+\H+dH+hH+lH+�H+�H+�H+�H+�H+�H+�H+�H+I+I+I+ I+LI+TI+XI+\I+�I+�I+�I+�I+�I+�I+�I+�I+J+J+J+J+<J+DJ+HJ+LJ+xJ+�J+�J+�J+�J+�J+�J+�J+�J+�J+�J+K+,K+4K+8K+<K+hK+pK+tK+xK+�K+�K+�K+�K+�K+�K+�K+�K+L+$L+(L+,L+XL+`L+dL+hL+�L+�L+�L+�L+�L+�L+�L+�L+M+M+M+M+HM+PM+TM+XM+�M+�M+�M+�M+�M+�M+�M+�M+�M+N+N+N+8N+@N+DN+HN+tN+|N+�N+�N+�N+�N+�N+�N+�N+�N+�N+�N+(O+0O+4O+8O+dO+lO+pO+tO+�O+�O+�O+�O+�O+�O+�O+�O+P+ P+$P+(P+TP+\P+`P+dP+�P+�P+�P+�P+�P+�P+�P+�P+Q+Q+Q+Q+DQ+LQ+PQ+TQ+�Q+�Q+�Q+�Q+�Q+�Q+�Q+�Q+�Q+R+R+R+4R+<R+@R+DR+pR+xR+|R+�R+�R+�R+�R+�R+�R+�R+�R+�R+$S+,S+0S+4S+8S+`S+hS+lS+pS+tS+�S+�S+�S+�S+�S+�S+�S+�S+�S+T+T+ T+$T+PT+XT+\T+`T+�T+�T+�T+�T+�T+�T+�T+�T+U+U+U+U+@U+HU+LU+PU+|U+�U+�U+�U+�U+�U+�U+�U+�U+�U+V+V+0V+8V+<V+@V+lV+tV+xV+|V+�V+�V+�V+�V+�V+�V+�V+�V+ W+(W+,W+0W+\W+dW+hW+lW+�W+�W+�W+�W+�W+�W+�W+�W+X+X+X+ X+LX+TX+XX+\X+�X+�X+�X+�X+�X+�X+�X+�X+Y+Y+Y+Y+<Y+DY+HY+LY+xY+�Y+�Y+�Y+�Y+�Y+�Y+�Y+�Y+�Y+�Y+Z+,Z+4Z+8Z+<Z+hZ+pZ+tZ+xZ+�Z+�Z+�Z+�Z+�Z+�Z+�Z+�Z+[+$[+([+,[+X[+`[+d[+h[+�[+�[+�[+�[+�[+�[+�[+�[+\+\+\+\+H\+P\+T\+X\+�\+�\+�\+�\+�\+�\+�\+�\+�\+]+]+]+8]+@]+D]+H]+t]+|]+�]+�]+�]+�]+�]+�]+�]+�]+�]+�]+(^+0^+4^+8^+d^+l^+p^+t^+�^+�^+�^+�^+�^+�^+�^+�^+_+ _+$_+(_+T_+\_+`_+d_+�_+�_+�_+�_+�_+�_+�_+�_+`+`+`+`+D`+L`+P`+T`+�`+�`+�`+�`+�`+�`+�`+�`+�`+a+a+a+4a+<a+@a+Da+pa+xa+|a+�a+�a+�a+�a+�a+�a+�a+�a+�a+$b+,b+0b+4b+`b+hb+lb+pb+�b+�b+�b+�b+�b+�b+�b+�b+c+c+ c+$c+Pc+Xc+\c+`c+�c+�c+�c+�c+�c+�c+�c+�c+d+d+d+d+@d+Hd+Ld+Pd+|d+�d+�d+�d+�d+�d+�d+�d+�d+�d+e+e+0e+8e+<e+@e+le+te+xe+|e+�e+�e+�e+�e+�e+�e+�e+�e+ f+(f+,f+0f+\f+df+hf+lf+�f+�f+�f+�f+�f+�f+�f+�f+g+g+g+ g+Lg+Tg+Xg+\g+�g+�g+�g+�g+�g+�g+�g+�g+h+h+h+h+<h+Dh+Hh+Lh+xh+�h+�h+�h+�h+�h+�h+�h+�h+�h+�h+i+,i+4i+8i+<i+hi+pi+ti+xi+�i+�i+�i+�i+�i+�i+�i+�i+j+$j+(j+,j+Xj+`j+dj+hj+�j+�j+�j+�j+�j+�j+�j+�j+k+k+k+k+Hk+Pk+Tk+Xk+�k+�k+�k+�k+�k+�k+�k+�k+�k+l+l+l+8l+@l+Dl+Hl+tl+|l+�l+�l+�l+�l+�l+�l+�l+�l+�l+�l+(m+0m+4m+8m+dm+lm+pm+tm+�m+�m+�m+�m+�m+�m+�m+�m+n+ n+$n+(n+Tn+\n+`n+dn+�n+�n+�n+�n+�n+�n+�n+�n+o+o+o+o+Do+Lo+Po+To+�o+�o+�o+�o+�o+�o+�o+�o+�o+p+p+p+4p+<p+@p+Dp+pp+xp+|p+�p+�p+�p+�p+�p+�p+�p+�p+�p+$q+,q+0q+4q+`q+hq+lq+pq+�q+�q+�q+�q+�q+�q+�q+�q+�q+r+r+ r+$r+Pr+Xr+\r+`r+dr+�r+�r+�r+�r+�r+�r+�r+�r+�r+s+s+s+s+@s+Hs+Ls+Ps+|s+�s+�s+�s+�s+�s+�s+�s+�s+�s+�s+t+t+t+0t+8t+<t+@t+lt+tt+xt+|t+�t+�t+�t+�t+�t+�t+�t+�t+ u+(u+,u+0u+\u+du+hu+lu+�u+�u+�u+�u+�u+�u+�u+�u+v+v+v+ v+Lv+Tv+Xv+\v+`v+�v+�v+�v+�v+�v+�v+�v+�v+w+w+w+w+w+<w+Dw+Hw+Lw+Pw+xw+�w+�w+�w+�w+�w+�w+�w+�w+�w+�w+�w+�w+�w+�w+�w+�w+x+x+x+x+,x+4x+8x+<x+Dx+Hx+Lx+hx+px+tx+xx+�x+�x+�x+�x+�x+�x+�x+�x+�x+y+$y+(y+,y+Xy+`y+dy+hy+ly+�y+�y+�y+�y+�y+�y+�y+�y+Hz+Pz+Tz+Xz+�z+�z+�z+�z+�z+�z+�z+�z+�z+{+{+{+8{+@{+D{+H{+t{+|{+�{+�{+�{+�{+�{+�{+�{+�{+�{+�{+(|+0|+4|+8|+d|+l|+p|+t|+�|+�|+�|+�|+�|+�|+�|+�|+}+ }+$}+(}+T}+\}+`}+d}+�}+�}+�}+�}+�}+�}+�}+�}+~+~+~+~+D~+L~+P~+T~+�~+�~+�~+�~+�~+�~+�~+�~+�~++++4+<+@+D+p+x+|+�+�+�+�+�+�+�+�+�+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+؀+�+�+�+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+ȁ+Ё+ԁ+؁+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+Ă+Ȃ+�+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+�+�+�+�+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+Ԅ+܄+�+�+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+ą+̅+Ѕ+ԅ+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+Ć+�+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+�+�+�+�+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+Ј+؈+܈+�+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+ȉ+̉+Љ+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+�+�+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+܋+�+�+�+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+̌+Ԍ+،+܌+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+č+ȍ+̍+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+�+�+�+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+؏+�+�+�+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+Ȑ+А+Ԑ+ؐ+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+đ+ȑ+�+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+�+�+�+�+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+ԓ+ܓ+�+�+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+Ĕ+̔+Д+Ԕ+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+ĕ+�+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+�+�+�+�+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+З+ؗ+ܗ+�+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+Ș+̘+И+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+�+�+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+ܚ+�+�+�+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+̛+ԛ+؛+ܛ+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+Ĝ+Ȝ+̜+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+�+�+�+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+؞+�+�+�+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+ȟ+П+ԟ+؟+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+Ġ+Ƞ+�+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+�+�+�+�+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+Ԣ+ܢ+�+�+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+ģ+̣+У+ԣ+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+Ĥ+�+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+�+�+�+�+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+Ц+ئ+ܦ+�+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+ȧ+̧+Ч+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+�+�+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+ܩ+�+�+�+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+̪+Ԫ+ت+ܪ+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+ī+ȫ+̫+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+�+�+�+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+ح+�+�+�+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+Ȯ+Ю+Ԯ+خ+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+į+ȯ+�+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+�+�+�+�+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+Ա+ܱ+�+�+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+IJ+̲+в+Բ+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+ij+�+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+�+�+�+�+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+е+ص+ܵ+�+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+ȶ+̶+ж+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+�+�+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+ܸ+�+�+�+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+̹+Թ+ع+ܹ+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+ĺ+Ⱥ+̺+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+�+�+�+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+ؼ+�+�+�+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+Ƚ+н+Խ+ؽ+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+ľ+Ⱦ+�+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+�+�+�+�+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+�+�+�+�+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+�+�+�+�+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+��+��+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+��+��+��+��+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+��+��+��+��+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+��+��+��+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+��+��+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+��+��+��+��+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+��+��+��+��+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+��+��+��+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+��+��+��+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+��+��+��+��+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+��+��+��+��+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+��+��+��+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+��+��+��+��+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+��+��+��+��+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+��+��+��+��+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+��+��+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+��+��+��+��+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+��+��+��+��+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+��+��+��+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+��+��+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+��+��+��+��+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+��+��+��+��+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+��+��+��+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+��+��+��+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+��+��+��+��+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+��+��+��+��+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+��+��+��+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+��+��+��+��+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+��+��+��+��+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+��+��+��+��+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+��+��+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+��+��+��+��+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+��+��+��+��+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+��+��+��+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+��+��+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+��+��+��+��+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+��+��+��+��+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+��+��+��+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+��+��+��+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+��+��+��+��+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+��+��+��+��+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+��+��+��+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+��+��+��+��+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+��+��+��+��+�+�+�+ �+L�+T�+X�+\�+��+��+��+��+��+��+��+��+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+��+��+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+��+��+��+��+�+$�+(�+,�+X�+`�+d�+h�+��+��+��+��+��+��+��+��+�+�+�+�+H�+P�+T�+X�+��+��+��+��+��+��+��+��+��+�+�+�+8�+@�+D�+H�+t�+|�+��+��+��+��+��+��+��+��+��+��+(�+0�+4�+8�+d�+l�+p�+t�+��+��+��+��+��+��+��+��+�+ �+$�+(�+T�+\�+`�+d�+��+��+��+��+�+�+�+�+�+�+�+�+D�+L�+P�+T�+��+��+��+��+��+�+�+�+��+�+�+�+4�+<�+@�+D�+p�+x�+|�+��+��+��+��+��+�+�+�+��+$�+,�+0�+4�+`�+h�+l�+p�+��+��+��+��+�+�+�+�+�+�+ �+$�+P�+X�+\�+`�+��+��+��+��+�+�+�+�+�+�+�+�+@�+H�+L�+P�+|�+��+��+��+��+��+�+�+�+��+�+�+0�+8�+<�+@�+l�+t�+x�+|�+��+��+��+��+�+�+�+�+ �+(�+,�+0�+\�+d�+h�+l�+��+��+��+��+�+�+�+�+L�+T�+X�+\�+��+��+��+��+�+�+�+�+�+�+�+�+<�+D�+H�+L�+x�+��+��+��+��+��+��+�+�+��+��+�+,�+4�+8�+<�+h�+p�+t�+x�+��+��+��+��+�+�+�+�+,$,(,,,X,`,d,h,�,�,�,�,�,�,�,�,,,,,H,P,T,X,�,�,�,�,�,�,�,�,�,,,,8,@,D,H,t,|,�,�,�,�,�,�,�,�,�,�,(,0,4,8,d,l,p,t,�,�,�,�,�,�,�,�,, ,$,(,T,\,`,d,�,�,�,�,�,�,�,�,,,,,D,L,P,T,�,�,�,�,�,�,�,�,�,,,,4,<,@,D,p,x,|,�,�,�,�,�,�,�,�,�,$,,,0,4,`,h,l,p,�,�,�,�,�,�,�,�,,, ,$,P,X,\,`,�,�,�,�,�,�,�,�,	,	,	,	,@	,H	,L	,P	,|	,�	,�	,�	,�	,�	,�	,�	,�	,�	,
,
,0
,8
,<
,@
,l
,t
,x
,|
,�
,�
,�
,�
,�
,�
,�
,�
, ,(,,,0,\,d,h,l,�,�,�,�,�,�,�,�,,,, ,L,T,X,\,�,�,�,�,�,�,�,�,
,
,
,
,<
,D
,H
,L
,x
,�
,�
,�
,�
,�
,�
,�
,�
,�
,�
,,,,4,8,<,h,p,t,x,�,�,�,�,�,�,�,�,,$,(,,,X,`,d,h,�,�,�,�,�,�,�,�,,,,,H,P,T,X,�,�,�,�,�,�,�,�,�,,,,8,@,D,H,t,|,�,�,�,�,�,�,�,�,�,�,(,0,4,8,d,l,p,t,�,�,�,�,�,�,�,�,, ,$,(,T,\,`,d,�,�,�,�,�,�,�,�,,,,,D,L,P,T,�,�,�,�,�,�,�,�,�,,,,4,<,@,D,p,x,|,�,�,�,�,�,�,�,�,�,$,,,0,4,`,h,l,p,�,�,�,�,�,�,�,�,,, ,$,P,X,\,`,�,�,�,�,�,�,�,�,,,,,@,H,L,P,|,�,�,�,�,�,�,�,�,�,,,0,8,<,@,l,t,x,|,�,�,�,�,�,�,�,�, ,(,,,0,\,d,h,l,�,�,�,�,�,�,�,�,,,, ,L,T,X,\,�,�,�,�,�,�,�,�,,,,,<,D,H,L,x,�,�,�,�,�,�,�,�,�,�,,,,4,8,<,h,p,t,x,�,�,�,�,�,�,�,�,,$,(,,,X,`,d,h,�,�,�,�,�,�,�,�,,,,,H,P,T,X,�,�,�,�,�,�,�,�,�, , , ,8 ,@ ,D ,H ,t ,| ,� ,� ,� ,� ,� ,� ,� ,� ,� ,� ,(!,0!,4!,8!,d!,l!,p!,t!,�!,�!,�!,�!,�!,�!,�!,�!,", ",$",(",T",\",`",d",�",�",�",�",�",�",�",�",#,#,#,#,D#,L#,P#,T#,�#,�#,�#,�#,�#,�#,�#,�#,�#,$,$,$,4$,<$,@$,D$,p$,x$,|$,�$,�$,�$,�$,�$,�$,�$,�$,�$,$%,,%,0%,4%,`%,h%,l%,p%,�%,�%,�%,�%,�%,�%,�%,�%,&,&, &,$&,P&,X&,\&,`&,�&,�&,�&,�&,�&,�&,�&,�&,',',',',@',H',L',P',|',�',�',�',�',�',�',�',�',�',(,(,0(,8(,<(,@(,l(,t(,x(,|(,�(,�(,�(,�(,�(,�(,�(,�(, ),(),,),0),\),d),h),l),�),�),�),�),�),�),�),�),*,*,*, *,L*,T*,X*,\*,�*,�*,�*,�*,�*,�*,�*,�*,+,+,+,+,<+,D+,H+,L+,x+,�+,�+,�+,�+,�+,�+,�+,�+,�+,�+,,,,,,4,,8,,<,,h,,p,,t,,x,,�,,�,,�,,�,,�,,�,,�,,�,,-,$-,(-,,-,X-,`-,d-,h-,�-,�-,�-,�-,�-,�-,�-,�-,.,.,.,.,H.,P.,T.,X.,�.,�.,�.,�.,�.,�.,�.,�.,�.,/,/,/,8/,@/,D/,H/,t/,|/,�/,�/,�/,�/,�/,�/,�/,�/,�/,�/,(0,00,40,80,d0,l0,p0,t0,�0,�0,�0,�0,�0,�0,�0,�0,1, 1,$1,(1,T1,\1,`1,d1,�1,�1,�1,�1,�1,�1,�1,�1,2,2,2,2,D2,L2,P2,T2,�2,�2,�2,�2,�2,�2,�2,�2,�2,3,3,3,43,<3,@3,D3,p3,x3,|3,�3,�3,�3,�3,�3,�3,�3,�3,�3,$4,,4,04,44,`4,h4,l4,p4,�4,�4,�4,�4,�4,�4,�4,�4,5,5, 5,$5,P5,X5,\5,`5,�5,�5,�5,�5,�5,�5,�5,�5,6,6,6,6,@6,H6,L6,P6,|6,�6,�6,�6,�6,�6,�6,�6,�6,�6,7,7,07,87,<7,@7,l7,t7,x7,|7,�7,�7,�7,�7,�7,�7,�7,�7, 8,(8,,8,08,\8,d8,h8,l8,�8,�8,�8,�8,�8,�8,�8,�8,9,9,9, 9,L9,T9,X9,\9,�9,�9,�9,�9,�9,�9,�9,�9,:,:,:,:,<:,D:,H:,L:,x:,�:,�:,�:,�:,�:,�:,�:,�:,�:,�:,;,,;,4;,8;,<;,h;,p;,t;,x;,�;,�;,�;,�;,�;,�;,�;,�;,<,$<,(<,,<,X<,`<,d<,h<,�<,�<,�<,�<,�<,�<,�<,�<,=,=,=,=,H=,P=,T=,X=,�=,�=,�=,�=,�=,�=,�=,�=,�=,>,>,>,8>,@>,D>,H>,t>,|>,�>,�>,�>,�>,�>,�>,�>,�>,�>,�>,(?,0?,4?,8?,d?,l?,p?,t?,�?,�?,�?,�?,�?,�?,�?,�?,@, @,$@,(@,T@,\@,`@,d@,�@,�@,�@,�@,�@,�@,�@,�@,A,A,A,A,DA,LA,PA,TA,�A,�A,�A,�A,�A,�A,�A,�A,�A,B,B,B,4B,<B,@B,DB,pB,xB,|B,�B,�B,�B,�B,�B,�B,�B,�B,�B,$C,,C,0C,4C,`C,hC,lC,pC,�C,�C,�C,�C,�C,�C,�C,�C,D,D, D,$D,PD,XD,\D,`D,�D,�D,�D,�D,�D,�D,�D,�D,E,E,E,E,@E,HE,LE,PE,|E,�E,�E,�E,�E,�E,�E,�E,�E,�E,F,F,0F,8F,<F,@F,lF,tF,xF,|F,�F,�F,�F,�F,�F,�F,�F,�F, G,(G,,G,0G,\G,dG,hG,lG,�G,�G,�G,�G,�G,�G,�G,�G,H,H,H, H,LH,TH,XH,\H,�H,�H,�H,�H,�H,�H,�H,�H,I,I,I,I,<I,DI,HI,LI,xI,�I,�I,�I,�I,�I,�I,�I,�I,�I,�I,J,,J,4J,8J,<J,hJ,pJ,tJ,xJ,�J,�J,�J,�J,�J,�J,�J,�J,K,$K,(K,,K,XK,`K,dK,hK,�K,�K,�K,�K,�K,�K,�K,�K,L,L,L,L,HL,PL,TL,XL,�L,�L,�L,�L,�L,�L,�L,�L,�L,M,M,M,8M,@M,DM,HM,tM,|M,�M,�M,�M,�M,�M,�M,�M,�M,�M,�M,(N,0N,4N,8N,dN,lN,pN,tN,�N,�N,�N,�N,�N,�N,�N,�N,O, O,$O,(O,TO,\O,`O,dO,�O,�O,�O,�O,�O,�O,�O,�O,P,P,P,P,DP,LP,PP,TP,�P,�P,�P,�P,�P,�P,�P,�P,�P,Q,Q,Q,4Q,<Q,@Q,DQ,pQ,xQ,|Q,�Q,�Q,�Q,�Q,�Q,�Q,�Q,�Q,�Q,$R,,R,0R,4R,`R,hR,lR,pR,�R,�R,�R,�R,�R,�R,�R,�R,S,S, S,$S,PS,XS,\S,`S,�S,�S,�S,�S,�S,�S,�S,�S,T,T,T,T,@T,HT,LT,PT,|T,�T,�T,�T,�T,�T,�T,�T,�T,�T,U,U,0U,8U,<U,@U,lU,tU,xU,|U,�U,�U,�U,�U,�U,�U,�U,�U, V,(V,,V,0V,\V,dV,hV,lV,�V,�V,�V,�V,�V,�V,�V,�V,W,W,W, W,LW,TW,XW,\W,�W,�W,�W,�W,�W,�W,�W,�W,X,X,X,X,<X,DX,HX,LX,xX,�X,�X,�X,�X,�X,�X,�X,�X,�X,�X,Y,,Y,4Y,8Y,<Y,hY,pY,tY,xY,�Y,�Y,�Y,�Y,�Y,�Y,�Y,�Y,Z,$Z,(Z,,Z,XZ,`Z,dZ,hZ,�Z,�Z,�Z,�Z,�Z,�Z,�Z,�Z,[,[,[,[,H[,P[,T[,X[,�[,�[,�[,�[,�[,�[,�[,�[,�[,\,\,\,8\,@\,D\,H\,t\,|\,�\,�\,�\,�\,�\,�\,�\,�\,�\,�\,(],0],4],8],d],l],p],t],�],�],�],�],�],�],�],�],^, ^,$^,(^,T^,\^,`^,d^,�^,�^,�^,�^,�^,�^,�^,�^,_,_,_,_,D_,L_,P_,T_,�_,�_,�_,�_,�_,�_,�_,�_,�_,`,`,`,4`,<`,@`,D`,p`,x`,|`,�`,�`,�`,�`,�`,�`,�`,�`,�`,$a,,a,0a,4a,`a,ha,la,pa,�a,�a,�a,�a,�a,�a,�a,�a,b,b, b,$b,Pb,Xb,\b,`b,�b,�b,�b,�b,�b,�b,�b,�b,c,c,c,c,@c,Hc,Lc,Pc,|c,�c,�c,�c,�c,�c,�c,�c,�c,�c,d,d,0d,8d,<d,@d,ld,td,xd,|d,�d,�d,�d,�d,�d,�d,�d,�d, e,(e,,e,0e,\e,de,he,le,�e,�e,�e,�e,�e,�e,�e,�e,f,f,f, f,Lf,Tf,Xf,\f,�f,�f,�f,�f,�f,�f,�f,�f,g,g,g,g,<g,Dg,Hg,Lg,xg,�g,�g,�g,�g,�g,�g,�g,�g,�g,�g,h,,h,4h,8h,<h,hh,ph,th,xh,�h,�h,�h,�h,�h,�h,�h,�h,i,$i,(i,,i,Xi,`i,di,hi,�i,�i,�i,�i,�i,�i,�i,�i,j,j,j,j,Hj,Pj,Tj,Xj,�j,�j,�j,�j,�j,�j,�j,�j,�j,k,k,k,8k,@k,Dk,Hk,tk,|k,�k,�k,�k,�k,�k,�k,�k,�k,�k,�k,(l,0l,4l,8l,dl,ll,pl,tl,�l,�l,�l,�l,�l,�l,�l,�l,m, m,$m,(m,Tm,\m,`m,dm,�m,�m,�m,�m,�m,�m,�m,�m,n,n,n,n,Dn,Ln,Pn,Tn,�n,�n,�n,�n,�n,�n,�n,�n,�n,o,o,o,4o,<o,@o,Do,po,xo,|o,�o,�o,�o,�o,�o,�o,�o,�o,�o,$p,,p,0p,4p,`p,hp,lp,pp,�p,�p,�p,�p,�p,�p,�p,�p,q,q, q,$q,Pq,Xq,\q,`q,�q,�q,�q,�q,�q,�q,�q,�q,r,r,r,r,@r,Hr,Lr,Pr,|r,�r,�r,�r,�r,�r,�r,�r,�r,�r,s,s,0s,8s,<s,@s,ls,ts,xs,|s,�s,�s,�s,�s,�s,�s,�s,�s, t,(t,,t,0t,\t,dt,ht,lt,�t,�t,�t,�t,�t,�t,�t,�t,u,u,u, u,Lu,Tu,Xu,\u,�u,�u,�u,�u,�u,�u,�u,�u,v,v,v,v,<v,Dv,Hv,Lv,xv,�v,�v,�v,�v,�v,�v,�v,�v,�v,�v,w,,w,4w,8w,<w,hw,pw,tw,xw,�w,�w,�w,�w,�w,�w,�w,�w,x,$x,(x,,x,Xx,`x,dx,hx,�x,�x,�x,�x,�x,�x,�x,�x,y,y,y,y,Hy,Py,Ty,Xy,�y,�y,�y,�y,�y,�y,�y,�y,�y,z,z,z,8z,@z,Dz,Hz,tz,|z,�z,�z,�z,�z,�z,�z,�z,�z,�z,�z,({,0{,4{,8{,d{,l{,p{,t{,�{,�{,�{,�{,�{,�{,�{,�{,|, |,$|,(|,T|,\|,`|,d|,�|,�|,�|,�|,�|,�|,�|,�|,},},},},D},L},P},T},�},�},�},�},�},�},�},�},�},~,~,~,4~,<~,@~,D~,p~,x~,|~,�~,�~,�~,�~,�~,$,,,0,4,8,`,h,l,p,t,�,�,�,�,�,�,�,�,�,�,�,�, �,$�,P�,X�,\�,`�,d�,��,��,��,��,Ȁ,Ѐ,Ԁ,؀,�,�,�,�,@�,H�,L�,P�,|�,��,��,��,��,��,��,ā,ȁ,́,�,��,�,�,�,0�,8�,<�,@�,D�,l�,t�,x�,|�,��,��,��,��,��,��,�,�,�,�,��, �,(�,,�,0�,4�,\�,d�,h�,l�,p�,��,��,��,��,��,ԃ,܃,�,�,�,�,�,�, �,$�,L�,T�,X�,\�,`�,��,��,��,��,��,Ą,̄,Є,Ԅ,؄,�,�,�,�,�,<�,D�,H�,L�,P�,x�,��,��,��,��,��,��,��,ą,ȅ,�,��,��,�,,�,4�,8�,<�,h�,p�,t�,x�,|�,��,��,��,��,�,�,�,�,�,�,$�,(�,,�,0�,X�,`�,d�,h�,l�,��,��,��,��,��,Ї,؇,܇,�,�,�,�,�,�, �,H�,P�,T�,X�,��,��,��,��,��,��,Ȉ,̈,Ј,��,�,�,�,�,8�,@�,D�,H�,L�,t�,|�,��,��,��,��,��,��,��,ĉ,�,�,��,��,�,(�,0�,4�,8�,<�,d�,l�,p�,t�,x�,��,��,��,��,��,܊,�,�,�,�,�, �,$�,(�,,�,T�,\�,`�,d�,h�,��,��,��,��,��,̋,ԋ,؋,܋,�,�,�,�,�,�,D�,L�,P�,T�,X�,��,��,��,��,��,��,Č,Ȍ,̌,Ќ,��,�,�,�,�,4�,<�,@�,D�,H�,p�,x�,|�,��,��,��,��,��,��,�,�,�,��,��,$�,,�,0�,4�,8�,`�,h�,l�,p�,��,��,��,��,؎,�,�,�,�,�, �,$�,P�,X�,\�,`�,��,��,��,��,ȏ,Џ,ԏ,؏,�,�,�,�,@�,H�,L�,P�,|�,��,��,��,��,��,Đ,Ȑ,�,��,�,�,0�,8�,<�,@�,l�,t�,x�,|�,��,��,��,��,�,�,�,�, �,(�,,�,0�,\�,d�,h�,l�,��,��,��,��,Ԓ,ܒ,�,�,�,�,�, �,L�,T�,X�,\�,��,��,��,��,ē,̓,Г,ԓ,�,�,�,�,<�,D�,H�,L�,x�,��,��,��,��,��,��,Ĕ,�,��,��,�,,�,4�,8�,<�,h�,p�,t�,x�,��,��,��,��,�,�,�,�,�,$�,(�,,�,X�,`�,d�,h�,��,��,��,��,Ж,ؖ,ܖ,�,�,�,�,�,H�,P�,T�,X�,��,��,��,��,��,ȗ,̗,З,��,�,�,�,8�,@�,D�,H�,t�,|�,��,��,��,��,��,��,�,�,��,��,(�,0�,4�,8�,d�,l�,p�,t�,��,��,��,��,ܙ,�,�,�,�, �,$�,(�,T�,\�,`�,d�,��,��,��,��,̚,Ԛ,ؚ,ܚ,�,�,�,�,D�,L�,P�,T�,��,��,��,��,��,ě,ț,̛,��,�,�,�,4�,<�,@�,D�,p�,x�,|�,��,��,��,��,��,�,�,�,��,$�,,�,0�,4�,`�,h�,l�,p�,��,��,��,��,؝,�,�,�,�,�, �,$�,P�,X�,\�,`�,��,��,��,��,Ȟ,О,Ԟ,؞,�,�,�,�,@�,H�,L�,P�,|�,��,��,��,��,��,ğ,ȟ,�,��,�,�,0�,8�,<�,@�,l�,t�,x�,|�,��,��,��,��,�,�,�,�, �,(�,,�,0�,\�,d�,h�,l�,��,��,��,��,ԡ,ܡ,�,�,�,�,�, �,L�,T�,X�,\�,��,��,��,��,Ģ,̢,Т,Ԣ,�,�,�,�,<�,D�,H�,L�,x�,��,��,��,��,��,��,ģ,�,��,��,�,,�,4�,8�,<�,h�,p�,t�,x�,��,��,��,��,�,�,�,�,�,$�,(�,,�,X�,`�,d�,h�,��,��,��,��,Х,إ,ܥ,�,�,�,�,�,H�,P�,T�,X�,��,��,��,��,��,Ȧ,̦,Ц,��,�,�,�,8�,@�,D�,H�,t�,|�,��,��,��,��,��,��,�,�,��,��,(�,0�,4�,8�,d�,l�,p�,t�,��,��,��,��,ܨ,�,�,�,�, �,$�,(�,T�,\�,`�,d�,��,��,��,��,̩,ԩ,ة,ܩ,�,�,�,�,D�,L�,P�,T�,��,��,��,��,��,Ī,Ȫ,̪,��,�,�,�,4�,<�,@�,D�,p�,x�,|�,��,��,��,��,��,�,�,�,��,$�,,�,0�,4�,`�,h�,l�,p�,��,��,��,��,ج,�,�,�,�,�, �,$�,P�,X�,\�,`�,��,��,��,��,ȭ,Э,ԭ,ح,�,�,�,�,@�,H�,L�,P�,|�,��,��,��,��,��,Į,Ȯ,�,��,�,�,0�,8�,<�,@�,l�,t�,x�,|�,��,��,��,��,�,�,�,�, �,(�,,�,0�,\�,d�,h�,l�,��,��,��,��,԰,ܰ,�,�,�,�,�, �,L�,T�,X�,\�,��,��,��,��,ı,̱,б,Ա,�,�,�,�,<�,D�,H�,L�,x�,��,��,��,��,��,��,IJ,�,��,��,�,,�,4�,8�,<�,h�,p�,t�,x�,��,��,��,��,�,�,�,�,�,$�,(�,,�,X�,`�,d�,h�,��,��,��,��,д,ش,ܴ,�,�,�,�,�,H�,P�,T�,X�,��,��,��,��,��,ȵ,̵,е,��,�,�,�,8�,@�,D�,H�,t�,|�,��,��,��,��,��,��,Ķ,�,�,��,��,(�,0�,4�,8�,d�,l�,p�,t�,��,��,��,��,ܷ,�,�,�,�,�, �,$�,(�,,�,T�,\�,`�,d�,h�,��,��,��,��,��,̸,Ը,ظ,ܸ,�,�,�,�,�,�,D�,L�,P�,T�,X�,��,��,��,��,��,��,Ĺ,ȹ,̹,й,��,�,�,�,�,4�,<�,@�,D�,H�,p�,x�,|�,��,��,��,��,��,��,��,�,�,�,��,$�,,�,0�,4�,`�,h�,l�,p�,t�,��,��,��,��,ػ,�,�,�,�,�,�, �,$�,P�,X�,\�,`�,��,��,��,��,��,ȼ,м,Լ,ؼ,�,�,�,�,@�,H�,L�,P�,T�,|�,��,��,��,��,��,Ľ,Ƚ,�,��,�,�,0�,8�,<�,@�,D�,l�,t�,x�,|�,��,��,��,��,��,��,�,�,�,�, �,(�,,�,0�,\�,d�,h�,l�,��,��,��,��,��,Կ,ܿ,�,�,�,�,�,�, �,$�,L�,T�,X�,\�,`�,��,��,��,��,��,�,�,�,�,�,�,�,�,�,�,<�,D�,H�,L�,P�,x�,��,��,��,��,��,��,��,�,�,�,��,��,�,�,,�,4�,8�,<�,@�,h�,p�,t�,x�,|�,��,��,��,��,��,��,��,��,��,��,�,$�,(�,,�,X�,`�,d�,h�,l�,��,��,��,��,��,��,��,��,��,��,�,�,�,�,H�,P�,T�,X�,��,��,��,��,��,��,��,��,��,��,�,�,�,8�,@�,D�,H�,L�,t�,|�,��,��,��,��,��,��,��,��,��,��,�,(�,0�,4�,8�,<�,d�,l�,p�,t�,x�,��,��,��,��,��,��,��,��,��,��,�, �,$�,(�,,�,T�,\�,`�,d�,h�,��,��,��,��,��,��,��,��,��,��,�,�,�,�,�,D�,L�,P�,T�,X�,��,��,��,��,��,��,��,��,��,��,�,�,�,4�,<�,@�,D�,p�,x�,|�,��,��,��,��,��,��,��,��,��,$�,,�,0�,4�,`�,h�,l�,p�,��,��,��,��,��,��,��,��,�,�, �,$�,P�,X�,\�,`�,��,��,��,��,��,��,��,��,�,�,�,�,@�,H�,L�,P�,|�,��,��,��,��,��,��,��,��,��,�,�,0�,8�,<�,@�,l�,t�,x�,|�,��,��,��,��,��,��,��,��, �,(�,,�,0�,\�,d�,h�,l�,��,��,��,��,��,��,��,��,�,�,�, �,L�,T�,X�,\�,��,��,��,��,��,��,��,��,�,�,�,�,<�,D�,H�,L�,x�,��,��,��,��,��,��,��,��,��,��,�,,�,4�,8�,<�,h�,p�,t�,x�,��,��,��,��,��,��,��,��,�,$�,(�,,�,X�,`�,d�,h�,��,��,��,��,��,��,��,��,�,�,�,�,H�,P�,T�,X�,��,��,��,��,��,��,��,��,��,�,�,�,8�,@�,D�,H�,t�,|�,��,��,��,��,��,��,��,��,��,��,(�,0�,4�,8�,d�,l�,p�,t�,��,��,��,��,��,��,��,��,�, �,$�,(�,T�,\�,`�,d�,��,��,��,��,��,��,��,��,�,�,�,�,D�,L�,P�,T�,��,��,��,��,��,��,��,��,��,�,�,�,4�,<�,@�,D�,p�,x�,|�,��,��,��,��,��,��,��,��,��,$�,,�,0�,4�,`�,h�,l�,p�,��,��,��,��,��,��,��,��,�,�, �,$�,P�,X�,\�,`�,��,��,��,��,��,��,��,��,�,�,�,�,@�,H�,L�,P�,|�,��,��,��,��,��,��,��,��,��,�,�,0�,8�,<�,@�,l�,t�,x�,|�,��,��,��,��,��,��,��,��, �,(�,,�,0�,\�,d�,h�,l�,��,��,��,��,��,��,��,��,�,�,�, �,$�,L�,T�,X�,\�,`�,��,��,��,��,��,��,��,��,��,��,�,�,�,�,<�,D�,H�,L�,P�,x�,��,��,��,��,��,��,��,��,��,��,��,��,�,,�,4�,8�,<�,@�,h�,p�,t�,x�,|�,��,��,��,��,��,��,��,��,��,��,�,$�,(�,,�,X�,`�,d�,h�,l�,��,��,��,��,��,��,��,��,��,��,�,�,�,�, �,H�,P�,T�,X�,\�,��,��,��,��,��,��,��,��,��,��,��,�,�,�,8�,@�,D�,H�,t�,|�,��,��,��,��,��,��,��,��,��,��,��,��,�,(�,0�,4�,8�,<�,d�,l�,p�,t�,x�,��,��,��,��,��,��,��,��,��,��,�, �,$�,(�,,�,T�,\�,`�,d�,h�,��,��,��,��,��,��,��,��,��,��,�,�,�,�,�,D�,L�,P�,T�,X�,��,��,��,��,��,��,��,��,��,�,�,�,4�,<�,@�,D�,p�,x�,|�,��,��,��,��,��,��,��,��,��,��,$�,,�,0�,4�,`�,h�,l�,p�,t�,��,��,��,��,��,��,��,��,��,�,�, �,$�,(�,P�,X�,\�,`�,��,��,��,��,��,��,��,��,�,�,�,�,@�,H�,L�,P�,|�,��,��,��,��,��,��,��,��,��,�,�,0�,8�,<�,@�,D�,l�,t�,x�,|�,��,��,��,��,��,��,��,��,��,��,��, �,(�,,�,0�,\�,d�,h�,l�,p�,��,��,��,��,��,��,��,��,��,��,�,�,�, �,$�,L�,T�,X�,\�,`�,��,��,��,��,��,��,��,��,��,�,�,�,�,�,<�,D�,H�,L�,x�,��,��,��,��,��,��,��,��,��,��,��,�,�,,�,4�,8�,<�,@�,h�,p�,t�,x�,|�,��,��,��,��,��,��,��,��,��,��,�,$�,(�,,�,0�,X�,`�,d�,h�,l�,��,��,��,��,��,��,��,��,��,�,�,�,�,H�,P�,T�,X�,\�,��,��,��,��,��,��,��,��,��,��,��,�,�,�,�,8�,@�,D�,H�,t�,|�,��,��,��,��,��,��,��,��,��,��,��,��,�,(�,0�,4�,8�,<�,d�,l�,p�,t�,x�,��,��,��,��,��,��,��,��,��,��,�, �,$�,(�,,�,T�,\�,`�,d�,h�,��,��,��,��,��,��,��,��,��,��,�,�,�,�,D�,L�,P�,T�,��,��,��,��,��,�,�,�,��,�,�,�,4�,<�,@�,D�,H�,p�,x�,|�,��,��,��,��,��,��,��,�,�,�,��,��,$�,,�,0�,4�,8�,`�,h�,l�,p�,t�,��,��,��,��,��,�,�,�,�,�,�,�, �,$�,P�,X�,\�,`�,d�,��,��,��,��,��,�,�,�,�,�,�,�,�,�,�,@�,H�,L�,P�,T�,|�,��,��,��,��,�,�,�,�,�,�,�,�,�,�, �,$�,<�,@�,D�,X�,\�,`�,x�,|�,��,��,��,��,��,��,��,�,�,�,�,�,�,��,�,�,�,,�,0�,4�,H�,L�,P�,h�,l�,p�,��,��,��,��,��,��,�,�,�,�,��,�,�,�, �,8�,<�,@�,X�,\�,t�,x�,|�,��,��,��,��,��,�,�,�,�,�,�,�,(�,,�,0�,H�,L�,d�,h�,l�,��,��,��,��,��,��,�,�,�,�,��,�,�,�, �,8�,<�,T�,X�,\�,t�,x�,��,��,��,��,��,�,�,�,�,�,�,�,�,(�,,�,D�,H�,L�,d�,h�,��,��,��,��,��,��,��,�,�,�,��,��,---4-8-<-T-X-p-t-x-�-�-�-�-�-�-�-�-�-�-�---$-(-,-D-H-`-d-h-�-�-�-�-�-�-�-�-�-�-�-�-�-----4-8-P-T-X-p-t-�-�-�-�-�-�-�-�-�-�----$-(-@-D-H-`-d-|-�-�-�-�-�-�-�-�-�-�-�-�---0-4-8-P-T-l-p-t-�-�-�-�-�-�-�-�-�-�---- -$-(-@-D-H-\-`-d-|-�-�-�-�-�-�-�-�-�-�-�-�-�-�----0-4-L-P-T-l-p-�-�-�-�-�-�-�-�-�-�---- -$-<-@-D-\-`-x-|-�-�-�-�-�-�-�-�-�-�-�---,-0-4-L-P-h-l-p-�-�-�-�-�-�-�-�-�-�-�-	-	-	- 	-$	-<	-@	-X	-\	-`	-x	-|	-�	-�	-�	-�	-�	-�	-�	-�	-�	-�	-
-
-
-,
-0
-H
-L
-P
-h
-l
-�
-�
-�
-�
-�
-�
-�
-�
-�
-�
-�
---- -8-<-@-X-\-t-x-|-�-�-�-�-�-�-�-�-�-�---(-,-0-8-H-L-d-h-l-�-�-�-�-�-�-�-�-�-�-�-
-
-
- 
-8
-<
-T
-X
-\
-t
-x
-�
-�
-�
-�
-�
-�
-�
-�
-�
-�
----(-,-D-H-L-d-h-�-�-�-�-�-�-�-�-�-�-�-�----4-8-<-T-X-p-t-x-�-�-�-�-�-�-�-�-�-�---$-(-,-D-H-`-d-h-�-�-�-�-�-�-�-�-�-�-�-�-�-�---(-0-4-8-<-@-d-l-p-t-x-|-�-�-�-�-�-�-�-�-�-�-�-�-- -$-(-,-0-T-\-`-d-h-l-�-�-�-�-�-�-�-�-�-�-�-�------ -D-L-P-T-X-\-�-�-�-�-�-�-�-�-�-�-�-�-�------4-<-@-D-H-L-p-x-|-�-�-�-�-�-�-�-�-�-�-�-�-�-�--$-,-0-4-8-<-`-h-l-p-t-x-�-�-�-�-�-�-�-�-�-�-�--- -$-(-P-X-\-`-d-�-�-�-�-�-�-�-�-�-�------@-H-L-P-T-|-�-�-�-�-�-�-�-�-�-�-�----0-8-<-@-D-l-t-x-|-�-�-�-�-�-�-�-�-�-�-�- -(-,-0-4-\-d-h-l-p-�-�-�-�-�-�-�-�-�-�---- -$-L-T-X-\-�-�-�-�-�-�-�-�-�-�------<-D-H-L-P-x-�-�-�-�-�-�-�-�-�-�-�-�---,-4-8-<-@-�-�-�-�-�-�-�-�-�-�--$-(-,-0-X-`-d-h-l-�-�-�-�-�-�-�-�-�-�----- -H-P-T-X-\-�-�-�-�-�-�-�-�-�-�-�-----8-@-D-H-L-t-|-�-�-�-�-�-�-�-�-�-�-�-( -0 -4 -8 -d -l -p -t -� -� -� -� -� -� -� -� -!- !-$!-(!-T!-\!-`!-d!-�!-�!-�!-�!-�!-�!-�!-�!-"-"-"-"-D"-L"-P"-T"-�"-�"-�"-�"-�"-�"-�"-�"-�"-#-#-#-4#-<#-@#-D#-p#-x#-|#-�#-�#-�#-�#-�#-�#-�#-�#-�#-$$-,$-0$-4$-`$-h$-l$-p$-�$-�$-�$-�$-�$-�$-�$-�$-%-%- %-$%-P%-X%-\%-`%-�%-�%-�%-�%-�%-�%-�%-�%-&-&-&-&-@&-H&-L&-P&-|&-�&-�&-�&-�&-�&-�&-�&-�&-�&-'-'-0'-8'-<'-@'-l'-t'-x'-|'-�'-�'-�'-�'-�'-�'-�'-�'- (-((-,(-0(-\(-d(-h(-l(-�(-�(-�(-�(-�(-�(-�(-�(-)-)-)- )-L)-T)-X)-\)-�)-�)-�)-�)-�)-�)-�)-�)-*-*-*-*-<*-D*-H*-L*-x*-�*-�*-�*-�*-�*-�*-�*-�*-�*-�*-+-,+-4+-8+-<+-h+-p+-t+-x+-�+-�+-�+-�+-�+-�+-�+-�+-,-$,-(,-,,-X,-`,-d,-h,-�,-�,-�,-�,-�,-�,-�,-�,-�,--------- --H--P--T--X--\--�--�--�--�--�--�--�--�--�--�--�--.-.-.-.-8.-@.-D.-H.-L.-t.-|.-�.-�.-�.-�.-�.-�.-�.-�.-�.-�.-�.-�.-/-(/-0/-4/-8/-</-d/-l/-p/-t/-x/-�/-�/-�/-�/-�/-�/-�/-�/-�/-�/-0- 0-$0-(0-,0-T0-\0-`0-d0-h0-�0-�0-�0-�0-�0-�0-�0-�0-�0-�0-1-1-1-1-1-D1-L1-P1-T1-X1-�1-�1-�1-�1-�1-�1-�1-�1-�1-�1-�1-2-2-2-2-42-<2-@2-D2-H2-p2-x2-|2-�2-�2-�2-�2-�2-�2-�2-�2-�2-�2-�2-�2-$3-,3-03-43-83-`3-h3-l3-p3-t3-�3-�3-�3-�3-�3-�3-�3-�3-�3-�3-4-4- 4-$4-(4-P4-X4-\4-`4-d4-�4-�4-�4-�4-�4-�4-�4-�4-�4-�4-5-5-5-5-5-@5-H5-L5-P5-T5-|5-�5-�5-�5-�5-�5-�5-�5-�5-�5-�5-�5-6-6-6-06-86-<6-@6-D6-l6-t6-x6-|6-�6-�6-�6-�6-�6-�6-�6-�6-�6-�6-�6- 7-(7-,7-07-47-\7-d7-h7-l7-p7-�7-�7-�7-�7-�7-�7-�7-�7-�7-�7-8-8-8- 8-L8-T8-X8-\8-`8-�8-�8-�8-�8-�8-�8-�8-�8-�8-9-9-9-9-9-<9-D9-H9-L9-P9-x9-�9-�9-�9-�9-�9-�9-�9-�9-�9-�9-�9-�9-:-:-,:-4:-8:-<:-@:-h:-p:-t:-x:-|:-�:-�:-�:-�:-�:-�:-�:-�:-�:-�:-X;-`;-d;-h;-�;-�;-�;-�;-�;-�;-�;-�;-<-<-<-<- <-H<-P<-T<-X<-�<-�<-�<-�<-�<-�<-�<-�<-�<-=-=-=-8=-@=-D=-H=-t=-|=-�=-�=-�=-�=-�=-�=-�=-�=-�=-�=-(>-0>-4>-8>-d>-l>-p>-t>-�>-�>-�>-�>-�>-�>-�>-�>-�>-?- ?-$?-(?-T?-\?-`?-d?-�?-�?-�?-�?-�?-�?-�?-�?-@-@-@-@-D@-L@-P@-T@-�@-�@-�@-�@-�@-�@-�@-�@-�@-A-A-A-4A-<A-@A-DA-pA-xA-|A-�A-�A-�A-�A-�A-�A-�A-�A-�A-$B-,B-0B-4B-`B-hB-lB-pB-�B-�B-�B-�B-�B-�B-�B-�B-C-C- C-$C-PC-XC-\C-`C-�C-�C-�C-�C-�C-�C-�C-�C-D-D-D-D-@D-HD-LD-PD-|D-�D-�D-�D-�D-�D-�D-�D-�D-�D-E-E-0E-8E-<E-@E-lE-tE-xE-|E-�E-�E-�E-�E-�E-�E-�E-�E- F-(F-,F-0F-\F-dF-hF-lF-�F-�F-�F-�F-�F-�F-�F-�F-�F-G-G-G- G-$G-LG-TG-XG-\G-�G-�G-�G-�G-�G-�G-�G-�G-�G-H-H-H-H-<H-DH-HH-LH-xH-�H-�H-�H-�H-�H-�H-�H-�H-�H-�H-�H-I-I-,I-4I-8I-<I-@I-hI-pI-tI-xI-�I-�I-�I-�I-�I-�I-�I-�I-�I-�I-J-$J-(J-,J-XJ-`J-dJ-hJ-�J-�J-�J-�J-�J-�J-�J-�J-K-K-K-K-HK-PK-TK-XK-�K-�K-�K-�K-�K-�K-�K-�K-�K-L-L-L-8L-@L-DL-HL-tL-|L-�L-�L-�L-�L-�L-�L-�L-�L-�L-�L-(M-0M-4M-8M-dM-lM-pM-tM-�M-�M-�M-�M-�M-�M-�M-�M-N- N-$N-(N-TN-\N-`N-dN-�N-�N-�N-�N-�N-�N-�N-�N-O-O-O-O-DO-LO-PO-TO-�O-�O-�O-�O-�O-�O-�O-�O-�O-P-P-P-4P-<P-@P-DP-pP-xP-|P-�P-�P-�P-�P-�P-�P-�P-�P-�P-$Q-,Q-0Q-4Q-`Q-hQ-lQ-pQ-�Q-�Q-�Q-�Q-�Q-�Q-�Q-�Q-R-R- R-$R-PR-XR-\R-`R-�R-�R-�R-�R-�R-�R-�R-�R-S-S-S-S-@S-HS-LS-PS-|S-�S-�S-�S-�S-�S-�S-�S-�S-�S-T-T-0T-8T-<T-@T-lT-tT-xT-|T-�T-�T-�T-�T-�T-�T-�T-�T- U-(U-,U-0U-\U-dU-hU-lU-�U-�U-�U-�U-�U-�U-�U-�U-V-V-V- V-LV-TV-XV-\V-�V-�V-�V-�V-�V-�V-�V-�V-W-W-W-W-<W-DW-HW-LW-xW-�W-�W-�W-�W-�W-�W-�W-�W-�W-�W-X-,X-4X-8X-<X-hX-pX-tX-xX-�X-�X-�X-�X-�X-�X-�X-�X-Y-$Y-(Y-,Y-XY-`Y-dY-hY-�Y-�Y-�Y-�Y-�Y-�Y-�Y-�Y-Z-Z-Z-Z-HZ-PZ-TZ-XZ-�Z-�Z-�Z-�Z-�Z-�Z-�Z-�Z-�Z-[-[-[-8[-@[-D[-H[-t[-|[-�[-�[-�[-�[-�[-�[-�[-�[-�[-�[-(\-0\-4\-8\-d\-l\-p\-t\-�\-�\-�\-�\-�\-�\-�\-�\-]- ]-$]-(]-T]-\]-`]-d]-�]-�]-�]-�]-�]-�]-�]-�]-^-^-^-^-D^-L^-P^-T^-�^-�^-�^-�^-�^-�^-�^-�^-�^-_-_-_-4_-<_-@_-D_-p_-x_-|_-�_-�_-�_-�_-�_-�_-�_-�_-�_-$`-,`-0`-4`-``-h`-l`-p`-�`-�`-�`-�`-�`-�`-�`-�`-a-a- a-$a-Pa-Xa-\a-`a-�a-�a-�a-�a-�a-�a-�a-�a-b-b-b-b-@b-Hb-Lb-Pb-|b-�b-�b-�b-�b-�b-�b-�b-�b-�b-c-c-0c-8c-<c-@c-lc-tc-xc-|c-�c-�c-�c-�c-�c-�c-�c-�c- d-(d-,d-0d-\d-dd-hd-ld-�d-�d-�d-�d-�d-�d-�d-�d-e-e-e- e-Le-Te-Xe-\e-�e-�e-�e-�e-�e-�e-�e-�e-f-f-f-f-<f-Df-Hf-Lf-xf-�f-�f-�f-�f-�f-�f-�f-�f-�f-�f-g-,g-4g-8g-<g-hg-pg-tg-xg-�g-�g-�g-�g-�g-�g-�g-�g-�g-h-$h-(h-,h-Xh-`h-dh-hh-�h-�h-�h-�h-�h-�h-�h-�h-i-i-i-i- i-Hi-Pi-Ti-Xi-�i-�i-�i-�i-�i-�i-�i-�i-�i-j-j-j-8j-@j-Dj-Hj-tj-|j-�j-�j-�j-�j-�j-�j-�j-�j-�j-�j-(k-0k-4k-8k-dk-lk-pk-tk-�k-�k-�k-�k-�k-�k-�k-�k-�k-l- l-$l-(l-Tl-\l-`l-dl-�l-�l-�l-�l-�l-�l-�l-�l-m-m-m-m-Dm-Lm-Pm-Tm-Xm-�m-�m-�m-�m-�m-�m-�m-�m-�m-n-n-n-4n-<n-@n-Dn-pn-xn-|n-�n-�n-�n-�n-�n-�n-�n-�n-�n-�n-$o-,o-0o-4o-`o-ho-lo-po-�o-�o-�o-�o-�o-�o-�o-�o-p-p- p-$p-Pp-Xp-\p-`p-�p-�p-�p-�p-�p-�p-�p-�p-q-q-q-q-@q-Hq-Lq-Pq-|q-�q-�q-�q-�q-�q-�q-�q-�q-�q-r-r-0r-8r-<r-@r-lr-tr-xr-|r-�r-�r-�r-�r-�r-�r-�r-�r- s-(s-,s-0s-\s-ds-hs-ls-ps-�s-�s-�s-�s-�s-�s-�s-�s-�s-t-t-t- t-Lt-Tt-Xt-\t-�t-�t-�t-�t-�t-�t-�t-�t-u-u-u-u-<u-Du-Hu-Lu-xu-�u-�u-�u-�u-�u-�u-�u-�u-�u-�u-v-,v-4v-8v-<v-hv-pv-tv-xv-�v-�v-�v-�v-�v-�v-�v-�v-w-$w-(w-,w-Xw-`w-dw-hw-�w-�w-�w-�w-�w-�w-�w-�w-x-x-x-x-Hx-Px-Tx-Xx-�x-�x-�x-�x-�x-�x-�x-�x-�x-y-y-y-8y-@y-Dy-Hy-Ly-ty-|y-�y-�y-�y-�y-�y-�y-�y-�y-�y-�y-(z-0z-4z-8z-dz-lz-pz-tz-�z-�z-�z-�z-�z-�z-�z-�z-{- {-${-({-T{-\{-`{-d{-�{-�{-�{-�{-�{-�{-�{-�{-|-|-|-|-D|-L|-P|-T|-�|-�|-�|-�|-�|-�|-�|-�|-�|-}-}-}-4}-<}-@}-D}-p}-x}-|}-�}-�}-�}-�}-�}-�}-�}-�}-�}-$~-,~-0~-4~-`~-h~-l~-p~-�~-�~-�~-�~-�~-�~-�~-�~--- -$-P-X-\-`-�-�-�-�-�-�-�-�-�-�-�-�-@�-H�-L�-P�-|�-��-��-��-��-��-Ā-Ȁ-�-��-�-�-0�-8�-<�-@�-l�-t�-x�-|�-��-��-��-��-��-�-�-�-�- �-(�-,�-0�-\�-d�-h�-l�-��-��-��-��-Ԃ-܂-�-�-�-�-�- �-L�-T�-X�-\�-��-��-��-��-��-ă-̃-Ѓ-ԃ-؃-�-�-�-�-<�-D�-H�-L�-P�-x�-��-��-��-��-��-��-Ą-�-��-��-�-,�-4�-8�-<�-h�-p�-t�-x�-��-��-��-��-�-�-�-�-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-І-؆-܆-�-�-�-�-�-�-H�-P�-T�-X�-��-��-��-��-��-ȇ-̇-Ї-��-�-�-�-8�-@�-D�-H�-t�-|�-��-��-��-��-��-��-�-�-��-��-(�-0�-4�-8�-<�-d�-l�-p�-t�-x�-��-��-��-��-܉-�-�-�-�- �-$�-(�-T�-\�-`�-d�-��-��-��-��-̊-Ԋ-؊-܊-�-�-�-�-D�-L�-P�-T�-X�-��-��-��-��-��-��-ċ-ȋ-̋-��-�-�-�-4�-<�-@�-D�-H�-p�-x�-|�-��-��-��-��-��-��-�-�-�-��-$�-,�-0�-4�-`�-h�-l�-p�-��-��-��-��-؍-�-�-�-�-�- �-$�-P�-X�-\�-`�-��-��-��-��-Ȏ-Ў-Ԏ-؎-�-�-�-�-@�-H�-L�-P�-|�-��-��-��-��-��-ď-ȏ-�-��-�-�-0�-8�-<�-@�-l�-t�-x�-|�-��-��-��-��-�-�-�-�- �-(�-,�-0�-\�-d�-h�-l�-��-��-��-��-ԑ-ܑ-�-�-�-�-�- �-L�-T�-X�-\�-��-��-��-��-Ē-̒-В-Ԓ-�-�-�-�-<�-D�-H�-L�-x�-��-��-��-��-��-��-ē-�-��-��-�-,�-4�-8�-<�-h�-p�-t�-x�-��-��-��-��-�-�-�-�-�-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-��-Е-ؕ-ܕ-�-�-�-�-�- �-H�-P�-T�-X�-��-��-��-��-̖-Ж-Ԗ-�-�-�-�-�-�-$�-(�-,�-D�-H�-L�-`�-d�-h�-��-��-��-��-��-��-��-��-ė-ؗ-ܗ-�-�-��-��-�-�-�-�-4�-8�-<�-P�-T�-X�-p�-t�-x�-��-��-��-��-��-Ș-̘-И-�-�-�-�-�-$�-(�-@�-D�-H�-`�-d�-|�-��-��-��-��-��-��-��-ؙ-ܙ-�-��-��-�-�-�-0�-4�-8�-P�-T�-l�-p�-t�-��-��-��-��-��-Ț-̚-�-�-�-�-�- �-$�-(�-@�-D�-\�-`�-d�-|�-��-��-��-��-��-��-ԛ-؛-ܛ-�-��-�-�-�-0�-4�-L�-P�-T�-l�-p�-��-��-��-��-��-Ĝ-Ȝ-̜-�-�-�-�-�- �-$�-<�-@�-D�-\�-`�-x�-|�-��-��-��-��-��-��-ԝ-؝-�-�-��-�-�-,�-0�-4�-L�-P�-h�-l�-p�-��-��-��-��-��-Ğ-Ȟ-�-�-�-�-�-�- �-$�-<�-@�-X�-\�-`�-x�-|�-��-��-��-��-��-П-ԟ-؟-�-�-�-�-�-,�-0�-H�-L�-P�-\�-d�-h�-l�-p�-t�-|�-��-��-��-��-��-��-��-Ԡ-ܠ-�-�-�-�-�-�-�-�- �-$�-(�-0�-L�-T�-X�-\�-`�-d�-l�-��-��-��-��-��-��-��-ġ-̡-С-ԡ-ء-ܡ-�-�-�-�-�-�-�- �-<�-D�-H�-L�-P�-T�-\�-x�-��-��-��-��-��-��-��-��-��-Ģ-Ȣ-̢-Ԣ-�-��-��-�-�-�-�-,�-4�-8�-<�-@�-D�-L�-h�-p�-t�-x�-|�-��-��-��-��-��-��-��-��-ģ-�-�-�-�-�-��-�-�-$�-(�-,�-0�-4�-<�-X�-`�-d�-h�-l�-p�-x�-��-��-��-��-��-��-��-Ф-ؤ-ܤ-�-�-�-�-�-�-�-�- �-$�-,�-H�-P�-T�-X�-\�-`�-h�-��-��-��-��-��-��-��-��-ȥ-̥-Х-ԥ-إ-�-��-�-�-�-�-�-�-8�-@�-D�-H�-L�-P�-X�-t�-|�-��-��-��-��-��-��-��-��-��-Ħ-Ȧ-Ц-�-�-��-��-�-�-�-(�-0�-4�-8�-<�-@�-H�-d�-l�-p�-t�-x�-|�-��-��-��-��-��-��-��-��-ܧ-�-�-�-�-�-��-�- �-$�-(�-,�-0�-8�-T�-\�-`�-d�-h�-l�-��-��-��-��-��-��-̨-Ԩ-ب-ܨ-�-�-�-�-�-�-�- �-D�-L�-P�-T�-X�-\�-��-��-��-��-��-��-��-ĩ-ȩ-̩-Щ-ԩ-��-�-�-�-�-�-4�-<�-@�-D�-H�-L�-p�-x�-|�-��-��-��-��-��-��-��-��-Ī-�-�-�-��-��-�-$�-,�-0�-4�-8�-<�-`�-h�-l�-p�-t�-x�-��-��-��-��-��-��-ث-�-�-�-�-�-�-�- �-$�-(�-,�-P�-X�-\�-`�-d�-h�-��-��-��-��-��-��-Ȭ-Ь-Ԭ-ج-ܬ-�-�-�-�-�-�-�-@�-H�-L�-P�-T�-X�-|�-��-��-��-��-��-��-��-ĭ-ȭ-̭-Э-�-��-�-�-�-�-0�-8�-<�-@�-D�-H�-l�-t�-x�-|�-��-��-��-��-��-�-�-�-�- �-(�-,�-0�-4�-\�-d�-h�-l�-��-��-��-��-ԯ-ܯ-�-�-�-�-�- �-L�-T�-X�-\�-��-��-��-��-İ-̰-а-԰-�-�-�-�-<�-D�-H�-L�-x�-��-��-��-��-��-��-��-ı-�-��-��-�-h�-p�-t�-x�-��-��-��-��-�-�-�-�-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-г-س-ܳ-�-�-�-�-�-H�-P�-T�-X�-��-��-��-��-��-ȴ-̴-д-��-�-�-�-8�-@�-D�-H�-t�-|�-��-��-��-��-��-��-�-�-��-��-(�-0�-4�-8�-d�-l�-p�-t�-��-��-��-��-ܶ-�-�-�-�- �-$�-(�-T�-\�-`�-d�-��-��-��-��-̷-Է-ط-ܷ-�-�-�-�-D�-L�-P�-T�-��-��-��-��-��-ĸ-ȸ-̸-��-�-�-�-4�-<�-@�-D�-p�-x�-|�-��-��-��-��-��-�-�-�-��-$�-,�-0�-4�-`�-h�-l�-p�-��-��-��-��-غ-�-�-�-�-�- �-$�-P�-X�-\�-`�-��-��-��-��-Ȼ-л-Ի-ػ-�-�-�-�-@�-H�-L�-P�-|�-��-��-��-��-��-ļ-ȼ-�-��-�-�-0�-8�-<�-@�-l�-t�-x�-|�-��-��-��-��-�-�-�-�- �-(�-,�-0�-\�-d�-h�-l�-��-��-��-��-Ծ-ܾ-�-�-�-�-�- �-L�-T�-X�-\�-��-��-��-��-Ŀ-̿-п-Կ-�-�-�-�-<�-D�-H�-L�-x�-��-��-��-��-��-��-�-�-��-��-�-,�-4�-8�-<�-h�-p�-t�-x�-��-��-��-��-�-�-�-�-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-��-��-��-��-�-�-�-�-H�-P�-T�-X�-��-��-��-��-��-��-��-��-��-�-�-�-8�-@�-D�-H�-t�-|�-��-��-��-��-��-��-��-��-��-��-(�-0�-4�-8�-d�-l�-p�-t�-��-��-��-��-��-��-��-��-�- �-$�-(�-T�-\�-`�-d�-��-��-��-��-��-��-��-��-�-�-�-�-D�-L�-P�-T�-��-��-��-��-��-��-��-��-��-�-�-�-4�-<�-@�-D�-p�-x�-|�-��-��-��-��-��-��-��-��-��-$�-,�-0�-4�-`�-h�-l�-p�-��-��-��-��-��-��-��-��-�-�- �-$�-P�-X�-\�-`�-��-��-��-��-��-��-��-��-�-�-�-�-@�-H�-L�-P�-|�-��-��-��-��-��-��-��-��-��-�-�-0�-8�-<�-@�-l�-t�-x�-|�-��-��-��-��-��-��-��-��- �-(�-,�-0�-\�-d�-h�-l�-��-��-��-��-��-��-��-��-�-�-�- �-L�-T�-X�-\�-��-��-��-��-��-��-��-��-�-�-�-�-<�-D�-H�-L�-x�-��-��-��-��-��-��-��-��-��-��-�-,�-4�-8�-<�-h�-p�-t�-x�-��-��-��-��-��-��-��-��-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-��-��-��-��-�-�-�-�-H�-P�-T�-X�-��-��-��-��-��-��-��-��-��-�-�-�-8�-@�-D�-H�-t�-|�-��-��-��-��-��-��-��-��-��-��-(�-0�-4�-8�-d�-l�-p�-t�-��-��-��-��-��-��-��-��-�- �-$�-(�-T�-\�-`�-d�-��-��-��-��-��-��-��-��-�-�-�-�-D�-L�-P�-T�-��-��-��-��-��-��-��-��-��-�-�-�-4�-<�-@�-D�-p�-x�-|�-��-��-��-��-��-��-��-��-��-$�-,�-0�-4�-`�-h�-l�-p�-��-��-��-��-��-��-��-��-�-�- �-$�-P�-X�-\�-`�-��-��-��-��-��-��-��-��-�-�-�-�-@�-H�-L�-P�-|�-��-��-��-��-��-��-��-��-��-�-�-0�-8�-<�-@�-l�-t�-x�-|�-��-��-��-��-��-��-��-��- �-(�-,�-0�-\�-d�-h�-l�-��-��-��-��-��-��-��-��-�-�-�- �-L�-T�-X�-\�-��-��-��-��-��-��-��-��-�-�-�-�-<�-D�-H�-L�-x�-��-��-��-��-��-��-��-��-��-��-�-,�-4�-8�-<�-h�-p�-t�-x�-��-��-��-��-��-��-��-��-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-��-��-��-��-�-�-�-�-H�-P�-T�-X�-��-��-��-��-��-��-��-��-��-�-�-�-8�-@�-D�-H�-t�-|�-��-��-��-��-��-��-��-��-��-��-(�-0�-4�-8�-d�-l�-p�-t�-��-��-��-��-��-��-��-��-�- �-$�-(�-T�-\�-`�-d�-��-��-��-��-��-��-��-��-�-�-�-�-D�-L�-P�-T�-��-��-��-��-��-��-��-��-��-�-�-�-4�-<�-@�-D�-p�-x�-|�-��-��-��-��-��-��-��-��-��-$�-,�-0�-4�-`�-h�-l�-p�-��-��-��-��-��-��-��-��-�-�- �-$�-P�-X�-\�-`�-��-��-��-��-��-��-��-��-�-�-�-�-@�-H�-L�-P�-|�-��-��-��-��-��-��-��-��-��-�-�-0�-8�-<�-@�-l�-t�-x�-|�-��-��-��-��-��-��-��-��- �-(�-,�-0�-\�-d�-h�-l�-��-��-��-��-��-��-��-��-�-�-�- �-L�-T�-X�-\�-��-��-��-��-��-��-��-��-�-�-�-�-<�-D�-H�-L�-x�-��-��-��-��-��-��-��-��-��-��-�-,�-4�-8�-<�-h�-p�-t�-x�-��-��-��-��-��-��-��-��-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-��-��-��-��-�-�-�-�-H�-P�-T�-X�-��-��-��-��-��-��-��-��-��-�-�-�-8�-@�-D�-H�-t�-|�-��-��-��-��-��-��-��-��-��-��-(�-0�-4�-8�-d�-l�-p�-t�-��-��-��-��-��-��-��-��-�- �-$�-(�-T�-\�-`�-d�-��-��-��-��-��-��-��-��-�-�-�-�-D�-L�-P�-T�-��-��-��-��-��-��-��-��-��-�-�-�-4�-<�-@�-D�-p�-x�-|�-��-��-��-��-��-�-�-�-��-$�-,�-0�-4�-`�-h�-l�-p�-��-��-��-��-�-�-�-�-�-�- �-$�-P�-X�-\�-`�-��-��-��-��-�-�-�-�-�-�-�-�-@�-H�-L�-P�-|�-��-��-��-��-��-�-�-�-��-�-�-0�-8�-<�-@�-l�-t�-x�-|�-��-��-��-��-�-�-�-�- �-(�-,�-0�-\�-d�-h�-l�-��-��-��-��-�-�-�-�-�-�-�- �-L�-T�-X�-\�-��-��-��-��-�-�-�-�-�-�-�-�-<�-D�-H�-L�-x�-��-��-��-��-��-��-�-�-��-��-�-,�-4�-8�-<�-h�-p�-t�-x�-��-��-��-��-�-�-�-�-�-$�-(�-,�-X�-`�-d�-h�-��-��-��-��-�-�-�-�-�-�-�-�-H�-P�-T�-X�-��-��-��-��-��-�-�-�-��-...8.@.D.H.t.|.�.�.�.�.�.�.�.�.�.�.(.0.4.8.d.l.p.t.�.�.�.�.�.�.�.�.. .$.(.T.\.`.d.�.�.�.�.�.�.�.�.....D.L.P.T.�.�.�.�.�.�.�.�.�....4.<.@.D.p.x.|.�.�.�.�.�.�.�.�.�.$.,.0.4.`.h.l.p.�.�.�.�.�.�.�.�... .$.P.X.\.`.�.�.�.�.�.�.�.�.....@.H.L.P.|.�.�.�.�.�.�.�.�.�...0.8.<.@.l.t.x.|.�.�.�.�.�.�.�.�. 	.(	.,	.0	.\	.d	.h	.l	.�	.�	.�	.�	.�	.�	.�	.�	.
.
.
. 
.L
.T
.X
.\
.�
.�
.�
.�
.�
.�
.�
.�
.....<.D.H.L.x.�.�.�.�.�.�.�.�.�.�..,.4.8.<.h.p.t.x.�.�.�.�.�.�.�.�.
.$
.(
.,
.X
.`
.d
.h
.�
.�
.�
.�
.�
.�
.�
.�
.....H.P.T.X.�.�.�.�.�.�.�.�.�....8.@.D.H.t.|.�.�.�.�.�.�.�.�.�.�.(.0.4.8.d.l.p.t.�.�.�.�.�.�.�.�.. .$.(.T.\.`.d.�.�.�.�.�.�.�.�.....D.L.P.T.�.�.�.�.�.�.�.�.�....4.<.@.D.p.x.|.�.�.�.�.�.�.�.�.�.$.,.0.4.`.h.l.p.�.�.�.�.�.�.�.�... .$.P.X.\.`.�.�.�.�.�.�.�.�.....@.H.L.P.|.�.�.�.�.�.�.�.�.�...0.8.<.@.l.t.x.|.�.�.�.�.�.�.�.�. .(.,.0.\.d.h.l.�.�.�.�.�.�.�.�.... .L.T.X.\.�.�.�.�.�.�.�.�.....<.D.H.L.x.�.�.�.�.�.�.�.�.�.�..,.4.8.<.h.p.t.x.�.�.�.�.�.�.�.�..$.(.,.X.`.d.h.�.�.�.�.�.�.�.�.....H.P.T.X.�.�.�.�.�.�.�.�.�....8.@.D.H.t.|.�.�.�.�.�.�.�.�.�.�.(.0.4.8.d.l.p.t.�.�.�.�.�.�.�.�. .  .$ .( .T .\ .` .d .� .� .� .� .� .� .� .� .!.!.!.!.D!.L!.P!.T!.�!.�!.�!.�!.�!.�!.�!.�!.�!.".".".4".<".@".D".p".x".|".�".�".�".�".�".�".�".�".�".$#.,#.0#.4#.`#.h#.l#.p#.�#.�#.�#.�#.�#.�#.�#.�#.$.$. $.$$.P$.X$.\$.`$.�$.�$.�$.�$.�$.�$.�$.�$.%.%.%.%.@%.H%.L%.P%.|%.�%.�%.�%.�%.�%.�%.�%.�%.�%.&.&.0&.8&.<&.@&.l&.t&.x&.|&.�&.�&.�&.�&.�&.�&.�&.�&. '.('.,'.0'.\'.d'.h'.l'.�'.�'.�'.�'.�'.�'.�'.�'.(.(.(. (.L(.T(.X(.\(.�(.�(.�(.�(.�(.�(.�(.�(.).).).).<).D).H).L).x).�).�).�).�).�).�).�).�).�).�).*.,*.4*.8*.<*.h*.p*.t*.x*.�*.�*.�*.�*.�*.�*.�*.�*.+.$+.(+.,+.X+.`+.d+.h+.�+.�+.�+.�+.�+.�+.�+.�+.,.,.,.,.H,.P,.T,.X,.�,.�,.�,.�,.�,.�,.�,.�,.�,.-.-.-.8-.@-.D-.H-.t-.|-.�-.�-.�-.�-.�-.�-.�-.�-.�-.�-.(..0..4..8..d..l..p..t..�..�..�..�..�..�..�..�../. /.$/.(/.T/.\/.`/.d/.�/.�/.�/.�/.�/.�/.�/.�/.0.0.0.0.D0.L0.P0.T0.�0.�0.�0.�0.�0.�0.�0.�0.�0.1.1.1.41.<1.@1.D1.�1.�1.�1.�1.�1.�1.�1.�1.$2.,2.02.42.`2.h2.l2.p2.�2.�2.�2.�2.�2.�2.�2.�2.3.3. 3.$3.P3.X3.\3.`3.�3.�3.�3.�3.�3.�3.�3.�3.4.4.4.4.@4.H4.L4.P4.|4.�4.�4.�4.�4.�4.�4.�4.�4.�4.5.5.05.85.<5.@5.l5.t5.x5.|5.�5.�5.�5.�5.�5.�5.�5.�5. 6.(6.,6.06.\6.d6.h6.l6.�6.�6.�6.�6.�6.�6.�6.�6.7.7.7. 7.L7.T7.X7.\7.�7.�7.�7.�7.�7.�7.�7.�7.8.8.8.8.<8.D8.H8.L8.x8.�8.�8.�8.�8.�8.�8.�8.�8.�8.�8.9.,9.49.89.<9.h9.p9.t9.x9.�9.�9.�9.�9.�9.�9.�9.�9.:.$:.(:.,:.X:.`:.d:.h:.�:.�:.�:.�:.�:.�:.�:.�:.;.;.;.;.H;.P;.T;.X;.�;.�;.�;.�;.�;.�;.�;.�;.�;.<.<.<.8<.@<.D<.H<.t<.|<.�<.�<.�<.�<.�<.�<.�<.�<.�<.�<.(=.0=.4=.8=.d=.l=.p=.t=.�=.�=.�=.�=.�=.�=.�=.�=.>. >.$>.(>.T>.\>.`>.d>.�>.�>.�>.�>.�>.�>.�>.�>.?.?.?.?.D?.L?.P?.T?.�?.�?.�?.�?.�?.�?.�?.�?.�?.@.@.@.4@.<@.@@.D@.p@.x@.|@.�@.�@.�@.�@.�@.�@.�@.�@.�@.$A.,A.0A.4A.`A.hA.lA.pA.�A.�A.�A.�A.�A.�A.�A.�A.B.B. B.$B.PB.XB.\B.`B.�B.�B.�B.�B.�B.�B.�B.�B.C.C.C.C.@C.HC.LC.PC.|C.�C.�C.�C.�C.�C.�C.�C.�C.�C.D.D.0D.8D.<D.@D.lD.tD.xD.|D.�D.�D.�D.�D.�D.�D.�D.�D. E.(E.,E.0E.\E.dE.hE.lE.�E.�E.�E.�E.�E.�E.�E.�E.F.F.F. F.LF.TF.XF.\F.�F.�F.�F.�F.�F.�F.�F.�F.G.G.G.G.<G.DG.HG.LG.xG.�G.�G.�G.�G.�G.�G.�G.�G.�G.�G.H.,H.4H.8H.<H.hH.pH.tH.xH.�H.�H.�H.�H.�H.�H.�H.�H.I.$I.(I.,I.XI.`I.dI.hI.�I.�I.�I.�I.�I.�I.�I.�I.J.J.J.J.HJ.PJ.TJ.XJ.�J.�J.�J.�J.�J.�J.�J.�J.�J.K.K.K.8K.@K.DK.HK.tK.|K.�K.�K.�K.�K.�K.�K.�K.�K.�K.�K.(L.0L.4L.8L.dL.lL.pL.tL.�L.�L.�L.�L.�L.�L.�L.�L.M. M.$M.(M.TM.\M.`M.dM.�M.�M.�M.�M.�M.�M.�M.�M.N.N.N.N.DN.LN.PN.TN.�N.�N.�N.�N.�N.�N.�N.�N.�N.O.O.O.4O.<O.@O.DO.pO.xO.|O.�O.�O.�O.�O.�O.�O.�O.�O.�O.$P.,P.0P.4P.`P.hP.lP.pP.�P.�P.�P.�P.�P.�P.�P.�P.Q.Q. Q.$Q.PQ.XQ.\Q.`Q.�Q.�Q.�Q.�Q.�Q.�Q.�Q.�Q.R.R.R.R.@R.HR.LR.PR.|R.�R.�R.�R.�R.�R.�R.�R.�R.�R.S.S.0S.8S.<S.@S.lS.tS.xS.|S.�S.�S.�S.�S.�S.�S.�S.�S. T.(T.,T.0T.\T.dT.hT.lT.�T.�T.�T.�T.�T.�T.�T.�T.U.U.U. U.LU.TU.XU.\U.�U.�U.�U.�U.�U.�U.�U.�U.V.V.V.V.<V.DV.HV.LV.xV.�V.�V.�V.�V.�V.�V.�V.�V.�V.�V.W.,W.4W.8W.<W.hW.pW.tW.xW.�W.�W.�W.�W.�W.�W.�W.�W.X.$X.(X.,X.XX.`X.dX.hX.�X.�X.�X.�X.�X.�X.�X.�X.Y.Y.Y.Y.HY.PY.TY.XY.�Y.�Y.�Y.�Y.�Y.�Y.�Y.�Y.�Y.Z.Z.Z.8Z.@Z.DZ.HZ.tZ.|Z.�Z.�Z.�Z.�Z.�Z.�Z.�Z.�Z.�Z.�Z.([.0[.4[.8[.d[.l[.p[.t[.�[.�[.�[.�[.�[.�[.�[.�[.\. \.$\.(\.T\.\\.`\.d\.�\.�\.�\.�\.�\.�\.�\.�\.].].].].D].L].P].T].�].�].�].�].�].�].�].�].�].^.^.^.4^.<^.@^.D^.p^.x^.|^.�^.�^.�^.�^.�^.�^.�^.�^.�^.$_.,_.0_.4_.`_.h_.l_.p_.�_.�_.�_.�_.�_.�_.�_.�_.`.`. `.$`.P`.X`.\`.``.�`.�`.�`.�`.�`.�`.�`.�`.a.a.a.a.@a.Ha.La.Pa.|a.�a.�a.�a.�a.�a.�a.�a.�a.�a.b.b.0b.8b.<b.@b.lb.tb.xb.|b.�b.�b.�b.�b.�b.�b.�b.�b. c.(c.,c.0c.\c.dc.hc.lc.�c.�c.�c.�c.�c.�c.�c.�c.d.d.d. d.Ld.Td.Xd.\d.�d.�d.�d.�d.�d.�d.�d.�d.e.e.e.e.<e.De.He.Le.xe.�e.�e.�e.�e.�e.�e.�e.�e.�e.�e.f.,f.4f.8f.<f.hf.pf.tf.xf.�f.�f.�f.�f.�f.�f.�f.�f.g.$g.(g.,g.Xg.`g.dg.hg.�g.�g.�g.�g.�g.�g.�g.�g.h.h.h.h.Hh.Ph.Th.Xh.�h.�h.�h.�h.�h.�h.�h.�h.�h.i.i.i.8i.@i.Di.Hi.ti.|i.�i.�i.�i.�i.�i.�i.�i.�i.�i.�i.(j.0j.4j.8j.dj.lj.pj.tj.�j.�j.�j.�j.�j.�j.�j.�j.k. k.$k.(k.Tk.\k.`k.dk.�k.�k.�k.�k.�k.�k.�k.�k.l.l.l.l.Dl.Ll.Pl.Tl.�l.�l.�l.�l.�l.�l.�l.�l.�l.m.m.m.4m.<m.@m.Dm.pm.xm.|m.�m.�m.�m.�m.�m.�m.�m.�m.�m.$n.,n.0n.4n.`n.hn.ln.pn.�n.�n.�n.�n.�n.�n.�n.�n.o.o. o.$o.Po.Xo.\o.`o.�o.�o.�o.�o.�o.�o.�o.�o.p.p.p.p.@p.Hp.Lp.Pp.|p.�p.�p.�p.�p.�p.�p.�p.�p.�p.q.q.0q.8q.<q.@q.lq.tq.xq.|q.�q.�q.�q.�q.�q.�q.�q.�q. r.(r.,r.0r.\r.dr.hr.lr.�r.�r.�r.�r.�r.�r.�r.�r.s.s.s. s.Ls.Ts.Xs.\s.�s.�s.�s.�s.�s.�s.�s.�s.t.t.t.t.<t.Dt.Ht.Lt.xt.�t.�t.�t.�t.�t.�t.�t.�t.�t.�t.u.,u.4u.8u.<u.hu.pu.tu.xu.�u.�u.�u.�u.�u.�u.�u.�u.v.$v.(v.,v.Xv.`v.dv.hv.�v.�v.�v.�v.�v.�v.�v.�v.w.w.w.w.Hw.Pw.Tw.Xw.�w.�w.�w.�w.�w.�w.�w.�w.�w.x.x.x.8x.@x.Dx.Hx.tx.|x.�x.�x.�x.�x.�x.�x.�x.�x.�x.�x.(y.0y.4y.8y.dy.ly.py.ty.�y.�y.�y.�y.�y.�y.�y.�y.z. z.$z.(z.Tz.\z.`z.dz.�z.�z.�z.�z.�z.�z.�z.�z.{.{.{.{.D{.L{.P{.T{.�{.�{.�{.�{.�{.�{.�{.�{.�{.|.|.|.4|.<|.@|.D|.p|.x|.||.�|.�|.�|.�|.�|.�|.�|.�|.�|.$}.,}.0}.4}.`}.h}.l}.p}.�}.�}.�}.�}.�}.�}.�}.�}.~.~. ~.$~.P~.X~.\~.`~.�~.�~.�~.�~.�~.�~.�~.�~.....@.H.L.P.|.�.�.�.�.�.�.�.�.�.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.�.�.�.�. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.ԁ.܁.�.�.�.�.�. �.L�.T�.X�.\�.��.��.��.��.Ă.̂.Ђ.Ԃ.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.ă.�.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.�.�.�.�.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.Ѕ.؅.܅.�.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.Ȇ.̆.І.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.�.�.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.܈.�.�.�.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.̉.ԉ.؉.܉.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.Ċ.Ȋ.̊.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.��.��.��.��.�.�.�.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.،.�.�.�.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.ȍ.Ѝ.ԍ.؍.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.Ď.Ȏ.�.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.�.�.�.�. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.Ԑ.ܐ.�.�.�.�.�. �.L�.T�.X�.\�.��.��.��.��.đ.̑.Б.ԑ.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.Ē.�.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.�.�.�.�.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.Д.ؔ.ܔ.�.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.ȕ.̕.Е.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.�.�.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.ܗ.�.�.�.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.̘.Ԙ.ؘ.ܘ.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.ę.ș.̙.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.��.��.��.��.�.�.�.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.؛.�.�.�.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.Ȝ.М.Ԝ.؜.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.ĝ.ȝ.�.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.�.�.�.�. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.ԟ.ܟ.�.�.�.�.�. �.L�.T�.X�.\�.��.��.��.��.Ġ.̠.Р.Ԡ.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.ġ.�.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.�.�.�.�.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.У.أ.ܣ.�.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.Ȥ.̤.Ф.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.�.�.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.ܦ.�.�.�.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.̧.ԧ.ا.ܧ.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.Ĩ.Ȩ.̨.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.��.��.��.��.�.�.�.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.ت.�.�.�.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.ȫ.Ы.ԫ.ث.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.Ĭ.Ȭ.�.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.�.�.�.�. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.Ԯ.ܮ.�.�.�.�.�. �.L�.T�.X�.\�.��.��.��.��.į.̯.Я.ԯ.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.İ.�.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.�.�.�.�.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.в.ز.ܲ.�.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.ȳ.̳.г.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.�.�.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.ܵ.�.�.�.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.̶.Զ.ض.ܶ.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.ķ.ȷ.̷.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.�.�.�.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.ع.�.�.�.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.Ⱥ.к.Ժ.غ.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.Ļ.Ȼ.�.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.�.�.�.�. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.Խ.ܽ.�.�.�.�.�. �.L�.T�.X�.\�.��.��.��.��.ľ.̾.о.Ծ.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.Ŀ.�.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.�.�.�.�.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.�.�.�.�.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.��.��.��.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.��.��.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.��.��.��.��.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.��.��.��.��.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.��.��.��.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.��.��.��.��.��.��.��.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.��.��.��.��.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.��.��.��.��.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.��.��.��.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.��.��.��.��. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.��.��.��.��.�.�.�. �.L�.T�.X�.\�.��.��.��.��.��.��.��.��.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.��.��.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.��.��.��.��.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.��.��.��.��.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.��.��.��.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.��.��.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.��.��.��.��.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.��.��.��.��.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.��.��.��.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.��.��.��.��.��.��.��.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.��.��.��.��.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.��.��.��.��.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.��.��.��.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.��.��.��.��. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.��.��.��.��.�.�.�. �.L�.T�.X�.\�.��.��.��.��.��.��.��.��.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.��.��.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.��.��.��.��.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.��.��.��.��.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.��.��.��.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.��.��.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.��.��.��.��.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.��.��.��.��.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.��.��.��.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.��.��.��.��.��.��.��.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.��.��.��.��.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.��.��.��.��.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.��.��.��.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.��.��.��.��. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.��.��.��.��.�.�.�. �.L�.T�.X�.\�.��.��.��.��.��.��.��.��.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.��.��.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.��.��.��.��.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.��.��.��.��.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.��.��.��.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.��.��.��.��.(�.0�.4�.8�.d�.l�.p�.t�.��.��.��.��.��.��.��.��.�. �.$�.(�.T�.\�.`�.d�.��.��.��.��.��.��.��.��.�.�.�.�.D�.L�.P�.T�.��.��.��.��.��.��.��.��.��.�.�.�.4�.<�.@�.D�.p�.x�.|�.��.��.��.��.��.��.��.��.��.$�.,�.0�.4�.`�.h�.l�.p�.��.��.��.��.�.�.�.�.�.�. �.$�.P�.X�.\�.`�.��.��.��.��.�.�.�.�.�.�.�.�.@�.H�.L�.P�.|�.��.��.��.��.��.�.�.�.��.�.�.0�.8�.<�.@�.l�.t�.x�.|�.��.��.��.��.�.�.�.�. �.(�.,�.0�.\�.d�.h�.l�.��.��.��.��.�.�.�.�.�.�.�. �.L�.T�.X�.\�.��.��.��.��.�.�.�.�.�.�.�.�.<�.D�.H�.L�.x�.��.��.��.��.��.��.�.�.��.��.�.,�.4�.8�.<�.h�.p�.t�.x�.��.��.��.��.�.�.�.�.�.$�.(�.,�.X�.`�.d�.h�.��.��.��.��.�.�.�.�.�.�.�.�.H�.P�.T�.X�.��.��.��.��.��.�.�.�.��.�.�.�.8�.@�.D�.H�.t�.|�.��.��.��.��.��.��.�.�.��.��.(/0/4/8/d/l/p/t/�/�/�/�/�/�/�/�// /$/(/T/\/`/d/�/�/�/�/�/�/�/�/////D/L/P/T/�/�/�/�/�/�/�/�/�////4/</@/D/p/x/|/�/�/�/�/�/�/�/�/�/$/,/0/4/`/h/l/p/�/�/�/�/�/�/�/�/// /$/P/X/\/`/�/�/�/�/�/�/�/�/////@/H/L/P/|/�/�/�/�/�/�/�/�/�///0/8/</@/l/t/x/|/�/�/�/�/�/�/�/�/ /(/,/0/\/d/h/l/�/�/�/�/�/�/�/�/	/	/	/ 	/L	/T	/X	/\	/�	/�	/�	/�	/�	/�	/�	/�	/
/
/
/
/<
/D
/H
/L
/x
/�
/�
/�
/�
/�
/�
/�
/�
/�
/�
//,/4/8/</h/p/t/x/�/�/�/�/�/�/�/�//$/(/,/X/`/d/h/�/�/�/�/�/�/�/�/
/
/
/
/H
/P
/T
/X
/�
/�
/�
/�
/�
/�
/�
/�
/�
////8/@/D/H/t/|/�/�/�/�/�/�/�/�/�/�/(/0/4/8/d/l/p/t/�/�/�/�/�/�/�/�// /$/(/T/\/`/d/�/�/�/�/�/�/�/�/////D/L/P/T/�/�/�/�/�/�/�/�/�////4/</@/D/p/x/|/�/�/�/�/�/�/�/�/�/$/,/0/4/`/h/l/p/�/�/�/�/�/�/�/�/// /$/P/X/\/`/�/�/�/�/�/�/�/�/////@/H/L/P/|/�/�/�/�/�/�/�/�/�///0/8/</@/l/t/x/|/�/�/�/�/�/�/�/�/ /(/,/0/\/d/h/l/�/�/�/�/�/�/�/�//// /L/T/X/\/�/�/�/�/�/�/�/�/////</D/H/L/x/�/�/�/�/�/�/�/�/�/�//,/4/8/</h/p/t/x/�/�/�/�/�/�/�/�//$/(/,/X/`/d/h/�/�/�/�/�/�/�/�/////H/P/T/X/�/�/�/�/�/�/�/�/�////8/@/D/H/t/|/�/�/�/�/�/�/�/�/�/�/(/0/4/8/d/l/p/t/�/�/�/�/�/�/�/�// /$/(/T/\/`/d/�/�/�/�/�/�/�/�/ / / / /D /L /P /T /� /� /� /� /� /� /� /� /� /!/!/!/4!/<!/@!/D!/p!/x!/|!/�!/�!/�!/�!/�!/�!/�!/�!/�!/$"/,"/0"/4"/`"/h"/l"/p"/�"/�"/�"/�"/�"/�"/�"/�"/#/#/ #/$#/P#/X#/\#/`#/�#/�#/�#/�#/�#/�#/�#/�#/$/$/$/$/@$/H$/L$/P$/|$/�$/�$/�$/�$/�$/�$/�$/�$/�$/%/%/0%/8%/<%/@%/l%/t%/x%/|%/�%/�%/�%/�%/�%/�%/�%/�%/ &/(&/,&/0&/\&/d&/h&/l&/�&/�&/�&/�&/�&/�&/�&/�&/'/'/'/ '/L'/T'/X'/\'/�'/�'/�'/�'/�'/�'/�'/�'/(/(/(/(/<(/D(/H(/L(/x(/�(/�(/�(/�(/�(/�(/�(/�(/�(/�(/)/,)/4)/8)/<)/h)/p)/t)/x)/�)/�)/�)/�)/�)/�)/�)/�)/*/$*/(*/,*/X*/`*/d*/h*/�*/�*/�*/�*/�*/�*/�*/�*/+/+/+/+/H+/P+/T+/X+/�+/�+/�+/�+/�+/�+/�+/�+/�+/,/,/,/8,/@,/D,/H,/t,/|,/�,/�,/�,/�,/�,/�,/�,/�,/�,/�,/(-/0-/4-/8-/d-/l-/p-/t-/�-/�-/�-/�-/�-/�-/�-/�-/./ ./$./(./T./\./`./d./�./�./�./�./�./�./�./�./////////D//L//P//T//�//�//�//�//�//�//�//�//�//0/0/0/40/<0/@0/D0/p0/x0/|0/�0/�0/�0/�0/�0/�0/�0/�0/�0/$1/,1/01/41/`1/h1/l1/p1/�1/�1/�1/�1/�1/�1/�1/�1/2/2/ 2/$2/P2/X2/\2/`2/�2/�2/�2/�2/�2/�2/�2/�2/3/3/3/3/@3/H3/L3/P3/|3/�3/�3/�3/�3/�3/�3/�3/�3/�3/4/4/04/84/<4/@4/l4/t4/x4/|4/�4/�4/�4/�4/�4/�4/�4/�4/ 5/(5/,5/05/\5/d5/h5/l5/�5/�5/�5/�5/�5/�5/�5/�5/6/6/6/ 6/L6/T6/X6/\6/�6/�6/�6/�6/�6/�6/�6/�6/7/7/7/7/<7/D7/H7/L7/x7/�7/�7/�7/�7/�7/�7/�7/�7/�7/�7/8/,8/48/88/<8/h8/p8/t8/x8/�8/�8/�8/�8/�8/�8/�8/�8/9/$9/(9/,9/X9/`9/d9/h9/�9/�9/�9/�9/�9/�9/�9/�9/:/:/:/:/H:/P:/T:/X:/�:/�:/�:/�:/�:/�:/�:/�:/�:/;/;/;/8;/@;/D;/H;/t;/|;/�;/�;/�;/�;/�;/�;/�;/�;/�;/�;/(</0</4</8</d</l</p</t</�</�</�</�</�</�</�</�</=/ =/$=/(=/T=/\=/`=/d=/�=/�=/�=/�=/�=/�=/�=/�=/>/>/>/>/D>/L>/P>/T>/�>/�>/�>/�>/�>/�>/�>/�>/�>/?/?/?/4?/<?/@?/D?/p?/x?/|?/�?/�?/�?/�?/�?/�?/�?/�?/�?/$@/,@/0@/4@/`@/h@/l@/p@/�@/�@/�@/�@/�@/A/A/ A/$A/(A/PA/XA/\A/`A/dA/�A/�A/�A/�A/�A/�A/�A/�A/�A/B/B/B/B/B/@B/HB/LB/PB/TB/|B/�B/�B/�B/�B/�B/�B/�B/�B/�B/C/C/(C/0C/8C/<C/@C/DC/lC/tC/xC/|C/�C/�C/�C/�C/�C/�C/�C/�C/�C/�C/�C/ D/(D/,D/0D/4D/\D/dD/hD/lD/pD/�D/�D/�D/�D/�D/�D/�D/�D/�D/�D/E/E/E/ E/$E/LE/TE/XE/\E/`E/�E/�E/�E/�E/�E/�E/�E/�E/�E/�E/F/F/F/F/F/<F/DF/HF/LF/PF/xF/�F/�F/�F/�F/�F/�F/�F/�F/�F/�F/�F/�F/G/G/,G/4G/8G/<G/@G/hG/pG/tG/xG/|G/�G/�G/�G/�G/�G/�G/�G/�G/H/$H/(H/,H/0H/XH/`H/dH/hH/�H/�H/�H/�H/�H/�H/�H/�H/�H/�H/I/I/I/I/ I/HI/PI/TI/XI/\I/�I/�I/�I/�I/�I/�I/�I/�I/�I/�I/�I/J/J/J/8J/@J/DJ/HJ/LJ/tJ/|J/�J/�J/�J/�J/�J/�J/�J/�J/�J/�J/�J/K/(K/0K/4K/8K/<K/dK/lK/pK/tK/�K/�K/�K/�K/�K/�K/�K/�K/�K/�K/L/ L/$L/(L/,L/TL/\L/`L/dL/hL/�L/�L/�L/�L/�L/�L/�L/�L/�L/�L/M/M/M/M/M/DM/LM/PM/TM/XM/�M/�M/�M/�M/�M/�M/�M/�M/�M/�M/�M/N/N/N/N/4N/<N/@N/DN/HN/pN/xN/|N/�N/�N/�N/�N/�N/�N/�N/�N/�N/�N/�N/�N/$O/,O/0O/4O/8O/`O/hO/lO/pO/tO/�O/�O/�O/�O/�O/�O/�O/�O/�O/�O/P/P/ P/$P/(P/PP/XP/\P/`P/�P/�P/�P/�P/�P/�P/�P/�P/�P/�P/Q/Q/Q/Q/@Q/HQ/LQ/PQ/|Q/�Q/�Q/�Q/�Q/�Q/�Q/�Q/�Q/�Q/R/R/0R/8R/<R/@R/lR/tR/xR/|R/�R/�R/�R/�R/�R/�R/�R/�R/ S/(S/,S/0S/\S/dS/hS/lS/�S/�S/�S/�S/�S/�S/�S/�S/T/T/T/ T/LT/TT/XT/\T/�T/�T/�T/�T/�T/�T/�T/�T/U/U/U/U/<U/DU/HU/LU/xU/�U/�U/�U/�U/�U/�U/�U/�U/�U/�U/V/,V/4V/8V/<V/hV/pV/tV/xV/�V/�V/�V/�V/�V/�V/�V/�V/W/$W/(W/,W/XW/`W/dW/hW/�W/�W/�W/�W/�W/�W/�W/�W/X/X/X/X/HX/PX/TX/XX/�X/�X/�X/�X/�X/�X/�X/�X/�X/Y/Y/Y/8Y/@Y/DY/HY/tY/|Y/�Y/�Y/�Y/�Y/�Y/�Y/�Y/�Y/�Y/�Y/(Z/0Z/4Z/8Z/dZ/lZ/pZ/tZ/�Z/�Z/�Z/�Z/�Z/�Z/�Z/�Z/[/ [/$[/([/T[/\[/`[/d[/�[/�[/�[/�[/�[/�[/�[/�[/\/\/\/\/D\/L\/P\/T\/�\/�\/�\/�\/�\/�\/�\/�\/�\/]/]/]/4]/<]/@]/D]/p]/x]/|]/�]/�]/�]/�]/�]/�]/�]/�]/�]/$^/,^/0^/4^/`^/h^/l^/p^/�^/�^/�^/�^/�^/�^/�^/�^/_/_/ _/$_/P_/X_/\_/`_/�_/�_/�_/�_/�_/�_/�_/�_/`/`/`/`/@`/H`/L`/P`/|`/�`/�`/�`/�`/�`/�`/�`/�`/�`/a/a/0a/8a/<a/@a/la/ta/xa/|a/�a/�a/�a/�a/�a/�a/�a/�a/�a/ b/(b/,b/0b/\b/db/hb/lb/�b/�b/�b/�b/�b/�b/�b/�b/c/c/c/ c/$c/Lc/Tc/Xc/\c/`c/�c/�c/�c/�c/�c/�c/�c/�c/�c/�c/d/d/d/d/d/<d/Dd/Hd/Ld/Pd/xd/�d/�d/�d/�d/�d/�d/�d/�d/�d/�d/�d/�d/e/e/,e/4e/8e/<e/@e/he/pe/te/xe/|e/�e/�e/�e/�e/�e/�e/�e/�e/�e/�e/f/$f/(f/,f/0f/Xf/`f/df/hf/lf/�f/�f/�f/�f/�f/�f/�f/�f/g/g/g/g/ g/Hg/Pg/Tg/Xg/�g/�g/�g/�g/�g/�g/�g/�g/�g/�g/h/h/h/8h/@h/Dh/Hh/Lh/th/|h/�h/�h/�h/�h/�h/�h/�h/�h/�h/�h/i/(i/0i/4i/8i/di/li/pi/ti/�i/�i/�i/�i/�i/�i/�i/�i/�i/j/ j/$j/(j/,j/Tj/\j/`j/dj/hj/�j/�j/�j/�j/�j/�j/�j/�j/k/k/k/k/Dk/Lk/Pk/Tk/Xk/�k/�k/�k/�k/�k/�k/�k/�k/�k/�k/�k/l/l/l/l/4l/<l/@l/Dl/Hl/pl/xl/|l/�l/�l/�l/�l/�l/�l/�l/�l/�l/�l/�l/�l/$m/,m/0m/4m/8m/`m/hm/lm/pm/tm/�m/�m/�m/�m/�m/�m/�m/�m/�m/�m/n/n/ n/$n/(n/Pn/Xn/\n/`n/dn/�n/�n/�n/�n/�n/�n/�n/�n/�n/o/o/o/o/o/@o/Ho/Lo/Po/To/|o/�o/�o/�o/�o/�o/�o/�o/�o/�o/�o/p/p/0p/8p/<p/@p/Dp/lp/tp/xp/|p/�p/�p/�p/�p/�p/�p/�p/�p/�p/ q/(q/,q/0q/\q/dq/hq/lq/�q/�q/�q/�q/�q/�q/�q/�q/�q/�q/r/r/r/ r/$r/Lr/Tr/Xr/\r/`r/�r/�r/�r/�r/�r/�r/�r/�r/�r/�r/s/s/s/s/s/<s/Ds/Hs/Ls/Ps/xs/�s/�s/�s/�s/�s/�s/�s/�s/�s/�s/�s/�s/t/t/,t/4t/8t/<t/@t/ht/pt/tt/xt/�t/�t/�t/�t/�t/�t/�t/�t/u/$u/(u/,u/Xu/`u/du/hu/�u/�u/�u/�u/�u/�u/�u/�u/v/v/v/v/Hv/Pv/Tv/Xv/�v/�v/�v/�v/�v/�v/�v/�v/�v/w/w/w/8w/@w/Dw/Hw/tw/|w/�w/�w/�w/�w/�w/�w/�w/�w/�w/�w/(x/0x/4x/8x/dx/lx/px/tx/�x/�x/�x/�x/�x/�x/�x/�x/y/ y/$y/(y/Ty/\y/`y/dy/�y/�y/�y/�y/�y/�y/�y/�y/z/z/z/z/Dz/Lz/Pz/Tz/�z/�z/�z/�z/�z/�z/�z/�z/�z/{/{/{/4{/<{/@{/D{/p{/x{/|{/�{/�{/�{/�{/�{/�{/�{/�{/�{/$|/,|/0|/4|/`|/h|/l|/p|/�|/�|/�|/�|/�|/�|/�|/�|/}/}/ }/$}/P}/X}/\}/`}/�}/�}/�}/�}/�}/�}/�}/�}/~/~/~/~/@~/H~/L~/P~/|~/�~/�~/�~/�~/�~/�~/�~/�~/�~///0/8/</@/l/t/x/|/�/�/�/�/�/�/�/�/ �/(�/,�/0�/\�/d�/h�/l�/��/��/��/��/Ԁ/܀/�/�/�/�/�/ �/L�/T�/X�/\�/��/��/��/��/ā/́/Ё/ԁ/�/�/�/�/<�/D�/H�/L�/x�/��/��/��/��/��/��/Ă/�/��/��/�/,�/4�/8�/<�/h�/p�/t�/x�/��/��/��/��/�/�/�/�/�/$�/(�/,�/X�/`�/d�/h�/��/��/��/��/Є/؄/܄/�/�/�/�/�/H�/P�/T�/X�/��/��/��/��/��/ȅ/̅/Ѕ/��/�/�/�/�/8�/@�/D�/H�/L�/t�/|�/��/��/��/��/��/��/��/Ć/�/�/��/��/(�/0�/4�/8�/<�/d�/l�/p�/t�/x�/��/��/��/��/��/܇/�/�/�/�/ �/$�/(�/,�/T�/\�/`�/d�/h�/��/��/��/��/��/̈/Ԉ/؈/܈/�/�/�/�/�/D�/L�/P�/T�/X�/��/��/��/��/��/��/ĉ/ȉ/̉/Љ/��/�/�/�/�/4�/<�/@�/D�/H�/p�/x�/|�/��/��/��/��/��/��/��/�/�/�/��/��/$�/,�/0�/4�/8�/`�/h�/l�/p�/t�/��/��/��/��/��/؋/�/�/�/�/�/�/ �/$�/(�/P�/X�/\�/`�/��/��/��/��/Ȍ/Ќ/Ԍ/،/܌/�/�/�/�/�/@�/H�/L�/P�/T�/|�/��/��/��/��/��/��/č/ȍ/̍/�/��/�/�/�/0�/8�/<�/@�/D�/l�/t�/x�/|�/��/��/��/��/��/��/�/�/�/�/��/ �/(�/,�/0�/4�/\�/d�/h�/l�/p�/��/��/��/��/��/ԏ/܏/�/�/�/�/�/ �/L�/T�/X�/\�/��/��/��/��/Đ/̐/А/Ԑ/�/�/�/�/�/<�/D�/H�/L�/x�/��/��/��/��/��/��/đ/ȑ/�/��/��/�/�/,�/4�/8�/<�/h�/p�/t�/x�/|�/��/��/��/��/�/�/�/�/�/$�/(�/,�/X�/`�/d�/h�/��/��/��/��/Г/ؓ/ܓ/�/�/�/�/�/H�/P�/T�/X�/��/��/��/��/��/��/Ȕ/̔/Д/Ԕ/��/�/�/�/�/8�/@�/D�/H�/L�/t�/|�/��/��/��/��/��/��/ĕ/�/�/��/��/�/(�/0�/4�/8�/<�/d�/l�/p�/t�/x�/��/��/��/��/��/ܖ/�/�/�/�/ �/$�/(�/,�/T�/\�/`�/d�/h�/��/��/��/��/̗/ԗ/ؗ/ܗ/�/�/�/�/�/D�/L�/P�/T�/X�/��/��/��/��/��/��/Ę/Ș/̘/И/��/�/�/�/�/4�/<�/@�/D�/H�/p�/x�/|�/��/��/��/��/��/��/��/�/�/�/��/$�/,�/0�/4�/8�/`�/h�/l�/p�/��/��/��/��/��/ؚ/�/�/�/�/�/�/ �/$�/(�/P�/X�/\�/`�/d�/��/��/��/��/ț/Л/ԛ/؛/ܛ/�/�/�/�/�/@�/H�/L�/P�/T�/|�/��/��/��/��/��/��/Ĝ/Ȝ/̜/�/��/�/�/�/0�/8�/<�/@�/D�/l�/t�/x�/|�/��/��/��/��/��/��/�/�/�/�/��/ �/(�/,�/0�/4�/\�/d�/h�/l�/��/��/��/��/Ԟ/ܞ/�/�/�/�/�/ �/L�/T�/X�/\�/��/��/��/��/��/ğ/̟/П/ԟ/؟/�/�/�/�/�/<�/D�/H�/L�/P�/x�/��/��/��/��/��/��/��/Ġ/Ƞ/�/��/��/�/�/,�/4�/8�/<�/@�/h�/p�/t�/x�/��/��/��/��/��/�/�/�/�/�/�/$�/(�/,�/0�/X�/`�/d�/h�/l�/��/��/��/��/��/Т/آ/ܢ/�/�/�/�/ �/4�/8�/<�/T�/X�/\�/p�/t�/x�/��/��/��/��/��/��/̣/У/ԣ/�/�/�/�/�/�/$�/(�/,�/4�/D�/H�/L�/`�/d�/h�/��/��/��/��/��/��/��/��/Ĥ/ؤ/ܤ/�/��/��/�/�/�/4�/8�/P�/T�/X�/p�/t�/��/��/��/��/��/ȥ/̥/Х/�/�/�/�/�/$�/(�/@�/D�/H�/`�/d�/|�/��/��/��/��/��/��/��/ئ/ܦ/�/��/��/�/�/0�/4�/8�/P�/T�/l�/p�/t�/��/��/��/��/��/ȧ/̧/�/�/�/�/�/ �/$�/(�/@�/D�/\�/`�/d�/|�/��/��/��/��/��/��/Ԩ/ب/ܨ/�/��/�/�/�/0�/4�/L�/P�/T�/l�/p�/��/��/��/��/��/ĩ/ȩ/̩/�/�/�/�/�/�/ �/$�/<�/@�/D�/\�/`�/x�/|�/��/��/��/��/��/��/Ԫ/ت/�/�/��/�/�/�/,�/0�/4�/L�/P�/T�/h�/l�/p�/��/��/��/��/��/ī/ȫ/�/�/�/�/�/�/ �/$�/<�/@�/X�/\�/`�/x�/|�/��/��/��/��/��/Ь/Ԭ/ج/�/�/�/�/�/,�/0�/H�/L�/P�/h�/l�/��/��/��/��/��/��/ĭ/ȭ/�/�/��/�/�/�/ �/8�/<�/@�/X�/\�/`�/t�/x�/|�/��/��/��/��/��/��/Ю/Ԯ/خ/�/�/�/�/�/�/(�/,�/0�/H�/L�/P�/d�/h�/l�/��/��/��/��/��/��/��/į/ܯ/�/�/��/�/�/�/ �/8�/<�/T�/X�/\�/t�/x�/��/��/��/��/��/̰/а/԰/�/�/�/�/�/(�/,�/D�/H�/L�/d�/h�/��/��/��/��/��/��/��/ı/ܱ/�/��/��/�/�/�/ �/4�/8�/<�/T�/X�/p�/t�/x�/��/��/��/��/��/̲/в/�/�/�/�/�/$�/(�/,�/D�/H�/`�/d�/h�/��/��/��/��/��/��/��/س/ܳ/�/��/��/�/�/�/4�/8�/P�/T�/X�/p�/t�/��/��/��/��/��/ȴ/̴/д/�/�/�/�/�/$�/(�/@�/D�/H�/`�/d�/|�/��/��/��/��/��/��/��/ȵ/ص/ܵ/�/��/��/�/�/0�/4�/8�/P�/T�/l�/p�/t�/��/��/��/��/��/ȶ/̶/�/�/�/�/�/ �/$�/(�/@�/D�/\�/`�/d�/|�/��/��/��/��/��/��/��/��/Է/ط/ܷ/�/��/��/�/�/�/0�/4�/8�/L�/P�/T�/l�/p�/t�/��/��/��/��/��/��/ĸ/ȸ/̸/�/�/�/�/�/ �/$�/<�/@�/D�/\�/`�/x�/|�/��/��/��/��/��/��/Թ/ع/�/�/��/�/�/,�/0�/4�/L�/P�/h�/l�/p�/��/��/��/��/��/ĺ/Ⱥ/�/�/�/�/�/�/ �/$�/<�/@�/X�/\�/`�/x�/|�/��/��/��/��/��/л/Ի/ػ/�/�/�/�/�/,�/0�/H�/L�/P�/h�/l�/��/��/��/��/��/��/ļ/ȼ/�/�/��/�/�/�/�/�/ �/(�/,�/0�/L�/T�/X�/\�/d�/h�/l�/��/��/��/��/��/��/��/Ľ/̽/н/Խ/ܽ/�/�/�/�/�/�/�/ �/<�/D�/H�/L�/T�/\�/d�/h�/x�/��/��/��/��/��/��/��/��/��/��/ľ/̾/Ծ/�/��/��/�/�/�/,�/4�/8�/<�/@�/D�/L�/h�/p�/t�/x�/|�/��/��/��/��/��/��/��/��/Ŀ/�/�/�/�/�/��/�/�/$�/(�/,�/0�/4�/<�/X�/`�/d�/h�/l�/p�/x�/��/��/��/��/��/��/��/�/�/�/�/�/�/�/�/�/�/�/ �/$�/,�/H�/P�/T�/X�/`�/h�/��/��/��/��/��/��/��/�/�/�/�/�/��/�/�/�/�/�/8�/@�/D�/H�/P�/X�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/(�/0�/4�/8�/@�/D�/H�/d�/l�/p�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/ �/$�/(�/0�/8�/T�/\�/`�/d�/l�/t�/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/ �/$�/(�/D�/L�/P�/T�/\�/`�/d�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/4�/<�/@�/D�/L�/P�/T�/p�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/$�/,�/0�/4�/<�/@�/D�/`�/h�/l�/p�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/ �/$�/,�/0�/4�/<�/@�/P�/X�/\�/`�/d�/h�/p�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/$�/@�/H�/L�/P�/T�/X�/`�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/0�/8�/<�/@�/D�/H�/P�/l�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/ �/(�/,�/0�/4�/8�/@�/\�/d�/h�/l�/p�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/ �/$�/(�/0�/L�/T�/X�/\�/`�/d�/l�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/ �/<�/D�/H�/L�/P�/T�/\�/x�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/,�/4�/8�/<�/@�/D�/L�/h�/p�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/$�/(�/,�/0�/4�/<�/X�/`�/d�/h�/l�/p�/x�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/ �/$�/,�/H�/P�/T�/X�/\�/`�/h�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/8�/@�/D�/H�/L�/P�/X�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/(�/0�/4�/8�/<�/@�/H�/d�/l�/p�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/ �/$�/(�/,�/0�/8�/T�/\�/`�/d�/h�/l�/t�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/ �/(�/D�/L�/P�/T�/X�/\�/d�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/4�/<�/@�/D�/H�/L�/T�/p�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/$�/,�/0�/4�/8�/<�/D�/`�/h�/l�/p�/t�/x�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/ �/$�/(�/,�/4�/P�/X�/\�/`�/d�/h�/p�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/$�/@�/H�/L�/P�/T�/X�/`�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/0�/8�/<�/@�/D�/H�/P�/l�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/ �/(�/,�/0�/4�/8�/@�/\�/d�/h�/l�/p�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/ �/$�/(�/0�/L�/T�/X�/\�/`�/d�/l�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/ �/<�/D�/H�/L�/P�/T�/\�/x�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/,�/4�/8�/<�/@�/D�/L�/h�/p�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/$�/(�/,�/0�/4�/<�/X�/`�/d�/h�/l�/p�/x�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/ �/$�/,�/H�/P�/T�/X�/\�/`�/h�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/8�/@�/D�/H�/L�/P�/X�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/(�/0�/4�/8�/<�/@�/H�/d�/l�/p�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/ �/$�/(�/,�/0�/8�/T�/\�/`�/d�/h�/l�/t�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/ �/(�/D�/L�/P�/T�/X�/\�/d�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/4�/<�/@�/D�/H�/L�/T�/p�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/$�/,�/0�/4�/8�/<�/D�/`�/h�/l�/p�/t�/x�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/ �/$�/(�/,�/4�/P�/X�/\�/`�/d�/h�/p�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/$�/@�/H�/L�/P�/T�/X�/`�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/0�/8�/<�/@�/D�/H�/P�/l�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/ �/(�/,�/0�/4�/8�/@�/\�/d�/h�/l�/p�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/ �/$�/(�/0�/L�/T�/X�/\�/`�/d�/l�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/�/ �/<�/D�/H�/L�/P�/T�/\�/x�/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/,�/4�/8�/<�/@�/D�/L�/h�/p�/t�/x�/|�/��/��/��/��/��/��/��/��/��/��/�/$�/(�/,�/0�/X�/`�/d�/h�/l�/��/��/��/��/��/��/��/��/��/��/�/�/�/�/ �/H�/P�/T�/X�/\�/��/��/��/��/��/��/��/��/��/��/��/�/�/�/�/8�/@�/D�/H�/L�/t�/|�/��/��/��/��/��/��/��/��/��/��/��/��/�/(�/0�/4�/8�/<�/d�/l�/p�/t�/x�/��/��/��/��/��/��/��/��/��/�/ �/$�/(�/,�/T�/\�/`�/d�/h�/��/��/��/��/��/��/��/��/��/��/�/�/�/�/�/D�/L�/P�/T�/X�/��/��/��/��/��/��/��/��/��/��/4�/<�/@�/D�/H�/p�/x�/|�/��/��/��/��/��/��/��/��/��/��/��/��/$�/,�/0�/4�/`�/h�/l�/p�/t�/��/��/��/��/��/��/��/��/��/�/�/ �/$�/P�/X�/\�/`�/��/��/��/��/��/��/�/�/�/�/�/�/�/�/�/�/@�/H�/L�/P�/T�/|�/��/��/��/��/��/��/�/�/�/�/��/�/�/�/0�/8�/<�/@�/D�/l�/t�/x�/|�/��/��/��/��/��/��/�/�/�/�/��/ �/(�/,�/0�/4�/\�/d�/h�/l�/p�/��/��/��/��/��/�/�/�/�/�/�/�/�/ �/$�/L�/T�/X�/\�/`�/��/��/��/��/��/�/�/�/�/�/�/�/�/�/<�/D�/H�/L�/x�/��/��/��/��/��/��/��/�/�/��/��/�/�/,�/4�/8�/<�/@�/h�/p�/t�/x�/|�/��/��/��/��/��/�/�/�/�/�/�/$�/(�/,�/0�/X�/`�/d�/h�/��/��/��/��/��/�/�/�/�/�/�/�/�/�/H�/P�/T�/X�/\�/��/��/��/��/��/��/�/�/�/�/��/�/�/�/8�/@�/D�/H�/L�/t�/|�/��/��/��/��/��/��/��/�/�/�/��/��/�/(�/0�/4�/8�/<�/d�/l�/p�/t�/x�/��/��/��/��/��/�/�/�/�/�/0 0$0(0T0\0`0d0�0�0�0�0�0�0�0�00000D0L0P0T0�0�0�0�0�0�0�0�0�000040<0@0D0p0x0|0�0�0�0�0�0�0�0�0�0$0,00040`0h0l0p0�0�0�0�0�0�0�0�000 0$0P0X0\0`0�0�0�0�0�0�0�0�00000@0H0L0P0|0�0�0�0�0�0�0�0�0�0000080<0@0l0t0x0|0�0�0�0�0�0�0�0�0 0(0,000\0d0h0l0�0�0�0�0�0�0�0�0000 0L0T0X0\0�0�0�0�0�0�0�0�0	0	0	0	0<	0D	0H	0L	0x	0�	0�	0�	0�	0�	0�	0�	0�	0�	0�	0
0,
04
08
0<
0h
0p
0t
0x
0�
0�
0�
0�
0�
0�
0�
0�
00$0(0,0X0`0d0h0�0�0�0�0�0�0�0�00000H0P0T0X0�0�0�0�0�0�0�0�0�0
0
0
08
0@
0D
0H
0t
0|
0�
0�
0�
0�
0�
0�
0�
0�
0�
0�
0(0004080d0l0p0t0�0�0�0�0�0�0�0�00 0$0(0T0\0`0d0�0�0�0�0�0�0�0�00000D0L0P0T0�0�0�0�0�0�0�0�0�000040<0@0D0p0x0|0�0�0�0�0�0�0�0�0�0$0,00040`0h0l0p0�0�0�0�0�0�0�0�000 0$0P0X0\0`0�0�0�0�0�0�0�0�00000@0H0L0P0|0�0�0�0�0�0�0�0�0�0000080<0@0l0t0x0|0�0�0�0�0�0�0�0�0 0(0,000\0d0h0l0�0�0�0�0�0�0�0�0000 0L0T0X0\0�0�0�0�0�0�0�0�00000<0D0H0L0x0�0�0�0�0�0�0�0�0�0�00,04080<0h0p0t0x0�0�0�0�0�0�0�0�00$0(0,0X0`0d0h0�0�0�0�0�0�0�0�00000H0P0T0X0�0�0�0�0�0�0�0�0�000080@0D0H0t0|0�0�0�0�0�0�0�0�0�0�0(0004080d0l0p0t0�0�0�0�0�0�0�0�00 0$0(0T0\0`0d0�0�0�0�0�0�0�0�00000D0L0P0T0�0�0�0�0�0�0�0�0�0 0 0 04 0< 0@ 0D 0p 0x 0| 0� 0� 0� 0� 0� 0� 0� 0� 0� 0$!0,!00!04!0`!0h!0l!0p!0�!0�!0�!0�!0�!0�!0�!0�!0"0"0 "0$"0P"0X"0\"0`"0�"0�"0�"0�"0�"0�"0�"0�"0#0#0#0#0@#0H#0L#0P#0|#0�#0�#0�#0�#0�#0�#0�#0�#0�#0$0$00$08$0<$0@$0l$0t$0x$0|$0�$0�$0�$0�$0�$0�$0�$0�$0 %0(%0,%00%0\%0d%0h%0l%0�%0�%0�%0�%0�%0�%0�%0�%0&0&0&0 &0L&0T&0X&0\&0�&0�&0�&0�&0�&0�&0�&0�&0'0'0'0'0<'0D'0H'0L'0x'0�'0�'0�'0�'0�'0�'0�'0�'0�'0�'0(0,(04(08(0<(0h(0p(0t(0x(0�(0�(0�(0�(0�(0�(0�(0�(0)0$)0()0,)0X)0`)0d)0h)0�)0�)0�)0�)0�)0�)0�)0�)0*0*0*0*0H*0P*0T*0X*0�*0�*0�*0�*0�*0�*0�*0�*0�*0+0+0+08+0@+0D+0H+0t+0|+0�+0�+0�+0�+0�+0�+0�+0�+0�+0�+0(,00,04,08,0d,0l,0p,0t,0�,0�,0�,0�,0�,0�,0�,0�,0-0 -0$-0(-0T-0\-0`-0d-0�-0�-0�-0�-0�-0�-0�-0�-0.0.0.0.0D.0L.0P.0T.0�.0�.0�.0�.0�.0�.0�.0�.0�.0/0/0/04/0</0@/0D/0p/0x/0|/0�/0�/0�/0�/0�/0�/0�/0�/0�/0$00,00000400`00h00l00p00�00�00�00�00�00�00�00�001010 10$10P10X10\10`10�10�10�10�10�10�10�10�1020202020@20H20L20P20|20�20�20�20�20�20�20�20�20�203030030830<30@30l30t30x30|30�30�30�30�30�30�30�30�30 40(40,40040\40d40h40l40�40�40�40�40�40�40�40�40505050 50L50T50X50\50�50�50�50�50�50�50�50�5060606060<60D60H60L60x60�60�60�60�60�60�60�60�60�60�6070,70470870<70h70p70t70x70�70�70�70�70�70�70�70�7080$80(80,80X80`80d80h80�80�80�80�80�80�80�80�8090909090H90P90T90X90�90�90�90�90�90�90�90�90�90:0:0:08:0@:0D:0H:0t:0|:0�:0�:0�:0�:0�:0�:0�:0�:0�:0�:0(;00;04;08;0d;0l;0p;0t;0�;0�;0�;0�;0�;0�;0�;0�;0<0 <0$<0(<0T<0\<0`<0d<0�<0�<0�<0�<0�<0�<0�<0�<0=0=0=0=0D=0L=0P=0T=0�=0�=0�=0�=0�=0�=0�=0�=0�=0>0>0>04>0<>0@>0D>0p>0x>0|>0�>0�>0�>0�>0�>0�>0�>0�>0�>0$?0,?00?04?0`?0h?0l?0p?0�?0�?0�?0�?0�?0�?0�?0�?0@0@0 @0$@0P@0X@0\@0`@0�@0�@0�@0�@0�@0�@0�@0�@0A0A0A0A0@A0HA0LA0PA0|A0�A0�A0�A0�A0�A0�A0�A0�A0�A0B0B00B08B0<B0@B0lB0tB0xB0|B0�B0�B0�B0�B0�B0�B0�B0�B0 C0(C0,C00C0\C0dC0hC0lC0�C0�C0�C0�C0�C0�C0�C0�C0D0D0D0 D0LD0TD0XD0\D0�D0�D0�D0�D0�D0�D0�D0�D0E0E0E0E0<E0DE0HE0LE0xE0�E0�E0�E0�E0�E0�E0�E0�E0�E0�E0F0,F04F08F0<F0hF0pF0tF0xF0�F0�F0�F0�F0�F0�F0�F0�F0G0$G0(G0,G0XG0`G0dG0hG0�G0�G0�G0�G0�G0�G0�G0�G0H0H0H0H0HH0PH0TH0XH0�H0�H0�H0�H0�H0�H0�H0�H0�H0I0I0I08I0@I0DI0HI0tI0|I0�I0�I0�I0�I0�I0�I0�I0�I0�I0�I0(J00J04J08J0dJ0lJ0pJ0tJ0�J0�J0�J0�J0�J0�J0�J0�J0K0 K0$K0(K0TK0\K0`K0dK0�K0�K0�K0�K0�K0�K0�K0�K0L0L0L0L0DL0LL0PL0TL0�L0�L0�L0�L0�L0�L0�L0�L0�L0M0M0M04M0<M0@M0DM0pM0xM0|M0�M0�M0�M0�M0�M0�M0�M0�M0�M0$N0,N00N04N0`N0hN0lN0pN0�N0�N0�N0�N0�N0�N0�N0�N0O0O0 O0$O0PO0XO0\O0`O0�O0�O0�O0�O0�O0�O0�O0�O0P0P0P0P0@P0HP0LP0PP0|P0�P0�P0�P0�P0�P0�P0�P0�P0�P0Q0Q00Q08Q0<Q0@Q0lQ0tQ0xQ0|Q0�Q0�Q0�Q0�Q0�Q0�Q0�Q0�Q0 R0(R0,R00R0\R0dR0hR0lR0�R0�R0�R0�R0�R0�R0�R0�R0S0S0S0 S0LS0TS0XS0\S0�S0�S0�S0�S0�S0�S0�S0�S0T0T0T0T0<T0DT0HT0LT0xT0�T0�T0�T0�T0�T0�T0�T0�T0�T0�T0U0,U04U08U0<U0hU0pU0tU0xU0�U0�U0�U0�U0�U0�U0�U0�U0V0$V0(V0,V0XV0`V0dV0hV0�V0�V0�V0�V0�V0�V0�V0�V0W0W0W0W0HW0PW0TW0XW0�W0�W0�W0�W0�W0�W0�W0�W0�W0X0X0X08X0@X0DX0HX0tX0|X0�X0�X0�X0�X0�X0�X0�X0�X0�X0�X0(Y00Y04Y08Y0dY0lY0pY0tY0�Y0�Y0�Y0�Y0�Y0�Y0�Y0�Y0Z0 Z0$Z0(Z0TZ0\Z0`Z0dZ0�Z0�Z0�Z0�Z0�Z0�Z0�Z0�Z0[0[0[0[0D[0L[0P[0T[0�[0�[0�[0�[0�[0�[0�[0�[0�[0\0\0\04\0<\0@\0D\0p\0x\0|\0�\0�\0�\0�\0�\0�\0�\0�\0�\0$]0,]00]04]0`]0h]0l]0p]0�]0�]0�]0�]0�]0�]0�]0�]0^0^0 ^0$^0P^0X^0\^0`^0�^0�^0�^0�^0�^0�^0�^0�^0_0_0_0_0@_0H_0L_0P_0|_0�_0�_0�_0�_0�_0�_0�_0�_0�_0`0`00`08`0<`0@`0l`0t`0x`0|`0�`0�`0�`0�`0�`0�`0�`0�`0 a0(a0,a00a0\a0da0ha0la0�a0�a0�a0�a0�a0�a0�a0�a0b0b0b0 b0Lb0Tb0Xb0\b0�b0�b0�b0�b0�b0�b0�b0�b0c0c0c0c0<c0Dc0Hc0Lc0xc0�c0�c0�c0�c0�c0�c0�c0�c0�c0�c0d0,d04d08d0<d0hd0pd0td0xd0�d0�d0�d0�d0�d0�d0�d0�d0e0$e0(e0,e0Xe0`e0de0he0�e0�e0�e0�e0�e0�e0�e0�e0f0f0f0f0Hf0Pf0Tf0Xf0�f0�f0�f0�f0�f0�f0�f0�f0�f0g0g0g08g0@g0Dg0Hg0tg0|g0�g0�g0�g0�g0�g0�g0�g0�g0�g0�g0(h00h04h08h0dh0lh0ph0th0�h0�h0�h0�h0�h0�h0�h0�h0i0 i0$i0(i0Ti0\i0`i0di0�i0�i0�i0�i0�i0�i0�i0�i0j0j0j0j0Dj0Lj0Pj0Tj0�j0�j0�j0�j0�j0�j0�j0�j0�j0k0k0k04k0<k0@k0Dk0pk0xk0|k0�k0�k0�k0�k0�k0�k0�k0�k0�k0$l0,l00l04l0`l0hl0ll0pl0�l0�l0�l0�l0�l0�l0�l0�l0m0m0 m0$m0Pm0Xm0\m0`m0�m0�m0�m0�m0�m0�m0�m0�m0n0n0n0n0@n0Hn0Ln0Pn0|n0�n0�n0�n0�n0�n0�n0�n0�n0�n0o0o00o08o0<o0@o0lo0to0xo0|o0�o0�o0�o0�o0�o0�o0�o0�o0 p0(p0,p00p0\p0dp0hp0lp0�p0�p0�p0�p0�p0�p0�p0�p0q0q0q0 q0Lq0Tq0Xq0\q0�q0�q0�q0�q0�q0�q0�q0�q0r0r0r0r0<r0Dr0Hr0Lr0xr0�r0�r0�r0�r0�r0�r0�r0�r0�r0�r0s0,s04s08s0<s0hs0ps0ts0xs0�s0�s0�s0�s0�s0�s0�s0�s0t0$t0(t0,t0Xt0`t0dt0ht0�t0�t0�t0�t0�t0�t0�t0�t0u0u0u0u0Hu0Pu0Tu0Xu0�u0�u0�u0�u0�u0�u0�u0�u0�u0v0v0v08v0@v0Dv0Hv0Lv0tv0|v0�v0�v0�v0�v0�v0�v0�v0�v0�v0�v0�v0�v0w0(w00w04w08w0<w0dw0lw0pw0tw0xw0�w0�w0�w0�w0�w0�w0�w0�w0�w0�w0x0 x0$x0(x0,x0Tx0\x0`x0dx0hx0�x0�x0�x0�x0�x0�x0�x0�x0�x0�x0y0y0y0y0y0Dy0Ly0Py0Ty0�y0�y0�y0�y0�y0�y0�y0�y0�y0�y0�y0z0z0z0,z04z0<z0@z0Dz0hz0pz0xz0|z0�z0�z0�z0�z0�z0�z0�z0�z0�z0�z0�z0{0${0,{00{04{0X{0`{0h{0l{0p{0�{0�{0�{0�{0�{0�{0�{0�{0�{0�{0|0|0|0 |0$|0H|0P|0X|0\|0`|0�|0�|0�|0�|0�|0�|0�|0�|0�|0�|0�|0}0}0}0}08}0@}0H}0L}0P}0t}0|}0�}0�}0�}0�}0�}0�}0�}0�}0�}0�}0�}0~0~0(~00~08~0<~0@~0d~0l~0t~0x~0|~0�~0�~0�~0�~0�~0�~0�~0�~0�~0�~00 0(0,000T0\0d0h0l0�0�0�0�0�0�0�0�0�0�0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0Ā0̀0Ѐ0Ԁ0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0ā0�0�0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0؂0�0�0�0�0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0ȃ0Ѓ0؃0܃0�0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0Ȅ0̄0Є0�0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0�0�0�0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0Ԇ0܆0�0�0�0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0ć0̇0ԇ0؇0܇0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0Ĉ0Ȉ0̈0�0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0�0�0�0�0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0Њ0؊0�0�0�0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0ȋ0Ћ0ԋ0؋0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0Č0Ȍ0�0�0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0܍0�0�0�0�0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0̎0Ԏ0܎0�0�0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0ď0̏0Џ0ԏ0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0Đ0�0�0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0ؑ0�0�0�0�0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0Ȓ0В0ؒ0ܒ0�0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0ȓ0̓0Г0�0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0�0�0�0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0ԕ0ܕ0�0�0�0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0Ė0̖0Ԗ0ؖ0ܖ0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0ė0ȗ0̗0�0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0�0�0�0�0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0Й0ؙ0�0�0�0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0Ț0К0Ԛ0ؚ0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0ě0ț0�0�0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0ܜ0�0�0�0�0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0̝0ԝ0ܝ0�0�0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0Ğ0̞0О0Ԟ0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0ğ0�0�0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0ؠ0�0�0�0�0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0ȡ0С0ء0ܡ0�0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0Ȣ0̢0Т0�0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0�0�0�0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0Ԥ0ܤ0�0�0�0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0ĥ0̥0ԥ0إ0ܥ0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0Ħ0Ȧ0̦0�0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0�0�0�0�0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0Ш0ب0�0�0�0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0ȩ0Щ0ԩ0ة0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0Ī0Ȫ0�0�0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0ܫ0�0�0�0�0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0̬0Ԭ0ܬ0�0�0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0ĭ0̭0Э0ԭ0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0Į0�0�0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0د0�0�0�0�0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0Ȱ0а0ذ0ܰ0�0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0ȱ0̱0б0�0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0�0�0�0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0Գ0ܳ0�0�0�0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0Ĵ0̴0Դ0ش0ܴ0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0ĵ0ȵ0̵0�0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0�0�0�0�0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0з0ط0�0�0�0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0ȸ0и0Ը0ظ0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0Ĺ0ȹ0�0�0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0ܺ0�0�0�0�0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0̻0Ի0ܻ0�0�0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0ļ0̼0м0Լ0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0Ľ0�0�0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0ؾ0�0�0�0�0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0ȿ0п0ؿ0ܿ0�0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0�0�0�0�0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0�0�0�0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0��0��0��0��0��0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0��0��0��0��0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0��0��0��0��0��0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0��0��0��0��0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0��0��0��0��0��0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0��0��0��0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0��0��0��0��0��0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0��0��0��0��0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0��0��0��0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0��0��0��0��0��0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0��0��0��0��0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0��0��0��0��0��0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0��0��0��0��0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0��0��0��0��0��0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0��0��0��0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0��0��0��0��0��0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0��0��0��0��0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0��0��0��0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0��0��0��0��0��0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0��0��0��0��0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0��0��0��0��0��0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0��0��0��0��0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0��0��0��0��0��0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0��0��0��0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0��0��0��0��0��0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0��0��0��0��0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0��0��0��0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0��0��0��0��0��0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0��0��0��0��0��0�0�0�0�0�0<�0D�0L�0P�0T�0x�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0,�04�0<�0@�0D�0h�0p�0x�0|�0��0��0��0��0��0��0��0��0��0��0��0�0$�0,�00�04�0X�0`�0h�0l�0p�0��0��0��0��0��0��0��0��0��0��0�0�0�0 �0$�0H�0P�0X�0\�0`�0��0��0��0��0��0��0��0��0��0��0��0�0�0�0�08�0@�0H�0L�0P�0t�0|�0��0��0��0��0��0��0�0�0�0�0��0�0�0(�00�08�0<�0@�0d�0l�0t�0x�0|�0��0��0��0��0��0�0�0�0�0�0�0 �0(�0,�00�0T�0\�0d�0h�0l�0��0��0��0��0��0�0�0�0�0�0�0�0�0�0 �0D�0L�0T�0X�0\�0��0��0��0��0��0��0�0�0�0�0��0�0�0�0�04�0<�0D�0H�0L�0p�0x�0��0��0��0��0��0��0��0�0�0�0��0��0�0$�0,�04�08�0<�0`�0h�0p�0t�0x�0��0��0��0��0��0�0�0�0�0�0�0�0$�0(�0,�0P�0X�0`�0d�0h�0��0��0��0��0��0�0�0�0�0�0�0�0�0�0�0@�0H�0P�0T�0X�0|�0��0��0��0��0��0��0�0�0�0�0��0�0�0�00�08�0@�0D�0H�0l�0t�0|�0��0��0��0��0��0��0��0�0�0�0��0��0 �0(�00�04�08�0\�0d�0l�0p�0t�0��0��0��0��0��0�0�0�0�0�0�0�0 �0$�0(�0L�0T�0\�0`�0d�0��0��0��0��0��0�0�0�0�0�011111<1D1L1P1T1x1�1�1�1�1�1�1�1�1�1�1�1111,141<1@1D1h1p1x1|1�1�1�1�1�1�1�1�1�1�1�11$1,10141X1`1h1l1p1�1�1�1�1�1�1�1�1�1�1111 1$1H1P1X1\1`1�1�1�1�1�1�1�1�1�1�1�1111181@1H1L1P1t1|1�1�1�1�1�1�1�1�1�1�1�111(10181<1@1d1l1t1x1|1�1�1�1�1�1�1�1�1�1�11 1(1,101T1\1d1h1l1�1�1�1�1�1�1�1�1�1�11111 1D1L1T1X1\1�1�1�1�1�1�1�1�1�1�1�1111141<1D1H1L1p1x1�1�1�1�1�1�1�1�1�1�1�1�1	1$	1,	14	18	1<	1`	1h	1p	1t	1x	1�	1�	1�	1�	1�	1�	1�	1�	1�	1�	1
1
1$
1(
1,
1P
1X
1`
1d
1h
1�
1�
1�
1�
1�
1�
1�
1�
1�
1�
111111@1H1P1T1X1|1�1�1�1�1�1�1�1�1�1�1�11110181@1D1H1l1t1|1�1�1�1�1�1�1�1�1�1�1�1�1 
1(
10
14
18
1\
1d
1l
1p
1t
1�
1�
1�
1�
1�
1�
1�
1�
1�
1�
111 1$1(1L1T1\1`1d1�1�1�1�1�1�1�1�1�1�111111<1D1L1P1T1x1�1�1�1�1�1�1�1�1�1�1�1111,141<1@1D1h1p1x1|1�1�1�1�1�1�1�1�1�1�1�11$1,10141X1`1h1l1p1�1�1�1�1�1�1�1�1�1�1111 1$1H1P1X1\1`1�1�1�1�1�1�1�1�1�1�1�1111181@1H1L1P1t1|1�1�1�1�1�1�1�1�1�1�1�111(10181<1@1D1l1t1x1|1�1�1�1�1�1�1�1�1 1(1,101h1l1p1�1�1�1�1�1�1�1�1�1�1�1�1�1111 1$181<1@1X1\1`1t1x1|1�1�1�1�1�1�1�1�1�1�1�1�1�1111(1,101H1L1d1h1l1�1�1�1�1�1�1�1�1�1�1�1111 181<1T1X1\1t1x1�1�1�1�1�1�1�1�1�1�1111(1,1D1H1L1d1h1�1�1�1�1�1�1�1�1�1�1�1�11114181<1T1X1p1t1x1�1�1�1�1�1�1�1�1�1�111$1(1,1D1H1`1d1h1�1�1�1�1�1�1�1�1�1�1�1�11114181<1P1T1X1p1t1�1�1�1�1�1�1�1�1�1�1111$1(1@1D1H1`1d1h1|1�1�1�1�1�1�1�1�1�1�1�1�1�111014181P1T1l1p1t1�1�1�1�1�1�1�1�1�1�111 1$1(1@1D1\1`1d1|1�1�1�1�1�1�1�1�1�1�1�1 1 1 10 14 1L 1P 1T 1l 1p 1� 1� 1� 1� 1� 1� 1� 1� 1� 1� 1� 1� 1!1!1!1 !1$!1(!1<!1@!1D!1\!1`!1d!1x!1|!1�!1�!1�!1�!1�!1�!1�!1�!1�!1�!1�!1�!1�!1"1"1,"10"14"1L"1P"1h"1l"1p"1�"1�"1�"1�"1�"1�"1�"1�"1�"1�"1#1#1#1 #1$#1<#1@#1X#1\#1`#1x#1|#1�#1�#1�#1�#1�#1�#1�#1�#1�#1�#1$1$1$1,$10$1H$1L$1P$1h$1l$1�$1�$1�$1�$1�$1�$1�$1�$1�$1�$1�$1�$1%1%1%1 %18%1<%1@%1X%1\%1t%1x%1|%1�%1�%1�%1�%1�%1�%1�%1�%1�%1�%1&1&1(&1,&10&1H&1L&1d&1h&1l&1�&1�&1�&1�&1�&1�&1�&1�&1�&1�&1�&1'1'1'1 '18'1<'1T'1X'1\'1t'1x'1�'1�'1�'1�'1�'1�'1�'1�'1�'1�'1(1(1(1((1,(1D(1H(1L(1T(1d(1h(1�(1�(1�(1�(1�(1�(1�(1�(1�(1�(1�(1�(1)1)1)14)18)1<)1T)1X)1p)1t)1x)1�)1�)1�)1�)1�)1�)1�)1�)1�)1�)1*1*1*1$*1(*1,*1D*1H*1L*1`*1d*1h*1�*1�*1�*1�*1�*1�*1�*1�*1�*1�*1�*1�*1�*1�*1+1+1+1+14+18+1<+1P+1T+1X+1p+1t+1�+1�+1�+1�+1�+1�+1�+1�+1�+1�+1,1,1,1$,1(,1@,1D,1H,1`,1d,1|,1�,1�,1�,1�,1�,1�,1�,1�,1�,1�,1�,1�,1-1-10-14-18-1P-1T-1l-1p-1t-1�-1�-1�-1�-1�-1�-1�-1�-1�-1�-1.1.1 .1$.1(.1@.1D.1\.1`.1d.1|.1�.1�.1�.1�.1�.1�.1�.1�.1�.1�.1�.1/1/1/10/14/1L/1P/1T/1l/1p/1�/1�/1�/1�/1�/1�/1�/1�/1�/1�/1010101 01$01<01@01D01P01X01\01`01d01�01�01�01�01�01�01�01�01�01�011111111111@11H11L11P11T11|11�11�11�11�11�11�11�11�11�11�11�11212121021821<21@21D21l21t21x21|21�21�21�21�21�21�21�21�21�21�21�21 31(31,31031431\31d31h31l31p31�31�31�31�31�31�31�31�31�31�31414141 41$41L41T41X41\41`41�41�41�41�41�41�41�41�415151515151<51D51H51L51x51�51�51�51�51�51�51�51�51�51�51�5161,61461861<61@61h61p61t61x61�61�61�61�61�61�61�61�61�6171$71(71,71X71`71d71h71�71�71�71�71�71�71�71�71�7181818181 81H81P81T81X81\81�81�81�81�81�81�81�81�81�81919191891@91D91H91L91t91|91�91�91�91�91�91�91�91�91�91�91�91�91:1(:10:14:18:1<:1d:1l:1p:1t:1x:1�:1�:1�:1�:1�:1�:1�:1�:1�:1�:1;1 ;1$;1(;1,;1T;1\;1`;1d;1h;1�;1�;1�;1�;1�;1�;1�;1�;1�;1�;1<1<1<1<1<1D<1L<1P<1T<1X<1�<1�<1�<1�<1�<1�<1�<1�<1�<1�<1=1=1=1=14=1<=1@=1D=1H=1p=1x=1|=1�=1�=1�=1�=1�=1�=1�=1�=1�=1�=1$>1,>10>14>18>1`>1h>1l>1p>1�>1�>1�>1�>1�>1�>1�>1�>1�>1?1?1 ?1$?1P?1X?1\?1`?1�?1�?1�?1�?1�?1�?1�?1�?1�?1�?1@1@1@1@1@1@@1H@1L@1P@1T@1|@1�@1�@1�@1�@1�@1�@1�@1�@1�@1�@1�@1A1A1A10A18A1<A1@A1DA1lA1tA1xA1|A1�A1�A1�A1�A1�A1�A1�A1�A1�A1�A1�A1 B1(B1,B10B1\B1dB1hB1lB1�B1�B1�B1�B1�B1�B1�B1�B1C1C1C1 C1LC1TC1XC1\C1�C1�C1�C1�C1�C1�C1�C1�C1D1D1D1D1<D1DD1HD1LD1xD1�D1�D1�D1�D1�D1�D1�D1�D1�D1�D1E1,E14E18E1<E1hE1pE1tE1xE1�E1�E1�E1�E1�E1�E1�E1�E1F1$F1(F1,F1XF1`F1dF1hF1�F1�F1�F1�F1�F1�F1�F1�F1G1G1G1G1HG1PG1TG1XG1�G1�G1�G1�G1�G1�G1�G1�G1�G1H1H1H18H1@H1DH1HH1tH1|H1�H1�H1�H1�H1�H1�H1�H1�H1�H1�H1(I10I14I18I1dI1lI1pI1tI1�I1�I1�I1�I1�I1�I1�I1�I1J1 J1$J1(J1TJ1\J1`J1dJ1�J1�J1�J1�J1�J1�J1�J1�J1K1K1K1K1DK1LK1PK1TK1�K1�K1�K1�K1�K1�K1�K1�K1�K1L1L1L14L1<L1@L1DL1pL1xL1|L1�L1�L1�L1�L1�L1�L1�L1�L1�L1$M1,M10M14M1`M1hM1lM1pM1�M1�M1�M1�M1�M1�M1�M1�M1N1N1 N1$N1PN1XN1\N1`N1�N1�N1�N1�N1�N1�N1�N1�N1O1O1O1O1@O1HO1LO1PO1|O1�O1�O1�O1�O1�O1�O1�O1�O1�O1P1P10P18P1<P1@P1lP1tP1xP1|P1�P1�P1�P1�P1�P1�P1�P1�P1 Q1(Q1,Q10Q1\Q1dQ1hQ1lQ1�Q1�Q1�Q1�Q1�Q1�Q1�Q1�Q1R1R1R1 R1LR1TR1XR1\R1�R1�R1�R1�R1�R1�R1�R1�R1S1S1S1S1<S1DS1HS1LS1xS1�S1�S1�S1�S1�S1�S1�S1�S1�S1�S1T1,T14T18T1<T1hT1pT1tT1xT1�T1�T1�T1�T1�T1�T1�T1�T1U1$U1(U1,U1XU1`U1dU1hU1�U1�U1�U1�U1�U1�U1�U1�U1V1V1V1V1HV1PV1TV1XV1�V1�V1�V1�V1�V1�V1�V1�V1�V1W1W1W18W1@W1DW1HW1tW1|W1�W1�W1�W1�W1�W1�W1�W1�W1�W1�W1(X10X14X18X1dX1lX1pX1tX1�X1�X1�X1�X1�X1�X1�X1�X1Y1 Y1$Y1(Y1TY1\Y1`Y1dY1�Y1�Y1�Y1�Y1�Y1�Y1�Y1�Y1Z1Z1Z1Z1DZ1LZ1PZ1TZ1�Z1�Z1�Z1�Z1�Z1�Z1�Z1�Z1�Z1[1[1[14[1<[1@[1D[1p[1x[1|[1�[1�[1�[1�[1�[1�[1�[1�[1�[1$\1,\10\14\1`\1h\1l\1p\1�\1�\1�\1�\1�\1�\1�\1�\1]1]1 ]1$]1P]1X]1\]1`]1�]1�]1�]1�]1�]1�]1�]1�]1^1^1^1^1@^1H^1L^1P^1|^1�^1�^1�^1�^1�^1�^1�^1�^1�^1_1_10_18_1<_1@_1l_1t_1x_1|_1�_1�_1�_1�_1�_1�_1�_1�_1 `1(`1,`10`1\`1d`1h`1l`1�`1�`1�`1�`1�`1�`1�`1�`1a1a1a1 a1La1Ta1Xa1\a1�a1�a1�a1�a1�a1�a1�a1�a1b1b1b1b1<b1Db1Hb1Lb1xb1�b1�b1�b1�b1�b1�b1�b1�b1�b1�b1c1,c14c18c1<c1hc1pc1tc1xc1�c1�c1�c1�c1�c1�c1�c1�c1d1$d1(d1,d1Xd1`d1dd1hd1�d1�d1�d1�d1�d1�d1�d1�d1e1e1e1e1He1Pe1Te1Xe1�e1�e1�e1�e1�e1�e1�e1�e1�e1f1f1f18f1@f1Df1Hf1tf1|f1�f1�f1�f1�f1�f1�f1�f1�f1�f1�f1(g10g14g18g1dg1lg1pg1tg1�g1�g1�g1�g1�g1�g1�g1�g1h1 h1$h1(h1Th1\h1`h1dh1�h1�h1�h1�h1�h1�h1�h1�h1i1i1i1i1Di1Li1Pi1Ti1�i1�i1�i1�i1�i1�i1�i1�i1�i1j1j1j14j1<j1@j1Dj1pj1xj1|j1�j1�j1�j1�j1�j1�j1�j1�j1�j1$k1,k10k14k1`k1hk1lk1pk1�k1�k1�k1�k1�k1�k1�k1�k1l1l1 l1$l1Pl1Xl1\l1`l1�l1�l1�l1�l1�l1�l1�l1�l1m1m1m1m1@m1Hm1Lm1Pm1|m1�m1�m1�m1�m1�m1�m1�m1�m1�m1n1n10n18n1<n1@n1ln1tn1xn1|n1�n1�n1�n1�n1�n1�n1�n1�n1 o1(o1,o10o1\o1do1ho1lo1�o1�o1�o1�o1�o1�o1�o1�o1p1p1p1 p1Lp1Tp1Xp1\p1�p1�p1�p1�p1�p1�p1�p1�p1q1q1q1q1<q1Dq1Hq1Lq1xq1�q1�q1�q1�q1�q1�q1�q1�q1�q1�q1r1,r14r18r1<r1hr1pr1tr1xr1�r1�r1�r1�r1�r1�r1�r1�r1s1$s1(s1,s1Xs1`s1ds1hs1�s1�s1�s1�s1�s1�s1�s1�s1t1t1t1t1Ht1Pt1Tt1Xt1�t1�t1�t1�t1�t1�t1�t1�t1�t1u1u1u18u1@u1Du1Hu1tu1|u1�u1�u1�u1�u1�u1�u1�u1�u1�u1�u1(v10v14v18v1dv1lv1pv1tv1�v1�v1�v1�v1�v1�v1�v1�v1w1 w1$w1(w1Tw1\w1`w1dw1�w1�w1�w1�w1�w1�w1�w1�w1x1x1x1x1Dx1Lx1Px1Tx1�x1�x1�x1�x1�x1�x1�x1�x1�x1y1y1y14y1<y1@y1Dy1py1xy1|y1�y1�y1�y1�y1�y1�y1�y1�y1�y1$z1,z10z14z1`z1hz1lz1pz1�z1�z1�z1�z1�z1�z1�z1�z1{1{1 {1${1P{1X{1\{1`{1�{1�{1�{1�{1�{1�{1�{1�{1|1|1|1|1@|1H|1L|1P|1||1�|1�|1�|1�|1�|1�|1�|1�|1�|1}1}10}18}1<}1@}1l}1t}1x}1|}1�}1�}1�}1�}1�}1�}1�}1�}1 ~1(~1,~10~1\~1d~1h~1l~1�~1�~1�~1�~1�~1�~1�~1�~1111 1L1T1X1\1�1�1�1�1�1�1�1�1�1�1�1�1<�1D�1H�1L�1x�1��1��1��1��1��1��1Ā1�1��1��1�1,�14�18�1<�1h�1p�1t�1x�1��1��1��1��1�1�1�1�1�1$�1(�1,�1X�1`�1d�1h�1��1��1��1��1Ђ1؂1܂1�1�1�1�1�1H�1P�1T�1X�1��1��1��1��1��1ȃ1̃1Ѓ1��1�1�1�18�1@�1D�1H�1t�1|�1��1��1��1��1��1��1�1�1��1��1(�10�14�18�1d�1l�1p�1t�1��1��1��1��1܅1�1�1�1�1 �1$�1(�1T�1\�1`�1d�1��1��1��1��1̆1Ԇ1؆1܆1�1�1�1�1D�1L�1P�1T�1��1��1��1��1��1ć1ȇ1̇1��1�1�1�14�1<�1@�1D�1p�1x�1|�1��1��1��1��1��1�1�1�1��1$�1,�10�14�1`�1h�1l�1p�1��1��1��1��1؉1�1�1�1�1�1 �1$�1P�1X�1\�1`�1��1��1��1��1Ȋ1Њ1Ԋ1؊1�1�1�1�1@�1H�1L�1P�1|�1��1��1��1��1��1ċ1ȋ1�1��1�1�10�18�1<�1@�1l�1t�1x�1|�1��1��1��1��1�1�1�1�1 �1(�1,�10�1\�1d�1h�1l�1��1��1��1��1ԍ1܍1�1�1�1�1�1 �1L�1T�1X�1\�1��1��1��1��1Ď1̎1Ў1Ԏ1�1�1�1�1<�1D�1H�1L�1x�1��1��1��1��1��1��1ď1�1��1��1�1,�14�18�1<�1h�1p�1t�1x�1��1��1��1��1�1�1�1�1�1$�1(�1,�1X�1`�1d�1h�1��1��1��1��1Б1ؑ1ܑ1�1�1�1�1�1H�1P�1T�1X�1��1��1��1��1��1Ȓ1̒1В1��1�1�1�18�1@�1D�1H�1t�1|�1��1��1��1��1��1��1�1�1��1��1(�10�14�18�1<�1d�1l�1p�1t�1��1��1��1��1ܔ1�1�1�1�1 �1$�1(�1L�1T�1\�1`�1d�1��1��1��1��1��1ĕ1̕1ԕ1ؕ1ܕ1�1�1�1�1�1<�1D�1L�1P�1T�1x�1��1��1��1��1��1��1Ė1Ȗ1̖1�1��1�1�1�1,�14�1<�1@�1D�1h�1p�1x�1|�1��1��1��1��1��1��1�1�1�1�1��1�1$�1,�10�14�1X�1`�1h�1l�1p�1��1��1��1��1��1И1ؘ1�1�1�1�1�1�1 �1$�1H�1P�1X�1\�1`�1��1��1��1��1��1��1ș1Й1ԙ1ؙ1��1�1�1�1�18�1@�1H�1L�1P�1t�1|�1��1��1��1��1��1��1Ě1Ț1�1�1��1�1�1(�10�18�1<�1@�1d�1l�1t�1x�1|�1��1��1��1��1��1ܛ1�1�1�1�1�1 �1(�1,�10�1T�1\�1d�1h�1l�1��1��1��1��1��1̜1Ԝ1ܜ1�1�1�1�1�1�1 �1D�1L�1T�1X�1\�1��1��1��1��1��1��1ĝ1̝1Н1ԝ1��1�1�1�1�14�1<�1D�1H�1L�1p�1x�1��1��1��1��1��1��1��1Ğ1�1�1��1��1�1$�1,�14�18�1<�1`�1h�1p�1t�1x�1��1��1��1��1��1؟1�1�1�1�1�1�1$�1(�1,�1P�1X�1`�1d�1h�1��1��1��1��1��1Ƞ1Р1ؠ1ܠ1�1�1�1�1�1�1@�1H�1P�1T�1X�1|�1��1��1��1��1��1��1ȡ1̡1С1�1��1�1�1�10�18�1@�1D�1H�1l�1t�1|�1��1��1��1��1��1��1��1�1�1�1��1��1 �1(�10�14�18�1\�1d�1l�1p�1t�1��1��1��1��1��1ԣ1ܣ1�1�1�1�1�1 �1$�1(�1L�1T�1\�1`�1d�1��1��1��1��1��1Ĥ1̤1Ԥ1ؤ1ܤ1�1�1�1�1�1<�1D�1L�1P�1T�1x�1��1��1��1��1��1��1ĥ1ȥ1̥1�1��1�1�1�1,�14�1<�1@�1D�1h�1p�1x�1|�1��1��1��1��1��1��1�1�1�1�1��1�1$�1,�10�14�1X�1`�1h�1l�1p�1��1��1��1��1��1Ч1ا1�1�1�1�1�1�1 �1$�1H�1P�1X�1\�1`�1��1��1��1��1��1��1Ȩ1Ш1Ԩ1ب1��1�1�1�1�18�1@�1H�1L�1P�1t�1|�1��1��1��1��1��1��1ĩ1ȩ1�1�1��1�1�1(�10�18�1<�1@�1d�1l�1t�1x�1|�1��1��1��1��1��1ܪ1�1�1�1�1�1 �1(�1,�10�1T�1\�1d�1h�1l�1��1��1��1��1��1̫1ԫ1ܫ1�1�1�1�1�1�1 �1D�1L�1T�1X�1\�1��1��1��1��1��1��1Ĭ1̬1Ь1Ԭ1��1�1�1�1�14�1<�1D�1H�1L�1p�1x�1��1��1��1��1��1��1��1ĭ1�1�1��1��1�1$�1,�14�18�1<�1`�1h�1p�1t�1x�1��1��1��1��1��1خ1�1�1�1�1�1�1$�1(�1,�1P�1X�1`�1d�1h�1��1��1��1��1��1ȯ1Я1د1ܯ1�1�1�1�1�1�1@�1H�1P�1T�1X�1|�1��1��1��1��1��1��1Ȱ1̰1а1�1��1�1�1�10�18�1@�1D�1H�1l�1t�1|�1��1��1��1��1��1��1��1�1�1�1��1��1 �1(�10�14�18�1\�1d�1l�1p�1t�1��1��1��1��1��1Բ1ܲ1�1�1�1�1�1 �1$�1(�1L�1T�1\�1`�1d�1��1��1��1��1��1ij1̳1Գ1س1ܳ1�1�1�1�1�1<�1D�1L�1P�1T�1x�1��1��1��1��1��1��1Ĵ1ȴ1̴1�1��1�1�1�1,�14�1<�1@�1D�1h�1p�1x�1|�1��1��1��1��1��1��1�1�1�1�1��1�1$�1,�10�14�1X�1`�1h�1l�1p�1��1��1��1��1��1ж1ض1�1�1�1�1�1�1 �1$�1H�1P�1X�1\�1`�1��1��1��1��1��1��1ȷ1з1Է1ط1��1�1�1�1�18�1@�1H�1L�1P�1t�1|�1��1��1��1��1��1��1ĸ1ȸ1�1�1��1�1�1(�10�18�1<�1@�1d�1l�1t�1x�1|�1��1��1��1��1��1ܹ1�1�1�1�1�1 �1(�1,�10�1T�1\�1d�1h�1l�1��1��1��1��1��1̺1Ժ1ܺ1�1�1�1�1�1�1 �1D�1L�1T�1X�1\�1��1��1��1��1��1��1Ļ1̻1л1Ի1��1�1�1�1�14�1<�1D�1H�1L�1p�1x�1��1��1��1��1��1��1��1ļ1�1�1��1��1�1$�1,�14�18�1<�1`�1h�1p�1t�1x�1��1��1��1��1��1ؽ1�1�1�1�1�1�1$�1(�1,�1P�1X�1`�1d�1h�1��1��1��1��1��1Ⱦ1о1ؾ1ܾ1�1�1�1�1�1�1@�1H�1P�1T�1X�1|�1��1��1��1��1��1��1ȿ1̿1п1�1��1�1�1�10�18�1@�1D�1H�1l�1t�1|�1��1��1��1��1��1��1��1�1�1�1��1��1 �1(�10�14�18�1\�1d�1l�1p�1t�1��1��1��1��1��1�1�1�1�1�1�1�1 �1$�1(�1L�1T�1\�1`�1d�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1�1<�1D�1L�1P�1T�1x�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1,�14�1<�1@�1D�1h�1p�1x�1|�1��1��1��1��1��1��1��1��1��1��1��1�1$�1,�10�14�1X�1`�1h�1l�1p�1��1��1��1��1��1��1��1��1��1��1�1�1�1 �1$�1H�1P�1X�1\�1`�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1�18�1@�1H�1L�1P�1t�1|�1��1��1��1��1��1��1��1��1��1��1��1�1�1(�10�18�1<�1@�1d�1l�1t�1x�1|�1��1��1��1��1��1��1��1��1��1��1�1 �1(�1,�10�1T�1\�1d�1h�1l�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1 �1D�1L�1T�1X�1\�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1�14�1<�1D�1H�1L�1p�1x�1��1��1��1��1��1��1��1��1��1��1��1��1�1$�1,�14�18�1<�1`�1h�1p�1t�1x�1��1��1��1��1��1��1��1��1��1��1�1�1$�1(�1,�1P�1X�1`�1d�1h�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1�1@�1H�1P�1T�1X�1|�1��1��1��1��1��1��1��1��1��1��1��1�1�1�10�18�1@�1D�1H�1l�1t�1|�1��1��1��1��1��1��1��1��1��1��1��1��1 �1(�10�14�18�1\�1d�1l�1p�1t�1��1��1��1��1��1��1��1��1��1��1�1�1 �1$�1(�1L�1T�1\�1`�1d�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1�1<�1D�1L�1P�1T�1x�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1,�14�1<�1@�1D�1h�1p�1x�1|�1��1��1��1��1��1��1��1��1��1��1��1�1$�1,�10�14�1X�1`�1h�1l�1p�1��1��1��1��1��1��1��1��1��1��1�1�1�1 �1$�1H�1P�1X�1\�1`�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1�18�1@�1H�1L�1P�1t�1|�1��1��1��1��1��1��1��1��1��1��1��1�1�1(�10�18�1<�1@�1d�1l�1t�1x�1|�1��1��1��1��1��1��1��1��1��1��1�1 �1(�1,�10�1T�1\�1d�1h�1l�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1 �1D�1L�1T�1X�1\�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1�14�1<�1D�1H�1L�1p�1x�1��1��1��1��1��1��1��1��1��1��1��1��1�1$�1,�14�18�1<�1`�1h�1p�1t�1x�1��1��1��1��1��1��1��1��1��1��1�1�1$�1(�1,�1P�1X�1`�1d�1h�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1�1@�1H�1P�1T�1X�1|�1��1��1��1��1��1��1��1��1��1��1��1�1�1�10�18�1@�1D�1H�1l�1t�1|�1��1��1��1��1��1��1��1��1��1��1��1��1 �1(�10�14�18�1\�1d�1l�1p�1t�1��1��1��1��1��1��1��1��1��1��1�1�1 �1$�1(�1L�1T�1\�1`�1d�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1�1<�1D�1L�1P�1T�1x�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1,�14�1<�1@�1D�1h�1p�1x�1|�1��1��1��1��1��1��1��1��1��1��1��1�1$�1,�10�14�1X�1`�1h�1l�1p�1��1��1��1��1��1��1��1��1��1��1�1�1�1 �1$�1H�1P�1X�1\�1`�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1�18�1@�1H�1L�1P�1t�1|�1��1��1��1��1��1��1��1��1��1��1��1�1�1(�10�18�1<�1@�1d�1l�1t�1x�1|�1��1��1��1��1��1��1��1��1��1��1�1 �1(�1,�10�14�1\�1d�1h�1l�1p�1��1��1��1��1��1��1��1��1��1��1�1�1�1 �1L�1T�1X�1\�1`�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1<�1D�1H�1L�1P�1x�1��1��1��1��1��1��1��1��1��1��1��1��1�1�1,�14�18�1<�1h�1p�1t�1x�1|�1��1��1��1��1��1��1��1��1��1��1�1$�1(�1,�10�1X�1`�1d�1h�1l�1��1��1��1��1��1��1��1��1��1��1�1�1�1�1 �1H�1P�1T�1X�1\�1��1��1��1��1��1��1��1��1��1��1��1�1�1�1�18�1@�1D�1H�1L�1t�1|�1��1��1��1��1��1��1��1��1��1��1(�10�14�18�1d�1l�1p�1t�1��1��1��1��1��1��1��1��1�1 �1$�1(�1T�1\�1`�1d�1��1��1��1��1��1��1��1��1�1�1�1�1D�1L�1P�1T�1��1��1��1��1��1��1��1��1��1�1�1�14�1<�1@�1D�1p�1x�1|�1��1��1��1��1��1��1��1��1��1$�1,�10�14�1`�1h�1l�1p�1��1��1��1��1��1��1��1��1�1�1 �1$�1P�1X�1\�1`�1��1��1��1��1��1��1��1��1�1�1�1�1@�1H�1L�1P�1|�1��1��1��1��1��1��1��1��1��1�1�10�18�1<�1@�1l�1t�1x�1|�1��1��1��1��1�1�1�1�1 �1(�1,�10�1\�1d�1h�1l�1��1��1��1��1�1�1�1�1�1�1�1 �1L�1T�1X�1\�1��1��1��1��1�1�1�1�1�1�1�1�1<�1D�1H�1L�1x�1��1��1��1��1��1��1�1�1��1��1�1,�14�18�1<�1h�1p�1t�1x�1��1��1��1��1�1�1�1�1�1$�1(�1,�1X�1`�1d�1h�1��1��1��1��1�1�1�1�1�1�1�1�1H�1P�1T�1X�1��1��1��1��1��1�1�1�1��1�1�1�18�1@�1D�1H�1t�1|�1��1��1��1��1��1��1�1�1��1��1(�10�14�18�1d�1l�1p�1t�1��1��1��1��1�1�1�1�1�1 �1$�1(�1T�1\�1`�1d�1��1��1��1��1�1�1�1�1�1�1�1�1D�1L�1P�1T�1��1��1��1��1��1�1�1�1��122242<2@2D2p2x2|2�2�2�2�2�2�2�2�2�2$2,20242`2h2l2p2�2�2�2�2�2�2�2�222 2$2P2X2\2`2�2�2�2�2�2�2�2�22222@2H2L2P2|2�2�2�2�2�2�2�2�2�2220282<2@2l2t2x2|2�2�2�2�2�2�2�2�2 2(2,202\2d2h2l2�2�2�2�2�2�2�2�2222 2L2T2X2\2�2�2�2�2�2�2�2�22222<2D2H2L2x2�2�2�2�2�2�2�2�2�2�22,24282<2h2p2t2x2�2�2�2�2�2�2�2�2	2$	2(	2,	2X	2`	2d	2h	2�	2�	2�	2�	2�	2�	2�	2�	2
2
2
2
2H
2P
2T
2X
2�
2�
2�
2�
2�
2�
2�
2�
2�
222282@2D2H2t2|2�2�2�2�2�2�2�2�2�2�2(2024282d2l2p2t2�2�2�2�2�2�2�2�2
2 
2$
2(
2T
2\
2`
2d
2�
2�
2�
2�
2�
2�
2�
2�
22222D2L2P2T2�2�2�2�2�2�2�2�2�222242<2@2D2p2x2|2�2�2�2�2�2�2�2�2�2$2,20242`2h2l2p2�2�2�2�2�2�2�2�222 2$2P2X2\2`2�2�2�2�2�2�2�2�2�222222@2H2L2P2T2|2�2�2�2�2�2�2�2�2�2�2�22220282<2@2D2l2t2x2|2�2�2�2�2�2�2�2�2�2�2�2 2(2,20242\2d2h2l2p2�2�2�2�2�2�2�2�2�2222 2L2T2X2\2�2�2�2�2�2�2�2�222222<2D2H2L2x2�2�2�2�2�2�2�2�2�2�2�222,24282<2h2p2t2x2|2�2�2�2�2�2�2�2�22$2(2,2X2`2d2h2�2�2�2�2�2�2�2�22222H2P2T2X2�2�2�2�2�2�2�2�2�2�2�2222282@2D2H2L2t2|2�2�2�2�2�2�2�2�2�2�2�22(2024282<2d2l2p2t2x2�2�2�2�2�2�2�2�2�22 2$2(2,2T2\2`2d2h2�2�2�2�2�2�2�2�2�22222D2L2P2T2X2�2�2�2�2�2�2�2�2�2�2�2222242<2@2D2H2p2x2|2�2�2�2�2�2�2�2�2�2�2�2$2,2024282`2h2l2p2�2�2�2�2�2�2�2�2�2�2 2 2  2$ 2( 2P 2X 2\ 2` 2d 2� 2� 2� 2� 2� 2� 2� 2� 2� 2!2!2!2!2!2@!2H!2L!2P!2T!2|!2�!2�!2�!2�!2�!2�!2�!2�!2�!2�!2�!2"2"2"20"28"2<"2@"2D"2l"2t"2x"2|"2�"2�"2�"2�"2�"2�"2�"2�"2�"2�"2�"2 #2(#2,#20#24#2\#2d#2h#2l#2�#2�#2�#2�#2�#2�#2�#2�#2$2$2$2 $2L$2T$2X$2\$2�$2�$2�$2�$2�$2�$2�$2�$2�$2�$2%2%2%2%2%2<%2D%2H%2L%2P%2x%2�%2�%2�%2�%2�%2�%2�%2�%2�%2�%2�%2�%2&2&2,&24&28&2<&2@&2h&2p&2t&2x&2�&2�&2�&2�&2�&2�&2�&2�&2�&2�&2'2$'2('2,'20'2X'2`'2d'2h'2l'2�'2�'2�'2�'2�'2�'2�'2�'2�'2�'2(2(2(2(2$(2((2,(2H(2P(2T(2X(2`(2d(2h(2�(2�(2�(2�(2�(2�(2�(2�(2�(2�(2�(2�(2�(2�(2�(2)2)2)2)2)28)2@)2D)2H)2P)2X)2`)2d)2t)2|)2�)2�)2�)2�)2�)2�)2�)2�)2�)2�)2�)2�)2�)2�)2�)2�)2*2*2(*20*24*28*2@*2H*2d*2l*2p*2t*2|*2�*2�*2�*2�*2�*2�*2�*2�*2�*2�*2�*2�*2�*2�*2�*2+2 +2$+2(+20+24+28+2T+2\+2`+2d+2l+2p+2t+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2�+2,2,2,2,2 ,2(,2D,2L,2P,2T,2\,2`,2d,2l,2p,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2�,2-2-2-2-2-2-24-2<-2@-2D-2H-2L-2T-2p-2x-2|-2�-2�-2�-2�-2�-2�-2�-2�-2�-2�-2�-2�-2�-2�-2�-2�-2.2.2$.2,.20.24.28.2<.2D.2`.2h.2l.2p.2t.2x.2�.2�.2�.2�.2�.2�.2�.2�.2�.2�.2�.2�.2�.2�.2�.2/2/2 /2$/2(/2,/24/2P/2X/2\/2`/2d/2h/2p/2�/2�/2�/2�/2�/2�/2�/2�/2�/2�/2�/2�/2�/2�/2020202020202$02@02H02L02P02T02X02`02|02�02�02�02�02�02�02�02�02�02�02�02�02�02�0212121212012812<12@12H12P12l12t12x12|12�12�12�12�12�12�12�12�12�12�12�12�12�1222 22(22,22022822@22\22d22h22l22t22x22|22�22�22�22�22�22�22�22�22�22�22�22�22�22�22323232 32(32,32032L32T32X32\32d32l32�32�32�32�32�32�32�32�32�32�32�32�324242424242 42<42D42H42L42T42\42x42�42�42�42�42�42�42�42�42�42�42�42�42�42�42�42�4252525252,52452852<52D52H52L52h52p52t52x52�52�52�52�52�52�52�52�52�52�52�52�52�52�52�52�5262626262$62(62,62462862<62D62H62X62`62d62h62p62t62x62�62�62�62�62�62�62�62�62�62�62�62�62�62�6272727272$72(72,72H72P72T72X72`72d72h72p72t72�72�72�72�72�72�72�72�72�72�72�72�72�72�72�72�72�72828282828282882@82D82H82L82P82X82t82|82�82�82�82�82�82�82�82�82�82�82�82�82�82�82�82�82929292(92092492892<92@92H92d92l92p92t92x92|92�92�92�92�92�92�92�92�92�92�92�92�92�92�92�92:2 :2$:2(:2,:20:28:2T:2\:2`:2d:2h:2l:2t:2�:2�:2�:2�:2�:2�:2�:2�:2�:2�:2�:2�:2�:2�:2;2;2;2;2;2 ;2(;2D;2L;2P;2T;2X;2\;2d;2�;2�;2�;2�;2�;2�;2�;2�;2�;2�;2�;2�;2�;2�;2�;2<2<2<2<2<2<24<2<<2@<2D<2H<2L<2T<2p<2x<2|<2�<2�<2�<2�<2�<2�<2�<2�<2�<2�<2�<2�<2�<2�<2�<2�<2=2=2$=2,=20=24=28=2<=2D=2`=2h=2l=2p=2t=2x=2�=2�=2�=2�=2�=2�=2�=2�=2�=2�=2�=2�=2�=2�=2�=2>2>2 >2$>2(>2,>24>2P>2X>2\>2`>2d>2h>2p>2�>2�>2�>2�>2�>2�>2�>2�>2�>2�>2�>2�>2�>2�>2?2?2?2?2?2?2$?2@?2H?2L?2P?2T?2X?2`?2|?2�?2�?2�?2�?2�?2�?2�?2�?2�?2�?2�?2�?2�?2�?2�?2@2@2@2@2@20@28@2<@2@@2D@2H@2P@2l@2t@2x@2|@2�@2�@2�@2�@2�@2�@2�@2�@2�@2�@2�@2�@2�@2�@2�@2�@2A2 A2(A2,A20A24A28A2@A2\A2dA2hA2lA2pA2tA2|A2�A2�A2�A2�A2�A2�A2�A2�A2�A2�A2�A2�A2�A2�A2B2B2B2 B2$B2(B20B2LB2TB2XB2\B2`B2dB2lB2�B2�B2�B2�B2�B2�B2�B2�B2�B2�B2�B2�B2�B2�B2C2C2C2C2C2C2 C2<C2DC2HC2LC2PC2TC2\C2xC2�C2�C2�C2�C2�C2�C2�C2�C2�C2�C2�C2�C2�C2�C2�C2�C2D2D2D2D2,D24D28D2<D2@D2DD2LD2hD2pD2tD2xD2|D2�D2�D2�D2�D2�D2�D2�D2�D2�D2�D2�D2�D2�D2�D2�D2E2E2$E2(E2,E20E24E2<E2XE2`E2dE2hE2lE2pE2xE2�E2�E2�E2�E2�E2�E2�E2�E2�E2�E2�E2�E2�E2�E2F2F2F2F2 F2$F2,F2HF2PF2TF2XF2\F2`F2hF2�F2�F2�F2�F2�F2�F2�F2�F2�F2�F2�F2�F2�F2�F2�F2G2G2G2G2G2G28G2@G2DG2HG2LG2PG2XG2tG2|G2�G2�G2�G2�G2�G2�G2�G2�G2�G2�G2�G2�G2�G2�G2�G2�G2H2H2H2(H20H24H28H2<H2@H2HH2dH2lH2pH2tH2xH2|H2�H2�H2�H2�H2�H2�H2�H2�H2�H2�H2�H2�H2�H2�H2�H2I2 I2$I2(I2,I20I28I2TI2\I2`I2dI2hI2lI2tI2�I2�I2�I2�I2�I2�I2�I2�I2�I2�I2�I2�I2�I2�I2J2J2J2J2J2 J2(J2DJ2LJ2PJ2TJ2XJ2\J2dJ2�J2�J2�J2�J2�J2�J2�J2�J2�J2�J2�J2�J2�J2�J2�J2K2K2K2K2K2K24K2<K2@K2DK2HK2LK2TK2pK2xK2|K2�K2�K2�K2�K2�K2�K2�K2�K2�K2�K2�K2�K2�K2�K2�K2�K2L2L2$L2,L20L24L28L2<L2DL2`L2hL2lL2pL2tL2xL2�L2�L2�L2�L2�L2�L2�L2�L2�L2�L2�L2�L2�L2�L2�L2M2M2 M2$M2(M2,M24M2PM2XM2\M2`M2dM2hM2pM2�M2�M2�M2�M2�M2�M2�M2�M2�M2�M2�M2�M2�M2�M2N2N2N2N2N2N2$N2@N2HN2LN2PN2TN2XN2`N2|N2�N2�N2�N2�N2�N2�N2�N2�N2�N2�N2�N2�N2�N2�N2�N2O2O2O2O2O20O28O2<O2@O2DO2HO2PO2lO2tO2xO2|O2�O2�O2�O2�O2�O2�O2�O2�O2�O2�O2�O2�O2�O2�O2�O2�O2P2 P2(P2,P20P24P28P2@P2\P2dP2hP2lP2pP2tP2|P2�P2�P2�P2�P2�P2�P2�P2�P2�P2�P2�P2�P2�P2�P2Q2Q2Q2 Q2$Q2(Q20Q2LQ2TQ2XQ2\Q2`Q2dQ2lQ2�Q2�Q2�Q2�Q2�Q2�Q2�Q2�Q2�Q2�Q2�Q2�Q2�Q2�Q2R2R2R2R2R2R2 R2<R2DR2HR2LR2PR2TR2\R2xR2�R2�R2�R2�R2�R2�R2�R2�R2�R2�R2�R2�R2�R2�R2�R2�R2S2S2S2S2,S24S28S2<S2@S2DS2LS2hS2pS2tS2xS2|S2�S2�S2�S2�S2�S2�S2�S2�S2�S2�S2�S2�S2�S2�S2�S2T2T2$T2(T2,T20T24T2<T2XT2`T2dT2hT2lT2pT2xT2�T2�T2�T2�T2�T2�T2�T2�T2�T2�T2�T2�T2�T2�T2U2U2U2U2 U2$U2,U2HU2PU2TU2XU2\U2`U2hU2�U2�U2�U2�U2�U2�U2�U2�U2�U2�U2�U2�U2�U2�U2�U2V2V2V2V2V2V28V2@V2DV2HV2LV2PV2XV2tV2|V2�V2�V2�V2�V2�V2�V2�V2�V2�V2�V2�V2�V2�V2�V2�V2�V2W2W2W2(W20W24W28W2<W2@W2HW2dW2lW2pW2tW2xW2|W2�W2�W2�W2�W2�W2�W2�W2�W2�W2�W2�W2�W2�W2�W2�W2X2 X2$X2(X2,X20X28X2TX2\X2`X2dX2hX2lX2tX2�X2�X2�X2�X2�X2�X2�X2�X2�X2�X2�X2�X2�X2�X2Y2Y2Y2Y2Y2 Y2(Y2DY2LY2PY2TY2XY2\Y2dY2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2�Y2Z2Z2Z2Z2Z2Z24Z2<Z2@Z2DZ2HZ2LZ2TZ2pZ2xZ2|Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2�Z2[2[2$[2,[20[24[28[2<[2D[2`[2h[2l[2p[2t[2x[2�[2�[2�[2�[2�[2�[2�[2�[2�[2�[2�[2�[2�[2�[2�[2\2\2 \2$\2(\2,\24\2P\2X\2\\2`\2d\2h\2p\2�\2�\2�\2�\2�\2�\2�\2�\2�\2�\2�\2�\2�\2�\2]2]2]2]2]2]2$]2@]2H]2L]2P]2T]2X]2`]2|]2�]2�]2�]2�]2�]2�]2�]2�]2�]2�]2�]2^2^2^20^28^2<^2@^2D^2l^2t^2x^2|^2�^2�^2�^2�^2�^2�^2�^2�^2�^2�^2�^2 _2(_2,_20_24_2\_2d_2h_2l_2p_2�_2�_2�_2�_2�_2�_2�_2�_2�_2�_2`2`2`2 `2$`2L`2T`2X`2\`2``2�`2�`2�`2�`2�`2�`2�`2�`2�`2�`2a2a2a2a2a2<a2Da2Ha2La2Pa2xa2�a2�a2�a2�a2�a2�a2�a2�a2�a2�a2�a2�a2b2,b24b28b2<b2@b2hb2pb2tb2xb2|b2�b2�b2�b2�b2�b2�b2�b2�b2�b2�b2c2$c2(c2,c20c2Xc2`c2dc2hc2lc2�c2�c2�c2�c2�c2�c2�c2�c2�c2�c2Hd2Pd2Td2Xd2\d2�d2�d2�d2�d2�d2�d2�d2�d2�d2�d2�d2e2e2e2e28e2@e2De2He2Le2te2|e2�e2�e2�e2�e2�e2�e2�e2�e2�e2�e2�e2�e2f2(f20f24f28f2<f2df2lf2pf2tf2xf2�f2�f2�f2�f2�f2�f2�f2�f2�f2�f2g2 g2$g2(g2,g2Tg2\g2`g2dg2hg2�g2�g2�g2�g2�g2�g2�g2�g2�g2�g2h2h2h2h2h2Dh2Lh2Ph2Th2Xh2�h2�h2�h2�h2�h2�h2�h2�h2�h2�h2�h2i2i2i2i24i2<i2@i2Di2Hi2pi2xi2|i2�i2�i2�i2�i2�i2�i2�i2�i2�i2�i2�i2�i2$j2,j20j24j28j2`j2hj2lj2pj2tj2�j2�j2�j2�j2�j2�j2�j2�j2�j2�j2k2k2 k2$k2(k2Pk2Xk2\k2`k2dk2�k2�k2�k2�k2�k2�k2�k2�k2�k2�k2l2l2l2l2l2@l2Hl2Ll2Pl2Tl2|l2�l2�l2�l2�l2�l2�l2�l2�l2�l2�l2�l2m2m2m20m28m2<m2@m2Dm2lm2tm2xm2|m2�m2�m2�m2�m2�m2�m2�m2�m2�m2�m2�m2 n2(n2,n20n24n2\n2dn2hn2ln2pn2�n2�n2�n2�n2�n2�n2�n2�n2o2o2o2 o2Lo2To2Xo2\o2�o2�o2�o2�o2�o2�o2�o2�o2p2p2p2p2<p2Dp2Hp2Lp2xp2�p2�p2�p2�p2�p2�p2�p2�p2�p2�p2q2,q24q28q2<q2hq2pq2tq2xq2�q2�q2�q2�q2�q2�q2�q2�q2r2$r2(r2,r2Xr2`r2dr2hr2�r2�r2�r2�r2�r2�r2�r2�r2s2s2s2s2Hs2Ps2Ts2Xs2�s2�s2�s2�s2�s2�s2�s2�s2�s2t2t2t28t2@t2Dt2Ht2tt2|t2�t2�t2�t2�t2�t2�t2�t2�t2�t2�t2(u20u24u28u2du2lu2pu2tu2�u2�u2�u2�u2�u2�u2�u2�u2v2 v2$v2(v2Tv2\v2`v2dv2�v2�v2�v2�v2�v2�v2�v2�v2w2w2w2w2Dw2Lw2Pw2Tw2�w2�w2�w2�w2�w2�w2�w2�w2�w2�w2x2x2x2x24x2<x2@x2Dx2Hx2px2xx2|x2�x2�x2�x2�x2�x2�x2�x2�x2�x2�x2�x2�x2$y2,y20y24y28y2`y2hy2ly2py2ty2�y2�y2�y2�y2�y2�y2�y2�y2�y2�y2z2z2 z2$z2(z2Pz2Xz2\z2`z2dz2�z2�z2�z2�z2�z2�z2�z2�z2�z2�z2{2{2{2{2{2@{2H{2L{2P{2T{2|{2�{2�{2�{2�{2�{2�{2�{2�{2�{2�{2�{2|2|2|20|28|2<|2@|2D|2l|2t|2x|2||2�|2�|2�|2�|2�|2�|2�|2�|2�|2�|2�|2 }2(}2,}20}24}2\}2d}2h}2l}2p}2�}2�}2�}2�}2�}2�}2�}2�}2�}2�}2~2~2~2 ~2$~2L~2T~2X~2\~2`~2�~2�~2�~2�~2�~2�~2�~2�~2�~2�~222222<2D2H2L2P2x2�2�2�2�2�2�2�2�2�2�2�2�2�2�2,�24�28�2<�2@�2h�2p�2t�2x�2|�2��2��2��2��2��2�2�2�2�2�2�2$�2(�2,�20�2X�2`�2d�2h�2l�2��2��2��2��2��2Ё2؁2܁2�2�2�2�2�2�2 �2H�2P�2T�2X�2\�2��2��2��2��2��2��2Ȃ2̂2Ђ2Ԃ2��2�2�2�2�28�2@�2D�2H�2L�2t�2|�2��2��2��2��2��2��2��2ă2�2�2��2��2�2(�20�24�28�2<�2d�2l�2p�2t�2x�2��2��2��2��2��2܄2�2�2�2�2�2 �2$�2(�2,�2T�2\�2`�2d�2h�2��2��2��2��2��2؅2܅2�2��2��2�2�20�24�28�2P�2T�2l�2p�2t�2��2��2��2��2��2Ȇ2̆2�2�2�2�2�2 �2$�2(�2@�2D�2H�2\�2`�2d�2|�2��2��2��2��2��2��2ԇ2؇2܇2�2��2��2�2�2�20�24�2L�2P�2T�2l�2p�2��2��2��2��2��2Ĉ2Ȉ2̈2�2�2�2�2�2 �2$�2<�2@�2D�2\�2`�2x�2|�2��2��2��2��2��2��2ԉ2؉2�2�2��2�2�2,�20�24�2L�2P�2h�2l�2p�2��2��2��2��2��2��2Ċ2Ȋ2̊2�2�2�2�2�2�2�2 �2$�2<�2@�2D�2X�2\�2`�2x�2|�2��2��2��2��2��2��2��2Ћ2ԋ2؋2�2�2�2�2�2,�20�2H�2L�2P�2h�2l�2��2��2��2��2��2��2Č2Ȍ2�2�2��2�2�2�2 �28�2<�2@�2X�2\�2t�2x�2|�2��2��2��2��2��2Ѝ2ԍ2�2�2�2�2�2(�2,�20�2H�2L�2P�2d�2h�2l�2��2��2��2��2��2��2Ď2܎2�2�2��2�2�2�2 �28�2<�2T�2X�2\�2t�2x�2��2��2��2��2��2̏2Џ2ԏ2�2�2�2�2�2(�2,�2D�2H�2L�2d�2h�2��2��2��2��2��2��2��2Đ2ܐ2�2��2��2�2�2�24�28�2<�2T�2X�2p�2t�2x�2��2��2��2��2��2��2̑2Б2�2�2�2�2�2$�2(�2,�2D�2H�2`�2d�2h�2��2��2��2��2��2��2��2Ē2ؒ2ܒ2�2��2��2�2�2�2�24�28�2<�2P�2T�2X�2p�2t�2��2��2��2��2��2ȓ2̓2Г2�2�2�2�2�2$�2(�2@�2D�2H�2`�2d�2|�2��2��2��2��2��2��2��2ؔ2ܔ2�2��2��2�2�20�24�28�2P�2T�2l�2p�2t�2��2��2��2��2��2ȕ2̕2�2�2�2�2�2 �2$�2(�2@�2D�2\�2`�2d�2|�2��2��2��2��2��2��2��2��2�2�2�2��2$�2,�20�24�28�2`�2h�2l�2p�2t�2��2��2��2��2��2ؗ2�2�2�2�2�2�2 �2$�2(�2P�2X�2\�2`�2d�2��2��2��2��2��2Ș2И2Ԙ2ؘ2ܘ2�2�2�2�2�2@�2H�2L�2P�2T�2|�2��2��2��2��2��2��2ę2ș2̙2�2��2�2�2�20�28�2<�2@�2D�2l�2t�2x�2|�2��2��2��2��2��2�2�2�2�2 �2(�2,�20�2\�2d�2h�2l�2��2��2��2��2ԛ2ܛ2�2�2�2�2�2 �2L�2T�2X�2\�2��2��2��2��2Ĝ2̜2М2Ԝ2؜2�2�2�2�2�2<�2D�2H�2L�2P�2x�2��2��2��2��2��2��2��2ĝ2ȝ2�2��2��2�2�2,�24�28�2<�2@�2h�2p�2t�2x�2|�2��2��2��2��2��2�2�2�2�2�2�2$�2(�2,�20�2X�2`�2d�2h�2l�2��2��2��2��2��2П2؟2ܟ2�2�2�2�2�2�2 �2H�2P�2T�2X�2\�2��2��2��2��2��2��2Ƞ2̠2Р2Ԡ2��2�2�2�2�28�2@�2D�2H�2L�2t�2|�2��2��2��2��2��2��2��2�2�2��2��2(�20�24�28�2d�2l�2p�2t�2��2��2��2��2ܢ2�2�2�2�2 �2$�2(�2T�2\�2`�2d�2��2��2��2��2̣2ԣ2أ2ܣ2�2�2�2�2D�2L�2P�2T�2��2��2��2��2��2Ĥ2Ȥ2̤2��2�2�2�24�2<�2@�2D�2p�2x�2|�2��2��2��2��2��2�2�2�2��2$�2,�20�24�2`�2h�2l�2p�2��2��2��2��2ئ2�2�2�2�2�2 �2$�2P�2X�2\�2`�2d�2��2��2��2��2��2ȧ2Ч2ԧ2ا2ܧ2�2�2�2�2�2@�2H�2L�2P�2T�2|�2��2��2��2��2��2��2Ĩ2Ȩ2̨2�2��2�2�2�20�28�2<�2@�2D�2l�2t�2x�2|�2��2��2��2��2��2��2�2�2�2�2��2 �2(�2,�20�24�2\�2d�2h�2l�2p�2��2��2��2��2��2Ԫ2ܪ2�2�2�2�2�2�2 �2$�2L�2T�2X�2\�2`�2��2��2��2��2��2ī2̫2Ы2ԫ2ث2�2�2�2�2�2<�2D�2H�2L�2P�2x�2��2��2��2��2��2��2��2Ĭ2Ȭ2�2��2��2�2�2,�24�28�2<�2h�2p�2t�2x�2��2��2��2��2�2�2�2�2�2$�2(�2,�2X�2`�2d�2h�2��2��2��2��2Ю2خ2ܮ2�2�2�2�2�2�2 �2H�2P�2T�2X�2\�2��2��2��2��2��2��2ȯ2̯2Я2ԯ2��2�2�2�2�28�2@�2D�2H�2L�2t�2|�2��2��2��2��2��2��2��2İ2�2�2��2��2(�20�24�28�2<�2d�2l�2p�2t�2x�2��2��2��2��2��2ܱ2�2�2�2�2�2 �2$�2(�2,�2T�2\�2`�2d�2h�2��2��2��2��2��2̲2Բ2ز2ܲ2�2�2�2�2�2�2D�2L�2P�2T�2X�2��2��2��2��2��2��2ij2ȳ2̳2г2��2�2�2�2�24�2<�2@�2D�2H�2p�2x�2|�2��2��2��2��2��2��2��2�2�2�2��2��2$�2,�20�24�28�2`�2h�2l�2p�2��2��2��2��2��2ص2�2�2�2�2�2�2 �2$�2(�2P�2X�2\�2`�2d�2��2��2��2��2��2ȶ2ж2Զ2ض2ܶ2�2�2�2�2�2@�2H�2L�2P�2T�2|�2��2��2��2��2��2��2ķ2ȷ2̷2�2��2�2�2�20�28�2<�2@�2D�2l�2t�2x�2|�2��2��2��2��2��2��2�2�2�2�2��2 �2(�2,�20�24�2\�2d�2h�2l�2p�2��2��2��2��2��2Թ2ܹ2�2�2�2�2�2�2 �2$�2L�2T�2X�2\�2`�2��2��2��2��2��2ĺ2̺2к2Ժ2غ2�2�2�2�2�2<�2D�2H�2L�2x�2��2��2��2��2��2��2��2Ļ2Ȼ2�2��2��2�2�2,�24�28�2<�2@�2h�2p�2t�2x�2|�2��2��2��2��2��2�2�2�2�2�2�2$�2(�2,�20�2X�2`�2d�2h�2l�2��2��2��2��2��2н2ؽ2ܽ2�2�2�2�2�2�2 �2H�2P�2T�2X�2\�2��2��2��2��2��2��2Ⱦ2̾2о2Ծ2��2�2�2�2�28�2@�2D�2H�2L�2t�2|�2��2��2��2��2��2��2��2Ŀ2�2�2��2��2�2(�20�24�28�2<�2d�2l�2p�2t�2x�2��2��2��2��2��2�2�2�2�2�2�2 �2$�2(�2,�2T�2\�2`�2d�2h�2��2��2��2��2��2�2�2�2�2�2D�2L�2P�2T�2X�2��2��2��2��2��2��2��2��2��2��2��2�2�2�2�24�2<�2@�2D�2H�2p�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2��2$�2,�20�24�28�2`�2h�2l�2p�2t�2��2��2��2��2��2��2��2��2��2��2�2�2 �2$�2(�2P�2X�2\�2`�2d�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2@�2H�2L�2P�2T�2|�2��2��2��2��2��2��2��2��2��2��2��2�2�2�20�28�2<�2@�2D�2l�2t�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2 �2(�2,�20�24�2\�2d�2h�2l�2p�2��2��2��2��2��2��2��2��2��2��2�2�2�2 �2L�2T�2X�2\�2`�2��2��2��2��2��2��2��2��2��2�2�2�2�2�2<�2D�2H�2L�2P�2x�2��2��2��2��2��2��2��2��2��2��2��2��2�2�2,�24�28�2<�2@�2h�2p�2t�2x�2|�2��2��2��2��2��2��2��2��2��2��2�2$�2(�2,�20�2X�2`�2d�2h�2l�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2 �2H�2P�2T�2X�2\�2��2��2��2��2��2��2��2��2��2��2��2�2�2�2�28�2@�2D�2H�2L�2t�2|�2��2��2��2��2��2��2��2��2��2��2��2�2(�20�24�28�2<�2d�2l�2p�2t�2x�2��2��2��2��2��2��2��2��2��2��2�2 �2$�2(�2,�2T�2\�2`�2d�2h�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2D�2L�2P�2T�2X�2��2��2��2��2��2��2��2��2��2��2��2�2�2�2�24�2<�2@�2D�2H�2p�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2��2$�2,�20�24�28�2`�2h�2l�2p�2t�2��2��2��2��2��2��2��2��2��2��2�2�2 �2$�2(�2P�2X�2\�2`�2d�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2@�2H�2L�2P�2T�2|�2��2��2��2��2��2��2��2��2��2��2��2�2�2�20�28�2<�2@�2D�2l�2t�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2 �2(�2,�20�24�2\�2d�2h�2l�2p�2��2��2��2��2��2��2��2��2��2��2�2�2�2 �2$�2L�2T�2X�2\�2`�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2<�2D�2H�2L�2P�2x�2��2��2��2��2��2��2��2��2��2��2��2��2�2�2,�24�28�2<�2@�2h�2p�2t�2x�2|�2��2��2��2��2��2��2��2��2��2�2$�2(�2,�20�2X�2`�2d�2h�2l�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2 �2H�2P�2T�2X�2\�2��2��2��2��2��2��2��2��2��2��2��2�2�2�2�28�2@�2D�2H�2L�2t�2|�2��2��2��2��2��2��2��2��2��2��2��2��2�2(�20�24�28�2<�2d�2l�2p�2t�2x�2��2��2��2��2��2��2��2��2��2��2�2 �2$�2(�2,�2T�2\�2`�2d�2h�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2D�2L�2P�2T�2X�2��2��2��2��2��2��2��2��2��2��2��2�2�2�2�24�2<�2@�2D�2H�2p�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2��2$�2,�20�24�28�2`�2h�2l�2p�2t�2��2��2��2��2��2��2��2��2��2��2�2�2 �2$�2(�2P�2X�2\�2`�2d�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2@�2H�2L�2P�2T�2|�2��2��2��2��2��2��2��2��2��2��2��2�2�2�20�28�2<�2@�2D�2l�2t�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2 �2(�2,�20�24�2\�2d�2h�2l�2p�2��2��2��2��2��2��2��2��2��2�2�2�2 �2L�2T�2X�2\�2��2��2��2��2��2��2��2��2�2�2�2�2<�2D�2H�2L�2x�2��2��2��2��2��2��2��2��2��2��2��2��2�2�2,�24�28�2<�2@�2h�2p�2t�2x�2|�2��2��2��2��2��2��2��2��2��2��2�2$�2(�2,�20�2X�2`�2d�2h�2l�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2H�2P�2T�2X�2\�2��2��2��2��2��2��2��2��2��2��2��2�2�2�2�28�2@�2D�2H�2L�2t�2|�2��2��2��2��2��2��2��2��2��2��2��2��2�2(�20�24�28�2<�2d�2l�2p�2t�2x�2��2��2��2��2��2��2��2��2��2��2�2 �2$�2(�2,�2T�2\�2`�2d�2h�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2D�2L�2P�2T�2��2��2��2��2��2��2��2��2��2��2��2�2�2�2�24�2<�2@�2D�2H�2p�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2��2$�2,�20�24�28�2`�2h�2l�2p�2t�2��2��2��2��2��2��2��2��2��2��2�2�2 �2$�2(�2P�2X�2\�2`�2d�2��2��2��2��2��2��2��2��2��2��2�2�2�2�2�2@�2H�2L�2P�2T�2|�2��2��2��2��2��2��2��2��2��2��2��2�2�2�20�28�2<�2@�2D�2l�2t�2x�2|�2��2��2��2��2��2��2��2��2��2��2��2 �2(�2,�20�2\�2d�2h�2l�2p�2��2��2��2��2��2�2�2�2�2�2�2�2�2 �2$�2L�2T�2X�2\�2`�2��2��2��2��2��2�2�2�2�2�2�2�2(�2,�20�2H�2L�2d�2h�2l�2��2��2��2��2��2��2�2�2�2�2��2�2�2�2 �28�2<�2T�2X�2\�2t�2x�2��2��2��2��2��2�2�2�2�2�2�2�2�2(�2,�2D�2H�2L�2d�2h�2��2��2��2��2��2��2��2�2�2�2��2��2�2�2�24�28�2<�2T�2X�2p�2t�2x�2��2��2��2��2��2�2�2�2�2�2�2�2$�2(�2,�2D�2H�2`�2d�2h�2��2��2��2��2��2��2��2�2�2�2�2��2��2�2�2�24�28�2P�2T�2X�2p�2t�2��2��2��2��2��2�2�2�2�2�2�2�2�2$�2(�2@�2D�2H�2`�2d�2|�2��2��2��2��2��2��2��2�2�2�2��2��2�2�20�24�28�2P�2T�2l�2p�2t�2��2��2��2��2��2�2�2�2�2�2�2�2 �2$�2(�2@�2D�2\�2`�2d�2|�2��2��2��2��2��2��2�2�2�2�2�2�2��2��2$3,3034383`3h3l3p3t3�3�3�3�3�3�3�3�3�3�333 3$3(3P3X3\3`3d3�3�3�3�3�3�3�3�3�3�333333@3H3L3P3T3|3�3�3�3�3�3�3�3�3�3�3�33330383<3@3D3l3t3x3|3�3�3�3�3�3�3�3�3�3�3�3 3(3,30343\3d3h3l3p3�3�3�3�3�3333 3$3L3T3X3\3`3�3�3�3�3�3�3�3�3�3�333333<3D3H3L3P3x3�3�3�3�3�3�3�3�3�3�3�3�333,34383<3@3h3p3t3x3|3�3�3�3�3�3�3�3�3�3�33$3(3,303X3`3d3h3l3�3�3�3�3�3�3�3�3�3�3	3	3	3	3 	3H	3P	3T	3X	3\	3�	3�	3�	3�	3�	3�	3�	3�	3�	3�	3�	3
3
3
3
38
3@
3D
3H
3L
3t
3|
3�
3�
3�
3�
3�
3�
3�
3�
3�
3�
3�
3�
33(3034383<3d3l3p3t3x3�3�3�3�3�3�3�3�3�3�33 3$3(3,3T3\3`3d3h3�3�3�3�3�3�3�3�3
3
3
3
3
3D
3L
3P
3T
3X
3�
3�
3�
3�
3�
3�
3�
3�
3�
3�
3�
333343<3@3D3H3p3x3|3�3�3�3�3�3�3�3�3�3�3�3�3$3,3034383`3h3l3p3t3�3�3�3�3�3�3�3�3�3�333 3$3(3P3X3\3`3d3�3�3�3�3�3�3�3�3�3�33333@3H3L3P3|3�3�3�3�3�3�3�3�3�3330383<3@3l3t3x3|3�3�3�3�3�3�3�3�3 3(3,303\3d3h3l3�3�3�3�3�3�3�3�3333 3L3T3X3\3�3�3�3�3�3�3�3�33333<3D3H3L3x3�3�3�3�3�3�3�3�3�3�33,34383<3h3p3t3x3�3�3�3�3�3�3�3�33$3(3,3X3`3d3h3�3�3�3�3�3�3�3�3�33333 3H3P3T3X3\3�3�3�3�3�3�3�3�3�3�3�3333383@3D3H3L3t3|3�3�3�3�3�3�3�3�3�3�3�3�33(3034383<3d3l3p3t3x3�3�3�3�3�3�3�3�3�3�33 3$3(3,3T3\3`3d3h3�3�3�3�3�3�3�3�3�3�333333D3L3P3T3X3�3�3�3�3�3�3�3�3�3�3�3333343<3@3D3H3p3x3|3�3�3�3�3�3�3�3�3�3�3�3�3$3,3034383`3h3l3p3t3�3�3�3�3�3�3�3�3�3�333 3$3(3P3X3\3`3d3�3�3�3�3�3�3�3�3�3�3 3 3 3 3 3@ 3H 3L 3P 3T 3| 3� 3� 3� 3� 3� 3� 3� 3� 3� 3� 3� 3!3!3!30!38!3<!3@!3D!3l!3t!3x!3|!3�!3�!3�!3�!3�!3�!3�!3�!3�!3�!3�!3 "3("3,"30"34"3\"3d"3h"3l"3p"3�"3�"3�"3�"3�"3�"3�"3�"3�"3�"3#3#3#3 #3$#3L#3T#3X#3\#3`#3�#3�#3�#3�#3�#3�#3�#3�#3�#3�#3$3$3$3$3$3<$3D$3H$3L$3P$3x$3�$3�$3�$3�$3�$3�$3�$3�$3�$3�$3�$3�$3%3%3,%34%38%3<%3@%3h%3p%3t%3x%3|%3�%3�%3�%3�%3�%3�%3�%3&3&3&3(&3,&3D&3H&3L&3d&3h&3�&3�&3�&3�&3�&3�&3�&3�&3�&3�&3�&3�&3'3'3'34'38'3<'3T'3X'3p'3t'3x'3�'3�'3�'3�'3�'3�'3�'3�'3�'3�'3(3(3(3$(3((3,(3D(3H(3`(3d(3h(3�(3�(3�(3�(3�(3�(3�(3�(3�(3�(3�(3�(3�(3)3)3)34)38)3P)3T)3X)3p)3t)3�)3�)3�)3�)3�)3�)3�)3�)3�)3�)3*3*3*3$*3(*3@*3D*3H*3`*3d*3|*3�*3�*3�*3�*3�*3�*3�*3�*3�*3�*3�*3�*3+3+30+34+38+3P+3T+3X+3l+3p+3t+3�+3�+3�+3�+3�+3�+3�+3�+3�+3�+3�+3�+3,3,3,3 ,3$,3(,3@,3D,3H,3\,3`,3d,3|,3�,3�,3�,3�,3�,3�,3�,3�,3�,3�,3�,3�,3-3-3-30-34-3L-3P-3T-3l-3p-3�-3�-3�-3�-3�-3�-3�-3�-3�-3�-3.3.3.3 .3$.3<.3@.3D.3\.3`.3x.3|.3�.3�.3�.3�.3�.3�.3�.3�.3�.3�.3�.3/3/3/3,/30/34/3L/3P/3h/3l/3p/3�/3�/3�/3�/3�/3�/3�/3�/3�/3�/3030303 03$03<03@03X03\03`03x03|03�03�03�03�03�03�03�03�03�03�03131313,13013H13L13P13h13l13�13�13�13�13�13�13�13�13�13�13�13232323 23823<23@23X23\23t23x23|23�23�23�23�23�23�23�23�23�23�23�233333(33,33033H33L33d33h33l33�33�33�33�33�33�33�33�33�33�33�3343434343 43843<43@43T43X43\43t43x43|43�43�43�43�43�43�43�43�43�43�43535353(53,53D53H53L53d53h53�53�53�53�53�53�53�53�53�53�53�53�53636363463863<63T63X63p63t63x63�63�63�63�63�63�63�63�63�63�637373$73(73,73D73H73`73d73h73�73�73�73�73�73�73�73�73�73�73�73�73838383483883P83T83X83p83t83�83�83�83�83�83�83�83�83�83�83�83�8393 93$93(93,93T93\93`93d93h93�93�93�93�93�93�93�93�93�93�93:3:3:3:3:3D:3L:3P:3T:3X:3�:3�:3�:3�:3�:3�:3�:3�:3�:3�:3�:3;3;3;3;34;3<;3@;3D;3p;3x;3|;3�;3�;3�;3�;3�;3�;3�;3�;3�;3$<3,<30<34<3`<3h<3l<3p<3�<3�<3�<3�<3�<3�<3�<3�<3=3=3 =3$=3P=3X=3\=3`=3�=3�=3�=3�=3�=3�=3�=3�=3>3>3>3>3@>3H>3L>3P>3|>3�>3�>3�>3�>3�>3�>3�>3�>3�>3?3?30?38?3<?3@?3l?3t?3x?3|?3�?3�?3�?3�?3�?3�?3�?3�?3�?3 @3(@3,@30@34@3\@3d@3h@3l@3p@3�@3�@3�@3�@3�@3�@3�@3�@3�@3�@3A3A3A3 A3$A3LA3TA3XA3\A3`A3�A3�A3�A3�A3�A3�A3�A3�A3�A3�A3B3B3B3B3B3<B3DB3HB3LB3PB3xB3�B3�B3�B3�B3�B3�B3�B3�B3�B3�B3�B3�B3C3C3,C34C38C3<C3@C3hC3pC3tC3xC3|C3�C3�C3�C3�C3�C3�C3�C3�C3�C3�C3D3$D3(D3,D30D3XD3`D3dD3hD3lD3�D3�D3�D3�D3�D3�D3�D3�D3�D3�D3E3E3E3E3 E3HE3PE3TE3XE3\E3�E3�E3�E3�E3�E3�E3�E3�E3�E3�E3F3F3F38F3@F3DF3HF3tF3|F3�F3�F3�F3�F3�F3�F3�F3�F3�F3�F3(G30G34G38G3dG3lG3pG3tG3�G3�G3�G3�G3�G3�G3�G3�G3H3 H3$H3(H3TH3\H3`H3dH3�H3�H3�H3�H3�H3�H3�H3�H3I3I3I3I3DI3LI3PI3TI3�I3�I3�I3�I3�I3�I3�I3�I3�I3J3J3J34J3<J3@J3DJ3pJ3xJ3|J3�J3�J3�J3�J3�J3�J3�J3�J3�J3$K3,K30K34K3`K3hK3lK3pK3�K3�K3�K3�K3�K3�K3�K3�K3L3L3 L3$L3PL3XL3\L3`L3�L3�L3�L3�L3�L3�L3�L3�L3M3M3M3M3@M3HM3LM3PM3|M3�M3�M3�M3�M3�M3�M3�M3�M3�M3N3N30N38N3<N3@N3lN3tN3xN3|N3�N3�N3�N3�N3�N3�N3�N3�N3 O3(O3,O30O34O3\O3dO3hO3lO3pO3�O3�O3�O3�O3�O3�O3�O3�O3�O3�O3P3P3P3 P3$P3LP3TP3XP3\P3`P3�P3�P3�P3�P3�P3�P3�P3�P3�P3�P3Q3Q3Q3Q3Q3<Q3DQ3HQ3LQ3PQ3xQ3�Q3�Q3�Q3�Q3�Q3�Q3�Q3�Q3�Q3�Q3�Q3�Q3R3R3,R34R38R3<R3@R3hR3pR3tR3xR3|R3�R3�R3�R3�R3�R3�R3�R3�R3�R3�R3S3$S3(S3,S30S3XS3`S3dS3hS3lS3�S3�S3�S3�S3�S3�S3�S3�S3�S3�S3T3T3T3T3 T3HT3PT3TT3XT3\T3�T3�T3�T3�T3�T3�T3�T3�T3�T3�T3�T3U3U3U38U3@U3DU3HU3tU3|U3�U3�U3�U3�U3�U3�U3�U3�U3�U3�U3(V30V34V38V3dV3lV3pV3tV3�V3�V3�V3�V3�V3�V3�V3�V3�V3�V3W3 W3$W3(W3,W3TW3\W3`W3dW3hW3�W3�W3�W3�W3�W3�W3�W3�W3�W3�W3X3X3X3X3X3DX3LX3PX3TX3XX3�X3�X3�X3�X3�X3�X3�X3�X3�X3�X3Y3Y3Y3Y34Y3<Y3@Y3DY3HY3pY3xY3|Y3�Y3�Y3�Y3�Y3�Y3�Y3�Y3�Y3�Y3�Y3�Y3�Y3$Z3,Z30Z34Z38Z3`Z3hZ3lZ3pZ3tZ3�Z3�Z3�Z3�Z3�Z3�Z3�Z3�Z3�Z3�Z3[3[3 [3$[3([3P[3X[3\[3`[3d[3�[3�[3�[3�[3�[3�[3�[3�[3�[3�[3\3\3\3\3\3@\3H\3L\3P\3T\3|\3�\3�\3�\3�\3�\3�\3�\3�\3�\3�\3�\3]3]30]38]3<]3@]3D]3l]3t]3x]3|]3�]3�]3�]3�]3�]3�]3�]3�]3�]3�]3�]3 ^3(^3,^30^34^3\^3d^3h^3l^3p^3�^3�^3�^3�^3�^3�^3�^3�^3�^3�^3_3_3_3 _3$_3L_3T_3X_3\_3`_3�_3�_3�_3�_3�_3�_3�_3�_3�_3�_3`3`3`3`3`3<`3D`3H`3L`3P`3x`3�`3�`3�`3�`3�`3�`3�`3�`3�`3�`3�`3�`3a3a3,a34a38a3<a3@a3ha3pa3ta3xa3|a3�a3�a3�a3�a3�a3�a3�a3�a3�a3�a3b3$b3(b3,b30b3Xb3`b3db3hb3�b3�b3�b3�b3�b3�b3�b3�b3�b3�b3c3c3c3c3 c3Hc3Pc3Tc3Xc3\c3�c3�c3�c3�c3�c3�c3�c3�c3�c3�c3�c3d3d3d3d38d3@d3Dd3Hd3Pd3Xd3td3|d3�d3�d3�d3�d3�d3�d3�d3�d3�d3�d3�d3�d3�d3�d3e3e3(e30e34e38e3@e3He3de3le3pe3te3|e3�e3�e3�e3�e3�e3�e3�e3�e3�e3�e3�e3�e3�e3f3 f3$f3(f30f38f3Tf3\f3`f3df3lf3tf3�f3�f3�f3�f3�f3�f3�f3�f3�f3�f3�f3�f3g3g3g3g3 g3(g3Dg3Lg3Pg3Tg3\g3dg3�g3�g3�g3�g3�g3�g3�g3�g3�g3�g3�g3�g3�g3h3h3h3h3h34h3<h3@h3Dh3Lh3Th3ph3xh3|h3�h3�h3�h3�h3�h3�h3�h3�h3�h3�h3�h3�h3�h3i3i3$i3,i30i34i3<i3Di3`i3hi3li3pi3xi3�i3�i3�i3�i3�i3�i3�i3�i3�i3�i3�i3�i3�i3j3j3 j3$j3,j34j3Pj3Xj3\j3`j3hj3pj3�j3�j3�j3�j3�j3�j3�j3�j3�j3�j3�j3�j3k3k3k3k3k3$k3@k3Hk3Lk3Pk3Xk3`k3|k3�k3�k3�k3�k3�k3�k3�k3�k3�k3�k3�k3�k3�k3l3l3l3l30l38l3<l3@l3Hl3Pl3ll3tl3xl3|l3�l3�l3�l3�l3�l3�l3�l3�l3�l3�l3�l3�l3�l3m3 m3(m3,m30m38m3@m3\m3dm3hm3lm3tm3|m3�m3�m3�m3�m3�m3�m3�m3�m3�m3�m3�m3�m3n3n3n3 n3(n30n3Ln3Tn3Xn3\n3dn3ln3�n3�n3�n3�n3�n3�n3�n3�n3�n3�n3�n3�n3o3o3o3o3o3 o3<o3Do3Ho3Lo3To3\o3xo3�o3�o3�o3�o3�o3�o3�o3�o3�o3�o3�o3�o3�o3�o3p3p3p3,p34p38p3<p3Dp3Lp3hp3pp3tp3xp3�p3�p3�p3�p3�p3�p3�p3�p3�p3�p3�p3�p3�p3q3q3$q3(q3,q34q3<q3Xq3`q3dq3hq3pq3xq3�q3�q3�q3�q3�q3�q3�q3�q3�q3�q3�q3�q3r3r3r3r3$r3,r3Hr3Pr3Tr3Xr3`r3hr3�r3�r3�r3�r3�r3�r3�r3�r3�r3�r3�r3�r3�r3s3s3s3s3s38s3@s3Ds3Hs3Ps3Xs3ts3|s3�s3�s3�s3�s3�s3�s3�s3�s3�s3�s3�s3�s3�s3�s3t3t3(t30t34t38t3@t3Ht3dt3lt3pt3tt3|t3�t3�t3�t3�t3�t3�t3�t3�t3�t3�t3�t3�t3�t3u3 u3$u3(u30u38u3Tu3\u3`u3du3lu3tu3�u3�u3�u3�u3�u3�u3�u3�u3�u3�u3�u3�u3v3v3v3v3 v3(v3Dv3Lv3Pv3Tv3\v3dv3�v3�v3�v3�v3�v3�v3�v3�v3�v3�v3�v3�v3�v3w3w3w3w3w34w3<w3@w3Dw3Lw3Tw3pw3xw3|w3�w3�w3�w3�w3�w3�w3�w3�w3�w3�w3�w3�w3�w3x3x3$x3,x30x34x3<x3Dx3`x3hx3lx3px3xx3�x3�x3�x3�x3�x3�x3�x3�x3�x3�x3�x3�x3�x3y3y3 y3$y3,y34y3Py3Xy3\y3`y3hy3py3�y3�y3�y3�y3�y3�y3�y3�y3�y3�y3�y3�y3z3z3z3z3z3$z3@z3Hz3Lz3Pz3Xz3`z3|z3�z3�z3�z3�z3�z3�z3�z3�z3�z3�z3�z3�z3�z3{3{3{3{30{38{3<{3@{3H{3P{3l{3t{3x{3|{3�{3�{3�{3�{3�{3�{3�{3�{3�{3�{3�{3�{3�{3|3 |3(|3,|30|38|3@|3\|3d|3h|3l|3t|3||3�|3�|3�|3�|3�|3�|3�|3�|3�|3�|3�|3�|3}3}3}3 }3(}30}3L}3T}3X}3\}3d}3l}3�}3�}3�}3�}3�}3�}3�}3�}3�}3�}3�}3�}3~3~3~3~3~3 ~3<~3D~3H~3L~3T~3\~3x~3�~3�~3�~3�~3�~3�~3�~3�~3�~3�~3�~3�~3�~3�~3333,34383<3D3L3h3p3t3x3�3�3�3�3�3�3�3�3�3�3�3�3�3�3�3$�3(�3,�34�3<�3X�3`�3d�3h�3p�3x�3��3��3��3��3��3��3Ѐ3؀3܀3�3�3�3�3�3�3�3$�3,�3H�3P�3T�3X�3`�3h�3��3��3��3��3��3��3��3ȁ3́3Ё3؁3�3��3�3�3�3�3�38�3@�3D�3H�3P�3X�3t�3|�3��3��3��3��3��3��3��3��3Ȃ3Ђ3�3�3��3��3�3�3(�30�34�38�3@�3H�3d�3l�3p�3t�3|�3��3��3��3��3��3��3��3܃3�3�3�3�3��3�3 �3$�3(�30�38�3T�3\�3`�3d�3l�3t�3��3��3��3��3��3��3̄3Ԅ3؄3܄3�3�3�3�3�3�3 �3(�3D�3L�3P�3T�3\�3d�3��3��3��3��3��3��3��3ą3ȅ3̅3ԅ3܅3��3�3�3�3�3�34�3<�3@�3D�3L�3T�3p�3x�3|�3��3��3��3��3��3��3��3Ć3̆3�3�3�3��3�3�3$�3,�30�34�3<�3D�3`�3h�3l�3p�3x�3��3��3��3��3��3��3��3؇3�3�3�3�3��3�3�3 �3$�3,�34�3P�3X�3\�3`�3h�3p�3��3��3��3��3��3��3Ȉ3Ј3Ԉ3؈3�3�3�3�3�3�3�3$�3@�3H�3L�3P�3X�3`�3|�3��3��3��3��3��3��3��3ĉ3ȉ3Љ3؉3�3��3�3�3�3�30�38�3<�3@�3H�3P�3l�3t�3x�3|�3��3��3��3��3��3��3��3Ȋ3�3�3�3�3��3�3 �3(�3,�30�38�3@�3\�3d�3h�3l�3t�3|�3��3��3��3��3��3��3ԋ3܋3�3�3�3�3�3�3�3 �3(�30�3L�3T�3X�3\�3d�3l�3��3��3��3��3��3��3Č3̌3Ќ3Ԍ3܌3�3�3�3�3�3�3 �3<�3D�3H�3L�3T�3\�3x�3��3��3��3��3��3��3��3��3č3̍3ԍ3�3��3��3�3�3�3,�34�38�3<�3D�3L�3h�3p�3t�3x�3��3��3��3��3��3��3��3Ď3�3�3�3�3��3�3�3$�3(�3,�34�3<�3X�3`�3d�3h�3p�3x�3��3��3��3��3��3��3Џ3؏3܏3�3�3�3�3�3�3�3$�3,�3H�3P�3T�3X�3`�3h�3��3��3��3��3��3��3��3Ȑ3̐3А3ؐ3�3��3�3�3�3�3�38�3@�3D�3H�3P�3X�3t�3|�3��3��3��3��3��3��3��3��3ȑ3Б3�3�3��3��3�3�3(�30�34�38�3@�3H�3d�3l�3p�3t�3|�3��3��3��3��3��3��3��3ܒ3�3�3�3�3��3�3 �3$�3(�30�38�3T�3\�3`�3d�3l�3t�3��3��3��3��3��3��3̓3ԓ3ؓ3ܓ3�3�3�3�3�3�3 �3(�3D�3L�3P�3T�3\�3d�3��3��3��3��3��3��3��3Ĕ3Ȕ3̔3Ԕ3ܔ3��3�3�3�3�3�34�3<�3@�3D�3L�3T�3p�3x�3|�3��3��3��3��3��3��3��3ĕ3̕3�3�3�3��3�3�3$�3,�30�34�3<�3D�3`�3h�3l�3p�3x�3��3��3��3��3��3��3��3ؖ3�3�3�3�3��3�3�3 �3$�3,�34�3P�3X�3\�3`�3h�3p�3��3��3��3��3��3��3ȗ3З3ԗ3ؗ3�3�3�3�3�3�3�3$�3@�3H�3L�3P�3X�3`�3|�3��3��3��3��3��3��3��3Ę3Ș3И3ؘ3�3��3�3�3�3�30�38�3<�3@�3H�3P�3l�3t�3x�3|�3��3��3��3��3��3��3��3ș3�3�3�3�3��3�3 �3(�3,�30�38�3@�3\�3d�3h�3l�3t�3|�3��3��3��3��3��3��3Ԛ3ܚ3�3�3�3�3�3�3�3 �3(�30�3L�3T�3X�3\�3d�3l�3��3��3��3��3��3��3ě3̛3Л3ԛ3ܛ3�3�3�3�3�3�3 �3<�3D�3H�3L�3T�3\�3x�3��3��3��3��3��3��3��3��3Ĝ3̜3Ԝ3�3��3��3�3�3�3,�34�38�3<�3D�3L�3h�3p�3t�3x�3��3��3��3��3��3��3��3ĝ3�3�3�3�3��3�3�3$�3(�3,�34�3<�3X�3`�3d�3h�3p�3x�3��3��3��3��3��3��3О3؞3ܞ3�3�3�3�3�3�3�3$�3,�3H�3P�3T�3X�3`�3h�3��3��3��3��3��3��3��3ȟ3̟3П3؟3�3��3�3�3�3�3�38�3@�3D�3H�3P�3X�3t�3|�3��3��3��3��3��3��3��3��3Ƞ3Р3�3�3��3��3�3�3�3(�30�34�38�3<�3@�3H�3d�3l�3p�3t�3x�3|�3��3��3��3��3��3��3��3��3ܡ3�3�3�3�3�3��3�3 �3$�3(�30�38�3T�3\�3`�3d�3h�3��3��3��3��3��3̢3Ԣ3آ3ܢ3�3�3�3�3�3�3D�3L�3P�3T�3X�3��3��3��3��3��3��3ģ3ȣ3̣3У3��3�3�3�3�34�3<�3@�3D�3H�3p�3x�3|�3��3��3��3��3��3��3��3�3�3�3��3��3$�3,�30�34�38�3`�3h�3l�3p�3t�3��3��3��3��3��3إ3�3�3�3�3�3�3 �3$�3(�3P�3X�3\�3`�3d�3��3��3��3��3��3Ȧ3Ц3Ԧ3ئ3ܦ3�3�3�3�3�3@�3H�3L�3P�3T�3��3��3ħ3ȧ3̧3�3��3�3�3�30�38�3<�3@�3D�3l�3t�3x�3|�3��3��3��3��3��3��3�3�3�3�3��3 �3(�3,�30�34�3\�3d�3h�3l�3p�3��3��3��3��3��3ԩ3ܩ3�3�3�3�3�3 �3$�3L�3T�3X�3\�3`�3��3��3��3��3��3Ī3̪3Ъ3Ԫ3�3�3�3�3<�3D�3H�3L�3x�3��3��3��3��3��3��3��3ī3�3��3��3�3,�34�38�3<�3h�3p�3t�3x�3��3��3��3��3�3�3�3�3�3$�3(�3,�3X�3`�3d�3h�3��3��3��3��3Э3ح3ܭ3�3�3�3�3�3�3 �3H�3P�3T�3X�3��3��3��3��3��3Ȯ3̮3Ю3��3�3�3�38�3@�3D�3H�3t�3|�3��3��3��3��3��3��3�3�3��3��3(�30�34�38�3d�3l�3p�3t�3��3��3��3��3ܰ3�3�3�3�3 �3$�3(�3T�3\�3`�3d�3��3��3��3��3̱3Ա3ر3ܱ3�3�3�3�3D�3L�3P�3T�3��3��3��3��3��3IJ3Ȳ3̲3��3�3�3�34�3<�3@�3D�3p�3x�3|�3��3��3��3��3��3�3�3�3��3$�3,�30�34�3`�3h�3l�3p�3��3��3��3��3ش3�3�3�3�3�3 �3$�3P�3X�3\�3`�3��3��3��3��3ȵ3е3Ե3ص3�3�3�3�3@�3H�3L�3P�3|�3��3��3��3��3��3Ķ3ȶ3�3��3�3�30�38�3<�3@�3l�3t�3x�3|�3��3��3��3��3�3�3�3�3 �3(�3,�30�3\�3d�3h�3l�3��3��3��3��3Ը3ܸ3�3�3�3�3�3 �3L�3T�3X�3\�3��3��3��3��3Ĺ3̹3й3Թ3�3�3�3�3<�3D�3H�3L�3x�3��3��3��3��3��3��3ĺ3�3��3��3�3,�34�38�3<�3h�3p�3t�3x�3��3��3��3��3�3�3�3�3�3$�3(�3,�3X�3`�3d�3h�3��3��3��3��3м3ؼ3ܼ3�3�3�3�3�3H�3P�3T�3X�3��3��3��3��3��3Ƚ3̽3н3��3�3�3�38�3@�3D�3H�3t�3|�3��3��3��3��3��3��3�3�3��3��3(�30�34�38�3d�3l�3p�3t�3��3��3��3��3��3ܿ3�3�3�3�3 �3$�3(�3,�3T�3\�3`�3d�3��3��3��3��3�3�3�3�3�3�3�3�3D�3L�3P�3T�3��3��3��3��3��3��3�3�3�3�3��3�3�3�3�34�3<�3@�3D�3p�3x�3|�3��3��3��3��3��3��3��3��3��3$�3,�30�34�3`�3h�3l�3p�3��3��3��3��3��3��3��3��3�3�3 �3$�3P�3X�3\�3`�3��3��3��3��3��3��3��3��3�3�3�3�3@�3H�3L�3P�3|�3��3��3��3��3��3��3��3��3��3�3�30�38�3<�3@�3l�3t�3x�3|�3��3��3��3��3��3��3��3��3 �3(�3,�30�3\�3d�3h�3l�3��3��3��3��3��3��3��3��3�3�3�3 �3L�3T�3X�3\�3��3��3��3��3��3��3��3��3�3�3�3�3<�3D�3H�3L�3x�3��3��3��3��3��3��3��3��3��3��3�3,�34�38�3<�3h�3p�3t�3x�3��3��3��3��3��3��3��3��3�3$�3(�3,�3X�3`�3d�3h�3��3��3��3��3��3��3��3��3��3�3�3�3�3H�3P�3T�3X�3��3��3��3��3��3��3��3��3��3��3��3�3�3�38�3@�3D�3H�3t�3|�3��3��3��3��3��3��3��3��3��3��3��3�3(�30�34�38�3d�3l�3p�3t�3��3��3��3��3��3��3��3��3�3 �3$�3(�3T�3\�3`�3d�3��3��3��3��3��3��3��3��3�3�3�3�3D�3L�3P�3T�3��3��3��3��3��3��3��3��3��3�3�3�34�3<�3@�3D�3p�3x�3|�3��3��3��3��3��3��3��3��3��3$�3,�30�34�3`�3h�3l�3p�3��3��3��3��3��3��3��3��3�3�3 �3$�3P�3X�3\�3`�3��3��3��3��3��3��3��3��3�3�3�3�3@�3H�3L�3P�3|�3��3��3��3��3��3��3��3��3��3�3�30�38�3<�3@�3l�3t�3x�3|�3��3��3��3��3��3��3��3��3 �3(�3,�30�3\�3d�3h�3l�3��3��3��3��3��3��3��3��3�3�3�3 �3L�3T�3X�3\�3��3��3��3��3��3��3��3��3�3�3�3�3<�3D�3H�3L�3x�3��3��3��3��3��3��3��3��3��3��3�3,�34�38�3<�3h�3p�3t�3x�3��3��3��3��3��3��3��3��3��3�3$�3(�3,�3X�3`�3d�3h�3��3��3��3��3��3��3��3��3�3�3�3�3H�3P�3T�3X�3��3��3��3��3��3��3��3��3��3�3�3�3�38�3@�3D�3H�3t�3|�3��3��3��3��3��3��3��3��3��3��3(�30�34�38�3d�3l�3p�3t�3x�3��3��3��3��3��3��3��3��3��3��3�3 �3$�3(�3,�3T�3\�3`�3d�3h�3��3��3��3��3��3��3��3��3��3�3�3�3�3D�3L�3P�3T�3X�3��3��3��3��3��3��3��3��3��3��3�3�3�34�3<�3@�3D�3p�3x�3|�3��3��3��3��3��3��3��3��3��3$�3,�30�34�3`�3h�3l�3p�3��3��3��3��3��3��3��3��3��3�3�3 �3$�3P�3X�3\�3`�3��3��3��3��3��3��3��3��3�3�3�3�3@�3H�3L�3P�3|�3��3��3��3��3��3��3��3��3��3�3�30�38�3<�3@�3l�3t�3x�3|�3��3��3��3��3��3��3��3��3��3 �3(�3,�30�34�3\�3d�3h�3l�3��3��3��3��3��3��3��3��3�3�3�3 �3L�3T�3X�3\�3`�3��3��3��3��3��3��3��3��3�3�3�3�3<�3D�3H�3L�3P�3x�3��3��3��3��3��3��3��3��3��3��3�3�3,�34�38�3<�3h�3p�3t�3x�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3$�3(�3,�34�38�3<�3X�3`�3d�3h�3p�3x�3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3�3 �3$�3,�3H�3P�3T�3X�3\�3`�3h�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3�3�3�38�3@�3D�3H�3L�3P�3X�3t�3|�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3(�30�34�38�3<�3@�3H�3d�3l�3p�3t�3x�3|�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3 �3$�3(�3,�30�38�3T�3\�3`�3d�3h�3l�3t�3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3�3 �3(�3D�3L�3P�3T�3\�3d�3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3�3�34�3<�3@�3D�3L�3T�3p�3x�3|�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3$�3,�30�34�3<�3@�3D�3`�3h�3l�3p�3x�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3�3 �3$�3,�30�34�3<�3@�3P�3X�3\�3`�3h�3l�3p�3x�3|�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3�3�3 �3$�3@�3H�3L�3P�3X�3\�3`�3|�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3�3�30�38�3<�3@�3D�3H�3P�3l�3t�3x�3|�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3 �3(�3,�30�34�38�3@�3\�3d�3h�3l�3p�3t�3|�3��3��3��3��3��3��3��3��3��3��3��3��3��3��3�3�3�3 �3$�3(�30�3L�3T�3X�3\�3`�3d�3l�3��3��3��3��3��3��3��3�3�3�3�3�3�3�3�3�3�3�3�3�3 �3<�3D�3H�3L�3P�3T�3\�3x�3��3��3��3��3��3��3��3��3��3�3�3�3�3�3��3��3�3�3�3�3,�34�38�3<�3@�3D�3L�3h�3p�3t�3x�3|�3��3��3��3��3��3��3��3��3�3�3�3�3�3�3��3�3�3$�3(�3,�30�34�3<�3X�3`�3d�3h�3l�3p�3x�3��3��3��3��3��3��3��3�3�3�3�3�3�3�3�3�3�3�3 �3$�3,�3H�3P�3T�3X�3\�3`�3h�3��3��3��3��3��3��3��3��3�3�3�3�3�3�3��3�3�3�3�3�3�38�3@�3D�3H�3L�3P�3X�3t�3|�3��3��3��3��3��3��3��3��3��3�3�3�3�3�3��3��3�3�3�3(�30�34�38�3<�3@�3H�3d�3l�3p�3t�3x�3|�3��3��3��3��3��3��3��3��3�3�3�3�3�3�3��3�3 �3$�3(�3,�30�38�3T�3\�3`�3d�3h�3l�3t�3��3��3��3��3��3��3��3�3�3�3�3�3�3�3�3�3�3�3�3 �3(�3D�3L�3P�3T�3X�3\�3d�3��3��3��3��3��3��3��3��3�3�3�3�3�3�3��3�3�3�3�3�3�34�3<�3@�3D�3H�3L�3T�3p�3x�3|�3��3��3��3��3��3��3��3��3��3�3�3�3�3�3��3��3�3�3$�3,�30�34�38�3<�3D�3`�3h�3l�3p�3t�3x�3��3��3��3��3��3��3��3��3�3�3�3�3�3�3��344 4$4(4,444P4X4\4`4d4h4p4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444444$4@4H4L4P4T4X4`4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444440484<4@4D4H4P4l4t4x4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�44 4(4,4044484@4\4d4h4l4p4t4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444 4$4(404L4T4X4\4`4d4l4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444444 4<4D4H4L4P4T4\4x4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�44444,44484<4@4D4L4h4p4t4x4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�444$4(4,40444<4X4`4d4h4l4p4x4�4�4�4�4�4�4�4�4�4�4�4�4�4�44444 4$4,4H4P4T4X4\4`4h4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4	4	4	4	4	4	48	4@	4D	4H	4L	4P	4X	4t	4|	4�	4�	4�	4�	4�	4�	4�	4�	4�	4�	4�	4�	4�	4�	4�	4�	4
4
4
4(
40
44
48
4<
4@
4H
4d
4l
4p
4t
4x
4|
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
44 4$4(4,40484T4\4`4d4h4l4t4�4�4�4�4�4�4�4�4�4�4�4�4�4�444444 4(4D4L4P4T4X4\4d4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4
4
4
4
4
4
44
4<
4@
4D
4H
4L
4T
4p
4x
4|
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
4�
444$4,4044484<4D4`4h4l4p4t4x4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�444 4$4(4,444P4X4\4`4d4h4p4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444444$4@4H4L4P4T4X4`4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444440484<4@4D4H4P4l4t4x4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�44 4(4,4044484@4\4d4h4l4p4t4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444 4$4(404L4T4X4\4`4d4l4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444444 4<4D4H4L4P4T4\4x4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�44444,44484<4@4D4L4h4p4t4x4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�444$4(4,40444<4X4`4d4h4l4p4x4�4�4�4�4�4�4�4�4�4�4�4�4�4�44444 4$4,4H4P4T4X4\4`4h4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�444444484@4D4H4L4P4X4t4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�4444(4044484<4@4H4d4l4p4t4x4|4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�44 4$4(4,40484T4\4`4d4h4l4t4�4�4�4�4�4�4�4�4�4�4�4�4�4�444444 4(4D4L4P4T4X4\4d4�4�4�4�4�4�4�4�4�4�4�4�4�4�4�444444<4@4D4p4x4|4�4�4�4�4�4�4�4�4�4�4�4�4$4,4044484`4h4l4p4t4�4�4�4�4�4�4�4�4�444 4$4(4P4X4\4`4d4�4�4�4�4�4�4�4�4�4�444444@4H4L4P4T4|4�4�4�4�4�4�4�4�4�4�4�4 4 4 40 48 4< 4@ 4D 4l 4t 4x 4| 4� 4� 4� 4� 4� 4� 4� 4� 4� 4 !4(!4,!40!44!4\!4d!4h!4l!4p!4�!4�!4�!4�!4�!4�!4�!4�!4�!4�!4"4"4"4 "4$"4L"4T"4X"4\"4�"4�"4�"4�"4�"4�"4�"4�"4#4#4#4#4x#4�#4�#4�#4�#4�#4�#4�#4�#4�#4�#4�#4�#4$4$4,$44$48$4<$4@$4h$4p$4t$4x$4|$4�$4�$4�$4�$4�$4�$4�$4�$4�$4%4$%4(%4,%40%4X%4`%4d%4h%4l%4�%4�%4�%4�%4�%4�%4�%4�%4&4&4&4&4H&4P&4T&4X&4�&4�&4�&4�&4�&4�&4�&4�&4�&4'4'4'48'4@'4D'4H'4L'4t'4|'4�'4�'4�'4�'4�'4�'4�'4�'4�'4(4(44(48(4P(4T(4p(4t(4�(4�(4�(4�(4�(4�(4�(4�(4)4)4$)4()4@)4D)4`)4d)4|)4�)4�)4�)4�)4�)4�)4�)4�)4�)4*4*40*44*4P*4T*4l*4p*4�*4�*4�*4�*4�*4�*4�*4�*4�*4+4+4+4+44+4<+4@+4D+4p+4x+4|+4�+4�+4�+4�+4�+4�+4�+4�+4�+4$,4,,40,44,4`,4h,4l,4p,4t,4�,4�,4�,4�,4�,4�,4�,4�,4�,4�,4-4-4 -4$-4(-4P-4X-4\-4`-4d-4�-4�-4�-4�-4�-4�-4�-4�-4�-4�-4.4.4.4.4.4@.4H.4L.4P.4T.4|.4�.4�.4�.4�.4�.4�.4�.4�.4�.4�.4/4/40/48/4</4@/4l/4t/4x/4|/4�/4�/4�/4�/4�/4�/4�/4�/4�/4�/4 04(04,04004\04d04h04l04�04�04�04�04�04�04�04�04�04�04141414 14L14T14X14\14`14�14�14�14�14�14�14�14�14�14�142424242424<24D24H24L24x24�24�24�24�24�24�24�24�24�24�2434,34434834<34h34p34t34x34�34�34�34�34�34�34�34�3444$44(44,44044X44`44d44h44�44�44�44�44�44�44�44�4454545454H54P54T54X54\54�54�54�54�54�54�54�54�54�54�54�5464646464864@64D64H64L64t64|64�64�64�64�64�64�64�64�64�64�64�64(74074474874<74d74l74p74t74x74�74�74�74�74�74�74�74�74�7484 84$84(84T84\84`84d84�84�84�84�84�84�84�84�8494949494D94L94P94T94�94�94�94�94�94�94�94�94�94:4:4:44:4<:4@:4D:4p:4x:4|:4�:4�:4�:4�:4�:4�:4�:4�:4�:4$;4,;40;44;48;4`;4h;4l;4p;4t;4�;4�;4�;4�;4�;4�;4�;4�;4<4<4 <4$<4(<4P<4X<4\<4`<4�<4�<4�<4�<4�<4�<4�<4�<4=4=4=4=4=4@=4H=4L=4P=4|=4�=4�=4�=4�=4�=4�=4�=4�=4�=4�=4>4>4>40>48>4<>4@>4D>4l>4t>4x>4|>4�>4�>4�>4�>4�>4�>4�>4�>4�>4�>4�>4 ?4(?4,?40?44?4\?4d?4h?4l?4p?4�?4�?4�?4�?4�?4�?4�?4�?4�?4�?4@4@4@4 @4$@4L@4T@4X@4\@4`@4�@4�@4�@4�@4�@4�@4�@4�@4�@4A4A4A4A4<A4DA4HA4LA4PA4xA4�A4�A4�A4�A4�A4�A4�A4�A4�A4�A4�A4�A4B4,B44B48B4<B4hB4pB4tB4xB4�B4�B4�B4C4C4C4(C4,C44C4DC4HC4LC4dC4hC4pC4�C4�C4�C4�C4�C4�C4�C4�C4�C4�C4�C4�C4�C4�C4D4D4D4$D44D48D4<D4TD4XD4\D4`D4pD4tD4xD4�D4�D4�D4�D4�D4�D4�D4�D4�D4�D4�D4�D4E4E4E4E4$E4(E4,E4DE4HE4PE4`E4dE4hE4�E4�E4�E4�E4�E4�E4�E4�E4�E4�E4�E4�E4�E4�E4F4F4F4F44F48F4@F4PF4TF4XF4pF4tF4|F4�F4�F4�F4�F4�F4�F4�F4�F4�F4�F4�F4�F4G4G4G4$G4(G40G4@G4DG4HG4`G4dG4lG4|G4�G4�G4�G4�G4�G4�G4�G4�G4�G4�G4�G4�G4�G4�G4�G4�G4H4H4H4 H40H44H48H4PH4TH4XH4\H4lH4pH4tH4�H4�H4�H4�H4�H4�H4�H4�H4�H4�H4�H4�H4�H4I4I4I4 I4$I4(I4@I4DI4LI4\I4`I4dI4|I4�I4�I4�I4�I4�I4�I4�I4�I4�I4�I4�I4�I4�I4J4J4J4J40J44J4<J4LJ4PJ4TJ4lJ4pJ4xJ4�J4�J4�J4�J4�J4�J4�J4�J4�J4�J4�J4�J4K4K4K4 K4$K4(K4,K4<K4@K4DK4\K4`K4hK4xK4|K4�K4�K4�K4�K4�K4�K4�K4�K4�K4�K4�K4�K4�K4L4L4L4,L40L44L4LL4PL4XL4hL4lL4pL4�L4�L4�L4�L4�L4�L4�L4�L4�L4�L4�L4�L4M4M4M4M4 M4$M4<M4@M4HM4XM4\M4`M4xM4|M4�M4�M4�M4�M4�M4�M4�M4�M4�M4�M4�M4�M4�M4N4N4N4,N40N48N4HN4LN4PN4hN4lN4tN4�N4�N4�N4�N4�N4�N4�N4�N4�N4�N4�N4�N4�N4�N4O4O4O4 O4(O48O4<O4@O4XO4\O4dO4tO4xO4|O4�O4�O4�O4�O4�O4�O4�O4�O4�O4�O4�O4�O4P4P4P4(P4,P40P4HP4LP4TP4dP4hP4lP4�P4�P4�P4�P4�P4�P4�P4�P4�P4�P4�P4�P4�P4Q4Q4Q4Q4 Q48Q4<Q4DQ4TQ4XQ4\Q4tQ4xQ4|Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4�Q4R4R4R4(R4,R44R4DR4HR4LR4dR4hR4pR4�R4�R4�R4�R4�R4�R4�R4�R4�R4�R4�R4�R4S4S4$S44S48S4TS4XS4`S4pS4tS4�S4�S4�S4�S4�S4�S4�S4�S4�S4�S4T4T4T4$T4(T4DT4HT4PT4`T4dT4�T4�T4�T4�T4�T4�T4�T4�T4�T4�T4�T4�T4U4U4U44U48U4@U4PU4TU4pU4tU4|U4�U4�U4�U4�U4�U4�U4�U4�U4�U4�U4V4V4$V4(V40V4@V4DV4`V4dV4lV4|V4�V4�V4�V4�V4�V4�V4�V4�V4�V4�V4�V4W4W4 W40W44W4PW4TW4\W4lW4pW4�W4�W4�W4�W4�W4�W4�W4�W4�W4�W4X4X4X4 X4$X4@X4DX4LX4\X4`X4|X4�X4�X4�X4�X4�X4�X4�X4�X4�X4�X4�X4Y4Y4Y40Y44Y4<Y4LY4PY4lY4pY4xY4�Y4�Y4�Y4�Y4�Y4�Y4�Y4�Y4�Y4Z4Z4Z4 Z4$Z4<Z4@Z4DZ4\Z4`Z4xZ4|Z4�Z4�Z4�Z4�Z4�Z4�Z4�Z4�Z4�Z4�Z4�Z4[4[4,[40[44[4L[4P[4h[4l[4p[4�[4�[4�[4�[4�[4�[4�[4�[4�[4�[4\4\4\4 \4$\40\48\4<\4@\4H\4l\4t\4x\4|\4�\4�\4�\4�\4�\4�\4�\4�\4�\4�\4�\4 ]4(]4,]40]48]4\]4d]4h]4l]4t]4�]4�]4�]4�]4�]4�]4�]4�]4�]4�]4^4^4^4 ^4(^4L^4T^4X^4\^4d^4�^4�^4�^4�^4�^4�^4�^4�^4�^4�^4_4_4_4_4_4<_4D_4H_4L_4T_4x_4�_4�_4�_4�_4�_4�_4�_4�_4�_4�_4�_4�_4`4`4,`44`48`4<`4D`4h`4p`4t`4x`4�`4�`4�`4�`4�`4�`4�`4�`4�`4�`4�`4a4$a4(a4,a44a4Xa4`a4da4ha4pa4�a4�a4�a4�a4�a4�a4�a4�a4�a4�a4b4b4b4b4$b4Hb4Pb4Tb4Xb4`b4�b4�b4�b4�b4�b4�b4�b4�b4�b4�b4�b4c4c4c4c48c4@c4Dc4Hc4Pc4tc4|c4�c4�c4�c4�c4�c4�c4�c4�c4�c4�c4�c4�c4d4(d40d44d48d4@d4dd4ld4pd4td4|d4�d4�d4�d4�d4�d4�d4�d4�d4�d4�d4e4 e4$e4(e40e4Te4\e4`e4de4le4�e4�e4�e4�e4�e4�e4�e4�e4�e4�e4f4f4f4f4 f4Df4Lf4Pf4Tf4\f4�f4�f4�f4�f4�f4�f4�f4�f4�f4�f4�f4g4g4g4g44g4<g4@g4Dg4Lg4pg4xg4|g4�g4�g4�g4�g4�g4�g4�g4�g4�g4�g4�g4h4$h4,h40h44h4<h4`h4hh4lh4ph4xh4�h4�h4�h4�h4�h4�h4�h4�h4�h4�h4i4i4 i4$i4,i4Pi4Xi4\i4`i4hi4�i4�i4�i4�i4�i4�i4�i4�i4�i4�i4j4j4j4j4j4@j4Hj4Lj4Pj4Xj4|j4�j4�j4�j4�j4�j4�j4�j4�j4�j4�j4�j4k4k4k40k48k4<k4@k4Hk4lk4tk4xk4|k4�k4�k4�k4�k4�k4�k4�k4�k4�k4�k4�k4 l4(l4,l40l48l4\l4dl4hl4ll4tl4�l4�l4�l4�l4�l4�l4�l4�l4�l4�l4m4m4m4 m4(m4Lm4Tm4Xm4\m4dm4�m4�m4�m4�m4�m4�m4�m4�m4�m4�m4n4n4n4n4n4<n4Dn4Hn4Ln4Tn4xn4�n4�n4�n4�n4�n4�n4�n4�n4�n4�n4�n4�n4o4o4,o44o48o4<o4Do4ho4po4to4xo4�o4�o4�o4�o4�o4�o4�o4�o4�o4�o4�o4p4$p4(p4,p44p4Xp4`p4dp4hp4pp4�p4�p4�p4�p4�p4�p4�p4�p4�p4�p4q4q4q4q4$q4Hq4Pq4Tq4Xq4`q4�q4�q4�q4�q4�q4�q4�q4�q4�q4�q4�q4r4r4r4r48r4@r4Dr4Hr4Pr4tr4|r4�r4�r4�r4�r4�r4�r4�r4�r4�r4�r4�r4�r4s4(s40s44s48s4@s4ds4ls4ps4ts4|s4�s4�s4�s4�s4�s4�s4�s4�s4�s4�s4t4 t4$t4(t40t4Tt4\t4`t4dt4lt4�t4�t4�t4�t4�t4�t4�t4�t4�t4�t4u4u4u4u4 u4Du4Lu4Pu4Tu4\u4�u4�u4�u4�u4�u4�u4�u4�u4�u4�u4�u4v4v4v4v44v4<v4@v4Dv4Lv4pv4xv4|v4�v4�v4�v4�v4�v4�v4�v4�v4�v4�v4�v4w4$w4,w40w44w4<w4`w4hw4lw4pw4xw4�w4�w4�w4�w4�w4�w4�w4�w4�w4�w4x4x4 x4$x4,x4Px4Xx4\x4`x4hx4�x4�x4�x4�x4�x4�x4�x4�x4�x4�x4y4y4y4y4y4@y4Hy4Ly4Py4Xy4|y4�y4�y4�y4�y4�y4�y4�y4�y4�y4�y4�y4z4z4z40z48z4<z4@z4Hz4lz4tz4xz4|z4�z4�z4�z4�z4�z4�z4�z4�z4�z4�z4�z4 {4({4,{40{48{4\{4d{4h{4l{4t{4�{4�{4�{4�{4�{4�{4�{4�{4�{4�{4|4|4|4 |4(|4L|4T|4X|4\|4d|4�|4�|4�|4�|4�|4�|4�|4�|4�|4�|4}4}4}4}4}4<}4D}4H}4L}4T}4x}4�}4�}4�}4�}4�}4�}4�}4�}4�}4�}4�}4�}4~4~4,~44~48~4<~4D~4h~4p~4t~4x~4�~4�~4�~4�~4�~4�~4�~4�~4�~4�~4�~44$4(4,444X4`4d4h4p4�4�4�4�4�4�4�4�4�4�4�4�4�4�4$�4H�4P�4T�4X�4`�4��4��4��4��4��4��4Ȁ4̀4Ѐ4؀4��4�4�4�4�48�4@�4D�4H�4P�4t�4|�4��4��4��4��4��4��4��4ȁ4�4�4��4��4�4(�40�44�48�4@�4d�4l�4p�4t�4|�4��4��4��4��4��4܂4�4�4�4�4�4 �4$�4(�40�4T�4\�4`�4d�4l�4��4��4��4��4��4̃4ԃ4؃4܃4�4�4�4�4�4 �4D�4L�4P�4T�4\�4��4��4��4��4��4��4Ą4Ȅ4̄4Ԅ4��4�4�4�4�44�4<�4@�4D�4L�4p�4x�4|�4��4��4��4��4��4��4ą4�4�4�4��4�4$�4,�40�44�4<�4`�4h�4l�4p�4x�4��4��4��4��4��4؆4�4�4�4�4�4�4 �4$�4,�4P�4X�4\�4`�4h�4��4��4��4��4��4ȇ4Ї4ԇ4؇4�4�4�4�4�4�4@�4H�4L�4P�4X�4|�4��4��4��4��4��4��4Ĉ4Ȉ4Ј4�4��4�4�4�40�48�4<�4@�4H�4l�4t�4x�4|�4��4��4��4��4��4��4�4�4�4�4��4 �4(�4,�40�48�4\�4d�4h�4l�4t�4��4��4��4��4��4Ԋ4܊4�4�4�4�4�4�4 �4(�4L�4T�4X�4\�4d�4��4��4��4��4��4ċ4̋4Ћ4ԋ4܋4�4�4�4�4�4<�4D�4H�4L�4T�4x�4��4��4��4��4��4��4��4Č4̌4�4��4��4�4�4,�44�48�4<�4D�4h�4p�4t�4x�4��4��4��4��4��4��4�4�4�4�4��4�4$�4(�4,�44�4X�4`�4d�4h�4p�4��4��4��4��4��4Ў4؎4܎4�4�4�4�4�4�4$�4H�4P�4T�4X�4`�4��4��4��4��4��4��4ȏ4̏4Џ4؏4��4�4�4�4�48�4@�4D�4H�4P�4t�4|�4��4��4��4��4��4��4��4Ȑ4�4�4��4��4�4(�40�44�48�4@�4d�4l�4p�4t�4|�4��4��4��4��4��4ܑ4�4�4�4�4�4 �4$�4(�40�4T�4\�4`�4d�4l�4��4��4��4��4��4̒4Ԓ4ؒ4ܒ4�4�4�4�4�4 �4D�4L�4P�4T�4\�4��4��4��4��4��4��4ē4ȓ4̓4ԓ4��4�4�4�4�44�4<�4@�4D�4L�4p�4x�4|�4��4��4��4��4��4��4Ĕ4�4�4�4��4�4$�4,�40�44�4<�4`�4h�4l�4p�4x�4��4��4��4��4��4ؕ4�4�4�4�4�4�4 �4$�4,�4P�4X�4\�4`�4h�4��4��4��4��4��4Ȗ4Ж4Ԗ4ؖ4�4�4�4�4�4�4@�4H�4L�4P�4X�4|�4��4��4��4��4��4��4ė4ȗ4З4�4��4�4�4�40�48�4<�4@�4H�4l�4t�4x�4|�4��4��4��4��4��4��4�4�4�4�4��4 �4(�4,�40�48�4\�4d�4h�4l�4t�4��4��4��4��4��4ԙ4ܙ4�4�4�4�4�4�4 �4(�4L�4T�4X�4\�4d�4��4��4��4��4��4Ě4̚4К4Ԛ4ܚ4�4�4�4�4�4<�4D�4H�4L�4T�4x�4��4��4��4��4��4��4��4ě4̛4�4��4��4�4�4,�44�48�4<�4D�4h�4p�4t�4x�4��4��4��4��4��4��4�4�4�4�4��4�4$�4(�4,�44�4X�4`�4d�4h�4p�4��4��4��4��4��4Н4؝4ܝ4�4�4�4�4�4�4$�4H�4P�4T�4X�4`�4��4��4��4��4��4��4Ȟ4̞4О4؞4��4�4�4�4�48�4@�4D�4H�4P�4t�4|�4��4��4��4��4��4��4��4ȟ4�4�4��4��4�4(�40�44�48�4@�4d�4l�4p�4t�4|�4��4��4��4��4��4ܠ4�4�4�4�4�4 �4$�4(�40�4T�4\�4`�4d�4l�4��4��4��4��4��4̡4ԡ4ء4ܡ4�4�4�4�4�4 �4D�4L�4P�4T�4\�4��4��4��4��4��4��4Ģ4Ȣ4̢4Ԣ4��4�4�4�4�44�4<�4@�4D�4L�4p�4x�4|�4��4��4��4��4��4��4ģ4�4�4�4��4�4$�4,�40�44�4<�4`�4h�4l�4p�4x�4��4��4��4��4��4ؤ4�4�4�4�4�4�4 �4$�4,�4P�4X�4\�4`�4h�4��4��4��4��4��4ȥ4Х4ԥ4إ4�4�4�4�4�4�4@�4H�4L�4P�4X�4|�4��4��4��4��4��4��4Ħ4Ȧ4Ц4�4��4�4�4�40�48�4<�4@�4H�4l�4t�4x�4|�4��4��4��4��4��4��4�4�4�4�4��4 �4(�4,�40�48�4\�4d�4h�4l�4t�4��4��4��4��4��4Ԩ4ܨ4�4�4�4�4�4�4 �4(�4L�4T�4X�4\�4d�4��4��4��4��4��4ĩ4̩4Щ4ԩ4ܩ4�4�4�4�4�4<�4D�4H�4L�4T�4x�4��4��4��4��4��4��4��4Ī4̪4Ԫ4�4��4��4�4�4�4,�44�48�4<�4D�4L�4h�4p�4t�4x�4��4��4��4��4��4��4��4ī4�4�4�4�4��4�4�4$�4(�4,�44�4<�4X�4`�4d�4h�4p�4x�4��4��4��4��4��4��4Ь4ج4ܬ4�4�4�4�4�4�4�4$�4,�4H�4P�4T�4X�4`�4h�4��4��4��4��4��4��4��4ȭ4̭4Э4ح4�4��4�4�4�4�4�48�4@�4D�4H�4P�4X�4t�4|�4��4��4��4��4��4��4��4��4Ȯ4Ю4�4�4��4��4�4�4(�40�44�48�4@�4H�4d�4l�4p�4t�4|�4��4��4��4��4��4��4��4ܯ4�4�4�4�4��4�4 �4$�4(�40�48�4T�4\�4`�4d�4l�4t�4��4��4��4��4��4��4̰4԰4ذ4ܰ4�4�4�4�4�4�4 �4(�4D�4L�4P�4T�4\�4d�4��4��4��4��4��4��4ı4ȱ4̱4Ա4��4�4�4�4�44�4<�4@�4D�4L�4p�4x�4|�4��4��4��4��4��4��4IJ4�4�4�4��4�4$�4,�40�44�4<�4`�4h�4l�4p�4x�4��4��4��4��4��4س4�4�4�4�4�4�4 �4$�4,�4P�4X�4\�4`�4h�4��4��4��4��4��4ȴ4д4Դ4ش4�4�4�4�4�4�4@�4H�4L�4P�4X�4|�4��4��4��4��4��4��4ĵ4ȵ4е4�4��4�4�4�40�48�4<�4@�4H�4l�4t�4x�4|�4��4��4��4��4��4��4 �4(�4,�40�44�4\�4d�4h�4l�4p�4��4��4��4��4��4Է4ܷ4�4�4�4�4�4 �4$�4L�4T�4X�4\�4`�4��4��4��4��4��4ĸ4̸4и4Ը4ظ4�4�4�4�4<�4D�4H�4L�4P�4x�4��4��4��4��4��4��4��4Ĺ4ȹ4�4��4��4�4�4,�44�48�4<�4@�4h�4p�4t�4x�4|�4��4��4��4��4��4�4�4�4�4�4�4$�4(�4,�40�4X�4`�4d�4h�4l�4��4��4��4��4��4л4ػ4ܻ4�4�4�4�4�4�4 �4H�4P�4T�4X�4\�4��4��4��4��4��4��4ȼ4̼4м4Լ4��4�4�4�4�48�4@�4D�4H�4L�4t�4|�4��4��4��4��4��4��4Ľ4�4�4��4��4�4(�40�44�48�4<�4d�4l�4p�4t�4x�4��4��4��4��4��4ܾ4�4�4�4�4�4 �4$�4(�4,�4T�4\�4`�4d�4h�4��4��4��4��4��4̿4Կ4ؿ4ܿ4�4�4�4�4�4�4D�4L�4P�4T�4X�4��4��4��4��4��4��4�4�4�4�4��4�4�4�44�4<�4@�4D�4H�4p�4x�4|�4��4��4��4��4��4��4�4�4�4��4��4$�4,�40�44�48�4`�4h�4l�4p�4t�4��4��4��4��4��4��4��4��4��4��4�4�4 �4$�4(�4P�4X�4\�4`�4d�4��4��4��4��4��4��4��4��4��4��4�4�4�4�4@�4H�4L�4P�4T�4|�4��4��4��4��4��4��4��4��4��4�4�40�48�4<�4@�4l�4t�4x�4|�4��4��4��4��4��4��4��4��4 �4(�4,�40�4\�4d�4h�4l�4��4��4��4��4��4��4��4��4�4�4�4 �4L�4T�4X�4\�4��4��4��4��4��4��4��4��4�4�4�4�4<�4D�4H�4L�4x�4��4��4��4��4��4��4��4��4��4��4�4,�44�48�4<�4h�4p�4t�4x�4��4��4��4��4��4��4��4��4�4$�4(�4,�4X�4`�4d�4h�4��4��4��4��4��4��4��4��4�4�4�4�4 �4H�4P�4T�4X�4\�4��4��4��4��4��4��4��4��4��4�4�4�48�4@�4D�4H�4t�4|�4��4��4��4��4��4��4��4��4��4��4��4��4�4(�40�44�48�4<�4d�4l�4p�4t�4��4��4��4��4��4��4��4��4��4�4 �4$�4(�4T�4\�4`�4d�4��4��4��4��4��4��4��4��4��4�4�4�4�4�4D�4L�4P�4T�4X�4��4��4��4��4��4��4��4��4��4��4��4�4�4�4�44�4<�4@�4D�4H�4p�4x�4|�4��4��4��4��4��4��4��4��4��4��4$�4,�40�44�4`�4h�4l�4p�4��4��4��4��4��4��4��4��4��4�4�4 �4$�4(�4P�4X�4\�4`�4d�4��4��4��4��4��4��4��4��4��4��4�4�4�4�4�4@�4H�4L�4P�4T�4|�4��4��4��4��4��4��4��4��4��4��4��4�4�4�40�48�4<�4@�4D�4l�4t�4x�4|�4��4��4��4��4��4��4��4��4��4��4��4 �4(�4,�40�44�4\�4d�4h�4l�4p�4��4��4��4��4��4��4��4��4��4��4�4�4�4 �4$�4L�4T�4X�4\�4`�4��4��4��4��4��4��4��4��4��4�4�4�4�4<�4D�4H�4L�4P�4x�4��4��4��4��4��4��4��4��4��4��4��4��4�4�4,�44�48�4<�4@�4h�4p�4t�4x�4|�4��4��4��4��4��4��4��4��4��4��4�4$�4(�4,�40�4X�4`�4d�4h�4l�4��4��4��4��4��4��4��4��4��4��4�4�4�4�4 �4H�4P�4T�4X�4\�4��4��4��4��4��4��4��4��4��4��4�4�4�48�4@�4D�4H�4t�4|�4��4��4��4��4��4��4��4��4��4��4�4(�40�44�48�4<�4d�4l�4p�4t�4x�4��4��4��4��4��4��4��4��4��4��4�4 �4$�4(�4,�4T�4\�4`�4d�4h�4��4��4��4��4��4��4��4��4��4�4�4�4�4D�4L�4P�4T�4��4��4��4��4��4��4��4��4��4�4�4�44�4<�4@�4D�4p�4x�4|�4��4��4��4��4��4��4��4��4��4$�4,�40�44�48�4`�4h�4l�4p�4t�4��4��4��4��4��4��4��4��4��4��4�4�4 �4$�4(�4P�4X�4\�4`�4d�4��4��4��4��4��4��4��4��4��4��4�4�4�4�4�4@�4H�4L�4P�4T�4|�4��4��4��4��4��4��4��4��4��4�4�40�48�4<�4@�4l�4t�4x�4|�4��4��4��4��4��4��4��4��4 �4(�4,�40�4\�4d�4h�4l�4��4��4��4��4��4��4��4��4�4�4�4 �4L�4T�4X�4\�4��4��4��4��4��4��4��4��4�4�4�4�4<�4D�4H�4L�4x�4��4��4��4��4��4��4��4��4��4��4�4,�44�48�4<�4h�4p�4t�4x�4��4��4��4��4��4��4��4��4�4$�4(�4,�4X�4`�4d�4h�4��4��4��4��4��4��4��4��4�4�4�4�4H�4P�4T�4X�4��4��4��4��4��4��4��4��4��4��4�4�4�48�4@�4D�4H�4t�4|�4��4��4��4��4��4��4��4��4��4��4�4(�40�44�48�4d�4l�4p�4t�4x�4��4��4��4��4��4��4��4��4��4�4 �4$�4(�4,�4T�4\�4`�4d�4h�4��4��4��4��4��4��4��4��4��4��4�4�4�4�4�4D�4L�4P�4T�4X�4��4��4��4��4��4��4��4��4��4��4��4�4�4�4�44�4<�4@�4D�4H�4p�4x�4|�4��4��4��4��4��4��4��4��4��4��4��4��4$�4,�40�44�48�4`�4h�4l�4p�4t�4��4��4��4��4��4��4��4��4��4��4�4�4 �4$�4(�4P�4X�4\�4`�4d�4��4��4��4��4��4��4��4��4��4��4�4�4�4�4�4@�4H�4L�4P�4|�4��4��4��4��4��4��4��4��4��4��4�4�4�40�48�4<�4@�4D�4l�4t�4x�4|�4��4��4��4��4��4��4��4��4��4 �4(�4,�40�4\�4d�4h�4l�4��4��4��4��4��4��4��4��4�4�4�4 �4$�4L�4T�4X�4\�4`�4��4��4��4��4��4��4��4��4��4��4�4�4�4�4<�4D�4H�4L�4P�4x�4��4��4��4��4��4��4��4�4�4�4��4��4�4�4,�44�48�4<�4@�4h�4p�4t�4x�4��4��4��4��4�4�4�4�4�4�4$�4(�4,�4X�4`�4d�4h�4l�4��4��4��4��4�4�4�4�4�4�4�4�4�4H�4P�4T�4X�4\�4��4��4��4��4��4�4�4�4��4�4�4�4�48�4@�4D�4H�4L�4t�4|�4��4��4��4��4��4��4�4�4��4��4(�40�44�48�4d�4l�4p�4t�4x�4��4��4��4��4��4�4�4�4�4�4�4 �4$�4(�4,�4T�4\�4`�4d�4h�4��4��4��4��4��4�4�4�4�4�4�4�4�4�4�4D�4L�4P�4T�4X�4��4��4��4��4��4�4�4�4�4��4�4�4�4�44�4<�4@�4D�4p�4x�4|�4��4��4��4��4��4��4��4�4�4�4��4��4$�4,�40�44�4`�4h�4l�4p�4t�4��4��4��4��4��4�4�4�4�4�4�4 �4$�4(�4P�4X�4\�4`�4d�4��4��4��4��4��4�4�4�4�4�455555@5H5L5P5T5|5�5�5�5�5�5�5�5�5�5�5�55550585<5@5D5l5t5x5|5�5�5�5�5�5�5�5�5�5�5�5 5(5,50545\5d5h5l5p5�5�5�5�5�5�5�5�5�5�5555 5L5T5X5\5�5�5�5�5�5�5�5�5�5�555555<5D5H5L5x5�5�5�5�5�5�5�5�5�5�5�555,54585<5h5p5t5x5�5�5�5�5�5�5�5�5�5�5�5�5�555$5(5,545<5X5`5d5h5p5x5�5�5�5�5�5�5�5�5�5�5�5�55555$5,5H5P5T5X5`5h5�5�5�5�5�5�5�5�5�5�5�5�5�55555585@5D5H5P5X5t5|5�5�5�5�5�5�5�5�5�5�5�5�5�5�5	5	5(	50	54	58	5@	5H	5d	5l	5p	5t	5|	5�	5�	5�	5�	5�	5�	5�	5�	5�	5�	5�	5�	5�	5
5 
5$
5(
50
58
5T
5\
5`
5d
5l
5t
5�
5�
5�
5�
5�
5�
5�
5�
5�
5�
5�
5�
55555 5(5D5L5P5T5\5d5�5�5�5�5�5�5�5�5�5�5�5�5�55555545<5@5D5L5T5p5x5|5�5�5�5�5�5�5�5�5�5�5�5�5�5
5
5$
5,
50
54
5<
5D
5`
5h
5l
5p
5x
5�
5�
5�
5�
5�
5�
5�
5�
5�
5�
5�
5�
5�
555 5$5,545P5X5\5`5h5p5�5�5�5�5�5�5�5�5�5�5�5�555555$5@5H5L5P5X5`5|5�5�5�5�5�5�5�5�5�5�5�5�5�555550585<5@5H5P5l5t5x5|5�5�5�5�5�5�5�5�5�5�5�5�5�55 5(5,50585@5\5d5h5l5t5|5�5�5�5�5�5�5�5�5�5�5�5�5555 5(505L5T5X5\5d5l5�5�5�5�5�5�5�5�5�5�5�5�555555 5<5D5H5L5T5\5x5�5�5�5�5�5�5�5�5�5�5�5�5�5�5555,54585<5D5L5h5p5t5x5�5�5�5�5�5�5�5�5�5�5�5�5�555$5(5,545<5X5`5d5h5p5x5�5�5�5�5�5�5�5�5�5�5�5�55555$5,5H5P5T5X5`5h5�5�5�5�5�5�5�5�5�5�5�5�5�55555585@5D5H5P5X5t5|5�5�5�5�5�5�5�5�5�5�5�5�5�5�555(5054585@5H5d5l5p5t5|5�5�5�5�5�5�5�5�5�5�5�5�5�55 5$5(50585T5\5`5d5l5t5�5�5�5�5�5�5�5�5�5�5�5�55555 5(5D5L5P5T5\5d5�5�5�5�5�5�5�5�5�5�5�5�5�55555545<5@5D5L5T5p5x5|5�5�5�5�5�5�5�5�5�5�5�5�5�555$5,50545<5D5`5h5l5p5x5�5�5�5�5�5�5�5�5�5�5�5�5�555 5$5,545P5X5\5`5h5p5�5�5�5�5�5�5�5�5�5�5�5�555555$5@5H5L5P5X5`5|5�5�5�5�5�5�5�5�5�5�5�5�5�555550585<5@5H5P5l5t5x5|5�5�5�5�5�5�5�5�5�5�5�5�5�5 5  5( 5, 50 58 5@ 5\ 5d 5h 5l 5t 5| 5� 5� 5� 5� 5� 5� 5� 5� 5� 5� 5� 5� 5!5!5!5 !5(!50!5L!5T!5X!5\!5d!5l!5�!5�!5�!5�!5�!5�!5�!5�!5�!5�!5�!5�!5"5"5"5"5"5 "5<"5D"5H"5L"5T"5\"5x"5�"5�"5�"5�"5�"5�"5�"5�"5�"5�"5�"5�"5�"5�"5#5#5#5,#54#58#5<#5D#5L#5h#5p#5t#5x#5�#5�#5�#5�#5�#5�#5�#5�#5�#5�#5�#5�#5�#5$5$5$$5($5,$54$5<$5X$5`$5d$5h$5p$5x$5�$5�$5�$5�$5�$5�$5�$5�$5�$5�$5�$5�$5%5%5%5%5$%5,%5H%5P%5T%5X%5`%5h%5�%5�%5�%5�%5�%5�%5�%5�%5�%5�%5�%5�%5�%5&5&5&5&5&58&5@&5D&5H&5P&5X&5t&5|&5�&5�&5�&5�&5�&5�&5�&5�&5�&5�&5�&5�&5�&5�&5'5'5('50'54'58'5@'5H'5d'5l'5p'5t'5|'5�'5�'5�'5�'5�'5�'5�'5�'5�'5�'5�'5�'5�'5(5 (5$(5((50(58(5T(5\(5`(5d(5l(5t(5�(5�(5�(5�(5�(5�(5�(5�(5�(5�(5�(5�(5)5)5)5)5 )5()5D)5L)5P)5T)5\)5d)5�)5�)5�)5�)5�)5�)5�)5�)5�)5�)5�)5�)5�)5*5*5*5*5*54*5<*5@*5D*5L*5T*5p*5x*5|*5�*5�*5�*5�*5�*5�*5�*5�*5�*5�*5�*5�*5�*5+5+5$+5,+50+54+5<+5D+5`+5h+5l+5p+5x+5�+5�+5�+5�+5�+5�+5�+5�+5�+5�+5�+5�+5�+5,5,5 ,5$,5,,54,5P,5X,5\,5`,5h,5p,5�,5�,5�,5�,5�,5�,5�,5�,5�,5�,5�,5�,5-5-5-5-5-5$-5@-5H-5L-5P-5X-5`-5|-5�-5�-5�-5�-5�-5�-5�-5�-5�-5�-5�-5�-5�-5.5.5.5.50.58.5<.5@.5H.5P.5l.5t.5x.5|.5�.5�.5�.5�.5�.5�.5�.5�.5�.5�.5�.5�.5�.5/5 /5(/5,/50/58/5@/5\/5d/5h/5l/5t/5|/5�/5�/5�/5�/5�/5�/5�/5�/5�/5�/5�/5�/5050505 05(05005L05T05X05\05d05l05�05�05�05�05�05�05�05�05�05�05�05�051515151515 15<15D15H15L15T15\15x15�15�15�15�15�15�15�15�15�15�15�15�15�15�15252525,25425825<25D25L25h25p25t25x25�25�25�25�25�25�25�25�25�25�25�25�25�253535$35(35,35435<35X35`35d35h35p35x35�35�35�35�35�35�35�35�35�35�35�35�3545454545$45,45H45P45T45X45`45h45�45�45�45�45�45�45�45�45�45�45�45�45�455555555555855@55D55H55P55X55t55|55�55�55�55�55�55�55�55�55�55�55�55�55�55�556565(65065465865@65H65d65l65p65t65|65�65�65�65�65�65�65�65�65�65�65�65�65�6575 75$75(75075875T75\75`75d75l75t75�75�75�75�75�75�75�75�75�75�75�75�7585858585 85(85D85L85P85T85\85d85�85�85�85�85�85�85�85�85�85�85�85�85�859595959595495<95@95D95L95T95p95x95|95�95�95�95�95�95�95�95�95�95�95�95�95�95:5:5$:5,:50:54:5<:5D:5`:5h:5l:5p:5x:5�:5�:5�:5�:5�:5�:5�:5�:5�:5�:5�:5�:5�:5;5;5 ;5$;5,;54;5P;5X;5\;5`;5h;5p;5�;5�;5�;5�;5�;5�;5�;5�;5�;5�;5�;5�;5<5<5<5<5<5$<5@<5H<5L<5P<5X<5`<5|<5�<5�<5�<5�<5�<5�<5�<5�<5�<5�<5�<5�<5�<5=5=5=5=50=58=5<=5@=5H=5P=5l=5t=5x=5|=5�=5�=5�=5�=5�=5�=5�=5�=5�=5�=5�=5�=5�=5>5 >5(>5,>50>58>5@>5\>5d>5h>5l>5t>5|>5�>5�>5�>5�>5�>5�>5�>5�>5�>5�>5�>5�>5?5?5?5 ?5(?50?5L?5T?5X?5\?5d?5l?5�?5�?5�?5�?5�?5�?5�?5�?5�?5�?5�?5�?5@5@5@5@5@5 @5<@5D@5H@5L@5T@5\@5x@5�@5�@5�@5�@5�@5�@5�@5�@5�@5�@5�@5�@5�@5�@5A5A5A5,A54A58A5<A5DA5LA5hA5pA5tA5xA5�A5�A5�A5�A5�A5�A5�A5�A5�A5�A5�A5�A5�A5B5B5$B5(B5,B54B5<B5XB5`B5dB5hB5pB5xB5�B5�B5�B5�B5�B5�B5�B5�B5�B5�B5�B5�B5C5C5C5C5$C5,C5HC5PC5TC5XC5`C5hC5�C5�C5�C5�C5�C5�C5�C5�C5�C5�C5�C5�C5�C5D5D5D5D5D58D5@D5DD5HD5PD5XD5tD5|D5�D5�D5�D5�D5�D5�D5�D5�D5�D5�D5�D5�D5�D5�D5E5E5(E50E54E58E5@E5HE5dE5lE5pE5tE5|E5�E5�E5�E5�E5�E5�E5�E5�E5�E5�E5�E5�E5�E5F5 F5$F5(F50F58F5TF5\F5`F5dF5lF5tF5�F5�F5�F5�F5�F5�F5�F5�F5�F5�F5�F5�F5G5G5G5G5 G5(G5DG5LG5PG5TG5\G5dG5�G5�G5�G5�G5�G5�G5�G5�G5�G5�G5�G5�G5�G5H5H5H5H5H54H5<H5@H5DH5LH5TH5pH5xH5|H5�H5�H5�H5�H5�H5�H5�H5�H5�H5�H5�H5�H5�H5I5I5$I5,I50I54I5<I5DI5`I5hI5lI5pI5xI5�I5�I5�I5�I5�I5�I5�I5�I5�I5�I5�I5�I5�I5J5J5 J5$J5,J54J5PJ5XJ5\J5`J5hJ5pJ5�J5�J5�J5�J5�J5�J5�J5�J5�J5�J5�J5�J5K5K5K5K5K5$K5@K5HK5LK5PK5XK5`K5|K5�K5�K5�K5�K5�K5�K5�K5�K5�K5�K5�K5�K5�K5L5L5L5L50L58L5<L5@L5HL5PL5lL5tL5xL5|L5�L5�L5�L5�L5�L5�L5�L5�L5�L5�L5�L5�L5�L5M5 M5(M5,M50M58M5@M5\M5dM5hM5lM5tM5|M5�M5�M5�M5�M5�M5�M5�M5�M5�M5�M5�M5�M5N5N5N5 N5(N50N5LN5TN5XN5\N5dN5lN5�N5�N5�N5�N5�N5�N5�N5�N5�N5�N5�N5�N5O5O5O5O5O5 O5<O5DO5HO5LO5TO5\O5xO5�O5�O5�O5�O5�O5�O5�O5�O5�O5�O5�O5�O5�O5�O5P5P5P5,P54P58P5<P5DP5LP5hP5pP5tP5xP5�P5�P5�P5�P5�P5�P5�P5�P5�P5�P5�P5�P5�P5Q5Q5$Q5(Q5,Q54Q5<Q5XQ5`Q5dQ5hQ5pQ5xQ5�Q5�Q5�Q5�Q5�Q5�Q5�Q5�Q5�Q5�Q5�Q5�Q5R5R5R5R5$R5,R5HR5PR5TR5XR5`R5hR5�R5�R5�R5�R5�R5�R5�R5�R5�R5�R5�R5�R5�R5S5S5S5S5S58S5@S5DS5HS5PS5XS5tS5|S5�S5�S5�S5�S5�S5�S5�S5�S5�S5�S5�S5�S5�S5�S5T5T5(T50T54T58T5@T5HT5dT5lT5pT5tT5|T5�T5�T5�T5�T5�T5�T5�T5�T5�T5�T5�T5�T5�T5U5 U5$U5(U50U58U5TU5\U5`U5dU5lU5tU5�U5�U5�U5�U5�U5�U5�U5�U5�U5�U5�U5�U5V5V5V5V5 V5(V5DV5LV5PV5TV5\V5dV5�V5�V5�V5�V5�V5�V5�V5�V5�V5�V5�V5�V5�V5W5W5W5W5W54W5<W5@W5DW5LW5TW5pW5xW5|W5�W5�W5�W5�W5�W5�W5�W5�W5�W5�W5�W5�W5�W5X5X5$X5,X50X54X5<X5DX5`X5hX5lX5pX5xX5�X5�X5�X5�X5�X5�X5�X5�X5�X5�X5�X5�X5�X5Y5Y5 Y5$Y5,Y54Y5PY5XY5\Y5`Y5hY5pY5�Y5�Y5�Y5�Y5�Y5�Y5�Y5�Y5�Y5�Y5�Y5�Y5Z5Z5Z5Z5Z5$Z5@Z5HZ5LZ5PZ5XZ5`Z5|Z5�Z5�Z5�Z5�Z5�Z5�Z5�Z5�Z5�Z5�Z5�Z5�Z5�Z5[5[5[5[50[58[5<[5@[5H[5P[5l[5t[5x[5|[5�[5�[5�[5�[5�[5�[5�[5�[5�[5�[5�[5�[5�[5\5 \5(\5,\50\58\5@\5\\5d\5h\5l\5t\5|\5�\5�\5�\5�\5�\5�\5�\5�\5�\5�\5�\5�\5]5]5]5 ]5(]50]5L]5T]5X]5\]5d]5l]5�]5�]5�]5�]5�]5�]5�]5�]5�]5�]5�]5�]5^5^5^5^5^5 ^5<^5D^5H^5L^5T^5\^5x^5�^5�^5�^5�^5�^5�^5�^5�^5�^5�^5�^5�^5�^5�^5_5_5_5,_54_58_5<_5D_5L_5h_5p_5t_5x_5�_5�_5�_5�_5�_5�_5�_5�_5�_5�_5�_5�_5�_5`5`5$`5(`5,`54`5<`5X`5``5d`5h`5p`5x`5�`5�`5�`5�`5�`5�`5�`5�`5�`5�`5�`5�`5a5a5a5a5$a5,a5Ha5Pa5Ta5Xa5`a5ha5�a5�a5�a5�a5�a5�a5�a5�a5�a5�a5�a5�a5�a5b5b5b5b5b58b5@b5Db5Hb5Pb5Xb5tb5|b5�b5�b5�b5�b5�b5�b5�b5�b5�b5�b5�b5�b5�b5�b5c5c5(c50c54c58c5@c5Hc5dc5lc5pc5tc5|c5�c5�c5�c5�c5�c5�c5�c5�c5�c5�c5�c5�c5�c5d5 d5$d5(d50d58d5Td5\d5`d5dd5ld5td5�d5�d5�d5�d5�d5�d5�d5�d5�d5�d5�d5�d5e5e5e5e5 e5(e5De5Le5Pe5Te5\e5de5�e5�e5�e5�e5�e5�e5�e5�e5�e5�e5�e5�e5�e5f5f5f5f5f54f5<f5@f5Df5Lf5Tf5pf5xf5|f5�f5�f5�f5�f5�f5�f5�f5�f5�f5�f5�f5�f5�f5g5g5$g5,g50g54g5<g5Dg5`g5hg5lg5pg5xg5�g5�g5�g5�g5�g5�g5�g5�g5�g5�g5�g5�g5�g5h5h5 h5$h5,h54h5Ph5Xh5\h5`h5hh5ph5�h5�h5�h5�h5�h5�h5�h5�h5�h5�h5�h5�h5i5i5i5i5i5$i5@i5Hi5Li5Pi5Xi5`i5|i5�i5�i5�i5�i5�i5�i5�i5�i5�i5�i5�i5�i5�i5j5j5j5j50j58j5<j5@j5Hj5Pj5lj5tj5xj5|j5�j5�j5�j5�j5�j5�j5�j5�j5�j5�j5�j5�j5�j5k5 k5(k5,k50k58k5@k5\k5dk5hk5lk5tk5|k5�k5�k5�k5�k5�k5�k5�k5�k5�k5�k5�k5�k5l5l5l5 l5(l50l5Ll5Tl5Xl5\l5dl5ll5�l5�l5�l5�l5�l5�l5�l5�l5�l5�l5�l5�l5m5m5m5m5m5 m5<m5Dm5Hm5Lm5Tm5\m5xm5�m5�m5�m5�m5�m5�m5�m5�m5�m5�m5�m5�m5�m5�m5n5n5n5,n54n58n5<n5Dn5Ln5hn5pn5tn5xn5�n5�n5�n5�n5�n5�n5�n5�n5�n5�n5�n5�n5�n5o5o5$o5(o5,o54o5<o5Xo5`o5do5ho5po5xo5�o5�o5�o5�o5�o5�o5�o5�o5�o5�o5�o5�o5p5p5p5p5$p5,p5Hp5Pp5Tp5Xp5`p5hp5�p5�p5�p5�p5�p5�p5�p5�p5�p5�p5�p5�p5�p5q5q5q5q5q58q5@q5Dq5Hq5Pq5Xq5tq5|q5�q5�q5�q5�q5�q5�q5�q5�q5�q5�q5�q5�q5�q5�q5r5r5(r50r54r58r5@r5Hr5dr5lr5pr5tr5|r5�r5�r5�r5�r5�r5�r5�r5�r5�r5�r5�r5�r5�r5s5 s5$s5(s50s58s5Ts5\s5`s5ds5ls5ts5�s5�s5�s5�s5�s5�s5�s5�s5�s5�s5�s5�s5t5t5t5t5 t5(t5Dt5Lt5Pt5Tt5\t5dt5�t5�t5�t5�t5�t5�t5�t5�t5�t5�t5�t5�t5�t5u5u5u5u5u54u5<u5@u5Du5Lu5Tu5pu5xu5|u5�u5�u5�u5�u5�u5�u5�u5�u5�u5�u5�u5�u5�u5v5v5$v5,v50v54v5<v5Dv5`v5hv5lv5pv5xv5�v5�v5�v5�v5�v5�v5�v5�v5�v5�v5�v5�v5�v5w5w5 w5$w5,w54w5Pw5Xw5\w5`w5hw5pw5�w5�w5�w5�w5�w5�w5�w5�w5�w5�w5�w5�w5x5x5x5x5x5$x5@x5Hx5Lx5Px5Xx5`x5|x5�x5�x5�x5�x5�x5�x5�x5�x5�x5�x5�x5�x5�x5y5y5y5y50y58y5<y5@y5Hy5Py5ly5ty5xy5|y5�y5�y5�y5�y5�y5�y5�y5�y5�y5�y5�y5�y5�y5z5 z5(z5,z50z58z5@z5\z5dz5hz5lz5tz5|z5�z5�z5�z5�z5�z5�z5�z5�z5�z5�z5�z5�z5{5{5{5 {5({50{5L{5T{5X{5\{5d{5l{5�{5�{5�{5�{5�{5�{5�{5�{5�{5�{5�{5�{5|5|5|5|5|5 |5<|5D|5H|5L|5T|5\|5x|5�|5�|5�|5�|5�|5�|5�|5�|5�|5�|5�|5�|5�|5�|5}5}5}5,}54}58}5<}5D}5L}5h}5p}5t}5x}5�}5�}5�}5�}5�}5�}5�}5�}5�}5�}5�}5�}5�}5~5~5$~5(~5,~54~5<~5X~5`~5d~5h~5l~5�~5�~5�~5�~5�~5�~5�~5�~5�~5�~55555 5H5P5T5X5\5�5�5�5�5�5�5�5�5�5�5�5�5�5�5�58�5@�5D�5H�5L�5t�5|�5��5��5��5��5��5��5��5Ā5�5�5��5��5�5(�50�54�58�5<�5d�5l�5p�5t�5x�5��5��5��5��5��5܁5�5�5�5�5�5 �5$�5(�5,�5T�5\�5`�5d�5h�5��5��5��5��5��5̂5Ԃ5؂5܂5�5D�5L�5P�5T�5��5��5��5��5��5ă5ȃ5̃5Ѓ5��5�5�5�54�5<�5@�5D�5p�5x�5|�5��5��5��5��5��5�5�5�5��5$�5,�50�54�5`�5h�5l�5p�5��5��5��5��5؅5�5�5�5�5�5 �5$�5P�5X�5\�5`�5��5��5��5��5Ȇ5І5Ԇ5؆5�5�5�5�5@�5H�5L�5P�5|�5��5��5��5��5��5ć5ȇ5�5��5�5�50�58�5<�5@�5l�5t�5x�5|�5��5��5��5��5�5�5�5�5 �5(�5,�50�5\�5d�5h�5l�5��5��5��5��5ԉ5܉5�5�5�5�5�5 �5L�5T�5X�5\�5��5��5��5��5Ċ5̊5Њ5Ԋ5�5�5�5�5<�5D�5H�5L�5x�5��5��5��5��5��5��5ċ5�5��5��5�5,�54�58�5<�5h�5p�5t�5x�5��5��5��5��5�5�5�5�5�5$�5(�5,�5X�5`�5d�5h�5��5��5��5��5Ѝ5؍5܍5�5�5�5�5�5H�5P�5T�5X�5��5��5��5��5��5Ȏ5̎5Ў5��5�5�5�58�5@�5D�5H�5t�5|�5��5��5��5��5��5��5�5�5��5��5(�50�54�58�5@�5d�5l�5p�5t�5|�5��5��5��5��5��5ܐ5�5�5�5�5�5 �5$�5(�50�5T�5\�5`�5d�5l�5��5��5��5��5̑5ԑ5ؑ5ܑ5�5�5�5�5�5�5D�5L�5P�5T�5X�5��5��5��5��5��5��5Ē5Ȓ5̒5В5��5�5�5�5�54�5<�5@�5D�5H�5p�5x�5|�5��5��5��5��5��5��5��5�5�5�5��5$�5,�50�54�58�5`�5h�5l�5p�5t�5��5��5��5��5��5ؔ5�5�5�5�5�5 �5$�5P�5X�5\�5`�5��5��5��5��5ȕ5Е5ԕ5ؕ5�5�5�5�5@�5H�5L�5P�5|�5��5��5��5��5��5��5��5��5Ė5Ȗ5̖5Ж5ؖ5�5��5�5�5�5�5�50�58�5<�5@�5D�5H�5P�5l�5t�5x�5|�5��5��5��5��5��5��5��5��5��5ȗ5�5�5�5�5��5��5�5 �5(�5,�50�54�58�5@�5\�5d�5h�5l�5p�5t�5|�5��5��5��5��5��5��5��5Ԙ5ܘ5�5�5�5�5�5�5�5�5 �5$�5(�50�5L�5T�5X�5\�5`�5��5��5��5��5��5ę5̙5Й5ԙ5ؙ5�5�5�5�5�5<�5D�5H�5L�5P�5x�5��5��5��5��5��5��5��5Ě5Ț5�5��5��5�5�5,�54�58�5<�5@�5h�5p�5t�5x�5|�5��5��5��5��5�5�5�5�5�5�5$�5(�5,�50�5X�5`�5d�5h�5l�5��5��5��5��5��5М5؜5ܜ5�5�5�5�5�5�5 �5H�5P�5T�5X�5\�5��5��5��5��5��5��5ȝ5̝5Н5ԝ5��5�5�5�5�58�5@�5D�5H�5L�5t�5|�5��5��5��5��5��5��5��5Ğ5�5�5��5��5�5(�50�54�58�5<�5d�5l�5p�5t�5x�5��5��5��5��5��5ܟ5�5�5�5�5�5 �5$�5(�5,�5T�5\�5`�5d�5h�5��5��5��5��5��5̠5Ԡ5ؠ5ܠ5�5�5�5�5D�5L�5P�5T�5��5��5��5��5��5ġ5ȡ5̡5��5�5�5�54�5<�5@�5D�5p�5x�5|�5��5��5��5��5��5�5�5�5��5$�5,�50�54�5`�5h�5l�5p�5��5��5��5��5أ5�5�5�5�5�5 �5$�5P�5X�5\�5`�5��5��5��5��5Ȥ5Ф5Ԥ5ؤ5�5�5�5�5@�5H�5L�5P�5|�5��5��5��5��5��5ĥ5ȥ5�5��5�5�50�58�5<�5@�5l�5t�5x�5|�5��5��5��5��5�5�5�5�5 �5(�5,�50�5\�5d�5h�5l�5��5��5��5��5ԧ5ܧ5�5�5�5�5�5 �5L�5T�5X�5\�5��5��5��5��5Ĩ5̨5Ш5Ԩ5�5�5�5�5<�5D�5H�5L�5x�5��5��5��5��5��5��5ĩ5�5��5��5�58�5<�5H�5T�5X�5\�5h�5p�5t�5x�5��5��5̪5Ъ5Ԫ5�5�5�5�5�5$�5(�5,�5d�5h�5��5��5��5��5��5��5��5ī5Ы5ث5ܫ5�5�5�5�5�5T�5X�5p�5t�5x�5��5��5��5��5��5��5Ȭ5̬5Ь5��5�5�5�58�5@�5D�5H�5t�5|�5��5��5��5��5��5��5��5ȭ5��5��5�5�5�5�54�58�5D�5P�5T�5X�5p�5t�5��5��5��5��5��5��5��5ܮ5�5�5�5�5 �5$�5(�5T�5\�5`�5d�5��5��5��5��5̯5ԯ5د5ܯ5�5�5�5�5P�5T�5`�5d�5l�5p�5t�5��5��5��5��5��5��5��5��5�5�5�54�5<�5@�5D�5p�5x�5|�5��5��5��5��5��5��5�5�5�5��5$�5,�50�54�5`�5h�5l�5p�5��5��5��5��5ز5�5�5�5�5�5 �5$�5P�5X�5\�5`�5��5��5��5��5ȳ5г5Գ5س5�5�5�5�5@�5H�5L�5P�5|�5��5��5��5��5��5Ĵ5ȴ5�5��5�5�50�58�5<�5@�5l�5t�5x�5|�5��5��5��5��5�5�5�5�5 �5(�5,�50�5\�5d�5h�5l�5��5��5��5��5Զ5ܶ5�5�5�5�5�5 �5L�5T�5X�5\�5��5��5��5��5ķ5̷5з5Է5�5�5�5�5<�5D�5H�5L�5x�5��5��5��5��5��5��5ĸ5�5��5��5�5,�54�58�5<�5h�5p�5t�5x�5��5��5��5��5�5�5�5�5�5$�5(�5,�5X�5`�5d�5h�5��5��5��5��5к5غ5ܺ5�5�5�5�5�5H�5P�5T�5X�5��5��5��5��5��5Ȼ5̻5л5��5�5�5�58�5@�5D�5H�5t�5|�5��5��5��5��5��5��5�5�5��5��5(�50�54�58�5d�5l�5p�5t�5��5��5��5��5ܽ5�5�5�5�5 �5$�5(�5T�5\�5`�5d�5��5��5��5��5̾5Ծ5ؾ5ܾ5�5�5�5�5D�5L�5P�5T�5��5��5��5��5��5Ŀ5ȿ5̿5Կ5��5�5�5�5�54�5<�5@�5D�5L�5p�5x�5|�5��5��5��5��5��5��5�5�5�5�5��5�5$�5,�50�54�5`�5h�5l�5p�5t�5��5��5��5��5��5�5�5�5�5�5�5�5 �5$�5(�5P�5X�5\�5`�5d�5��5��5��5��5��5��5��5��5��5��5�5�5�5�5�5@�5H�5L�5P�5T�5|�5��5��5��5��5��5��5��5��5��5��5�5�5�50�58�5<�5@�5D�5l�5t�5x�5|�5��5��5��5��5��5��5��5��5 �5(�5,�50�5\�5d�5h�5l�5��5��5��5��5��5��5��5��5�5�5�5 �5L�5T�5X�5\�5`�5d�5l�5��5��5��5��5��5��5��5��5��5��5��5��5��5��5�5�5�5�5�5�5 �5<�5D�5H�5L�5P�5T�5\�5x�5��5��5��5��5��5��5��5��5��5��5��5��5��5��5��5��5�5�5�5�5,�54�58�5<�5@�5D�5L�5h�5p�5t�5x�5|�5��5��5��5��5��5��5��5��5��5��5��5��5��5��5��5�5�5$�5(�5,�50�5X�5`�5d�5h�5l�5��5��5��5��5��5��5��5��5��5��5�5�5�5�5 �5H�5P�5T�5X�5��5��5��5��5��5��5��5��5��5�5�5�58�5@�5D�5H�5t�5|�5��5��5��5��5��5��5��5��5��5��5(�50�54�58�5d�5l�5p�5t�5��5��5��5��5��5��5��5��5�5 �5$�5(�5T�5\�5`�5d�5��5��5��5��5��5��5��5��5�5�5�5�5D�5L�5P�5T�5X�5��5��5��5��5��5��5��5��5��5�5�5�54�5<�5@�5D�5p�5x�5|�5��5��5��5��5��5��5��5��5��5$�5,�50�54�5`�5h�5l�5p�5��5��5��5��5��5��5��5��5�5�5 �5$�5P�5X�5\�5`�5��5��5��5��5��5��5��5��5��5�5�5�5�5@�5H�5L�5P�5|�5��5��5��5��5��5��5��5��5��5�5�50�58�5<�5@�5l�5t�5x�5|�5��5��5��5��5��5��5��5��5 �5(�5,�50�5\�5d�5h�5l�5��5��5��5��5��5��5��5��5�5�5�5 �5L�5T�5X�5\�5��5��5��5��5��5��5��5��5�5�5�5�5<�5D�5H�5L�5x�5��5��5��5��5��5��5��5��5��5��5�5,�54�58�5<�5h�5p�5t�5x�5��5��5��5��5��5��5��5��5�5$�5(�5,�5X�5`�5d�5h�5��5��5��5��5��5��5��5��5�5�5�5�5H�5P�5T�5X�5��5��5��5��5��5��5��5��5��5�5�5�58�5@�5D�5H�5t�5|�5��5��5��5��5��5��5��5��5��5��5(�50�54�58�5d�5l�5p�5t�5��5��5��5��5��5��5��5��5�5 �5$�5(�5T�5\�5`�5d�5��5��5��5��5��5��5��5��5�5�5�5�5D�5L�5P�5T�5��5��5��5��5��5��5��5��5��5�5�5�54�5<�5@�5D�5p�5x�5|�5��5��5��5��5��5��5��5��5��5$�5,�50�54�5`�5h�5l�5p�5��5��5��5��5��5��5��5��5�5�5 �5$�5\�5`�5l�5x�5|�5��5��5��5��5��5��5��5��5��5��5�5�5�5�5@�5H�5L�5P�5��5��5��5��5��5��5��5��5��5��5��5��5�5�50�58�5<�5@�5x�5|�5��5��5��5��5��5��5��5��5��5��5��5��5 �5(�5,�50�5\�5d�5h�5l�5��5��5��5��5��5��5��5��5��5��5�5 �5,�58�5<�5@�5X�5\�5h�5t�5x�5|�5��5��5��5��5��5��5��5��5��5�5�5�5�5<�5D�5H�5L�5x�5��5��5��5��5��5��5��5��5��5��5�5,�54�58�5<�5t�5x�5��5��5��5��5��5��5��5��5��5��5��5��5�5$�5(�5,�5X�5`�5d�5h�5��5��5��5��5��5��5��5��5��5��5�5�5�5�5H�5P�5T�5X�5��5��5��5��5��5��5��5��5��5�5�5�58�5@�5D�5H�5t�5|�5��5��5��5��5��5��5��5��5��5��5(�50�54�58�5d�5l�5p�5t�5��5��5��5��5��5��5��5��5�5 �5$�5(�5T�5\�5`�5d�5��5��5��5��5��5��5��5��5�5�5�5�5D�5L�5P�5T�5��5��5��5��5��5��5��5��5��5�5�5�54�5<�5@�5D�5p�5x�5|�5��5��5��5��5��5��5��5��5��5$�5,�50�54�5`�5h�5l�5p�5��5��5��5��5��5��5��5��5�5�5 �5$�5P�5X�5\�5`�5��5��5��5��5��5��5��5��5�5�5�5�5@�5H�5L�5P�5|�5��5��5��5��5��5��5��5��5��5�5�50�58�5<�5@�5l�5t�5x�5|�5��5��5��5��5��5��5��5��5 �5(�5,�50�5\�5d�5h�5l�5��5��5��5��5��5��5��5��5�5�5�5 �5L�5T�5X�5\�5��5��5��5��5��5��5��5��5�5�5�5�5<�5D�5H�5L�5x�5��5��5��5��5��5��5��5��5��5��5�5,�54�58�5<�5h�5p�5t�5x�5��5��5��5��5�5�5�5�5�5$�5(�5,�5X�5`�5d�5h�5��5��5��5��5�5�5�5�5�5�5�5�5H�5P�5T�5X�5��5��5��5��5��5��5�5�5�5�5��5�5�5�5�58�5@�5D�5H�5P�5t�5|�5��5��5��5��5��5��5��5�5�5�5��5��5(�50�54�58�5<�5d�5l�5p�5t�5x�5��5��5��5��5��5�5�5�5�5�5�5 �5$�5(�5,�5T�5\�5`�5d�5h�5��5��5��5��5��5�5�5�5�5�5�5�5�5�5�5D�5L�5P�5T�5X�5��5��5��5��5��5��5�5�5�5�5��5�5�5�5�54�5<�5@�5D�5p�5x�5|�5��5��5��5��5��5�5�5�5��5$�5,�50�54�58�5`�5h�5l�5p�5��5��5��5��5�5�5�5�5�5�5 �5$�5P�5X�5\�5`�5d�5h�5p�5��5��5��5��5��5��5��5�5�5�5�5�5�5�5�5�5�5�5�5�5$�5@�5H�5L�5P�5T�5X�5`�5|�5��5��5��5��5��5��5��5��5�5�5�5�5�5�5��5666660686<6@6D6H6P6l6t6x6|6�6�6�6�6�6�6�6�6�6�6�6�6�6�6�6�66 6(6,60646\6d6h6l6p6�6�6�6�6�6�6�6�6�6�6666 6$6L6T6X6\6�6�6�6�6�6�6�6�66666<6D6H6L6x6�6�6�6�6�6�6�6�6�6�66,64686<6h6p6t6x6�6�6�6�6�6�6�6�66$6(6,6X6`6d6h6�6�6�6�6�6�6�6�66666H6P6T6X6\6�6�6�6�6�6�6�6�6�666686@6D6H6t6|6�6�6�6�6�6�6�6�6�6�6(6064686d6l6p6t6�6�6�6�6�6�6�6�6	6 	6$	6(	6T	6\	6`	6d	6�	6�	6�	6�	6�	6�	6�	6�	6�	6
6
6
6
6D
6L
6P
6T
6�
6�
6�
6�
6�
6�
6�
6�
6�
666646<6@6D6p6x6|6�6�6�6�6�6�6�6�6�6$6,60646`6h6l6p6�6�6�6�6�6�6�6�6
6
6 
6$
6P
6X
6\
6`
6�
6�
6�
6�
6�
6�
6�
6�
66666@6H6L6P6|6�6�6�6�6�6�6�6�6�6660686<6@6l6t6x6|6�6�6�6�6�6�6�6�6 6(6,606\6d6h6l6�6�6�6�6�6�6�6�6666 6L6T6X6\6�6�6�6�6�6�6�6�66666<6D6H6L6x6�6�6�6�6�6�6�6�6�6�66,64686<6h6p6t6x6�6�6�6�6�6�6�6�66$6(6,6X6`6d6h6�6�6�6�6�6�6�6�66666H6P6T6X6�6�6�6�6�6�6�6�6�666686@6D6H6t6|6�6�6�6�6�6�6�6�6�6�6�6(6064686d6l6p6t6�6�6�6�6�6�6�6�66 6$6(6T6\6`6d6�6�6�6�6�6�6�6�66666D6L6P6T6�6�6�6�6�6�6�6�6�666646<6@6D6p6x6|6�6�6�6�6�6�6�6�6�60646@6L6P6T6`6h6l6p6�6�6�6�6�6�6�6�6�666 6$6\6`6x6|6�6�6�6�6�6�6�6�6�6�66666L6P6h6l6p6�6�6�6�6�6�6�6�6�6�6�6660686<6@6l6t6x6|6�6�6�6�6�6�6�6�66666,606H6L6P6h6l6x6�6�6�6�6�6�6�6�6�6�6�6 6 6 6  6L 6T 6X 6\ 6� 6� 6� 6� 6� 6� 6� 6� 6� 6!6!6!6!6<!6D!6H!6L!6x!6�!6�!6�!6�!6�!6�!6�!6�!6"6"6"6 "6,"64"68"6<"6h"6p"6t"6x"6�"6�"6�"6�"6�"6�"6�"6�"6�"6�"6#6#6#6#6X#6`#6d#6h#6�#6�#6�#6�#6�#6�#6�#6�#6$6$6$6$6H$6P$6T$6X$6�$6�$6�$6�$6�$6�$6�$6�$6�$6%6%6%68%6@%6D%6H%6t%6|%6�%6�%6�%6�%6�%6�%6�%6�%6�%6�%6(&60&64&68&6d&6l&6p&6t&6�&6�&6�&6�&6�&6�&6�&6�&6'6 '6$'6('6T'6\'6`'6d'6�'6�'6�'6�'6�'6�'6�'6�'6(6(6(6(6D(6L(6P(6T(6�(6�(6�(6�(6�(6�(6�(6�(6�(6)6)6)64)6<)6@)6D)6p)6x)6|)6�)6�)6�)6�)6�)6�)6�)6�)6�)6$*6,*60*64*6`*6h*6l*6p*6�*6�*6�*6�*6�*6�*6�*6�*6+6+6 +6$+6P+6X+6\+6`+6�+6�+6�+6�+6�+6�+6�+6�+6,6,6,6,6@,6H,6L,6P,6|,6�,6�,6�,6�,6�,6�,6�,6�,6�,6-6-60-68-6<-6@-6l-6t-6x-6|-6�-6�-6�-6�-6�-6�-6�-6�-6 .6(.6,.60.6\.6d.6h.6l.6�.6�.6�.6�.6�.6�.6�.6�.6/6/6/6 /6L/6T/6X/6\/6�/6�/6�/6�/6�/6�/6�/6�/606060606<06D06H06L06x06�06�06�06�06�06�06�06�06�06�0616,16416816<16h16p16t16x16�16�16�16�16�16�16�16�1626$26(26,26X26`26d26h26�26�26�26�26�26�26�26�2636363636H36P36T36X36�36�36�36�36�36�36�36�36�36464646846@46D46H46t46|46�46�46�46�46�46�46�46�46�46�46(56056456856d56l56p56t56�56�56�56�56�56�56�56�5666 66$66(66T66\66`66d66�66�66�66�66�66�66�66�6676767676D76L76P76T76�76�76�76�76�76�76�76�76�76868686486<86@86D86p86x86|86�86�86�86�86�86�86�86�86�86$96,96096496`96h96l96p96�96�96�96�96�96�96�96�96:6:6 :6$:6P:6X:6\:6`:6�:6�:6�:6�:6�:6�:6�:6�:6;6;6;6;6@;6H;6L;6P;6|;6�;6�;6�;6�;6�;6�;6�;6�;6�;6<6<60<68<6<<6@<6l<6t<6x<6|<6�<6�<6�<6�<6�<6�<6�<6�<6 =6(=6,=60=6\=6d=6h=6l=6�=6�=6�=6�=6�=6�=6�=6�=6>6>6>6 >6L>6T>6X>6\>6�>6�>6�>6�>6�>6�>6�>6�>6?6?6?6?6<?6D?6H?6L?6x?6�?6�?6�?6�?6�?6�?6�?6�?6�?6�?6@6,@64@68@6<@6h@6p@6t@6x@6�@6�@6�@6�@6�@6�@6�@6�@6�@6�@6A6$A6(A6,A64A6XA6`A6dA6hA6pA6�A6�A6�A6�A6�A6�A6�A6�A6�A6�A6B6B6B6B6$B6HB6PB6TB6XB6`B6�B6�B6�B6�B6�B6�B6�B6�B6�B6�B6�B6C6C6C6C68C6@C6DC6HC6PC6tC6|C6�C6�C6�C6�C6�C6�C6�C6�C6�C6�C6�C6�C6D6(D60D64D68D6@D6dD6lD6pD6tD6|D6�D6�D6�D6�D6�D6�D6�D6�D6E6 E6$E6(E6TE6\E6`E6dE6�E6�E6�E6�E6�E6�E6�E6�E6F6F6F6F6DF6LF6PF6TF6�F6�F6�F6�F6�F6�F6�F6�F6�F6G6G6G64G6<G6@G6DG6pG6xG6|G6�G6�G6�G6�G6�G6�G6�G6�G6�G6$H6,H60H64H6`H6hH6lH6pH6�H6�H6�H6�H6�H6�H6�H6�H6I6I6 I6$I6PI6XI6\I6`I6�I6�I6�I6�I6�I6�I6�I6�I6�I6�I6�I6J6J6J6J6J6J6$J6@J6HJ6LJ6PJ6TJ6XJ6`J6|J6�J6�J6�J6�J6�J6�J6�J6�J6�J6�J6�J6�J6�J6�J6�J6K6K6K6K6K60K68K6<K6@K6DK6HK6PK6lK6tK6xK6|K6�K6�K6�K6�K6�K6�K6�K6�K6�K6�K6�K6�K6�K6�K6�K6�K6L6 L6(L6,L60L64L68L6@L6\L6dL6hL6lL6pL6tL6|L6�L6�L6�L6�L6�L6�L6�L6�L6�L6�L6�L6�L6�L6�L6M6M6M6 M6$M6(M60M6LM6TM6XM6\M6`M6dM6lM6�M6�M6�M6�M6�M6�M6�M6�M6�M6�M6�M6�M6�M6�M6N6N6N6N6N6N6 N6<N6DN6HN6LN6PN6TN6\N6xN6�N6�N6�N6�N6�N6�N6�N6�N6�N6�N6�N6�N6�N6�N6�N6�N6O6O6O6O6,O64O68O6<O6@O6DO6LO6hO6pO6tO6xO6|O6�O6�O6�O6�O6�O6�O6�O6�O6�O6�O6�O6�O6�O6�O6�O6P6P6$P6(P6,P60P64P6<P6XP6`P6dP6hP6lP6pP6xP6�P6�P6�P6�P6�P6�P6�P6�P6�P6�P6�P6�P6�P6�P6Q6Q6Q6Q6 Q6$Q6,Q6HQ6PQ6TQ6XQ6\Q6`Q6hQ6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6�Q6R6R6R6R6R6R68R6@R6DR6HR6LR6PR6XR6tR6|R6�R6�R6�R6�R6�R6�R6�R6�R6�R6�R6�R6�R6�R6�R6�R6�R6S6S6S6(S60S64S68S6<S6@S6HS6dS6lS6pS6tS6xS6|S6�S6�S6�S6�S6�S6�S6�S6�S6�S6�S6�S6�S6�S6�S6�S6T6 T6$T6(T6,T60T68T6TT6\T6`T6dT6hT6lT6tT6�T6�T6�T6�T6�T6�T6�T6�T6�T6�T6�T6�T6�T6�T6U6U6U6U6U6 U6(U6DU6LU6PU6TU6XU6\U6dU6�U6�U6�U6�U6�U6�U6�U6�U6�U6�U6�U6�U6�U6�U6�U6V6V6V6V6V6V64V6<V6@V6DV6HV6LV6TV6pV6xV6|V6�V6�V6�V6�V6�V6�V6�V6�V6�V6�V6�V6�V6�V6�V6�V6�V6W6W6$W6,W60W64W68W6<W6DW6`W6hW6lW6pW6tW6xW6�W6�W6�W6�W6�W6�W6�W6�W6�W6�W6�W6�W6�W6�W6�W6X6X6 X6$X6(X6,X64X6PX6XX6\X6`X6dX6hX6pX6�X6�X6�X6�X6�X6�X6�X6�X6�X6�X6�X6�X6�X6�X6Y6Y6Y6Y6Y6Y6$Y6@Y6HY6LY6PY6TY6XY6`Y6|Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6�Y6Z6Z6Z6Z6Z60Z68Z6<Z6@Z6DZ6HZ6PZ6lZ6tZ6xZ6|Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6�Z6[6 [6([6,[60[64[68[6@[6\[6d[6h[6l[6p[6t[6|[6�[6�[6�[6�[6�[6�[6�[6�[6�[6�[6�[6�[6�[6�[6\6\6\6 \6$\6(\60\6L\6T\6X\6\\6`\6d\6l\6�\6�\6�\6�\6�\6�\6�\6�\6�\6�\6�\6�\6�\6�\6]6]6]6]6]6]6 ]6<]6D]6H]6L]6P]6T]6\]6x]6�]6�]6�]6�]6�]6�]6�]6�]6�]6�]6�]6�]6�]6�]6�]6�]6^6^6^6^6,^64^68^6<^6@^6D^6L^6h^6p^6t^6x^6|^6�^6�^6�^6�^6�^6�^6�^6�^6�^6�^6�^6�^6�^6�^6�^6_6_6$_6(_6,_60_64_6<_6X_6`_6d_6h_6l_6p_6x_6�_6�_6�_6�_6�_6�_6�_6�_6�_6�_6�_6�_6�_6�_6`6`6`6`6 `6$`6,`6H`6P`6T`6X`6\`6``6h`6�`6�`6�`6�`6�`6�`6�`6�`6�`6�`6�`6�`6�`6�`6�`6a6a6a6a6a6a68a6@a6Da6Ha6La6Pa6Xa6ta6|a6�a6�a6�a6�a6�a6�a6�a6�a6�a6�a6�a6�a6�a6�a6�a6�a6b6b6b6(b60b64b68b6<b6@b6Hb6db6lb6pb6tb6xb6|b6�b6�b6�b6�b6�b6�b6�b6�b6�b6�b6�b6�b6�b6�b6�b6c6 c6$c6(c6,c60c68c6Tc6\c6`c6dc6hc6lc6tc6�c6�c6�c6�c6�c6�c6�c6�c6�c6�c6�c6�c6�c6�c6d6d6d6d6d6 d6(d6Dd6Ld6Pd6Td6Xd6\d6dd6�d6�d6�d6�d6�d6�d6�d6�d6�d6�d6�d6�d6�d6�d6�d6e6e6e6e6e6e64e6<e6@e6De6He6Le6Te6pe6xe6|e6�e6�e6�e6�e6�e6�e6�e6�e6�e6�e6�e6�e6�e6�e6�e6�e6f6f6$f6,f60f64f68f6<f6Df6`f6hf6lf6pf6tf6xf6�f6�f6�f6�f6�f6�f6�f6�f6�f6�f6�f6�f6�f6�f6�f6g6g6 g6$g6(g6,g64g6Pg6Xg6\g6`g6dg6hg6pg6�g6�g6�g6�g6�g6�g6�g6�g6�g6�g6�g6�g6�g6�g6h6h6h6h6h6h6$h6@h6Hh6Lh6Ph6Th6Xh6`h6|h6�h6�h6�h6�h6�h6�h6�h6�h6�h6�h6�h6�h6�h6�h6�h6i6i6i6i6i60i68i6<i6@i6Di6Hi6Pi6li6ti6xi6|i6�i6�i6�i6�i6�i6�i6�i6�i6�i6�i6�i6 j6(j6,j60j6\j6dj6hj6lj6�j6�j6�j6�j6�j6�j6�j6�j6k6k6k6 k6Lk6Tk6Xk6\k6�k6�k6�k6�k6�k6�k6�k6�k6l6l6l6l6<l6Dl6Hl6Ll6xl6�l6�l6�l6�l6�l6�l6�l6�l6�l6�l6m6,m64m68m6<m6hm6pm6tm6xm6�m6�m6�m6�m6�m6�m6�m6�m6n6$n6(n6,n6Xn6`n6dn6hn6�n6�n6�n6�n6�n6�n6�n6�n6o6o6o6o6Ho6Po6To6Xo6�o6�o6�o6�o6�o6�o6�o6�o6�o6p6p6p68p6@p6Dp6Hp6tp6|p6�p6�p6�p6�p6�p6�p6�p6�p6�p6�p6(q60q64q68q6dq6lq6pq6tq6�q6�q6�q6�q6�q6�q6�q6�q6r6 r6$r6(r6Tr6\r6`r6dr6�r6�r6�r6�r6�r6�r6�r6�r6s6s6s6s6Ds6Ls6Ps6Ts6�s6�s6�s6�s6�s6�s6�s6�s6�s6t6t6t64t6<t6@t6Dt6pt6xt6|t6�t6�t6�t6�t6�t6�t6�t6�t6�t6$u6,u60u64u6`u6hu6lu6pu6�u6�u6�u6�u6�u6�u6�u6�u6v6v6 v6$v6Pv6Xv6\v6`v6�v6�v6�v6�v6�v6�v6�v6�v6w6w6w6w6@w6Hw6Lw6Pw6|w6�w6�w6�w6�w6�w6�w6�w6�w6�w6x6x60x68x6<x6@x6lx6tx6xx6|x6�x6�x6�x6�x6�x6�x6�x6�x6 y6(y6,y60y6\y6dy6hy6ly6�y6�y6�y6�y6�y6�y6�y6�y6z6z6z6 z6Lz6Tz6Xz6\z6�z6�z6�z6�z6�z6�z6�z6�z6{6{6{6{6<{6D{6H{6L{6x{6�{6�{6�{6�{6�{6�{6�{6�{6�{6�{6|6,|64|68|6<|6h|6p|6t|6x|6�|6�|6�|6�|6�|6�|6�|6�|6}6$}6(}6,}6X}6`}6d}6h}6�}6�}6�}6�}6�}6�}6�}6�}6~6~6~6~6H~6P~6T~6X~6�~6�~6�~6�~6�~6�~6�~6�~6�~666686@6D6H6t6|6�6�6�6�6�6�6�6�6�6�6(�60�64�68�6d�6l�6p�6t�6��6��6��6��6܀6�6�6�6�6 �6$�6(�6T�6\�6`�6d�6��6��6��6��6́6ԁ6؁6܁6�6�6�6�6D�6L�6P�6T�6��6��6��6��6��6Ă6Ȃ6̂6��6�6�6�64�6<�6@�6D�6p�6x�6|�6��6��6��6��6��6�6�6�6��6$�6,�60�64�6`�6h�6l�6p�6��6��6��6��6؄6�6�6�6�6�6 �6$�6P�6X�6\�6`�6��6��6��6��6ȅ6Ѕ6ԅ6؅6�6�6�6�6@�6H�6L�6P�6|�6��6��6��6��6��6Ć6Ȇ6�6��6�6�60�68�6<�6@�6l�6t�6x�6|�6��6��6��6��6�6�6�6�6 �6(�6,�60�6\�6d�6h�6l�6��6��6��6��6Ԉ6܈6�6�6�6�6�6 �6L�6T�6X�6\�6��6��6��6��6��6��6��6ĉ6̉6Љ6ԉ6؉6܉6�6�6�6�6�6�6�6 �6<�6D�6H�6L�6P�6T�6\�6x�6��6��6��6��6��6��6��6��6��6Ċ6Ȋ6̊6Ԋ6�6��6��6�6�6�6�6,�64�68�6<�6@�6D�6L�6h�6p�6t�6x�6|�6��6��6��6��6��6��6��6��6ċ6�6�6�6�6�6��6�6�6$�6(�6,�60�64�6<�6X�6`�6d�6h�6l�6p�6x�6��6��6��6��6Ќ6،6܌6�6�6�6�6�6H�6P�6T�6X�6��6��6��6��6��6ȍ6̍6Ѝ6��6�6�6�68�6@�6D�6H�6t�6|�6��6��6��6��6��6��6�6�6��6��6(�60�64�68�6d�6l�6p�6t�6��6��6��6��6܏6�6�6�6�6 �6$�6(�6T�6\�6`�6d�6��6��6��6��6̐6Ԑ6ؐ6ܐ6�6�6�6�6D�6L�6P�6T�6��6��6��6��6��6đ6ȑ6̑6��6�6�6�64�6<�6@�6D�6p�6x�6|�6��6��6��6��6��6�6�6�6��6$�6,�60�64�6`�6h�6l�6p�6��6��6��6��6ؓ6�6�6�6�6�6 �6$�6P�6X�6\�6`�6��6��6��6��6Ȕ6Д6Ԕ6ؔ6�6�6�6�6@�6H�6L�6P�6|�6��6��6��6��6��6ĕ6ȕ6�6��6�6�60�68�6<�6@�6l�6t�6x�6|�6��6��6��6��6�6�6�6�6 �6(�6,�60�6\�6d�6h�6l�6��6��6��6��6ԗ6ܗ6�6�6�6�6�6 �6L�6T�6X�6\�6��6��6��6��6И6Ԙ6�6�6�6�6�6(�6,�60�6H�6L�6X�6d�6h�6l�6��6��6��6��6��6��6��6ę6Й6ܙ6�6�6��6�6�6�6�6 �68�6<�6H�6T�6X�6\�6t�6x�6��6��6��6��6��6��6��6̚6К6Ԛ6�6�6��6�6�6�6(�6,�68�6D�6H�6L�6d�6h�6t�6��6��6��6��6��6��6��6��6ě6ܛ6�6�6��6��6�6�6�6(�64�68�6<�6H�6P�6T�6X�6��6��6��6��6��6̜6М6�6�6�6�6�6$�6(�6,�6D�6H�6`�6d�6h�6��6��6��6��6��6��6��6؝6ܝ6�6��6��6�6�6�64�68�6P�6T�6X�6p�6t�6��6��6��6��6��6Ȟ6̞6О6�6�6�6�6�6$�6(�6@�6D�6H�6`�6d�6|�6��6��6��6��6��6��6��6؟6ܟ6�6�6��6��6�6�6$�60�64�6P�6T�6`�6l�6p�6��6��6��6��6��6Ƞ6̠6�6�6�6�6�6 �6$�6(�6@�6D�6\�6`�6d�6|�6��6��6��6��6��6��6ԡ6ء6ܡ6�6��6�6�6�60�64�6L�6P�6T�6l�6p�6��6��6��6��6��6Ģ6Ȣ6̢6�6�6�6�6�6 �6$�6<�6@�6D�6\�6`�6x�6|�6��6��6��6��6��6��6ԣ6أ6�6�6��6�6�6,�60�64�6L�6P�6h�6l�6p�6��6��6��6��6Ĥ6Ȥ6�6�6�6�6�6�6�6 �6$�6<�6@�6L�6P�6X�6\�6`�6x�6|�6��6��6��6��6��6��6��6ĥ6ȥ6Х6ԥ6إ6�6�6�6�6�6�6�6,�60�6<�6@�6H�6L�6P�6h�6l�6x�6|�6��6��6��6��6��6��6��6��6Ħ6Ȧ6�6�6�6�6��6�6�6�6 �6,�60�68�6<�6@�6X�6\�6h�6l�6t�6x�6|�6��6��6��6��6��6��6��6Ч6ԧ6�6�6�6�6�6�6�6�6 �6(�6,�60�6H�6L�6\�6d�6h�6l�6��6��6��6��6��6��6��6��6Ĩ6Ш6Ԩ6ܨ6�6�6��6�6�6�6�6�6 �68�6<�6H�6L�6T�6X�6\�6t�6x�6��6��6��6��6��6��6��6��6ĩ6̩6Щ6ԩ6�6�6�6�6�6�6(�6,�6<�6D�6H�6L�6d�6h�6x�6��6��6��6Ъ6ت6ܪ6�6�6�6�6�6H�6P�6T�6X�6\�6��6��6��6��6��6ȫ6̫6Ы6��6�6�6�68�6@�6D�6H�6L�6P�6t�6|�6��6��6��6��6��6��6��6��6Ĭ6Ȭ6�6�6��6��6�6�6(�60�64�68�6<�6@�6d�6l�6p�6t�6x�6|�6��6��6��6��6��6��6�6�6 �6$�6(�6,�60�64�68�6<�6@�6D�6H�6L�6P�6T�6X�6\�6`�6d�6h�6l�6p�6t�6x�6|�6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6Į6Ȯ6̮6Ю6Ԯ6خ6ܮ6�6�6�6�6�6�6��6��6�6�6�6�6�6�6�6�6 �6$�6(�6,�60�64�68�6<�6@�6D�6H�6L�6P�6T�6X�6\�6`�6d�6h�6l�6p�6t�6x�6|�6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6į6ȯ6̯6Я6ԯ6د6ܯ6�6�6�6�6�6�6��6��6�6�6�6�6�6�6�6�6 �6$�6(�6,�60�64�68�6<�6@�6D�6H�6L�6P�6T�6X�6\�6`�6d�6h�6l�6p�6t�6x�6|�6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6��6İ6Ȱ6̰6а6԰6ذ6ܰ6�6�6�6�6�6�6��6��6�6�6�6�6�6�6�6�6 �6$�6(�6,�60�64�68�6<�6@�6D�6H�6L�6P�6T�6X�6\�6`�6d�6x�6|�6��6��6��6��6��6��6��6��6ı6̱6,�60�6H�6P�6d�6h�6l�6p�6t�6x�6|�6��6��6��6��6��6��6��6��6��6��6��6��6�6 �6$�6(�6,�6#
0�6$
4�6<�6.@�6/D�63��648�6V��6��6��6��6��6I��6_��6��6��6��6	��6
��6�6�6�6�6�6�6�6X�6 �6$�6(�6,�60�64�68�6B<�6@�6D�6H�6L�6 P�6!T�6"X�6#\�6$`�6%d�6&h�6'l�6(p�6)t�6*x�6+|�6,��6-��60��61��62q��9���9�9�9�a�9�9�9�9�9�9���������9�,�9�9�������������������X�X�X�������������~����������������������v�������������������������������n��c�c�c�1�1�1�����k����������������1��������������������������1�1���������������1�	

��DDCPU_CLK_UNHALTED_CORECPU_CLK_UNHALTED_REFFR_DISPATCH_STALL_FROM_BRANCH_ABORT_TO_RETIREEVENT_1AHEVENT_4EHEVENT_65HEVENT_92HEVENT_A3HEVENT_DDHSOFTWARE_JAVA_BYTECODECOHERENT_LINEFILL_HITCSTREX_PASSEDEVENT_138HEVENT_160HEVENT_193HEVENT_198HEVENT_1CDHEVENT_1F5HEVENT_200HEVENT_208HEVENT_23AHEVENT_25FHEVENT_263HEVENT_26EHEVENT_279HEVENT_28AHEVENT_2B9HEVENT_2C4HEVENT_2CDHEVENT_305HEVENT_309HEVENT_32FHEVENT_373HEVENT_397HEVENT_3C1HEVENT_3DAHEVENT_3E9HEVENT_3FBHTTBR_WRITE_RETIREDLD_SPECLDST_SPECDMB_SPECdn_rxreq_dvmop_other_filteredcxha_snptrk_occcxla_tx_tlp_link0cxla_avg_tx_tlp_sz_dwscxla_avg_tx_tlp_sz_ccix_msgVFPU_INSTR_WAIT_CYCLESVSCR_SAT_SETL1_DATA_PUSHESL3_STORE_HITSEXTERNAL_PUSHESPREFETCH_ENGINE_FULLMARKED_GROUP_DISPATCHGCT_EMPTYFPU_MARKED_INSTR_COMPLETEDGROUP_DISPATCH_REJECTSTASH_L2_HITSSTASH_BUSY_3L2_CACHE_UPDATESINSTR_LFB_WENT_HIGH_PRIORITYmask=csdspmu: %s
perpkg: %s
metric_group: %s
LLC_MISS_RHITMperiod{"type": "dropnotify"}
GenuineIntel-6-3Fv20GenuineIntel-6-4DPipeline;MemIpMispredictIpTBInstructions per FP Arithmetic AVX* 256-bit instruction (lower number means higher occurrence rate). May undercount due to FMA double countingL1D_Cache_Fill_BW64 * LONGEST_LAT_CACHE.MISS / 1000000000 / duration_timeL2HPKI_AllMem;MemoryTLBMem;MemoryTLB_SMTCPU_UtilizationCPU_CLK_UNHALTED.THREAD_P:k / CPU_CLK_UNHALTED.THREADInstructions per Far Branch ( Far Branches apply upon transition from application to operating system, handling interrupts, exceptions) [lower number means higher occurrence rate]INST_RETIRED.ANY / BR_INST_RETIRED.FAR_BRANCH:u(cstate_core@c7\-residency@ / msr@tsc@) * 100Transactions accessing L2 pipeThis event counts core-originated cacheable demand requests that refer to the last level cache (LLC). Demand requests include loads, RFOs, and hardware prefetches from L1D, and instruction fetches from IFUmem_load_uops_l3_miss_retired.remote_fwdThis is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were load uops missed L1 but hit a fill buffer due to a preceding miss to the same cache line with the data not ready.
Note: Only two data-sources of L1/FB are applicable for AVX-256bit  even though the corresponding AVX load could be serviced by a deeper level in the memory hierarchy. Data source is reported for the Low-half load  Supports address when precise (Precise event)mem_load_uops_retired.l2_hitThis is a precise version (that is, uses PEBS) of the event that counts store uops retired to the architected path with a filter on bits 0 and 1 applied.
Note: This event ?ounts AVX-256bit load/store double-pump memory uops as a single uop at retirement  Supports address when precise (Precise event)offcore_requests.demand_data_rdThis event counts the number of offcore outstanding Code Reads transactions in the super queue every cycle. The Offcore outstanding state of the transaction lasts from the L2 miss until the sending transaction completion to requestor (SQ deallocation). See the corresponding Umask under OFFCORE_REQUESTS  Spec update: BDM76floating pointmove_elimination.simd_eliminatedevent=0x58,period=1000003,umask=0x2Cycles when uops are being delivered to Instruction Decode Queue (IDQ) from Decode Stream Buffer (DSB) pathidq.mite_all_uopsevent=0x79,period=2000003,umask=0x10This event counts the number of uops not delivered to Resource Allocation Table (RAT) per thread adding 4  x when Resource Allocation Table (RAT) is not stalled and Instruction Decode Queue (IDQ) delivers x uops to Resource Allocation Table (RAT) (where x belongs to {0,1,2,3}). Counting does not cover cases when:
 a. IDQ-Resource Allocation Table (RAT) pipe serves the other thread;
 b. Resource Allocation Table (RAT) is stalled for the thread (including uop drops and clear BE conditions); 
 c. Instruction Decode Queue (IDQ) delivers four uopshle_retired.aborted_misc3hle_retired.aborted_misc4event=0xcd,period=20011,umask=0x1,ldlat=0x10Unhalted core cycles when thread is in rings 1, 2, or 3This event counts both taken and not taken speculative and retired branch instructionsevent=0x88,period=200003,umask=0xc1br_inst_exec.all_direct_jmpThis event counts both taken and not taken mispredicted indirect branches excluding calls and returnsThis event counts taken speculative and retired mispredicted indirect branches excluding calls and returnscycle_activity.cycles_no_executeild_stall.lcpld_blocks.store_forwardevent=0x3,period=100003,umask=0x2ld_blocks_partial.address_aliasevent=0xa2,period=2000003,umask=0x8Number of uops executed on the coreuops_executed.core_cycles_ge_2uops_executed_port.port_3llc_references.pcie_ns_partial_writePCIe read current. Derived from unc_c_tor_inserts.opcode. Unit: uncore_cbox unc_h_requests.readsevent=0x21,umask=0x40event=0x4,umask=0x3event=0x4,umask=0xCCycles where DRAM ranks are in power down (CKE) mode. Unit: uncore_imc unc_m_power_critical_throttle_cyclesunc_m_pre_count.rduncore_pcufreq_max_os_cycles %event=0x85,period=100003,umask=0x8page_walker_loads.dtlb_l3Average latency of all requests to external memory (in Uncore cycles)This event counts load uops retired to the architected path with a filter on bits 0 and 1 applied.
Note: This event counts AVX-256bit load/store double-pump memory uops as a single uop at retirement. This event also counts SW prefetches  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020091offcore_response.all_pf_code_rd.supplier_none.snoop_noneCounts all prefetch data reads have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0090offcore_response.all_pf_data_rd.l3_hit.snoop_not_neededoffcore_response.all_pf_data_rd.supplier_none.snoop_hit_no_fwdoffcore_response.all_pf_data_rd.supplier_none.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0122offcore_response.all_rfo.supplier_none.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0001offcore_response.pf_l2_data_rd.any_responseoffcore_response.pf_l2_rfo.supplier_none.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020200offcore_response.pf_l3_data_rd.l3_hit.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000091offcore_response.all_pf_data_rd.l3_miss.snoop_missoffcore_response.all_pf_rfo.l3_miss.snoop_noneoffcore_response.corewb.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000001offcore_response.demand_rfo.l3_miss.snoop_hit_no_fwdoffcore_response.pf_l2_data_rd.l3_miss_local_dram.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000080offcore_response.pf_l3_data_rd.l3_miss_local_dram.snoop_hit_no_fwdConditional branch instructions retired (Precise event)event=0x3c,period=100003,umask=0x1event=0x34,umask=0x8fCounts all data/code/rfo reads (demand & prefetch) hit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts all demand & prefetch data reads miss the L3 and the data is returned from local dramevent=0,umask=0x4Number of non data (control) flits transmitted . Derived from unc_q_txl_flits_g0.non_data. Unit: uncore_qpi l2_ld.self.demand.s_stateevent=0x26,period=200000,umask=0x40l2_lines_out.self.prefetchevent=0x2b,period=200000,umask=0x48l2_m_lines_in.selfl2_m_lines_out.self.prefetchl2_reject_busq.self.prefetch.e_stateRetired computational Streaming SIMD Extensions (SSE) scalar-single instructionsevent=0xcf,period=2000000,umask=0x0SIMD micro-ops retired (excluding stores) (Must be precise)decode_stall.pfb_emptymisalign_mem_ref.rmw_splitprefetch.sw_l2bus_data_rcv.selfevent=0x63,period=200000,umask=0xe0bus_trans_pwr.selfevent=0x9,period=200000,umask=0x20event=0x77,period=200000,umask=0x1Hardware interrupts receivedNumber of segment register loadsdata_tlb_misses.dtlb_miss_stCounts the number of demand and prefetch transactions that the L2 XQ rejects due to a full or near full condition which likely indicates back pressure from the intra-die interconnect (IDI) fabric. The XQ may reject transactions from the L2Q (non-cacheable requests), L2 misses and L2 write-back victimsL2 cache request missesLoads retired that hit WCB (Precise event capable)  Supports address when precise (Must be precise)Counts data reads generated by L1 or L2 prefetchers that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data reads generated by L1 or L2 prefetchers that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000008Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that true miss for the L2 cache with a snoop miss in the other processor moduleCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that true miss for the L2 cache with a snoop miss in the other processor moduleCounts data cache line reads generated by hardware L1 data cache prefetcher that true miss for the L2 cache with a snoop miss in the other processor moduleCounts data cacheline reads generated by hardware L2 cache prefetcher that hit the L2 cacheCounts BACLEARS on return instructionsCycles hardware interrupts are maskedbr_misp_retired.ind_callevent=0xcd,period=2000003Counts core cycles if either divide unit is busyevent=0xc3,period=200003,umask=0x1event=0xd0,period=200003,umask=0x13Memory uops retired that missed the DTLB (Precise event capable)  Supports address when precise (Must be precise)mem_uops_retired.dtlb_miss_storesevent=0x5,period=200003,umask=0x2Counts data reads generated by L1 or L2 prefetchers have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000003010Counts data read, code read, and read for ownership (RFO) requests (demand & prefetch) hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000400Counts demand reads for ownership (RFO) requests generated by a write to full data cache line hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand reads for ownership (RFO) requests generated by a write to full data cache line miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes hit the L2 cacheoffcore_response.streaming_stores.l2_miss.snoop_miss_or_no_snoop_neededReference cycles when core is not halted.  This event uses a (_P)rogrammable general purpose performance counterevent=0x8,period=200003,umask=0x10Counts the number of store RFO requests that miss the L2 cacheAll retired store uops  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)Cycles when offcore outstanding cacheable Core Data Read transactions are present in SuperQueue (SQ), queue to uncore  Spec update: HSD62, HSD61, HSM63offcore_response.demand_code_rd.l3_hit.hit_other_core_no_fwdNumber of transitions from SSE to AVX-256 when penalty applicable  Spec update: HSD56, HSM57Counts cycles the IDQ is empty  Spec update: HSD135This event counts cycles during which the microcode sequencer assisted the Front-end in delivering uops.  Microcode assists are used for complex instructions or scenarios that can't be handled by the standard decoder.  Using other instructions, if possible, will usually improve performanceRandomly selected loads with latency value being above 4  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100400091arith.divider_uopsAll (macro) branch instructions retired (Must be precise)event=0xc4,period=100003,umask=0x2Precise instruction retired event with HW to reduce effect of PEBS shadow in IP distribution  Spec update: HSD140 (Must be precise)event=0x34,umask=0x41unc_cbo_xsnp_response.hit_evictionNumber of ITLB page walker loads that hit in the L3  Spec update: HSD25offcore_response.demand_rfo.llc_hit.hit_other_core_no_fwdevent=0x28,period=200003,umask=0x4Retired load uops which data sources were load uops missed L1 but hit FB due to preceding miss to the same cache line with data not ready (Precise event)All retired store uops. (Precise Event)Retired load uops that split across a cacheline boundary. (Precise Event)event=0x60,cmask=1,period=2000003,umask=0x2Offcore outstanding demand rfo reads transactions in SuperQueue (SQ), queue to uncore, every cycleOffcore outstanding Demand Code Read transactions in SQ to uncore. Set Cmask=1 to count cyclesOffcore outstanding Demand Data Read transactions in uncore queueCounts demand data reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0001number of AVX-256 Computational FP double precision uops issued this cyclenumber of GSSE-256 Computational FP single precision uops issued this cycleCounts the number of executed AVX masked load operations that refer to an illegal address range with the mask bits set to 0Cycles per core when uops are dispatched to port 1Cycles per thread when load or STA uops are dispatched to port 2unc_cbo_xsnp_response.hitmunc_cbo_xsnp_response.xcore_filterLLC lookup request that access cache and found line in E-state. Unit: uncore_cbox Filter on processor core initiated cacheable read requests. Unit: uncore_cbox event=0x8,period=100003,umask=0x88event=0x8,period=100003,umask=0x82event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c03f7Counts all prefetch (that bring data to L2) code reads that hit in the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x600400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc20010event=0xb,edge=1This event counts retired load uops that hit in the last-level cache (L3) and were found in a non-modified state in a neighboring core's private cache (same package).  Since the last level cache is inclusive, hits to the L3 may require snooping the private L2 caches of any cores on the same socket that have the line.  In this case, a snoop was required, and another L2 had the line in a non-modified stateRetired load uops with locked access (Precise event)Loads with latency value being above 4  (Must be precise)event=0x89,period=200003,umask=0x90ld_blocks_partial.all_sta_blockpartial_rat_stalls.flags_merge_uop_cyclesUops dispatched per threadevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010044Counts Demand code reads and prefetch code read requests  that accounts for responses which hit its own tile's L2 with data in E stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800183091Counts any Prefetch requests that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeoffcore_response.any_read.l2_hit_near_tileoffcore_response.any_rfo.l2_hit_near_tileoffcore_response.bus_locks.l2_hit_near_tileevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000004offcore_response.partial_reads.l2_hit_near_tileevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180080offcore_response.partial_writes.l2_hit_near_tileCounts L1 data HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeoffcore_response.pf_l1_data_rd.l2_hit_near_tile_e_fCounts L2 code HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts L2 code HW prefetches that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateoffcore_response.pf_l2_code_rd.outstandingCounts the number of scalar SSE, AVX, AVX2, AVX-512 micro-ops retired. More specifically, it counts scalar SSE, AVX, AVX2, AVX-512 micro-ops except for loads (memory-to-register mov-type micro ops), division, sqrtCounts the number of times the front end resteers for any branch as a result of another branch handling mechanism in the front endCounts Demand code reads and prefetch code read requests  that accounts for responses from MCDRAM (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080203091event=0xb7,period=100007,umask=0x1,offcore_rsp=0x01010032f7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000022offcore_response.pf_l1_data_rd.ddr_nearCounts the number of core cycles when no micro-ops are allocated and the ROB is fullCounts the number of micro-ops retired that are from the complex flows issued by the micro-sequencer (MS)unc_m_cas_count.wrevent=0x4,period=200003,umask=0x8l1d_cache_ld.e_statel1d_cache_lock_fb_hitL1 writebacks to L2 in E stateevent=0xf2,period=100000,umask=0x2l2_rqsts.ld_hitmem_inst_retired.latency_above_threshold_256Memory instructions retired above 4 clocks (Precise Event)event=0xb,period=2000,umask=0x10,ldlat=0x40event=0xb,period=2000000,umask=0x1Instructions retired which contains a store (Precise Event)mem_uncore_retired.remote_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x10FFOffcore RFO requests satisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3808offcore_response.data_ifetch.io_csr_mmioevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x133Offcore demand data requests satisfied by the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3804event=0xb7,period=100000,umask=0x1,offcore_rsp=0x3802offcore_response.other.llc_hit_no_other_coreoffcore_response.other.remote_cacheoffcore_response.pf_data_rd.llc_hit_other_core_hitmOffcore prefetch data reads satisfied by a remote cacheOffcore prefetch RFO requests satisfied by a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF70All offcore prefetch requestsoffcore_response.prefetch.remote_cache_hitmevent=0xcc,period=2000000,umask=0x2simd_int_64.packed_logicalevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4077event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4002offcore_response.other.any_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2080Offcore prefetch data requests satisfied by the local DRAMload_dispatch.anyBranch prediction unit missed call or returnevent=0x88,period=200000,umask=0x2Indirect non call branches executedbr_inst_exec.takenevent=0x1e,period=2000000,umask=0x1Retired MMX instructions (Precise Event)Load operations conflicting with software prefetchesMXCSR rename stall cyclesevent=0xd1,cmask=1,inv=1,period=2000000,umask=0x1event=0xb1,period=2000000,umask=0x40event=0x8,period=200000,umask=0x2mem_store_retired.dtlb_missevent=0x24,period=200003,umask=0xd8mem_inst_retired.split_storesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1001C0004offcore_response.demand_code_rd.l3_hit_s.snoop_missoffcore_response.demand_code_rd.l3_hit_s.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x801C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0040001offcore_response.demand_data_rd.l3_hit_m.snoop_hitmoffcore_response.demand_rfo.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040002offcore_response.demand_rfo.l3_hit_s.snoop_hit_no_fwdoffcore_response.demand_rfo.l4_hit_local_l4.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40020002offcore_response.other.l3_hit_m.any_snoopCounts the number of cache line split locks sent to the uncoreCounts once for most SIMD 256-bit packed double computational precision floating-point instructions retired. Counts twice for DPP and FM(N)ADD/SUB instructions retiredRetired instructions that are fetched after an interval where the front-end had at least 1 bubble-slot for a period of 2 cycles which was not interrupted by a back-end stall (Precise event)Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 32 cycles which was not interrupted by a back-end stall (Precise event)Cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularityicache_64b.iftag_stallCounts the number of memory ordering Machine Clears detected. Memory Ordering Machine Clears can result from one of the following:a. memory disambiguation,b. external snoop, orc. cross SMT-HW-thread snoop (stores) hitting load buffer  Spec update: SKL089event=0x60,period=2000003,umask=0x10offcore_response.demand_rfo.l3_miss_local_dram.snoop_hitmoffcore_response.demand_rfo.l3_miss_local_dram.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x7C408000exe_activity.4_ports_utilCycles total of 4 uops are executed on all ports and Reservation Station was not emptyCounts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 2Total pipeline cost of branch related instructions (used for program control-flow including function calls)Number of Instructions per non-speculative DSB missCounts completed page walks  (all page sizes) caused by demand data stores. This implies it missed in the DTLB and further levels of TLB. The page walk can end with or without a faultCounts the number of (demand and L1 prefetchers) core requests rejected by the L2Q due to a full or nearly full w condition which likely indicates back pressure from L2Q.  It also counts requests that would have gone directly to the XQ, but are rejected due to a full or nearly full condition, indicating back pressure from the IDI link.  The L2Q may also reject transactions  from a core to insure fairness between cores, or to delay a core?s dirty eviction when the address conflicts incoming external snoops.  (Note that L2 prefetcher requests that are dropped are not counted by this event.)This event counts the number of load ops retired that miss in the L2 (Precise event)Counts any code reads (demand & prefetch) that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000008offcore_response.pf_l2_code_rd.l2_miss.hit_other_core_no_fwdevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000020Counts the number of mispredicted JCC branch instructions retired (Precise event)mem_load_uops_misc_retired.llc_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0240Counts demand & prefetch RFOs that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.pf_l2_rfo.llc_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0020Counts all prefetch (that bring data to LLC only) RFOs that hit in the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400240offcore_response.data_in_socket.llc_miss_local.any_llc_hitevent=0xb0,period=100000,umask=0x1REQUEST = ANY_REQUEST and RESPONSE = LLC_HIT_OTHER_CORE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x10ffevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5077REQUEST = DEMAND_DATA and RESPONSE = IO_CSR_MMIOREQUEST = OTHER and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = PF_DATA and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f70offcore_response.any_data.any_dram_and_remote_fwdoffcore_response.demand_data_rd.any_dram_and_remote_fwdevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3004offcore_response.demand_ifetch.other_local_dramREQUEST = DEMAND_RFO and RESPONSE = ANY_DRAM AND REMOTE_FWDoffcore_response.pf_data_rd.any_dram_and_remote_fwdevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3070event=0xb3,cmask=1,period=2000000,umask=0x1offcore_requests.uncached_memCounts all demand & prefetch data reads that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.all_pf_data_rd.l3_hit.snoop_hit_with_fwdCounts all demand & prefetch RFOs that have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0004offcore_response.pf_l2_data_rd.l3_hit.hitm_other_coreCounts all prefetch (that bring data to LLC only) RFOs that have any response typeevent=0xc7,period=2000003,umask=0x80event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00490Counts all demand & prefetch RFOs that miss the L3 and the data is returned from local or remote dramoffcore_response.demand_code_rd.l3_miss.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000010offcore_response.pf_l2_rfo.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00100CORE_POWER.LVL2_TURBO_LICENSE / 2 / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )IO_Read_BWunc_cha_clockticksStreaming stores (full cache line). Derived from unc_cha_tor_inserts.ia_miss. Unit: uncore_cha event=0x50,umask=0x03uncore_iioPCI Express bandwidth reading at IIO, part 2. Unit: uncore_iio PCI Express bandwidth writing at IIO. Unit: uncore_iio event=0x5f,umask=0x01unc_iio_comp_buf_occupancy.cmpd.part1Read request for 4 bytes made by the CPU to IIO Part3. Unit: uncore_iio Counts every write request of 4 bytes of data made to the MMIO space of a card on IIO Part2 by  a unit on the main die (generally a core) or by another IIO unit. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busCounts ever peer to peer read request for 4 bytes of data made by a different IIO unit to the MMIO space of a card on IIO Part3. Does not include requests made by the same IIO unit. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busevent=0xc1,ch_mask=0x02,fc_mask=0x07,umask=0x04Counts every read request for up to a 64 byte transaction of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part3. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busunc_iio_txn_req_by_cpu.mem_write.part1Counts every peer to peer read request for up to a 64 byte transaction of data made by a different IIO unit to the MMIO space of a card on IIO Part2. Does not include requests made by the same IIO unit. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busevent=0xc1,ch_mask=0x02,fc_mask=0x07,umask=0x02Peer to peer write request of up to a 64 byte transaction is made to IIO Part2 by a different IIO unit. Unit: uncore_iio Counts every write request of up to a 64 byte transaction of data made by IIO Part0 to a unit on the main die (generally memory). In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busunc_iio_txn_req_of_cpu.mem_write.part1Counts when messages were sent direct to core (bypassing the CHA)unc_m2m_direct2upi_takenCounts when the M2M (Mesh to Memory) updates the multi-socket cacheline Directory state from from A (SnoopAll) to S (Shared)unc_m2m_txc_bl_inserts.allBL Egress (to CMS) Allocations; All. Unit: uncore_m2m BL Egress (to CMS) Allocations; Allunc_upi_txl_flits.dataocr.all_data_rd.l3_hit_e.snoop_noneOCR.ALL_DATA_RD.L3_HIT_M.SNOOP_NONEocr.all_data_rd.l3_hit_s.hitm_other_coreOCR.ALL_PF_DATA_RD.L3_HIT_F.HITM_OTHER_CORE  OCR.ALL_PF_DATA_RD.L3_HIT_F.HITM_OTHER_COREocr.all_pf_data_rd.l3_hit_f.hit_other_core_no_fwdOCR.ALL_PF_RFO.L3_HIT_F.HITM_OTHER_CORE  OCR.ALL_PF_RFO.L3_HIT_F.HITM_OTHER_COREOCR.ALL_PF_RFO.L3_HIT_F.SNOOP_NONEocr.all_pf_rfo.l3_hit_s.snoop_missocr.all_reads.l3_hit_e.hit_other_core_no_fwdOCR.ALL_READS.L3_HIT_E.SNOOP_MISSocr.all_reads.l3_hit_m.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8001007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4001007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1001007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2001007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080122ocr.all_rfo.l3_hit_e.hit_other_core_no_fwdocr.all_rfo.l3_hit_e.no_snoop_neededocr.all_rfo.l3_hit_f.snoop_missOCR.ALL_RFO.L3_HIT_M.HITM_OTHER_CORE  OCR.ALL_RFO.L3_HIT_M.HITM_OTHER_COREOCR.ALL_RFO.L3_HIT_M.NO_SNOOP_NEEDED  OCR.ALL_RFO.L3_HIT_M.NO_SNOOP_NEEDEDOCR.ALL_RFO.L3_HIT_M.SNOOP_MISSOCR.ALL_RFO.L3_HIT_M.SNOOP_NONEOCR.ALL_RFO.L3_HIT_S.HITM_OTHER_CORE  OCR.ALL_RFO.L3_HIT_S.HITM_OTHER_COREocr.demand_code_rd.l3_hit_e.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200004Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_M.NO_SNOOP_NEEDEDCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDocr.demand_data_rd.l3_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0001ocr.demand_data_rd.l3_hit_e.hit_other_core_fwdocr.demand_data_rd.l3_hit_s.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200002ocr.other.l3_hit_e.any_snoopocr.other.l3_hit_f.hit_other_core_no_fwdocr.other.l3_hit_s.snoop_noneocr.pf_l1d_and_sw.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0400ocr.pf_l2_data_rd.l3_hit.any_snoopocr.pf_l2_data_rd.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200010ocr.pf_l2_data_rd.l3_hit_m.hitm_other_coreocr.pf_l2_rfo.l3_hit.hit_other_core_fwdocr.pf_l2_rfo.l3_hit_f.any_snoopocr.pf_l3_data_rd.l3_hit_e.hitm_other_coreCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_S.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_hit_e.no_snoop_neededoffcore_response.all_pf_rfo.l3_hit_s.hit_other_core_fwdoffcore_response.all_pf_rfo.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x107F7offcore_response.all_reads.pmm_hit_local_pmm.snoop_not_neededoffcore_response.all_rfo.l3_hit_e.any_snoopoffcore_response.all_rfo.l3_hit_s.snoop_noneoffcore_response.all_rfo.pmm_hit_local_pmm.snoop_not_neededThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_F.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020004offcore_response.demand_code_rd.supplier_none.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_F.ANY_SNOOPoffcore_response.demand_rfo.l3_hit_f.snoop_noneoffcore_response.demand_rfo.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_F.SNOOP_NONEoffcore_response.pf_l1d_and_sw.l3_hit_m.snoop_missThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_S.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.SUPPLIER_NONE.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020400offcore_response.pf_l2_data_rd.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_F.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020080offcore_response.pf_l3_rfo.l3_hit_f.snoop_missoffcore_response.pf_l3_rfo.l3_hit_s.no_snoop_needed100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) * ( ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD) / #( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (( OFFCORE_REQUESTS_BUFFER.SQ_FULL / 2 ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) ) + ( (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( ((L1D_PEND_MISS.PENDING / ( MEM_LOAD_RETIRED.L1_MISS + MEM_LOAD_RETIRED.FB_HIT )) * cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ / CPU_CLK_UNHALTED.THREAD) / #(max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) ) MEM_PMM_Read_LatencyOCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDocr.all_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.all_pf_data_rd.l3_miss.snoop_missocr.all_pf_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdOCR.ALL_PF_RFO.L3_MISS.REMOTE_HITM OCR.ALL_PF_RFO.L3_MISS.REMOTE_HITMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000120ocr.all_pf_rfo.l3_miss_local_dram.any_snoopocr.all_pf_rfo.l3_miss_local_dram.no_snoop_neededocr.all_pf_rfo.l3_miss_remote_hop1_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C0007F7OCR.ALL_READS.L3_MISS.REMOTE_HIT_FORWARD OCR.ALL_READS.L3_MISS.REMOTE_HIT_FORWARDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4100007F7ocr.all_rfo.l3_miss.no_snoop_neededOCR.ALL_RFO.L3_MISS.SNOOP_NONE OCR.ALL_RFO.L3_MISS.SNOOP_NONEOCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C008000Counts any other requests OCR.OTHER.L3_MISS.SNOOP_NONEocr.other.l3_miss_local_dram.no_snoop_neededCounts any other requests  OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000400ocr.pf_l1d_and_sw.l3_miss.snoop_noneCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORECounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDocr.pf_l2_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000010ocr.pf_l2_data_rd.l3_miss_remote_hop1_dram.snoop_missocr.pf_l2_rfo.l3_miss.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000080ocr.pf_l3_data_rd.l3_miss_remote_hop1_dram.any_snoopCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.HIT_OTHER_CORE_FWD OCR.PF_L3_RFO.L3_MISS.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000100offcore_response.all_pf_data_rd.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREoffcore_response.demand_code_rd.l3_miss_remote_hop1_dram.snoop_noneoffcore_response.demand_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdoffcore_response.pf_l1d_and_sw.l3_miss.no_snoop_neededoffcore_response.pf_l1d_and_sw.l3_miss_local_dram.no_snoop_neededoffcore_response.pf_l3_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreocr.all_data_rd.supplier_none.snoop_missOCR.ALL_READS.PMM_HIT_LOCAL_PMM.SNOOP_NONE OCR.ALL_READS.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.demand_code_rd.any_responseocr.demand_code_rd.pmm_hit_local_pmm.snoop_noneCounts demand data reads OCR.DEMAND_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPocr.pf_l1d_and_sw.pmm_hit_local_pmm.snoop_noneUNC_M_PMM_RPQ_INSERTS + UNC_M_PMM_WPQ_INSERTSevent=0xe0,umask=0x1event=0xd3,umask=0x4Read requests to Intel Optane DC persistent memory issued to the iMC from M2M. Unit: uncore_m2m event=0x2c,umask=0x02Counts the retired load instructions whose data sources were L3 hit and cross-core snoop missed in on-pkg core cache  Supports address when precise (Precise event)Counts retired load instructions missed L2 cache as data sources  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C2380event=0x60,period=1000003,umask=0x1Counts the number of PREFETCHW instructions executedidq.mite_cycles_anyocr.demand_code_rd.l3_missocr.hwpf_l2_data_rd.dramocr.hwpf_l2_data_rd.local_dramCounts streaming stores that have any type of responseCounts the number of core cycles while the thread is not in a halt state. The thread enters the halt state when it is running the HLT instruction. This event is a component in many key event ratios. The core frequency may change from time to time due to transitions associated with Enhanced Intel SpeedStep Technology or TM2. For this reason this event may have a changing ratio with regards to time. When the core frequency is constant, this event can approximate elapsed time while the core was not in the halt state. It is counted on a dedicated fixed counter, leaving the eight programmable counters available for other eventsevent=0xa6,cmask=2,period=1000003,umask=0x40event=0xd,period=1000003,umask=0x10Counts the retirement slots used each cyclePage walks completed due to a demand data store to a 2M/4M pageActual Average Latency for L1 data-cache miss demand loads (in core cycles)CacheMissesThis event is deprecated. Refer to new event MEM_LOAD_L3_HIT_RETIRED.XSNP_NO_FWD  Supports address when precise (Precise event)ocr.demand_data_rd.remote_cache.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F003C0477C1 residency percent per coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC08000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000800Hit snoop reply with data, line kept in Shared stateHit snoop reply without sending the data, line kept in Shared stateCounts demand data reads that were supplied by PMMocr.demand_rfo.snc_pmmocr.reads_to_core.dramDRAM Precharge commands. : Precharge due to write. Unit: uncore_imc event=0x4,umask=0x3fevent=0x45,umask=0x01unc_m_dram_refresh.panicRemote write requests sent to the CHA's home agent. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_drd_prefevent=0x35,umask=0xC887FE01unc_iio_data_req_of_cpu.cmpd.part1event=0x83,ch_mask=0x04,fc_mask=0x07,umask=0x80unc_iio_data_req_by_cpu.mem_write.part5unc_iio_txn_req_by_cpu.mem_write.part4unc_iio_txn_req_by_cpu.mem_read.part5unc_iio_comp_buf_inserts.cmpd.part5PCIe Completion Buffer Occupancy of completions with data : Part 5. Unit: uncore_iio unc_i_faf_fullClockticks of the mesh to UPI (M3UPI). Unit: uncore_m3upi TOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices to remotely HOMed memory. Unit: uncore_cha Counts the number of demand and prefetch transactions that the External Queue (XQ) rejects due to a full or near full conditionCounts the number of cycles a core is stalled due to an instruction cache or translation lookaside buffer (TLB) access which hit in DRAM or MMIO (non-DRAM)event=0x34,period=200003,umask=0x4event=0x34,period=200003,umask=0x40event=0xd1,period=200003,umask=0x4Counts the number of memory uops retired that were splits  Supports address when precise (Precise event)Counts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were supplied by the L3 cache where a snoop was sent but the snoop missedocr.uc_rd.l3_hit.snoop_missCounts modified writeBacks from L2 cache that miss the L3 cache that were not supplied by the L3 cacheocr.prefetches.l3_missCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were not supplied by the L3 cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10044ocr.corewb_m.any_responseCounts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were supplied by DRAMevent=0x74,period=1000003,umask=0x20Counts the number of issue slots every cycle that were not consumed by the backend due to frontend stallsunc_cha_tor_inserts.ia_miss_drd_optevent=0x36,umask=0xC827FF01TOR Occupancy : PCIRdCurs issued by IO Devices : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsData requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 3Data requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Number Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 1Number Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Number Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Counts the number of times there was an ITLB miss and a new translation was filled into the ITLBNumber of Instructions per non-speculative Branch Misprediction (JEClear). Unit: cpu_core Number of Instructions per non-speculative DSB miss. Unit: cpu_core TOPDOWN_BE_BOUND.ALL / (5 * CPU_CLK_UNHALTED.CORE)Cycles Per Instruction. Unit: cpu_atom Load_SplitsFPDiv_Uop_RatioEstimated Pause cost. In percent. Unit: cpu_atom event=0x4,period=20003,umask=0x2mem_uops_retired.store_latencyevent=0x25,period=100003,umask=0x1fDemand Data Read miss L2, no rejects. Unit: cpu_core For every cycle where the core is waiting on at least 1 outstanding Demand RFO request, increments by 1. Unit: cpu_core TBD (Precise event). Unit: cpu_core Cycles where a code fetch is stalled due to L1 instruction cache miss. Unit: cpu_core memory_activity.stalls_l3_missxq.full_cyclesCounts the number of unhalted reference clock cycles at TSC frequency. (Fixed event). Unit: cpu_atom event=0xc3,period=20003,umask=0x6fCounts the number of uops that are from complex flows issued by the micro-sequencer (MS) (Precise event). Unit: cpu_atom event=0xb0,cmask=1,period=1000003,umask=0x8All mispredicted branch instructions retired (Precise event). Unit: cpu_core This event counts the number of mispredicted ret instructions retired. Non PEBS (Precise event). Unit: cpu_core TMA slots where no uops were being issued due to lack of back-end resources. Unit: cpu_core event=0xae,period=2000003,umask=0x1Incoming VC0 read request. Unit: uncore_imc Counts all L2 requests.[This event is alias to L2_REQUEST.ALL]event=0xb7,period=2000003,umask=0x2event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F3FFC0002Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM attached to another socketcpu_clk_unhalted.c02event=0x3,umask=0x0000000002IRP Clockticks. Unit: uncore_irp unc_iio_data_req_of_cpu.peer_write.part6event=0x50,umask=0x0000000001unc_upi_txl_flits.slot0event=0x30,umask=0x0000000001event=0x30,umask=0x0000000002RxQ Occupancy - All Packets : Slot 1. Unit: uncore_upi RxQ Occupancy - All Packets : Slot 2. Unit: uncore_upi event=0xc0,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000001Data requested by the CPU : Core writing to Cards MMIO space. Unit: uncore_iio event=0x1a,umask=0x07event=0x56,umask=0x000000000aTOR Inserts : All Snoops from Remote. Unit: uncore_cha TOR Inserts; CRd Pref from local IA. Unit: uncore_cha event=0x36,umask=0x00c001fd01TOR Occupancy : PRQ - IOSF. Unit: uncore_cha event=0x36,umask=0x00c897ff01event=0x35,umask=0x00C8978601TOR Inserts : CRd_Prefs issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00c8970601TOR Occupancy : ItoMs issued by iA Cores that Missed LLC. Unit: uncore_cha event=0x36,umask=0x00c86f8a01unc_cha_tor_occupancy.io_wbmtoiTOR Occupancy : CLFlushes issued by IO Devices. Unit: uncore_cha IC line invalidated due to overwriting fill response. The number of instruction cache lines invalidated. A non-SMC event is CMC (cross modifying code), either from the other thread of the core or another corel2_request_g1.all_no_prefetchCore to L2 cacheable request access status (not including L2 Prefetch). Data cache store or state change hit in L2event=0xcb,umask=0x02Remote Link Controller Outbound Packet Types: Data (32B): Remote Link Controller 0event=0x1c7,umask=0x38Total number multi-pipe uOps assigned to pipe 1Total number of fp uOps on pipe 2This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15Double precision divide/square root FLOPSSingle-precision add/subtract FLOPSls_tablewalker.dc_type0all_tlbs_flushedevent=0x94,umask=0xffbp_l1_tlb_fetch_hit.if1gx87 bottom-executing uOps retired. The number of serializing Ops retiredNumber of retired CPUID instructionsevent=0x43,umask=0x04ex_ret_uncond_brnch_instrls_any_fills_from_sys.alternate_memoriesevent=0x44,umask=0x80Hardware prefetch data cache fills from either DRAM or MMIO in the same NUMA nodeL2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L1Stream (fetch additional sequential lines into L1 cache)L3 cache requests for all coherent accesses. Unit: amd_l3 event=0xac,umask=0x08ex_no_retire.load_not_completeex_ret_ucode_opsWrite data beats (64 bytes) for local processor at Coherent Station (CS) 2Write data beats (64 bytes) for local processor at Coherent Station (CS) 10Write data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 3local_socket_inf0_inbound_data_beats_ccm2Data beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 2event=0x4de,umask=0x7ffData beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 5event=0x59f,umask=0x7ffData beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 3Data beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 7event=0x45f,umask=0xbfeData beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 1local_socket_outbound_data_beats_link0fp_ops_retired_by_width.mmx_uops_retiredevent=0x8,umask=0x10fp_ops_retired_by_type.scalar_macevent=0xa,umask=0x04fp_ops_retired_by_type.scalar_cmpevent=0xb,umask=0x10sse_avx_ops_retired.sse_avx_movRetired 128-bit packed floating-point add opsevent=0xc,umask=0x06event=0xc,umask=0x0bRetired 256-bit packed floating-point divide opsevent=0xc,umask=0xd0packed_int_op_type.int256_subRetired 256-bit packed integer pack opsevent=0xd,umask=0xe0Retired packed integer ops of all typesInstruction fetches that miss in both the L1 and L2 ITLBs (page-table walks are requested) for 2M pagesresyncs_or_nc_redirectsevent=0xaa,umask=0x04Number of ops dispatched to the floating-point unitfrontend_bound_latencyFraction of dispatched ops that were flushed due to branch mispredictsl2_cache_accesses_from_l1_ic_missesL1 data TLB missesRemote socket upstream DMA write datad_ratio(umc_data_slot_clks.all / 2, umc_mem_clk)uncore_imc.cache_hitsNB_MEMORY_CONTROLLER_DRAM_COMMAND_SLOTS_MISSEDPMNC_SW_INCREVENT_05HEVENT_25HEVENT_52HEVENT_6FHEVENT_72HEVENT_81HEVENT_AFHEVENT_C9HEVENT_D8HEVENT_DAHCHAINLD_RETIREDEVENT_106HEVENT_121HEVENT_123HEVENT_144HEVENT_15BHEVENT_1DBHEVENT_20DHEVENT_211HEVENT_21CHEVENT_22AHEVENT_252HEVENT_258HEVENT_274HEVENT_299HEVENT_2E7HEVENT_2F8HEVENT_300HEVENT_311HEVENT_35CHEVENT_361HEVENT_362HEVENT_365HEVENT_384HL1D_CACHE_REFILL_LDL1D_TLB_WRhnf_mc_reqshnf_pocq_atomic_addrhazrni_rxdat_flitsrni_txdat_flitsrni_rdb_ordcxra_snp_bcastscxra_req_chainsclkdiv2_ranks_in_srefCYCLES_THREE_INSTR_DISPATCHEDBRANCHES_COMPLETEDL1_DATA_SNOOP_HIT_ON_MODIFIEDVTQ_LINE_FETCH_HITFP_STORE_CAUSES_STALL_IN_LSULSU_CSQ_FORWARDINGL1_DATA_LOAD_ACCESS_MISSL1_DATA_CACHE_CASTOUTS_TO_L2FPR_ISSUE_STALLEDFPR_ISSUE_QUEUE_ENTRIESL3_DATA_CACHE_MISSESPREFETCH_ENGINE_COLLISION_VS_INSTR_FETCHMARKED_INSTR_FINISH_ANY_UNITCYCLES_ISSUE_STALLEDGUARDED_LOADS_TRANSLATEDLOAD_MISS_DLFB_FULLLOAD_GUARDED_MISSILFB_FETCH_MISS_CYCLEShscancelledIAFSTANDALONEevent: %s
LLC-REFERENCEGenuineIntel-6-36GenuineIntel-6-A7This category represents fraction of slots where the processor's Frontend undersupplies its BackendINST_RETIRED.ANY / CPU_CLK_UNHALTED.THREADCycles Per Instruction (per Logical Processor)FLOPc_SMTINST_RETIRED.ANY / BR_MISP_RETIRED.ALL_BRANCHESSMTInstructions per FP Arithmetic Scalar Double-Precision instruction (lower number means higher occurrence rate)L1MPKIPowerL1D miss oustandings duration in cyclesl2_rqsts.all_demand_referencesl2_rqsts.rfo_missThis is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were HitM responses from a core on same socket (shared L3)  Supports address when precise.  Spec update: BDM100 (Precise event)This is a precise version (that is, uses PEBS) of the event that counts load uops retired to the architected path with a filter on bits 0 and 1 applied.
Note: This event ?ounts AVX-256bit load/store double-pump memory uops as a single uop at retirement. This event also counts SW prefetches  Supports address when precise (Precise event)This is a precise version (that is, uses PEBS) of the event that counts line-splitted store uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K)  Supports address when precise (Precise event)offcore_requests.all_data_rdevent=0xb0,period=100003,umask=0x8offcore_requests_buffer.sq_fullCycles when offcore outstanding Demand Data Read transactions are present in SuperQueue (SQ), queue to uncore  Spec update: BDM76This event counts the number of offcore outstanding RFO (store) transactions in the super queue (SQ) every cycle. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation). See corresponding Umask under OFFCORE_REQUESTS  Spec update: BDM76Offcore response can be programmed only with a specific pair of event select and counter MSR, and with specific event codes and predefine mask bit value in a dedicated MSR to specify attributes of the offcore transactionevent=0xc7,period=2000003,umask=0x8other_assists.avx_to_sseDecode Stream Buffer (DSB)-to-MITE switch true penalty cyclesNumber of Instruction Cache, Streaming Buffer and Victim Cache Reads. both cacheable and noncacheable, including UC fetchesThis event counts the number of both cacheable and noncacheable Instruction Cache, Streaming Buffer and Victim Cache Reads including UC fetchesCycles MITE is delivering 4 Uopsevent=0x79,cmask=1,period=2000003,umask=0x8This event counts the number of uops delivered to Instruction Decode Queue (IDQ) from the MITE path. Counting includes uops that may bypass the IDQ. This also means that uops are not being delivered from the Decode Stream Buffer (DSB)Cycles when uops are being delivered to Instruction Decode Queue (IDQ) while Microcode Sequenser (MS) is busyDeliveries to Instruction Decode Queue (IDQ) initiated by Decode Stream Buffer (DSB) while Microcode Sequenser (MS) is busyNumber of times the TSX watchdog signaled an HLE abortNumber of times an HLE execution aborted due to incompatible memory typeNumber of times HLE caused a faultmem_trans_retired.load_latency_gt_16Loads with latency value being above 64  Spec update: BDM100, BDM35 (Must be precise)misalign_mem_ref.loadsNumber of times an RTM execution aborted due to various memory events (e.g., read/write capacity and conflicts)Counts the number of times a class of instructions that may cause a transactional abort was executed. Since this is the count of execution, it may not always cause a transactional abortevent=0x5d,period=2000003,umask=0x10event=0x54,period=2000003,umask=0x8Taken speculative and retired direct near callsThis is a precise version (that is, uses PEBS) of the event that counts taken branch instructions retired (Precise event)All mispredicted macro branch instructions retiredbr_misp_retired.near_takenCount XClk pulses when this thread is unhalted and the other thread is haltedThis event counts the number of cycles during which Resource Allocation Table (RAT) external stall is sent to Instruction Decode Queue (IDQ) for the current thread. This also includes the cycles during which the Allocator is serving another threadCore cycles the allocator was stalled due to recovery from earlier clear event for this thread (e.g. misprediction or memory nuke)This event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 4Cycles with no micro-ops executed from any thread on physical coreCycles where at least 2 uops were executed per-threadCycles per core when uops are exectuted in port 4This event counts the number of Uops issued by the Resource Allocation Table (RAT) to the reservation station (RS)uncore_cboxevent=0x35,umask=0x3,filter_opc=0x187,filter_nc=1PCIe writes (partial cache line). Derived from unc_c_tor_inserts.opcode. Unit: uncore_cbox Streaming stores (full cache line). Derived from unc_c_tor_inserts.opcode. Unit: uncore_cbox event=0x35,umask=0x1,filter_opc=0x18devent=0x1,umask=0x1event=0x1,umask=0x8Shared line forwarded from remote cache. Unit: uncore_ha power_critical_throttle_cycles %unc_p_power_state_occupancy.cores_c3event=0x74(UNC_P_FREQ_TRANS_CYCLES / UNC_P_CLOCKTICKS) * 100.Store miss in all TLB levels causes a page walk that completes. (1G)  Spec update: BDM69Number of ITLB page walker hits in the L2  Spec update: BDM69, BDM98DTLB flush attempts of the thread-specific entriesMEM_Parallel_RequestsAll retired load uops  Supports address when precise (Precise event)All retired store uops  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020091offcore_response.all_pf_code_rd.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0120offcore_response.demand_code_rd.l3_hit.snoop_missoffcore_response.demand_data_rd.l3_hit.any_snoopoffcore_response.demand_rfo.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020010offcore_response.pf_l2_data_rd.supplier_none.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0020offcore_response.pf_l2_rfo.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0080offcore_response.all_pf_data_rd.l3_miss_local_dram.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0120offcore_response.corewb.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000004offcore_response.demand_code_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0001offcore_response.demand_data_rd.l3_miss_local_dram.snoop_not_neededoffcore_response.demand_rfo.l3_miss_local_dram.any_snoopoffcore_response.other.l3_miss.snoop_noneoffcore_response.pf_l2_code_rd.l3_hit.snoop_non_dramoffcore_response.pf_l2_code_rd.l3_miss.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000040offcore_response.pf_l2_code_rd.supplier_none.snoop_non_dramoffcore_response.pf_l2_rfo.l3_miss.snoop_missoffcore_response.pf_l2_rfo.l3_miss_local_dram.snoop_missoffcore_response.pf_l2_rfo.l3_miss_local_dram.snoop_non_dramoffcore_response.pf_l3_rfo.l3_miss_local_dram.snoop_noneThis event counts all actually retired uops. Counting increments by two for micro-fused uops, and by one for macro-fused and other uops. Maximal increment value for one cycle is eight (Precise event)A cross-core snoop initiated by this Cbox due to processor core memory request which hits a non-modified line in some processor coreunc_arb_trk_occupancy.allevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0244offcore_response.demand_rfo.llc_miss.remote_hitmunc_q_clockticksAll data requests from the L1 data cachel2_data_rqsts.self.i_statel2_ld.self.demand.mesievent=0x30,period=200000,umask=0x4fevent=0xcb,period=10000,umask=0x2simd_uop_type_exec.mul.arevent=0xb3,period=2000000,umask=0x81event=0xb3,period=2000000,umask=0x8BACLEARS assertedmacro_insts.non_cisc_decodedNonzero segbase load 1 bubblemisalign_mem_ref.ld_split.arMemory references that cross an 8-byte boundarymisalign_mem_ref.st_split.arbus_bnr_drv.this_agentbus_hitm_drv.this_agentbus_request_outstanding.all_agentscycles_int_masked.cycles_int_maskedevent=0x7e,period=200000,umask=0x40br_inst_retired.mispredevent=0xc4,period=2000000,umask=0x4Retired taken branch instructionsevent=0x89,period=200000,umask=0x2Mispredicted return branchesmul.arreissue.anyNumber of page-walks executedevent=0x51,period=200003,umask=0x1Counts load uops retired that hit the L1 data cache  Supports address when precise (Must be precise)Counts load uops retired that miss the L1 data cache  Supports address when precise (Must be precise)offcore_response.any_data_rd.l2_hitevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000003010Counts data reads generated by L1 or L2 prefetchers that true miss for the L2 cache with a snoop miss in the other processor moduleoffcore_response.corewb.l2_miss.anyevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040004Counts demand cacheable data reads of full cache lines that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.demand_data_rd.l2_miss.snoop_miss_or_no_snoop_neededevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400002000Counts reads for ownership (RFO) requests generated by L2 prefetcher that miss the L2 cachemachine_clears.fp_assistFloating point divide uops retired. (Precise Event Capable) (Must be precise)event=0xe7,period=200003,umask=0x1event=0xcb,period=203,umask=0x1event=0xc4,period=200003,umask=0x7ebr_inst_retired.non_return_indReference cycles when core is not halted.  This event uses a programmable general purpose performance counterCounts data reads (demand & prefetch) outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data read, code read, and read for ownership (RFO) requests (demand & prefetch) outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts requests to the uncore subsystem outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts reads for ownership (RFO) requests (demand & prefetch) hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand cacheable data reads of full cache lines hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010020Counts reads for ownership (RFO) requests generated by L2 prefetcher true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Page walk completed due to a demand data store to a 2M or 4M pageevent=0x4f,period=200003,umask=0x10Retired load uops with locked access  Supports address when precise.  Spec update: HSD76, HSD29, HSM30 (Precise event)Number of SIMD move elimination candidate uops that were not eliminatedCounts cycles MITE is delivered at least one uop. Set Cmask = 1Increment each cycle # of uops delivered to IDQ from MITE path. Set Cmask = 1 to count cyclesoffcore_response.all_code_rd.l3_miss.any_responseoffcore_response.all_data_rd.l3_miss.local_dramoffcore_response.demand_rfo.l3_miss.local_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00020Number of times an RTM execution successfully committedMispredicted macro branch instructions retired (Must be precise)Execution stalls due to L1 data cache missesResource-related stall cycles  Spec update: HSD135event=0xb7,period=100003,umask=0x1,offcore_rsp=0x063F800091event=0xf2,period=100003,umask=0x8Requests from the L2 hardware prefetchers that miss L2 cacheRFOs that access cache lines in any stateL2 or LLC HW prefetches that access L2 cacheNumber of GSSE memory assist for stores. GSSE microcode assist is being invoked whenever the hardware is unable to properly handle GSSE-256b operationsSample stores and collect precise store operation via PEBS record. PMC3 only (Must be precise)page_walks.llc_missCycles with pending L2 miss loads. Set AnyThread to count per coreint_misc.recovery_stalls_countevent=0xd,cmask=1,edge=1,period=2000003,umask=0x3Cycles weighted by number of requests pending in Coherency Tracker. Unit: uncore_arb Counts cycles weighted by the number of requests waiting for data returning from the memory controller. Accounts for coherent and non-coherent requests initiated by IA cores, processor graphic units, or LLC. Unit: uncore_arb A snoop hits a non-modified line in some processor coreFilter on processor core initiated cacheable write requestsFilter on any IRQ or IPQ initiated requests including uncacheable, non-coherent requests. Unit: uncore_cbox Data forwarded from remote cachemem_load_uops_llc_miss_retired.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0080Counts prefetch (that bring data to LLC only) data reads that hit in the LLC and the snoops sent to sibling cores return clean responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc20091Counts prefetch (that bring data to L2) data reads that miss the LLC  and the data returned from local dramCounts prefetch (that bring data to L2) data reads that miss the LLC  and the data forwarded from remote cacheLLC prefetch misses for data reads. Derived from unc_c_tor_inserts.miss_opcode.data_read. Unit: uncore_cbox LLC misses for ItoM writes (as part of fast string memcpy stores). Derived from unc_c_tor_inserts.miss_opcode.itom_write. Unit: uncore_cbox Read requests to home agent. Unit: uncore_ha Number of data flits transmitted . Unit: uncore_qpi event=0xeunc_p_freq_ge_1200mhz_transitionsevent=0xc,edge=1,filter_band1=20This event counts L1D data line replacements.  Replacements occur when a new line is brought into the cache, causing eviction of a line loaded earlierMiss in last-level (L3) cache. Excludes Unknown data-sourceUops not delivered to Resource Allocation Table (RAT) per thread when backend of the machine is not stalled This event counts the number of uops not delivered to the back-end per cycle, per thread, when the back-end was not stalled.  In the ideal case 4 uops can be delivered each cycle.  The event counts the undelivered uops - so if 3 were delivered in one cycle, the counter would be incremented by 1 for that cycle (4 - 3). If the back-end is stalled, the count for this event is not incremented even when uops were not delivered, because the back-end would not have been able to accept them.  This event is used in determining the front-end bound category of the top-down pipeline slots characterizationhw_pre_req.dl1_missThis event counts the number of reference cycles when the core is not in a halt state. The core enters the halt state when it is running the HLT instruction or the MWAIT instruction. This event is not affected by core frequency changes (for example, P states, TM2 transitions) but has the same incrementing frequency as the time stamp counter. This event can approximate elapsed time while the core was not in a halt state. This event has a constant ratio with the CPU_CLK_UNHALTED.REF_XCLK event. It is counted on a dedicated fixed counter, leaving the four (eight when Hyperthreading is disabled) programmable counters available for other eventsCounts the number of times that the uncore transitioned a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band0=XXX with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Derived from unc_p_freq_band0_cycles. Unit: uncore_pcu Counts Demand code reads and prefetch code read requests  that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180044Counts Demand code reads and prefetch code read requests  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateCounts any Read request  that accounts for responses which hit its own tile's L2 with data in E stateoffcore_response.any_read.l2_hit_this_tile_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800408000Counts Bus locks and split lock requests that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts demand code reads and prefetch code reads that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeoffcore_response.demand_code_rd.l2_hit_far_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180004offcore_response.demand_code_rd.l2_hit_this_tile_mCounts Demand cacheable data writes that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000080event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000080Counts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800182000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400040Counts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses which hit its own tile's L2 with data in F stateoffcore_response.pf_software.l2_hit_this_tile_mCounts UC code reads (valid only for Outstanding response type)  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts UC code reads (valid only for Outstanding response type)  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180200event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080200Counts any request that accounts for data responses from DRAM Faroffcore_response.bus_locks.ddr_nearoffcore_response.demand_code_rd.ddr_nearCounts demand cacheable data and L1 prefetch data reads that accounts for data responses from DRAM Farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800040Counts the total number of instructions retiredCounts the number of micro-ops retiredThis event counts every cycle when an I-side (walks due to an instruction fetch) page walk is in progressCounts the total I-side page walks that are completedCounts the total page walks that are completed (I-side and D-side)l2_lines_in.s_stateL2 lines evicted by a prefetch requestevent=0x27,period=100000,umask=0x1l2_write.rfo.m_stateevent=0xcb,period=40000,umask=0x4offcore_response.any_data.llc_hit_no_other_coreoffcore_response.any_data.local_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1811Offcore requests satisfied by the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x822offcore_response.corewb.any_cache_dramOffcore code or data read requests satisfied by the LLC and not found in a sibling coreoffcore_response.demand_data_rd.llc_hit_other_core_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x801Offcore demand code reads satisfied by the IO, CSR, MMIO unitoffcore_response.demand_ifetch.local_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x704Offcore demand code reads satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x280event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F40event=0xb7,period=100000,umask=0x1,offcore_rsp=0x470SSE FP packed Uopssimd_int_128.packed_mpy128 bit SIMD integer shift operations128 bit SIMD integer unpack operationsoffcore_response.any_rfo.local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6080event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2030event=0xb7,period=100000,umask=0x1,offcore_rsp=0x6070event=0x4,period=200000,umask=0x7event=0x14,cmask=1,edge=1,inv=1,period=2000000,umask=0x1bpu_clears.earlyevent=0x88,period=200000,umask=0x7fbr_inst_exec.directIndirect call branches executedevent=0x87,period=2000000,umask=0x4Length Change Prefix stall cyclesevent=0xd2,period=2000000,umask=0x2Scoreboard stall cyclesResource related stall cyclesStore buffer stall cyclesuops_executed.core_active_cycles_no_port5event=0xc2,cmask=16,inv=1,period=2000000,umask=0x1event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40040004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40020004offcore_response.demand_data_rd.l3_hit_e.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0100002offcore_response.demand_rfo.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400028000event=0xc6,period=100007,umask=0x1,frontend=0x12Counts retired instructions that are delivered to the back-end after a front-end stall of at least 16 cycles. During this period the front-end delivered no uops (Precise event)Counts the number of cycles 4 uops were delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Count includes uops that may 'bypass' the IDQevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000400001event=0xd,period=2000003,umask=0x80Counts allocation stall cycles caused by the store buffer (SB) being full. This counts cycles that the pipeline back-end blocked uop delivery from the front-endevent=0xb1,period=2000003,umask=0x10100 * (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * ( (ICACHE_64B.IFTAG_STALL / CPU_CLK_UNHALTED.THREAD) + (( ICACHE_16B.IFDATA_STALL + 2 * cpu@ICACHE_16B.IFDATA_STALL\,cmask\=1\,edge@ ) / CPU_CLK_UNHALTED.THREAD) + (9 * BACLEARS.ANY / CPU_CLK_UNHALTED.THREAD) ) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))Loads hit L2 (Precise event)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000008008Counts RFO requests generated by L2 prefetchers that hit in the other module where modified copies were found in other core's L1 cacheThis event counts all instruction fetches from the instruction cacheCounts the number of cycles when no uops are allocated for any reasonRetired load uops with L1 cache hits as data sources. (Precise Event - PEBS) (Precise event)All retired store uops. (Precise Event - PEBS) (Precise event)offcore_response.all_pf_code_rd.llc_hit.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0240REQUEST = PF_RFO and RESPONSE = ANY_RESPONSECounts prefetch (that bring data to L2) code reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.pf_l2_code_rd.llc_hit.snoop_missREQUEST = PF_LLC_DATA_RD and RESPONSE = ANY_RESPONSECounts all prefetch code reads that miss the LLC  and the data returned from dramCounts all demand & prefetch RFOs that miss the LLC  and the data returned from dramoffcore_response.demand_ifetch.llc_miss_local.dramCounts demand data writes (RFOs) that miss the LLC and the data returned from dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1f80400080offcore_requests_outstanding.demand.read_dataREQUEST = ANY IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = ANY IFETCH and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HIToffcore_response.any_request.all_local_dram_and_remote_cache_hitREQUEST = ANY_REQUEST and RESPONSE = ANY_CACHE_DRAMREQUEST = CORE_WB and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DATA_IFETCH and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5003event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5001REQUEST = DEMAND_RFO and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = OTHER and RESPONSE = LLC_HIT_NO_OTHER_COREevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff70offcore_response.data_ifetch.any_dram_and_remote_fwdREQUEST = DATA_IN and RESPONSE = REMOTE_DRAMoffcore_response.other.other_local_dramSnoop data requestsdtlb_load_misses.walk_cyclesITLB miss large page walksevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2704mem_load_l3_miss_retired.local_drammem_load_l3_miss_retired.remote_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0490OFFCORE_RESPONSE.DEMAND_RFO.L3_HIT.SNOOP_HIT_WITH_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests that hit in the L3offcore_response.pf_l2_data_rd.l3_hit.snoop_hit_with_fwdCounts all demand & prefetch data reads that miss the L3 and the data is returned from local or remote dramoffcore_response.all_pf_data_rd.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00490Counts all prefetch data reads that miss the L3 and the data is returned from local or remote dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000122offcore_response.demand_code_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts all demand data writes (RFOs) that miss the L3 and the modified data is transferred from remote cacheoffcore_response.pf_l3_data_rd.l3_miss.any_snoopoffcore_response.pf_l3_data_rd.l3_miss.remote_hit_forwardoffcore_response.pf_l3_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdINST_RETIRED.ANY / ( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * ( FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE ) + 16 * FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE )L2_Evictions_NonSilent_PKIPower_License1_Utilization_SMTCounts the total number or DRAM Write CAS commands issued on this channel while in Write-Major-Modeevent=0x81unc_iio_data_req_of_cpu.mem_read.part1PCI Express bandwidth writing at IIO. Derived from unc_iio_data_req_of_cpu.mem_write.part0. Unit: uncore_iio event=0x54,umask=0x01Counts read requests from a remote socket which hit in the HitME cache (used to cache the multi-socket Directory state) to a line in the E(Exclusive) state.  This includes the following read opcodes (RdCode, RdData, RdDataMigratory, RdCur, RdInv*, Inv*)Counts when a normal (Non-Isochronous) read is issued to any of the memory controller channels from the CHAunc_cha_imc_writes_count.fullevent=0x37,umask=0x02Ingress (from CMS) Allocations; IRQ. Unit: uncore_cha Counts every write request of 4 bytes of data made to the MMIO space of a card on IIO Part3 by  a unit on the main die (generally a core) or by another IIO unit. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busunc_iio_data_req_by_cpu.peer_write.part1unc_iio_data_req_of_cpu.peer_read.part0event=0x83,ch_mask=0x01,fc_mask=0x07,umask=0x08Counts every peer to peer write request of 4 bytes of data made by IIO Part3 to the MMIO space of an IIO target. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busunc_iio_txn_req_by_cpu.mem_read.part1Write request of up to a 64 byte transaction is made to IIO Part1 by the CPU. Unit: uncore_iio Peer to peer read request for up to a 64 byte transaction is made by a different IIO unit to IIO Part0. Unit: uncore_iio Counts every peer to peer write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part3 by a different IIO unit. Does not include requests made by the same IIO unit. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busCounts every read request for up to a 64 byte transaction of data made by IIO Part1 to a unit on the main die (generally memory). In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busCounts every read request for up to a 64 byte transaction of data made by IIO Part3 to a unit on the main die (generally memory). In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busevent=0x84,ch_mask=0x08,fc_mask=0x07,umask=0x02Number of reads that a message sent direct2 Intel UPI was overridden. Unit: uncore_m2m event=0x2d,umask=0x2Counts when the M2M (Mesh to Memory) issues reads to the iMC (Memory Controller).  It only counts  normal priority non-isochronous readsunc_upi_rxl_bypassed.slot2unc_upi_txl_flits.non_dataRetired load instructions with local Intel Optane DC persistent memory as the data source where the data request missed all caches. Precise event  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080491OCR.ALL_DATA_RD.L3_HIT_F.HITM_OTHER_CORE  OCR.ALL_DATA_RD.L3_HIT_F.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080490OCR.ALL_PF_DATA_RD.L3_HIT_F.ANY_SNOOP  OCR.ALL_PF_DATA_RD.L3_HIT_F.ANY_SNOOPocr.all_pf_data_rd.l3_hit_m.any_snoopocr.all_pf_data_rd.l3_hit_s.snoop_missOCR.ALL_PF_RFO.L3_HIT.ANY_SNOOP OCR.ALL_PF_RFO.L3_HIT.ANY_SNOOP OCR.ALL_PF_RFO.L3_HIT.ANY_SNOOPocr.all_pf_rfo.l3_hit.no_snoop_neededocr.all_pf_rfo.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200120ocr.all_pf_rfo.l3_hit_s.snoop_noneocr.all_reads.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C07F7ocr.all_reads.l3_hit.snoop_missocr.all_reads.l3_hit_e.no_snoop_neededocr.all_rfo.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080122ocr.all_rfo.l3_hit_f.no_snoop_neededocr.demand_code_rd.l3_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200004ocr.demand_data_rd.l3_hit_e.hit_other_core_no_fwdocr.demand_data_rd.l3_hit_f.no_snoop_neededocr.demand_rfo.l3_hit_m.hit_other_core_no_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_S.HITM_OTHER_COREocr.other.l3_hit.hit_other_core_fwdocr.other.l3_hit_e.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800088000Counts any other requests  OCR.OTHER.L3_HIT_E.HIT_OTHER_CORE_NO_FWDCounts any other requests  OCR.OTHER.L3_HIT_M.HIT_OTHER_CORE_FWDocr.other.l3_hit_m.snoop_missocr.other.l3_hit_s.hit_other_core_no_fwdocr.pf_l1d_and_sw.l3_hit.hit_other_core_no_fwdocr.pf_l1d_and_sw.l3_hit.snoop_missCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_E.NO_SNOOP_NEEDEDocr.pf_l1d_and_sw.l3_hit_f.any_snoopCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_HIT.SNOOP_NONEocr.pf_l2_data_rd.l3_hit_f.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080020Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100020ocr.pf_l3_data_rd.l3_hit_e.hit_other_core_fwdCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDocr.pf_l3_rfo.l3_hit.snoop_noneocr.pf_l3_rfo.l3_hit_f.snoop_missCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDoffcore_response.all_data_rd.l3_hit_m.hit_other_core_fwdoffcore_response.all_data_rd.supplier_none.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDoffcore_response.all_pf_data_rd.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_S.ANY_SNOOPoffcore_response.all_pf_data_rd.l3_hit_s.hitm_other_coreoffcore_response.all_pf_data_rd.supplier_none.hitm_other_coreoffcore_response.all_pf_data_rd.supplier_none.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_M.SNOOP_NONEoffcore_response.all_reads.l3_hit.any_snoopoffcore_response.all_reads.l3_hit.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_F.ANY_SNOOPoffcore_response.all_reads.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.NO_SNOOP_NEEDEDoffcore_response.all_rfo.l3_hit_e.snoop_missoffcore_response.all_rfo.l3_hit_e.snoop_noneoffcore_response.all_rfo.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_M.SNOOP_MISSoffcore_response.all_rfo.l3_hit_s.snoop_missThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.demand_rfo.l3_hit_f.hit_other_core_no_fwdoffcore_response.other.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.SUPPLIER_NONE.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020400This event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDoffcore_response.pf_l2_rfo.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDoffcore_response.pf_l3_data_rd.l3_hit_f.hit_other_core_no_fwdoffcore_response.pf_l3_data_rd.l3_hit_m.snoop_missoffcore_response.pf_l3_data_rd.l3_hit_s.hitm_other_coreoffcore_response.pf_l3_rfo.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_M.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400100clx metrics100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) * ( ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , OFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DATA_RD ) / CPU_CLK_UNHALTED.THREAD - (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD)) / #( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (( (20.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) - (3.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) ) * MEM_LOAD_RETIRED.L3_HIT * (1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) / 2) / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) + ( (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD)) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) )Average 3DXP Memory Bandwidth Use for Writes [GB / sec]event=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000491OCR.ALL_PF_DATA_RD.L3_MISS.HITM_OTHER_CORE OCR.ALL_PF_DATA_RD.L3_MISS.HITM_OTHER_CORE OCR.ALL_PF_DATA_RD.L3_MISS.HITM_OTHER_COREOCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDOCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORE  OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.all_pf_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.all_reads.l3_miss_remote_hop1_dram.snoop_missocr.all_rfo.l3_miss.snoop_noneOCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOP  OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPocr.demand_code_rd.l3_miss.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000004ocr.demand_code_rd.l3_miss_local_dram.any_snoopocr.demand_code_rd.l3_miss_local_dram.hitm_other_coreCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.REMOTE_HITMocr.other.l3_miss_local_dram.any_snoopCounts any other requests  OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORECounts any other requests  OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.other.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.pf_l1d_and_sw.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000400ocr.pf_l1d_and_sw.l3_miss_remote_hop1_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000010ocr.pf_l3_data_rd.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000100offcore_response.all_data_rd.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.demand_data_rd.l3_miss.hit_other_core_no_fwdoffcore_response.demand_rfo.l3_miss.hitm_other_coreoffcore_response.other.l3_miss.remote_hit_forwardThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSoffcore_response.pf_l2_rfo.l3_miss_local_dram.hitm_other_coreoffcore_response.pf_l3_data_rd.l3_miss_local_dram.hitm_other_coreoffcore_response.pf_l3_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.SNOOP_NONEOCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONE OCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.all_pf_rfo.supplier_none.hit_other_core_no_fwdOCR.ALL_READS.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDED OCR.ALL_READS.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.demand_code_rd.supplier_none.any_snoopocr.demand_code_rd.supplier_none.hit_other_core_no_fwdocr.other.any_responseocr.pf_l3_data_rd.pmm_hit_local_pmm.snoop_not_neededIntel Optane DC persistent memory bandwidth read (MB/sec). Unit: uncore_imc Read Pending Queue Occupancy of all read requests for Intel Optane DC persistent memory. Unit: uncore_imc Intel Optane DC persistent memory read latency (ns). Derived from unc_m_pmm_rpq_occupancy.all. Unit: uncore_imc Retired load instructions whose data sources were HitM responses from shared L3  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1E003C0004ocr.demand_data_rd.l3_hit.snoop_not_neededocr.hwpf_l2_rfo.l3_hit.anyevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C0020Cycles when optimal number of uops was delivered to the back-end when the back-end is not stalled100 * (( BR_INST_RETIRED.COND + 3 * BR_INST_RETIRED.NEAR_CALL + (BR_INST_RETIRED.NEAR_TAKEN - BR_INST_RETIRED.COND_TAKEN - 2 * BR_INST_RETIRED.NEAR_CALL) ) / TOPDOWN.SLOTS)100 * (( 5 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE - INT_MISC.UOP_DROPPING ) / TOPDOWN.SLOTS) * ( (ICACHE_64B.IFTAG_STALL / CPU_CLK_UNHALTED.THREAD) + (ICACHE_16B.IFDATA_STALL / CPU_CLK_UNHALTED.THREAD) + (10 * BACLEARS.ANY / CPU_CLK_UNHALTED.THREAD) ) / #(( 5 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE - INT_MISC.UOP_DROPPING ) / TOPDOWN.SLOTS)CORE_POWER.LVL1_TURBO_LICENSE / CPU_CLK_UNHALTED.DISTRIBUTEDNumber of times an RTM execution abortedAll mispredicted branch instructions retired (Precise event)event=0xa3,cmask=20,period=1000003,umask=0x14Number of uops executed on port 5uops_dispatched.port_6uops_dispatched.port_7_8Uops that RAT issues to RSuops_retired.slotsPage walks completed due to a demand data load to a 2M/4M pagePage walks completed due to a demand data load to a 4K pageCounts the number of times a load got blocked due to false dependencies in MOB due to partial compare on addressocr.prefetches.l3_hit 1000000000 * ( UNC_CHA_TOR_OCCUPANCY.IA_MISS_DRD_DDR / UNC_CHA_TOR_INSERTS.IA_MISS_DRD_DDR ) / cha_0@event\=0x0@event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84400004Counts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallyCounts demand data reads that were supplied by PMM attached to another socketocr.demand_rfo.local_pmmCounts full cacheline writes (ItoM) that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline was homed in a remote socketunc_cha_llc_victims.allunc_cha_tor_inserts.ia_hitTOR Inserts : All requests from IO Devices. Unit: uncore_cha event=0x35,umask=0xC8177E01event=0x35,umask=0xC8C7FF01event=0x35,umask=0xCC57FF01TOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices that missed the LLC. Unit: uncore_cha event=0x35,umask=0xC8178A01event=0x35,umask=0xc867fe01TOR Inserts : LLCPrefData issued by iA Cores that missed the LLC. Unit: uncore_cha Number Transactions requested of the CPU : Card writing to DRAM. Unit: uncore_iio event=0x83,ch_mask=0x80,fc_mask=0x07,umask=0x01unc_iio_data_req_of_cpu.cmpd.part7event=0xc1,ch_mask=0x10,fc_mask=0x07,umask=0x04unc_m2m_cms_clockticksTOR Inserts : ItoMs issued by IO Devices : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsCounts the number of L2 Cache accesses that resulted in a miss. Counts on a per core basisCounts the number of cacheable memory requests that miss in the Last Level Cache (LLC). If the platform has an L3 cache, the LLC is the L3 cache, otherwise it is the L2 cache. Counts on a per core basisCounts the number of retired split store uops  Supports address when precise (Precise event)ocr.all_code_rd.l3_hitCounts all code reads that were supplied by the L3 cache where no snoop was needed to satisfy the requestThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_HITCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent but the snoop missedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0010Counts the number of misaligned store uops that are 4K page splits (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000010Counts L2 cache hardware prefetch code reads (written to the L2 cache only) that were supplied by DRAMocr.hwpf_l2_code_rd.local_dramCounts L2 cache hardware prefetch RFOs (written to the L2 cache only) that have any type of responsebtclear.anyCounts the total number of consumed retirement slots (Precise event)TOR Inserts : DRd_Opt_Prefs issued by iA Cores that missed the LLC. Unit: uncore_cha TOR Occupancy : All requests from iA Cores : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsPCIe Completion Buffer Inserts of completions with data : Part 2 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 7PCIe Completion Buffer Occupancy : Part 0-7Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 5Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Counts the number of page walks completed due to loads (including SW prefetches) whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 1GB pages. Includes page walks that page faultCounts the number of Extended Page Directory Entry hitsevent=0x4f,period=2000003,umask=0x8Counts the number Extended Page Directory Pointer Entry misses.  The Extended Page Directory cache is used by Virtual Machine operating systems while the guest operating systems use the standard TLB cachescycles / CPU_CLK_UNHALTED.REF_TSCFill Buffer (FB) true hits per kilo instructions for retired demand loads. Unit: cpu_core Fraction of cycles where both hardware Logical Processors were active. Unit: cpu_core Counts the number of load ops retired that hit in the L2 cache  Supports address when precise (Precise event). Unit: cpu_atom event=0xd0,period=1000003,umask=0x6Retired load instructions whose data sources were HitM responses from shared L3  Supports address when precise (Precise event). Unit: cpu_core Number of SSE/AVX computational 128-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP14 RSQRT14 SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. Unit: cpu_core Counts the total number of BACLEARS due to all branch types including conditional and unconditional jumps, returns, and indirect branches. Unit: cpu_atom DSB-to-MITE switch true penalty cycles. Unit: cpu_core Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 256 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core event=0xc6,period=100007,umask=0x1,frontend=0x602006event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3FBFC00002This event is deprecated. Refer to new event LD_BLOCKS.ADDRESS_ALIAS (Precise event). Unit: cpu_atom ld_blocks.address_aliasCounts the number of issue slots every cycle that were not consumed by the backend due to the reorder buffer being full (ROB stalls). Unit: cpu_atom Total execution stalls. Unit: cpu_core int_vec_retired.vnni_256unc_m_vc1_requests_rdCounts every 64B write request entering the Memory Controller 0 to DRAM (sum of all channels). Each write request counts as a new request incrementing this counter. However, same cache line write requests (both full and partial) are combined to a single 64 byte data transfer to DRAM. Unit: uncore_imc Read requests with true-miss in L2 cache.[This event is alias to L2_REQUEST.MISS]All L2 requests.[This event is alias to L2_REQUEST.ALL]event=0xd3,period=1000003,umask=0x2event=0x2a,period=100003,umask=0x1,offcore_rsp=0x808000004event=0x2a,period=100003,umask=0x1,offcore_rsp=0x1030004477event=0xcf,period=100003,umask=0x10event=0x2a,period=100003,umask=0x1,offcore_rsp=0x73C000002event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F33004477Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were not supplied by the local socket's L1, L2, or L3 caches and were supplied by a remote socketevent=0x5,umask=0x00000000cfevent=0xe0,umask=0x0000000001Precharge due to read, write, underfill, or PGT. Unit: uncore_imc DRAM Precharge commands. : Prechages from Page Table. Unit: uncore_imc Number Transactions requested of the CPU : Card writing to another Card (same or different stack). Unit: uncore_iio event=0x35,umask=0x00c816fe01TOR Inserts for DRds issued by iA Cores targeting PMM Mem that Missed the LLC. Unit: uncore_cha All Null Flits. Unit: uncore_upi unc_upi_rxl_flits.llctrlMatches on Receive path of a UPI Port : Non-Coherent Bypass, Match Opcode. Unit: uncore_upi Matches on Receive path of a UPI Port : Non-Coherent Standard. Unit: uncore_upi Direct packet attempts : D2C. Unit: uncore_upi event=0x32,umask=0x0000000001event=0xc0,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000001event=0x83,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000080event=0x21,umask=0x0304unc_m2m_prefcam_demand_drops.ch1_xptevent=0x34,umask=0x00001c19ffTOR Inserts; RFO misses from local IA. Unit: uncore_cha TOR Inserts : All from Local iA and IO. Unit: uncore_cha TOR Inserts : Match the Opcode in b[29:19] of the extended umask field. Unit: uncore_cha TOR Occupancy : All. Unit: uncore_cha event=0x36,umask=0x0000000080unc_cha_tor_occupancy.mmcfgunc_cha_tor_occupancy.ia_llcprefrfoTOR Occupancy; DRd from local IA. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_rfo_localTOR Inserts; LLCPrefData hits from local IA. Unit: uncore_cha event=0x36,umask=0x00cccffd01TOR Occupancy; LLCPrefCode from local IA. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_drd_pref_local_pmmTOR Inserts : WCiLFs issued by iA Cores targeting PMM that missed the LLC - HOMed remotely. Unit: uncore_cha event=0x35,umask=0x00C8678601event=0x35,umask=0x00C86E8A01unc_cha_tor_inserts.ia_miss_local_wcil_ddrTOR Inserts : WbMtoIs issued by IO Devices. Unit: uncore_cha unc_cha_tor_occupancy.ia_clflushoptTOR Occupancy : SpecItoMs issued by iA Cores. Unit: uncore_cha TOR Occupancy : ItoMs issued by iA Cores that Hit LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_itomevent=0x36,umask=0x00c8668a01TOR Occupancy : WCiLs issued by iA Cores targeting PMM that missed the LLC. Unit: uncore_cha event=0x8aThe number of 64 byte instruction cache line fulfilled from system memory or another cachel2_request_g1.change_to_xCore to L2 cacheable request access status (not including L2 Prefetch). Instruction cache request miss in L2 and Data cache request miss in L2 (all types)event=0x6,umask=0x01Retired Instructionsevent=0xc2ex_tagged_ibs_ops.ibs_tagged_opsThe number of x87 floating-point Ops that have retired. The number of events logged per cycle can vary from 0 to 8. Multiply OpsThis is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Double precision multiply-add FLOPS. Multiply-add counts as 2 FLOPSThis is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemes. Number of Scalar Ops optimizedThe number of serializing Ops retired. SSE control word mispredict traps due to mispredictions in RC, FTZ or DAZ, or changes in mask bitsSSE bottom-executing uOps retiredls_stlfls_dc_accessesevent=0x41,umask=0x02event=0x46,umask=0x01Misaligned loadsExecution-Time Branch Misprediction Ratio (Non-Speculative)event=0x64,umask=0x70Average L3 Read Miss Latency (in core clocks)sse_avx_stallsbp_l1_tlb_miss_l2_tlb_miss.if4kRetired Fused Instructions. The number of fuse-branch instructions retired per cycle. The number of events logged per cycle can vary from 0-8event=0x2cls_refills_from_sys.ls_mabresp_rmt_cacheevent=0x43,umask=0x02event=0x5a,umask=0x10event=0x78de_dis_dispatch_token_stalls0.sc_agu_dispatch_stallInstruction Cache Refills from System. The number of 64 byte instruction cache line fulfilled from system memory or another cacheevent=0x85,umask=0x08event=0x28f,umask=0x03x87 bottom-executing ops retired. The number of serializing Ops retiredls_any_fills_from_sys.mem_io_remoteAny Data Cache Fills by Data Source. From CCX Cache in different NodeL1 DTLB Miss. DTLB reload to a coalesced page that hit in the L2 TLBSoftware Prefetch Data Cache Fills by Data Source. From L3 or different L2 in same CCXCycles where a dispatch group is valid but does not get dispatched due to a Token Stall. Also counts cycles when the thread is not selected to dispatch but would have been stalled due to a Token Stall. Floating point register file resource stall. Applies to all FP ops that have a destination registerls_dmnd_fills_from_sys.dram_io_nearSoftware prefetch data cache fills from local L2 cacheevent=0x64,umask=0xf8Core to L2 cache requests (not including L2 prefetch) for data cache accessl2_pf_miss_l2_l3.l1_stridel3_xi_sampled_latency.far_cacheAverage sampled latency when data is sourced from extension memory (CXL) in the same NUMA node. Unit: amd_l3 event=0xad,umask=0x04event=0xad,umask=0x10Cycles with no retire caused by other reasons (retire breaks, traps, faults, etc.)local_processor_read_data_beats_cs11remote_processor_write_data_beats_cs1event=0x21f,umask=0xbffRead data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 2local_socket_inf0_inbound_data_beats_ccm7event=0x5de,umask=0x7feevent=0x55f,umask=0x7feData beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 3Data beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 1event=0x59e,umask=0xbfeRetired SSE and AVX floating-point divide and square root opsevent=0x5,umask=0x0fRetired x87 floating-point opsRetired SSE and AVX integer multiply-accumulate opsevent=0xb,umask=0x50Retired SSE and AVX integer MOV opsRetired 128-bit packed floating-point multiply opsevent=0xc,umask=0x70event=0xc,umask=0x90packed_int_op_type.int128_clmumc_cas_cmd.rdbp_l1_tlb_miss_l2_tlb_miss.allde_op_queue_emptyPipelineL1bad_speculationd_ratio(de_no_dispatch_per_slot.smt_contention, total_dispatch_slots)L2 cache accesses from L1 data cache misses (including prefetch)l2_cache_req_stat.ic_dc_hit_in_l2 + l2_pf_hit_l2.allDRAM write data for local processorevent=0xe0,umask=0x00UNC_CBO_HYPHENTotal read hitsTotal cache misses. Unit: uncore_imc_free_running sys_ddr_pmu.write_cyclesCPU_CYCLESWRITE_BUF_FULLEVENT_1FHEVENT_30HEVENT_43HEVENT_66HEVENT_8FHEVENT_BCHEVENT_CDHEVENT_FAHEXC_RETURNEVENT_104HEVENT_10FHEVENT_116HEVENT_117HEVENT_122HEVENT_136HEVENT_170HEVENT_173HEVENT_19CHEVENT_1ACHEVENT_20BHEVENT_227HEVENT_24BHEVENT_25AHEVENT_289HEVENT_2FDHEVENT_392HEVENT_3B0HEVENT_3BEHINST_SPECEXC_PABORTL1D_TLBhni_rrt_wr_allochni_txdat_stallsbsx_cmo_reqsbsx_cmo_req_trkr_occ_cnt_ovflrnd_rdb_unordcxra_dat_pcrd_stalls_lnk2cxla_tx_cxs_link2L1_INSTR_CACHE_MISSESSC_INSTR_COMPLETEDL1_DATA_CACHE_LOAD_MISS_CYCLES_OVER_THRESHOLDFP_ALL_FPSCR_RENAMES_BUSYSNOOP_VALIDL2_VALID_REQUESTBUS_READS_NOT_RETRIEDTOUCHES_TRANSLATED_ALLOCATED_TO_DLFBINSTR_L1_CACHE_FETCHESFPU_INPUT_DATA_STALLSfsdram-controller-interface-bypasscommandINTEL_COREI7ARMV7_CORTEX_A7ARMV8_CORTEX_A76PMUDEBUGGenuineIntelBR_INST_RETIRED.ALL_BRANCHESUNC_data fabricGenuineIntel-6-2DGenuineIntel-6-1AIDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))This category represents fraction of slots utilized by useful work i.e. issued uops that eventually get retired. Ideally; all pipeline slots would be attributed to the Retiring category.  Retiring of 100% would indicate the maximum Pipeline_Width throughput was achieved.  Maximizing Retiring typically increases the Instructions-per-cycle (see IPC metric). Note that a high Retiring value does not necessary mean there is no room for more performance.  For example; Heavy-operations or Microcode Assists are categorized under Retiring. They often indicate suboptimal performance and can often be optimized or avoided. SMT version; use when SMT is enabled and measuring per logical CPUThe ratio of Executed- by Issued-Uops. Ratio > 1 suggests high rate of uop micro-fusions. Ratio < 1 suggest high rate of "execute" at rename stageInstruction-Level-Parallelism (average number of uops executed when there is at least 1 uop executed)( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * (BR_MISP_RETIRED.ALL_BRANCHES * (12 * ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY ) / CPU_CLK_UNHALTED.THREAD) / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY )) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) ) * (4 * CPU_CLK_UNHALTED.THREAD) / BR_MISP_RETIRED.ALL_BRANCHESInstructions per FP Arithmetic instruction (lower number means higher occurrence rate). May undercount due to FMA double counting. Approximated prior to BDWINST_RETIRED.ANY / FP_ARITH_INST_RETIRED.SCALAR_DOUBLEL1D_PEND_MISS.PENDING / ( MEM_LOAD_UOPS_RETIRED.L1_MISS + mem_load_uops_retired.hit_lfb )Kernel_CPIMEM_Read_Latency(cstate_pkg@c7\-residency@ / msr@tsc@) * 100cacheThis event counts the total number of L2 code requestsl2_rqsts.demand_data_rd_missCycles when L1D is lockedRetired load uop whose Data Source was: remote DRAM either Snoop not needed or Snoop Miss (RspI) (Precise Event)  Supports address when precise.  Spec update: BDE70mem_load_uops_retired.l3_missmem_uops_retired.stlb_miss_storesOffcore outstanding demand rfo reads transactions in SuperQueue (SQ), queue to uncore, every cycle  Spec update: BDM76event=0x60,period=2000003,umask=0x1event=0xc7,period=2000003,umask=0x4This event counts the number of x87 floating point (FP) micro-code assist (numeric overflow/underflow, inexact result) when the output value (destination register) is invalidThis event counts the number of micro-operations cancelled after they were dispatched from the scheduler to the execution units when the total number of physical register read ports across all dispatch ports exceeds the read bandwidth of the physical register file.  The SIMD_PRF subevent applies to the following instructions: VDPPS, DPPS, VPCMPESTRI, PCMPESTRI, VPCMPESTRM, PCMPESTRM, VFMADD*, VFMADDSUB*, VFMSUB*, VMSUBADD*, VFNMADD*, VFNMSUB*.  See the Broadwell Optimization Guide for more informationevent=0xab,period=2000003,umask=0x2This event counts the number of instruction cache, streaming buffer and victim cache misses. Counting includes UC accessesThis event counts cycles during which uops are being delivered to Instruction Decode Queue (IDQ) from the MITE path. Counting includes uops that may bypass the IDQevent=0x79,period=2000003,umask=0x30br_inst_exec.taken_direct_jumpThis is a precise version (that is, uses PEBS) of the event that counts conditional branch instructions retired (Precise event)Mispredicted indirect branches excluding calls and returnsevent=0xc0,period=2000003,umask=0x2FP operations  retired. X87 FP operations that have no exceptions:Cycles checkpoints in Resource Allocation Table (RAT) are recovering from JEClear or machine clearFalse dependencies in MOB due to partial compareevent=0xc3,period=100003,umask=0x4Cycles per core when uops are dispatched to port 2unc_h_requests.writes_remote(UNC_P_POWER_STATE_OCCUPANCY.CORES_C3 / UNC_P_CLOCKTICKS) * 100.unc_p_freq_trans_cyclesevent=0x85,period=100003,umask=0x1event=0x85,period=100003,umask=0x2Number of DTLB page walker hits in Memory  Spec update: BDM69, BDM98Retired load uops which data sources were L3 hit and cross-core snoop missed in on-pkg core cache  Supports address when precise.  Spec update: BDM100 (Precise event)Miss in mid-level (L2) cache. Excludes Unknown data-source  Supports address when precise (Precise event)This event counts load uops with locked access retired to the architected path  Supports address when precise.  Spec update: BDM35 (Precise event)This event counts store uops with true STLB miss retired to the architected path. True STLB miss is an uop triggering page walk that gets completed without blocks, and later gets retired. This page walk can end up with or without a fault  Supports address when precise (Precise event)offcore_response.all_data_rd.supplier_none.snoop_noneoffcore_response.all_data_rd.supplier_none.snoop_not_neededoffcore_response.all_pf_data_rd.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0008offcore_response.corewb.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0040Counts all prefetch (that bring data to LLC only) RFOs have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020100Counts randomly selected loads with latency value being above 512  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)offcore_response.all_pf_code_rd.l3_miss.snoop_missoffcore_response.all_pf_code_rd.supplier_none.snoop_non_dramoffcore_response.all_pf_data_rd.l3_miss_local_dram.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000002offcore_response.other.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000020offcore_response.pf_l3_code_rd.l3_miss.snoop_noneoffcore_response.pf_l3_code_rd.supplier_none.snoop_non_dramoffcore_response.pf_l3_data_rd.l3_miss.snoop_missoffcore_response.pf_l3_rfo.l3_miss_local_dram.snoop_missNumber of times RTM abort was triggered (Precise event)Retirement slots used (Precise event)L3 Lookup read request that access cache and found line in M-stateevent=0x34,umask=0x88L3 Lookup read request that access cache and found line in any MESI-state. Unit: uncore_cbox unc_arb_trk_requests.writesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C07F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0604000091event=0x21,period=200000,umask=0x40l2_reject_busq.self.prefetch.s_statel2_st.self.m_stateSIMD packed micro-ops retiredSIMD packed micro-ops executedevent=0xb3,period=2000000,umask=0x82event=0xb3,period=2000000,umask=0x88event=0x10,period=2000000,umask=0x82FXCH uops retired (Must be precise)Icache hitevent=0xc4,period=2000000,umask=0xfevent=0xc5,period=200000,umask=0x0Retired branch instructions that were predicted not-takenevent=0x88,period=2000000,umask=0x41event=0x89,period=200000,umask=0x8Self-Modifying Code detectedMicro-ops retiredITLB flushesevent=0xd1,period=200003,umask=0x8Memory uops retired (Precise event capable)  Supports address when precise (Must be precise)Counts the number of memory uops retired that is either a loads or a store or both  Supports address when precise (Must be precise)event=0xd0,period=200003,umask=0x82Counts data reads (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredoffcore_response.any_pf_data_rd.l2_miss.anyCounts the number of writeback transactions caused by L1 or L2 cache evictions that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000001Counts demand reads for ownership (RFO) requests generated by a write to full data cache line that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredoffcore_response.partial_streaming_stores.l2_miss.hit_other_core_no_fwdevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200002000Counts reads for ownership (RFO) requests generated by L2 prefetcher that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)BACLEARs asserted for conditional branchfetch_stall.allCounts near CALL branch instructions retired (Must be precise)uops_retired.idivCounts every core cycle a page-walk is in progress due to either a data memory operation or an instruction fetchoffcore_response.any_data_rd.any_responseoffcore_response.any_rfo.any_responseCounts reads for ownership (RFO) requests (demand & prefetch) outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache have any transaction responses from the uncore subsystemCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts data cacheline reads generated by hardware L2 cache prefetcher miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredFloating point divide uops retired (Precise Event Capable) (Must be precise)machine_clears.page_faultdtlb_load_misses.walk_completed_1gbPage walk completed due to a demand data store to a 1GB pageevent=0xd0,period=100003,umask=0x21Offcore outstanding demand rfo reads transactions in SuperQueue (SQ), queue to uncore, every cycle  Spec update: HSD62, HSD61, HSM63Offcore outstanding RFO store transactions in SuperQueue (SQ), queue to uncore  Spec update: HSD62, HSD61, HSM63Counts demand data reads hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedNote that a whole rep string only counts AVX_INST.ALL onceNumber of X87 FP assists due to input valuesCycles with pending L2 miss loads. Set Cmask=2 to count cycle  Spec update: HSD78, HSM63, HSM80This event counts the number of instructions retired from execution. For instructions that consist of multiple micro-ops, this event counts the retirement of the last micro-op of the instruction. Counting continues during hardware interrupts, traps, and inside interrupt handlers. INST_RETIRED.ANY is counted by a designated fixed counter, leaving the programmable counters available for other events. Faulting executions of GETSEC/VM entry/VM Exit/MWait will not count as retired instructions  Spec update: HSD140, HSD143This event is incremented when self-modifying code (SMC) is detected, which causes a machine clear.  Machine clears can have a significant performance impact if they are happening frequentlyThis events counts the cycles where at least one uop was executed. It is counted per thread  Spec update: HSD144, HSD30, HSM31Cycles per core when uops are executed in port 1event=0xffCompleted page walks due to demand load misses that caused 2M/4M page walks in any TLB levelsThis event counts cycles when the  page miss handler (PMH) is servicing page walks caused by DTLB store missesCode miss in all TLB levels causes a page walk that completes. (2M/4M)Counts the number of Extended Page Table walks from the DTLB that hit in the L1 and FBCounts the number of Extended Page Table walks from the ITLB that hit in the L1 and FBevent=0xbc,period=2000003,umask=0x88Retired load uop whose Data Source was: Remote cache HITM  Supports address when precise.  Spec update: HSM30 (Precise event)offcore_response.pf_llc_data_rd.llc_miss.any_responseevent=0xf2,period=100003,umask=0xaevent=0x24,period=200003,umask=0xc0Counts demand data reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts demand data writes (RFOs) that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts 256-bit packed single-precision floating-point instructionsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400004Execution stalls due to L2 cache missesNumber of instructions retired. General Counter   - architectural eventCycles stalled due to no store buffers available (not including draining form sync)event=0xa1,period=2000003,umask=0x30event=0x22,umask=0x04unc_cbo_xsnp_response.external_filterevent=0x34,umask=0x02Counts load operations that missed 1st level DTLB but hit the 2nd levelMisses in all ITLB levels that cause page walksevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x67fc00001PCIe non-snoop reads. Derived from unc_c_tor_inserts.opcode.pcie_read. Unit: uncore_cbox Occupancy counter for LLC data reads (demand and L2 prefetch). Derived from unc_c_tor_occupancy.miss_opcode.llc_data_read. Unit: uncore_cbox event=0x36,umask=0x8Aevent=0x10(UNC_P_FREQ_GE_2000MHZ_CYCLES / UNC_P_CLOCKTICKS) * 100.This event counts all LLC misses for all demand and L2 prefetches. LLC prefetches are excludedagu_bypass_cancel.countbr_misp_exec.all_direct_near_callbr_misp_retired.not_takenevent=0x59,period=2000003,umask=0x80resource_stalls2.all_prf_controlunc_c_tor_occupancy.miss_allCounts the number of cycles that the uncore was running at a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band1=XXX with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Unit: uncore_pcu Counts Demand code reads and prefetch code read requests  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts Demand cacheable data and L1 prefetch data read requests  that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0offcore_response.any_pf_l2.l2_hit_near_tileoffcore_response.any_read.l2_hit_this_tile_mCounts any Read request  that accounts for responses which hit its own tile's L2 with data in M stateCounts Bus locks and split lock requests that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180002Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010100event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000082000offcore_response.pf_l1_data_rd.l2_hit_this_tile_moffcore_response.pf_l2_code_rd.l2_hit_far_tileevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080020Counts Software Prefetches that accounts for any responseoffcore_response.pf_software.l2_hit_far_tileCounts Software Prefetches that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010001000Counts all streaming stores (WC and should be programmed on PMC1) that accounts for any responseoffcore_response.any_request.mcdramoffcore_response.bus_locks.mcdram_nearCounts demand code reads and prefetch code reads that accounts for responses from MCDRAM (local and far)Counts demand cacheable data and L1 prefetch data reads that accounts for data responses from MCDRAM Localoffcore_response.demand_rfo.ddr_nearCounts Demand cacheable data writes that accounts for data responses from MCDRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800080event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080201000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800200Counts the number of near RET branch instructions retired (Precise event)br_misp_retired.rel_callCounts the number of occurences a retired load gets blocked because its address partially overlaps with a store  Supports address when precise (Precise event)This event counts the number of micro-ops retired that were supplied from MSROMCycles L1D and L2 lockedL1D snoop eviction of cache lines in M statel1d_cache_ld.mesil1d_cache_lock.hitL1 data cache load locks in M statel1d_cache_lock.s_stateevent=0x28,period=100000,umask=0x8L1 writebacks to L2 in S stateevent=0x26,period=200000,umask=0xfl2_data_rqsts.prefetch.e_stateL2 load hitsL2 prefetch hitsl2_rqsts.rfosl2_transactions.fillLongest latency cache referenceMemory instructions retired above 128 clocks (Precise Event)event=0xb,period=500,umask=0x10,ldlat=0x100Memory instructions retired above 8192 clocks (Precise Event)Load instructions retired remote cache HIT data source (Precise Event)event=0xb0,period=100000,umask=0x40offcore_response.any_ifetch.llc_hit_no_other_coreOffcore code reads that HITM in a remote cacheoffcore_response.any_request.llc_hit_other_core_hitmOffcore RFO requests satisfied by the LLC  and HITM in a sibling coreoffcore_response.data_ifetch.llc_hit_no_other_coreoffcore_response.data_in.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF33event=0xb7,period=100000,umask=0x1,offcore_rsp=0x433Offcore demand code reads satisfied by the LLCoffcore_response.pf_data.io_csr_mmiooffcore_response.pf_data.llc_hit_no_other_coreOffcore prefetch code reads satisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1820MMX UopsComputational floating-point operations executedTransitions from Floating Point to MMX instructions128 bit SIMD integer shuffle/move operationsSIMD integer 64 bit shift operationsevent=0xd0,period=2000000,umask=0x1event=0xb7,period=100000,umask=0x1,offcore_rsp=0x6011Offcore demand data requests satisfied by a remote DRAMoffcore_response.pf_data.remote_dramevent=0x80,period=2000000,umask=0x3All loads dispatchedseg_rename_stallsCycles no Uops issued on ports 0-4 (core count)Counts L2 cache misses when fetching instructionsRequests from the L1/L2/L3 hardware prefetchers or Load software prefetches that hit L2 cacheevent=0x24,period=200003,umask=0x38event=0xd4,period=100007,umask=0x4Retired load instructions missed L3 cache as data sources  Supports address when precise (Precise event)Counts cycles when offcore outstanding Demand Data Read transactions are present in the super queue (SQ). A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080001offcore_response.demand_data_rd.l4_hit_local_l4.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200048000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100408000Counts cycles during which uops are being delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Counting includes uops that may 'bypass' the IDQCycles with less than 2 uops delivered by the front-endhle_retired.aborted_unfriendlyevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C400002offcore_response.other.l3_hit_e.snoop_non_dramThis event counts not taken branch instructions retired  Spec update: SKL091event=0xa6,period=2000003,umask=0x8event=0xa6,period=2000003,umask=0x10Counts self-modifying code (SMC) detected, which causes a machine clearCounts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 31000 * MEM_LOAD_RETIRED.L2_MISS / INST_RETIRED.ANYCounts page walks of any page size (4K/2M/4M/1G) caused by a code fetch. This implies it missed in the ITLB and further levels of TLB, but the walk need not have completedThis event counts the total number of L2 cache references and the number of L2 cache misses respectivelyCross core or cross module hitm (Precise event)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000022Counts the number of baclearsThe BACLEARS event counts the number of times the front end is resteered, mainly when the Branch Prediction Unit cannot provide a correct prediction and this is corrected by the Branch Address Calculator at the front end.  The BACLEARS.RETURN event counts the number of RETURN baclearsFAR counts the number of far branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)IND_CALL counts the number of mispredicted near indirect CALL branch instructions retired.  This event counts the number of retired branch instructions that were mispredicted by the processor, categorized by type. A branch misprediction occurs when the processor predicts that the branch would be taken, but it is not, or vice-versa.  When the misprediction is discovered, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct path (Precise event)This event counts retired demand loads that missed the  last-level (L3) cache. This means that the load is usually satisfied from memory in a client system or possibly from the remote socket in a server. Demand loads are non speculative load uops. (Precise Event - PEBS) (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10433offcore_response.pf_l2_rfo.llc_hit.hitm_other_coreoffcore_response.pf_l2_rfo.llc_hit.snoop_missoffcore_response.any_request.llc_miss_local.dramCycles offcore reads busyREQUEST = ANY_DATA read and RESPONSE = REMOTE_CACHE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x50ffevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff77REQUEST = DATA_IFETCH and RESPONSE = LOCAL_CACHEevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5050REQUEST = PF_DATA and RESPONSE = LOCAL_CACHEREQUEST = PREFETCH and RESPONSE = LLC_HIT_NO_OTHER_COREREQUEST = ANY IFETCH and RESPONSE = OTHER_LOCAL_DRAMREQUEST = ANY RFO and RESPONSE = OTHER_LOCAL_DRAMREQUEST = OTHER and RESPONSE = ANY_DRAM AND REMOTE_FWDevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3040Cycles snoop data requests queuedevent=0x49,period=200000,umask=0x80Offcore uncached memory accessesevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5833event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0004offcore_response.pf_l1d_and_sw.l3_hit.snoop_hit_with_fwdfp_arith_inst_retired.512b_packed_singleevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000004Counts all demand code reads that miss the L3 and the data is returned from local dramCounts all demand data writes (RFOs) that miss the L3 and the data is returned from remote dramCounts L1 data cache hardware prefetch requests and software prefetch requests that miss the L3 and the data is returned from local dramoffcore_response.pf_l2_data_rd.l3_miss.remote_hit_forwardCounts all prefetch (that bring data to L2) RFOs that miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00080event=0x28,period=200003,umask=0x20IpArith_AVX5121000 * L2_LINES_OUT.NON_SILENT / INST_RETIRED.ANYCORE_POWER.LVL1_TURBO_LICENSE / CPU_CLK_UNHALTED.THREADMEM_DRAM_Read_LatencyIO_Write_BWCounts the number of writes requests allocated into the Write Pending Queue (WPQ).  The WPQ is used to schedule writes out to the memory controller and to track the requests.  Requests allocate into the WPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the CHA to the iMC (Memory Controller).  The write requests deallocate after being issued to DRAM.  Write requests themselves are able to complete (from the perspective of the rest of the system) as soon they have 'posted' to the iMCUncore cache clock ticks. Unit: uncore_cha Streaming stores (partial cache line). Derived from unc_cha_tor_inserts.ia_miss. Unit: uncore_cha UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART0 +UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART1 +UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART2 +UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART3Multi-socket cacheline Directory state lookups; Snoop Not Needed. Unit: uncore_cha Local requests for exclusive ownership of a cache line without receiving data. Unit: uncore_cha unc_cha_rxc_occupancy.irqunc_cha_snoop_resp.rspsfwdunc_iio_comp_buf_inserts.cmpd.part0event=0xc2,ch_mask=0x04,fc_mask=0x4,umask=0x03Peer to peer read request for 4 bytes made by a different IIO unit to IIO Part0. Unit: uncore_iio event=0x83,ch_mask=0x02,fc_mask=0x07,umask=0x02event=0xc1,ch_mask=0x04,fc_mask=0x07,umask=0x04event=0xc1,ch_mask=0x02,fc_mask=0x07,umask=0x08event=0x84,ch_mask=0x04,fc_mask=0x07,umask=0x01PCIITOM request issued by the IRP unit to the mesh with the intention of writing a full cacheline to coherent memory, without a RFO.  PCIITOM is a speculative Invalidate to Modified command that requests ownership of the cacheline and does not move data from the mesh to IRP cacheInbound write (fast path) requests to coherent memory, received by the IRP resulting in write ownership requests issued by IRP to the meshevent=0x28event=0x26Counts when the M2M (Mesh to Memory) updates the multi-socket cacheline Directory state from from I (Invalid) to S (Shared)Counts when the M2M (Mesh to Memory) updates the multi-socket cacheline Directory state from from S (Shared) to I (Invalid)Reads to iMC issued. Unit: uncore_m2m Null FLITs transmitted from any slot. Unit: uncore_upi OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200491ocr.all_pf_data_rd.l3_hit.any_snoopOCR.ALL_PF_DATA_RD.L3_HIT_E.HITM_OTHER_CORE  OCR.ALL_PF_DATA_RD.L3_HIT_E.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040490ocr.all_pf_data_rd.l3_hit_s.hitm_other_coreocr.all_pf_rfo.l3_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0120ocr.all_pf_rfo.l3_hit_e.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C07F7ocr.all_reads.l3_hit_f.snoop_noneocr.all_reads.l3_hit_m.no_snoop_neededocr.all_reads.l3_hit_s.snoop_noneOCR.ALL_RFO.L3_HIT_S.SNOOP_MISSCounts all demand code reads OCR.DEMAND_CODE_RD.L3_HIT.NO_SNOOP_NEEDED OCR.DEMAND_CODE_RD.L3_HIT.NO_SNOOP_NEEDEDocr.demand_code_rd.l3_hit_f.no_snoop_neededocr.demand_code_rd.l3_hit_s.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200001Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_S.HITM_OTHER_COREocr.demand_rfo.l3_hit_f.any_snoopocr.demand_rfo.l3_hit_m.hitm_other_coreCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_S.ANY_SNOOPocr.other.l3_hit.hit_other_core_no_fwdCounts any other requests  OCR.OTHER.L3_HIT_F.HIT_OTHER_CORE_FWDocr.other.l3_hit_f.snoop_missocr.pf_l1d_and_sw.l3_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080010ocr.pf_l2_data_rd.l3_hit_e.hit_other_core_no_fwdocr.pf_l2_data_rd.l3_hit_s.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200020Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0080Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDocr.pf_l3_rfo.l3_hit_e.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200100offcore_response.all_data_rd.l3_hit_e.any_snoopoffcore_response.all_data_rd.l3_hit_f.snoop_missThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDoffcore_response.all_data_rd.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_M.SNOOP_NONEoffcore_response.all_data_rd.pmm_hit_local_pmm.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020491offcore_response.all_pf_data_rd.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_M.ANY_SNOOPoffcore_response.all_pf_rfo.l3_hit_f.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_F.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_READS.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_M.HIT_OTHER_CORE_NO_FWDoffcore_response.all_reads.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_READS.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_READS.SUPPLIER_NONE.NO_SNOOP_NEEDEDoffcore_response.all_rfo.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_S.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDoffcore_response.demand_data_rd.l3_hit_e.hitm_other_coreoffcore_response.demand_data_rd.l3_hit_e.no_snoop_neededoffcore_response.demand_rfo.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.SNOOP_NONEThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_M.SNOOP_NONEoffcore_response.other.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_M.NO_SNOOP_NEEDEDoffcore_response.pf_l2_data_rd.l3_hit.hit_other_core_fwdoffcore_response.pf_l2_data_rd.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_F.SNOOP_NONEoffcore_response.pf_l2_rfo.supplier_none.hitm_other_coreoffcore_response.pf_l2_rfo.supplier_none.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_data_rd.l3_hit_f.hit_other_core_fwdoffcore_response.pf_l3_data_rd.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_data_rd.supplier_none.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L3_RFO.SUPPLIER_NONE.ANY_SNOOPFraction of Uops delivered by the LSD (Loop Stream Detector; aka Loop Cache)Counts once for each Intel AVX-512 computational 512-bit packed BFloat16 floating-point instruction retired. Applies to the ZMM based VDPBF16PS instruction.  Each count represents 64 computation operations. This event is only supported on products formerly named Cooper Lake and is not supported on products formerly named Cascade Lakeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000490ocr.all_pf_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdOCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEOCR.ALL_READS.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_READS.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_READS.L3_MISS.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10100007F7ocr.all_reads.l3_miss_remote_hop1_dram.hit_other_core_no_fwdOCR.ALL_RFO.L3_MISS.ANY_SNOOP OCR.ALL_RFO.L3_MISS.ANY_SNOOP OCR.ALL_RFO.L3_MISS.ANY_SNOOPOCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_FWDocr.all_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.all_rfo.l3_miss_remote_hop1_dram.snoop_noneocr.demand_code_rd.l3_miss.hit_other_core_no_fwdocr.demand_code_rd.l3_miss_local_dram.hit_other_core_no_fwdocr.demand_data_rd.l3_miss.remote_hit_forwardevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000002ocr.demand_rfo.l3_miss_local_dram.any_snoopocr.pf_l1d_and_sw.l3_miss_local_dram.hit_other_core_fwdocr.pf_l1d_and_sw.l3_miss_remote_hop1_dram.hitm_other_coreocr.pf_l2_data_rd.l3_miss.any_snoopCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.HITM_OTHER_CORE OCR.PF_L2_DATA_RD.L3_MISS.HITM_OTHER_CORECounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.REMOTE_HITMCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000020ocr.pf_l2_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000020ocr.pf_l2_rfo.l3_miss_remote_hop1_dram.snoop_missCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.SNOOP_MISSocr.pf_l3_rfo.l3_miss.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.SNOOP_MISSoffcore_response.all_pf_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreoffcore_response.all_pf_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.all_reads.l3_miss.hit_other_core_no_fwdoffcore_response.all_reads.l3_miss.remote_hitmThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.REMOTE_HITMoffcore_response.demand_data_rd.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREoffcore_response.pf_l3_rfo.l3_miss_remote_hop1_dram.any_snoopOCR.ALL_DATA_RD.SUPPLIER_NONE.SNOOP_NONEocr.all_pf_data_rd.supplier_none.any_snoopocr.pf_l2_rfo.pmm_hit_local_pmm.any_snoopocr.pf_l2_rfo.supplier_none.hit_other_core_fwdCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDunc_m_pmm_bandwidth.read6.103515625E-5MB/secunc_m_pmm_cmd1.ufill_rdTag Hit; Read Hit from NearMem, Dirty  Lineevent=0xd0,period=1000003,umask=0x81Counts demand instruction fetches and L1 instruction cache prefetches that hit a cacheline in the L3 where a snoop was sent or notocr.demand_code_rd.l3_hit.snoop_hit_no_fwdCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit a cacheline in the L3 where a snoop was sent or notocr.hwpf_l2_rfo.l3_hit.snoop_missCounts miscellaneous requests, such as I/O and un-cacheable accesses that hit a cacheline in the L3 where a snoop was not needed to satisfy the requestevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C0800Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of at least 1 cycle which was not interrupted by a back-end stall (Precise event)event=0xc6,period=100007,umask=0x1,frontend=0x520006event=0xc6,period=100007,umask=0x1,frontend=0x504006C8 residency percent per packageC8_Pkg_ResidencyCounts Core cycles where the core was running with power-delivery for baseline license level 0.  This includes non-AVX codes, SSE, AVX 128-bit, and low-current AVX 256-bit codesCore cycles where the core was running with power-delivery for license level 2 (introduced in Skylake Server microarchtecture).  This includes high current AVX 512-bit instructionsevent=0x14,cmask=1,period=1000003,umask=0x9Counts Core crystal clock cycles when current thread is unhalted and the other thread is haltedCounts the number of times a load got blocked due to false dependencies due to partial compare on addressCounts the number of uops delivered to the back-end by the LSD(Loop Stream Detector)TMA slots available for an unhalted logical processor. General counter - architectural eventCounts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to ports 7 and 8uops_executed.cycles_ge_1Counts the number of page walks outstanding for an outstanding code (instruction fetch) request in the PMH (Page Miss Handler) each cycle1 / IPCBrMispredictsBranchesCounts the number of lines that are evicted by the L2 cache due to L2 cache fills.  Evicted lines are delivered to the L3, which may or may not cache them, according to system load and prioritiesocr.hwpf_l3.l3_hitcore_snoop_response.i_fwd_feunc_m_dram_refresh.highevent=0xea,umask=0x08unc_cha_requests.writes_remoteTOR Inserts : DRd_Prefs issued by iA Cores that Missed the LLC. Unit: uncore_cha event=0x35,umask=0xCC43FF04TOR Occupancy : CRDs issued by iA Cores. Unit: uncore_cha event=0x35,umask=0xC8877E01event=0x35,umask=0xCD43FE04unc_cha_tor_inserts.ia_llcprefdataTOR Inserts : LLCPrefData issued by iA Cores. Unit: uncore_cha unc_iio_data_req_of_cpu.cmpd.part3Number Transactions requested of the CPU : CmpD - device sending completion to CPU request. Unit: uncore_iio unc_iio_txn_req_of_cpu.cmpd.part1event=0xc0,ch_mask=0x80,fc_mask=0x07,umask=0x04unc_iio_data_req_of_cpu.mem_write.part6unc_iio_txn_req_by_cpu.mem_read.part4Coherent Ops : WbMtoI. Unit: uncore_irp Counts the number of cycles the core is stalled due to an instruction cache or Translation Lookaside Buffer (TLB) access which hit in the Last Level Cache (LLC) or other core with HITE/F/MCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0040Counts L2 cache hardware prefetch data reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent but the snoop missedCounts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedCounts the number of requests to the instruction cache for one or more bytes of a cache lineevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184008000c0_stalls.load_l2_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400000010000Counts the number of unhalted reference clock cycles at TSC frequency. (Fixed event)event=0x3,period=1000003,umask=0x10Counts the number of issue slots every cycle that were not delivered by the frontend due to other common frontend stalls not categorizedtopdown_fe_bound.predecodeHalf clockticks for IMCTOR Inserts : DRd_Opt issued by iA Cores that missed the LLC. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_ucrdfTOR Occupancy : DRd_Opts issued by iA Cores that hit the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsNumber Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 5ept.epdpe_hitevent=0x85,period=200003,umask=0xeFraction of branches that are unconditional (direct or indirect) jumps. Unit: cpu_core Average per-core data fill bandwidth to the L3 cache [GB / sec]. Unit: cpu_core Unit: cpu_atom 100 * MEM_UOPS_RETIRED.SPLIT_LOADS / MEM_UOPS_RETIRED.ALL_LOADSIDiv_Uop_Ratio100 * UOPS_RETIRED.X87 / UOPS_RETIRED.ALLMEM_BOUND_STALLS.LOAD_DRAM_HIT / MEM_LOAD_UOPS_RETIRED.DRAM_HIT100 * MEM_BOUND_STALLS.IFETCH_DRAM_HIT / ( MEM_BOUND_STALLS.IFETCH )Counts the number of cycles the core is stalled due to a demand load miss which hit in the L2, LLC, DRAM or MMIO (Non-DRAM). Unit: cpu_atom Counts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 128 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom mem_uops_retired.load_latency_gt_4Counts the number of issue slots every cycle that were not delivered by the frontend due to instruction cache misses. Unit: cpu_atom All retired memory instructions  Supports address when precise (Precise event). Unit: cpu_core event=0x21,period=100003,umask=0x1Number of PREFETCHT0 instructions executed. Unit: cpu_core Uops delivered to Instruction Decode Queue (IDQ) from MITE path. Unit: cpu_core ld_head.any_at_retevent=0x5,period=1000003,umask=0xf4Number of occurrences where a microcode assist is invoked by hardware. Unit: cpu_core core_power.license_1Counts the number of retired loads that are blocked because its address exactly matches an older store whose data is not ready (Precise event). Unit: cpu_atom event=0xc0,period=2000003,umask=0x8Counts cycles after recovery from a branch misprediction or machine clear till the first uop is issued from the resteered path. Unit: cpu_core event=0xad,period=1000003,umask=0x40,frontend=0x7Uops executed on ports 4 and 9. Unit: cpu_core uops_retired.heavyevent=0x3event=0xd1,period=1000003,umask=0x80event=0x2a,period=100003,umask=0x1,offcore_rsp=0x8003C4477event=0x2a,period=100003,umask=0x1,offcore_rsp=0x90002380Counts Demand RFOs, ItoM's, PREFECTHW's, Hardware RFO Prefetches to the L1/L2 and Streaming stores that likely resulted in a store to Memory (DRAM or PMM)INT_MISC.MBA_STALLSUops executed on ports 5 and 11Number of uops dispatch to execution  ports 7 and 8Counts cycles where at least one uop has retiredUOPS_RETIRED.HEAVYevent=0x1,umask=0x0000000001unc_m_pmm_rpq_occupancy.all_sch1PMM Read Pending Queue inserts. Unit: uncore_imc event=0xe4,umask=0x03event=0x5,umask=0x00000000c8event=0x3,umask=0x0000000040unc_iio_data_req_of_cpu.peer_write.part5TOR Inserts for RdCur from local IO. Unit: uncore_cha event=0x2,umask=0x0000000008unc_upi_rxl_flits.idleRxQ Flit Buffer Allocations : Slot 0. Unit: uncore_upi FAF - request insert from TC. Unit: uncore_irp Read request for 4 bytes made by IIO Part3 to Memory. Unit: uncore_iio event=0x83,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000080event=0x33,umask=0x0000000001event=0x5d,umask=0x0000000005unc_m3upi_d2u_sentunc_m3upi_d2c_sentunc_cha_llc_lookup.remote_snpTOR Inserts; RFO hits from local IO. Unit: uncore_cha TOR Inserts; RFO prefetch misses from local IA. Unit: uncore_cha event=0x35,umask=0x00cc57ff01event=0x36,umask=0x00c807fe01TOR Occupancy : MMCFG Access. Unit: uncore_cha TOR Occupancy : Just Remote Targets. Unit: uncore_cha event=0x36,umask=0x00c8877e01TOR Occupancy; RdCur and FsRdCur misses from local IO. Unit: uncore_cha TOR Occupancy; RdCur and FsRdCur from local IO. Unit: uncore_cha TOR Occupancy; LLCPrefCode hits from local IA. Unit: uncore_cha event=0x35,umask=0x00c8168601event=0x35,umask=0x00C8968A01TOR Inserts : WCiLs issued by iA Cores targeting DDR that missed the LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_drd_remote_ddrTOR Occupancy : DRd_Prefs issued by iA Cores targeting PMM Mem that Missed the LLC. Unit: uncore_cha TOR Occupancy : WCiLF issued by iA Cores. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_local_wcilf_pmmTOR Occupancy : WCiLs issued by iA Cores that Missed the LLC. Unit: uncore_cha Dynamic Indirect Predictionsbp_l1_tlb_miss_l2_missInstruction Pipe Stall. IC pipe was stalled during this clock cycle (including IC to OC fetches) due to DQ emptyMiscellaneous events covered in more detail by l2_request_g1 (PMCx060)l2_request_g2.ls_rd_sized_ncAll L2 Cache Requests (Breakdown 2 - Rare). Bus locksl2_request_g2.bus_locks_responsesAll L2 Cache Requests (Breakdown 2 - Rare). Bus lock responseRetired UopsThe number of branch instructions retired, of any type, that were not correctly predicted. This includes those for which prediction is not attempted (far control transfers, exceptions and interrupts)Div Cycles Busy countremote_outbound_data_controller_1remote_outbound_data_controller_3DRAM Channel Controller Request Types: Requests with Data (64B): DRAM Channel Controller 0Multiply OpsThis is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Single-precision divide/square root FLOPSevent=0x46,umask=0x08Total Page Table Walks DC Type 0event=0xaf,umask=0x04Floating Point Dispatch Faults. YMM spill faultevent=0x25,umask=0x04ls_sw_pf_dc_fill.ls_mabresp_rmt_cacheevent=0x5a,umask=0x40de_dis_dispatch_token_stalls1.store_queue_token_stallDecode Redirectsbp_l1_tlb_miss_l2_tlb_hitevent=0x44,umask=0x04de_dis_dispatch_token_stalls2.retire_token_stallL1 Data Cache Fills: From Memorymacro_ops_dispatchedls_sw_pf_dc_fills.near_cachel2_pf_hit_l2.l2_streamL2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L2Burst (aggressively fetch additional sequential lines into L2 cache)Instruction cache accesses of all typesevent=0xac,umask=0x02Average sampled latency when data is sourced from DRAM in a different NUMA node. Unit: amd_l3 l3_xi_sampled_latency.ext_nearevent=0xad,umask=0x20L3 cache fill requests sourced from extension memory (CXL) in a different NUMA node. Unit: amd_l3 ex_no_retire.otherex_no_retire.alllocal_processor_write_data_beats_cs3event=0x29f,umask=0x7ffRead data beats (64 bytes) for remote processor at Coherent Station (CS) 7remote_processor_write_data_beats_cs11remote_socket_upstream_read_beats_iom0Read data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 3event=0x81f,umask=0xbfflocal_socket_inf1_outbound_data_beats_ccm2event=0x45e,umask=0xbferemote_socket_inf0_inbound_data_beats_ccm5Data beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 2fp_ops_retired_by_type.vector_macfp_ops_retired_by_type.vector_sqrtsse_avx_ops_retired.mmx_subsse_avx_ops_retired.mmx_packRetired SSE and AVX integer add opssse_avx_ops_retired.sse_avx_cmpfp_pack_ops_retired.fp128_subfp_pack_ops_retired.fp256_subRetired packed floating-point ops of all typesRetired 128-bit packed integer shuffle ops (may include instructions not necessarily thought of as including shuffles e.g. horizontal add, dot product, and certain MOV instructions)event=0xd,umask=0x0fevent=0xd,umask=0x10packed_int_op_type.int256_macpacked_int_op_type.int256_otherNumber of PRECHARGE commands sent for readsumc_pchg_cmd.wrStore-to-load-forward (STLF) hits64B misaligned (cacheline crossing) loadsbackend_bound_memoryPipelineL2;retiring_groupL2 instruction TLB misses and instruction page walksRemote socket outbound data from the CPU (e.g. write data)d_ratio(umc_cas_cmd.all * 1000, umc_mem_clk)umc_cas_cmd_rateRatio of memory controller CAS commands for readsd_ratio(umc_cas_cmd.wr, umc_cas_cmd.all)DC_ACCESSIC_RETURN_STACK_OVERFLOWNB_MEMORY_CONTROLLER_TURNAROUNDPMUEXTIN_EVTEVENT_10HEVENT_15HEVENT_35HEVENT_88HEVENT_9FHEVENT_E1HEVENT_EAHNEON_INSTRS_RENAMEDL1D_CACHE_REFILLINST_RETIREDL1I_CACHEEVENT_100HEVENT_129HEVENT_142HEVENT_145HEVENT_151HEVENT_156HEVENT_178HEVENT_199HEVENT_19BHEVENT_1B4HEVENT_1E3HEVENT_1EAHEVENT_1F1HEVENT_215HEVENT_259HEVENT_266HEVENT_29CHEVENT_2AAHEVENT_2C5HEVENT_2CAHEVENT_2D6HEVENT_2D7HEVENT_31EHEVENT_34EHEVENT_376HEVENT_3B5HL1D_TLB_REFILL_STL2D_CACHE_REFILL_STEXC_TRAP_IRQL2D_TLB_REFILL_WRhni_rrt_wr_occ_cnt_ovflhni_awready_no_awvalidrni_rdb_replaycxha_snppcrd_lnk2_stallcxra_dat_pcrd_stalls_lnk0cxla_avg_latency_process_rx_tlpcxla_avg_latency_form_tx_tlpclkdiv2_hazard_resolutionPMON_SIGBRANCH_LINK_STACK_CORRECTLY_RESOLVEDPREFETCH_ENGINE_REQUESTSTOP_COMPLETIONDISPATCH_SUCCESSSNOOP_PUSHESSTASH_REQUESTSscalar-sse-sse2probe-hit-dirty-no-memory-cancel+INTEL_XSCALEPPC_970AuthenticAMDUNHALTED_CORE_CYCLESuncoreERROR: unrecognized event type: %d
}
GenuineIntel-6-55-[56789ABCDEF]GenuineIntel-6-86This category represents fraction of slots where the processor's Frontend undersupplies its Backend. Frontend denotes the first part of the processor core responsible to fetch operations that are executed later on by the Backend part. Within the Frontend; a branch predictor predicts the next address to fetch; cache-lines are fetched from the memory subsystem; parsed into instructions; and lastly decoded into micro-operations (uops). Ideally the Frontend can issue Machine_Width uops every cycle to the Backend. Frontend Bound denotes unutilized issue-slots when there is no Backend stall; i.e. bubbles where Frontend delivered no uops while Backend could have accepted them. For example; stalls due to instruction-cache misses would be categorized under Frontend BoundThis category represents fraction of slots where the processor's Frontend undersupplies its Backend. SMT version; use when SMT is enabled and measuring per logical CPUUOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))UOPS_RETIRED.RETIRE_SLOTS / BR_INST_RETIRED.NEAR_TAKENTmaL1CORE_CLKSIpBranchInstructions per Floating Point (FP) Operation (lower number means higher occurrence rate)L2 cache true misses per kilo instruction for retired demand loadsTurbo_Utilization(cstate_core@c6\-residency@ / msr@tsc@) * 100C3 residency percent per packageC7 residency percent per packageL2 cache lines in S state filling L2L2 code requestsl2_rqsts.code_rd_hitDemand Data Read requests that hit L2 cacheevent=0x24,period=200003,umask=0x3fl2_trans.demand_data_rdlock_cycles.cache_lock_durationlongest_lat_cache.missRetired load uop whose Data Source was: forwarded from remote cache (Precise Event)  Supports address when precise.  Spec update: BDE70Retired load uops with L2 cache hits as data sources. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM35 (Precise event)move_elimination.simd_not_eliminatedNumber of transitions from AVX-256 to legacy SSE when penalty applicable  Spec update: BDM30icache.hitevent=0x80,period=2000003,umask=0x1Number of Instruction Cache, Streaming Buffer and Victim Cache Misses. Includes Uncacheable accessesCycles MITE is delivering any Uopevent=0xc8,period=2000003,umask=0x4event=0x5,period=2000003,umask=0x2rtm_retired.aborted_misc5RTM region detected inside HLEevent=0x88,period=200003,umask=0x41event=0x88,period=200003,umask=0x88br_misp_exec.all_branchesTaken speculative and retired mispredicted indirect branches excluding calls and returnscpu_clk_unhalted.ref_xclk_anyCounts number of cycles the CPU has at least one pending  demand load request (that is cycles with non-completed load waiting for its data from memory subsystem)This event increments by 1 for every cycle where there was no execute for this threadcycle_activity.stalls_mem_anyevent=0xc3,cmask=1,edge=1,period=100003,umask=0x1event=0xc3,period=100003,umask=0x20event=0x58,period=1000003,umask=0x4This event counts resource-related stall cycles. Reasons for stalls can be as follows:
 - *any* u-arch structure got full (LB, SB, RS, ROB, BOB, LM, Physical Register Reclaim Table (PRRT), or Physical History Table (PHT) slots)
 - *any* u-arch structure got empty (like INT/SIMD FreeLists)
 - FPU control word (FPCW), MXCSR
and others. This counts cycles that the pipeline backend blocked uop delivery from the front endevent=0xa1,period=2000003,umask=0x20uops_executed.coreuops_executed_port.port_5_coreevent=0xc2,cmask=1,inv=1,period=2000003,umask=0x1Cycles with less than 10 actually retired uopsStreaming stores (partial cache line). Derived from unc_c_tor_inserts.opcode. Unit: uncore_cbox Occupancy counter for LLC data reads (demand and L2 prefetch). Derived from unc_c_tor_occupancy.miss_opcode. Unit: uncore_cbox unc_h_requests.writesConflict requests (requests for same address from multiple agents simultaneously). Unit: uncore_ha unc_h_snoop_resp.rspifwdevent=0x43event=0x8,period=2000003,umask=0x40event=0x8,period=2000003,umask=0x20event=0x8,period=100003,umask=0xe64 * ( arb@event\=0x81\,umask\=0x1@ + arb@event\=0x84\,umask\=0x1@ ) / 1000000 / duration_time / 1000Retired load uops which data sources were HitM responses from shared L3  Supports address when precise.  Spec update: BDM100 (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0091offcore_response.all_data_rd.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020120offcore_response.corewb.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020004offcore_response.demand_data_rd.l3_hit.snoop_hitmoffcore_response.other.supplier_none.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0040offcore_response.pf_l3_data_rd.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020080offcore_response.pf_l3_rfo.l3_hit.any_snoopoffcore_response.pf_l3_rfo.supplier_none.any_snoopoffcore_response.all_data_rd.l3_miss.snoop_missoffcore_response.all_data_rd.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000010offcore_response.pf_l2_data_rd.l3_miss_local_dram.snoop_noneoffcore_response.pf_l3_code_rd.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0200offcore_response.pf_l3_code_rd.l3_miss.snoop_hit_no_fwdoffcore_response.pf_l3_data_rd.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0100unc_cbo_xsnp_response.miss_evictionL3 Lookup write request that access cache and found line in MESI-state. Unit: uncore_cbox Total number of Core outgoing entries allocated. Accounts for Coherent and non-coherent trafficRetired load uop whose Data Source was: remote DRAM either Snoop not needed or Snoop Miss (RspI)  Supports address when precise.  Spec update: BDE70 (Precise event)Counts all demand data writes (RFOs) hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.all_data_rd.llc_miss.remote_dramoffcore_response.all_data_rd.llc_miss.remote_hitmoffcore_response.all_rfo.llc_miss.local_dramevent=0x40,period=200000,umask=0x48event=0x27,period=200000,umask=0x50event=0xca,period=2000000,umask=0x1simd_uops_exec.sBus cycles when data is sent on the busbus_lock_clocks.selfBurst (full cache-line) bus transactionsbus_trans_io.all_agentsevent=0x6f,period=200000,umask=0x40External snoopsevent=0x77,period=200000,umask=0x28snoop_stall_drv.selfbr_inst_type_retired.ind_callevent=0x3,period=200000,umask=0x81Load uops retired that hit L1 data cache (Precise event capable)  Supports address when precise (Must be precise)Locked load uops retired (Precise event capable)  Supports address when precise (Must be precise)Counts data reads (demand & prefetch) that miss the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200003010offcore_response.any_request.any_responseCounts the number of writeback transactions caused by L1 or L2 cache evictions that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand cacheable data reads of full cache lines that miss the L2 cacheCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.pf_l2_rfo.l2_hitevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000041000Counts data cache lines requests by software prefetch instructions that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)baclears.condevent=0xc3,period=200003,umask=0x2All machine clearsCounts the number of times that the processor detects that a program is writing to a code section and has to perform a machine clear because of that modification.  Self-modifying code (SMC) causes a severe penalty in all Intel architecture processorsdl1.replacementevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040400Counts demand cacheable data reads of full cache lines have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand cacheable data reads of full cache lines outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts data cache line reads generated by hardware L1 data cache prefetcher have any transaction responses from the uncore subsystemCounts reads for ownership (RFO) requests generated by L2 prefetcher have any transaction responses from the uncore subsystemoffcore_response.streaming_stores.outstandingCounts data cache lines requests by software prefetch instructions outstanding, per cycle, from the time of the L2 miss to when any response is receivedevent=0x8,period=200003,umask=0x2STLB flushesData from local DRAM either Snoop not needed or Snoop Miss (RspI)  Supports address when precise.  Spec update: HSD74, HSD29, HSD25, HSM30 (Precise event)offcore_response.all_data_rd.l3_hit.hit_other_core_no_fwdCounts all demand code reads hit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.pf_l3_data_rd.l3_hit.any_responseCounts all demand code reads miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00100Number of times an RTM execution aborted due to any reasons (multiple categories may count as one) (Precise event)This event counts cycles during which no instructions were executed in the execution stage of the pipeline and there were memory instructions pending (waiting for data)Number of uops delivered by the LSDCycles with no micro-ops executed from any thread on physical core  Spec update: HSD30, HSM31L3 Lookup external snoop request that access cache and found line in M-state. Unit: uncore_cbox L3 Lookup external snoop request that access cache and found line in M-stateA cross-core snoop resulted from L3 Eviction which hits a non-modified line in some processor core. Unit: uncore_cbox event=0xd3,period=100003,umask=0x4event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0600400244event=0xb7,period=100003,umask=0x1,offcore_rsp=0x06004007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x063F8007F7offcore_response.demand_code_rd.llc_miss.any_responseNot rejected writebacks from L1D to L2 cache lines in M stateevent=0x24,period=200003,umask=0x10RFOs that miss cache linesoffcore_response.split_lock_uc_lock.any_responsesimd_fp_256.packed_singleCounts all demand & prefetch data reads that miss the LLC  and the data returned from dramuops_dispatched_port.port_3_coreuops_dispatched_port.port_4_coreCycles per core when uops are dispatched to port 5Counts total number of uops to be executed per-core each cycleunc_cbo_xsnp_response.missFilter on cross-core snoops initiated by this Cbox due to processor core memory requestevent=0x22,umask=0x80Filter on cross-core snoops initiated by this Cbox due to LLC eviction. Unit: uncore_cbox Retired load uops whose data source was remote DRAM (Snoop not needed, Snoop Miss, or Snoop Hit data not forwarded)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc203f7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x600400010PCIe read current. Derived from unc_c_tor_inserts.opcode.pcie_read_current. Unit: uncore_cbox PCIe non-snoop writes (full line). Derived from unc_c_tor_inserts.opcode.pcie_full_write. Unit: uncore_cbox Number of non data (control) flits transmitted . Unit: uncore_qpi unc_p_freq_band0_cyclesCounts the number of cycles that the uncore was running at a frequency greater than or equal to 3Ghz. Derived from unc_p_freq_band2_cycles. Unit: uncore_pcu event=0xd,period=2000003,umask=0x40event=0x1,umask=0xcevent=0x80event=0x4,period=200003,umask=0x40Counts Demand code reads and prefetch code read requests  that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0offcore_response.any_data_rd.l2_hit_this_tile_sevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400070Counts any request that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts Demand cacheable data write requests  that accounts for responses which hit its own tile's L2 with data in S stateoffcore_response.bus_locks.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080400offcore_response.bus_locks.l2_hit_this_tile_foffcore_response.bus_locks.l2_hit_this_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080004offcore_response.demand_code_rd.l2_hit_near_tile_moffcore_response.demand_data_rd.l2_hit_this_tile_moffcore_response.demand_rfo.l2_hit_this_tile_soffcore_response.partial_reads.l2_hit_near_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400100Counts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateoffcore_response.pf_l1_data_rd.l2_hit_far_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000200Counts the number of floating operations retired that required microcode assistsoffcore_response.any_pf_l2.ddr_nearCounts any Prefetch requests that accounts for data responses from DRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400400Counts demand code reads and prefetch code reads that accounts for data responses from MCDRAM LocalCounts demand cacheable data and L1 prefetch data reads that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for data responses from DRAM FarCounts L2 code HW prefetches that accounts for data responses from MCDRAM Far or Other tile L2 hit farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100401000Counts UC code reads (valid only for Outstanding response type)  that accounts for responses from DDR (local and far)offcore_response.uc_code_reads.mcdramCounts the number of branch instructions retired that were conditional jumps (Precise event)event=0xc5,period=200003,umask=0xf9This event counts every cycle when a data (D) page walk or instruction (I) page walk is in progressevent=0x5,edge=1,period=100003,umask=0x1event=0x51,period=2000000,umask=0x4l1d.m_replevent=0x28,period=100000,umask=0xfL2 data demand loads in I state (misses)event=0x26,period=200000,umask=0x2l2_rqsts.ifetch_missevent=0x24,period=200000,umask=0xffevent=0x24,period=200000,umask=0x8L2 prefetch transactionsevent=0x27,period=100000,umask=0xfevent=0x27,period=100000,umask=0x2Memory instructions retired above 32768 clocks (Precise Event)event=0xf,period=10000,umask=0x10Load instructions retired remote DRAM and remote home-remote cache HITM (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x411Offcore requests satisfied by the IO, CSR, MMIO unitOffcore requests satisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x47FFevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1022Offcore code or data read requests satisfied by the IO, CSR, MMIO unitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4733event=0xb7,period=100000,umask=0x1,offcore_rsp=0x3801offcore_response.demand_ifetch.remote_cache_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF02offcore_response.demand_rfo.llc_hit_other_core_hitOffcore other requests satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x430event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4730Offcore prefetch data reads satisfied by any cache or DRAMOffcore prefetch data reads satisfied by the IO, CSR, MMIO unitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF40Offcore prefetch RFO requests satisfied by the IO, CSR, MMIO unitoffcore_response.pf_rfo.remote_cache_dramevent=0xcc,period=2000000,umask=0x3128 bit SIMD integer arithmetic operationsevent=0x12,period=200000,umask=0x8event=0xfd,period=200000,umask=0x10offcore_response.any_rfo.any_llc_missevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2022Offcore RFO requests satisfied by a remote DRAMoffcore_response.corewb.local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2003event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4001event=0xb7,period=100000,umask=0x1,offcore_rsp=0x6030offcore_response.pf_data.any_llc_missOffcore prefetch data requests satisfied by a remote DRAMAll Store buffer stall cyclesThread responded HITM to snoopevent=0x14,period=2000000,umask=0x2event=0xd2,period=2000000,umask=0x1rat_stalls.registersevent=0xa2,period=2000000,umask=0x20event=0xb1,any=1,cmask=1,inv=1,period=2000000,umask=0x3fUops retired (Precise Event)event=0x49,period=200000,umask=0x1event=0xf1,period=100003,umask=0x1fl2_lines_out.silentCounts retired load instructions that split across a cacheline boundary  Supports address when precise (Precise event)Retired load instructions which data sources were L3 and cross-core snoop hits in on-pkg core cache  Supports address when precise (Precise event)Counts retired load instructions with at least one uop that hit in the L3 cache  Supports address when precise (Precise event)offcore_response.demand_code_rd.l3_hit_m.snoop_noneoffcore_response.demand_data_rd.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100002offcore_response.demand_rfo.l4_hit_local_l4.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC01C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100028000Counts once for most SIMD scalar computational double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SIMD scalar double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsfrontend_retired.dsb_missCounts retired instructions that are delivered to the back-end after a front-end stall of at least 32 cycles. During this period the front-end delivered no uops (Precise event)icache_16b.ifdata_stallevent=0x83,period=200003,umask=0x4offcore_response.demand_code_rd.l3_miss.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C408000Cycles when divide unit is busy executing divide or square root operations. Accounts for integer and floating-point operationsevent=0xa6,period=2000003,umask=0x2A version of INST_RETIRED that allows for a more unbiased distribution of samples across instructions retired. It utilizes the Precise Distribution of Instructions Retired (PDIR) feature to mitigate some bias in how retired instructions get sampled  Spec update: SKL091, SKL044 (Must be precise)Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector)Counts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 6Memory_Latency_SMTMemory_Data_TLBs_SMTCounts cycles for each PMH (Page Miss Handler) that is busy with an EPT (Extended Page Table) walk for any request typeThis event counts the number of load ops retired that had UTLB missThis event counts every cycle when a data (D) page walk or instruction (I) page walk is in progress.  Since a pagewalk implies a TLB miss, the approximate cost of a TLB miss can be determined from this eventRetired load uops which data sources were LLC hit and cross-core snoop missed in on-pkg core cache. (Precise Event - PEBS) (Precise event)Counts prefetch code reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.pf_l2_code_rd.llc_hit.no_snoop_neededoffcore_response.pf_l2_rfo.llc_miss.dramoffcore_response.pf_l_data_rd.llc_miss_local.dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5011REQUEST = ANY_REQUEST and RESPONSE = ANY_LOCATIONREQUEST = ANY_REQUEST and RESPONSE = LLC_HIT_NO_OTHER_COREoffcore_response.corewb.all_local_dram_and_remote_cache_hitREQUEST = DATA_IN and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DATA_IN and RESPONSE = IO_CSR_MMIOREQUEST = PF_DATA_RD and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5040REQUEST = PREFETCH and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HIToffcore_response.demand_data.any_dram_and_remote_fwdREQUEST = DEMAND_DATA_RD and RESPONSE = OTHER_LOCAL_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf804event=0xb7,period=100000,umask=0x1,offcore_rsp=0x3080snoopq_requests_outstanding.invalidate_not_emptyevent=0x49,period=2000000,umask=0x4Extended Page Table walk cyclesitlb_misses.large_walk_completedevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2711event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2708event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5870event=0xb7,period=100000,umask=0x1,offcore_rsp=0x6050Counts all demand code reads that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0400OFFCORE_RESPONSE.PF_L2_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.pf_l3_data_rd.l3_hit.snoop_hit_with_fwdoffcore_response.all_pf_rfo.l3_miss.remote_hitmCounts demand data reads that miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800001offcore_response.pf_l2_rfo.l3_miss.remote_hit_forwardCounts all prefetch (that bring data to L2) RFOs that miss the L3 and clean or shared data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000100Core cycles the out-of-order engine was throttled due to a pending power level requestMMIO reads. Unit: uncore_cha event=0x50,umask=0x02PCI Express bandwidth reading at IIO, part 0. Unit: uncore_iio unc_cha_dir_lookup.no_snpevent=0xa5,umask=0x02unc_cha_imc_reads_count.normalunc_cha_rxc_inserts.irqunc_iio_comp_buf_inserts.cmpd.part3PCIe Completion Buffer occupancy of completions with data: Part 3event=0xc0,ch_mask=0x08,fc_mask=0x07,umask=0x04Write request of 4 bytes made to IIO Part0 by the CPU. Unit: uncore_iio unc_iio_txn_req_by_cpu.mem_write.part3unc_iio_txn_req_by_cpu.peer_write.part1RFO request issued by the IRP unit to the mesh with the intention of writing a partial cacheline. Unit: uncore_irp unc_i_transactions.wr_prefCounts when the M2M (Mesh to Memory) looks into the multi-socket cacheline Directory state, and found the cacheline marked in Any State (A, I, S or unused)Partial Non-Isochronous writes to the iMC. Unit: uncore_m2m AD Ingress (from CMS) Occupancy. Unit: uncore_m2m unc_m2m_rxc_bl_occupancyunc_m2m_txc_ad_insertsevent=0x15,umask=0x03Data Response packets that go direct to core. Unit: uncore_upi FLITs received which bypassed the Slot0 Recieve Buffer. Unit: uncore_upi ocr.all_data_rd.l3_hit_e.hit_other_core_fwdocr.all_data_rd.l3_hit_e.snoop_missocr.all_data_rd.l3_hit_f.any_snoopocr.all_data_rd.l3_hit_f.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040491ocr.all_data_rd.l3_hit_s.no_snoop_neededOCR.ALL_PF_DATA_RD.L3_HIT_E.SNOOP_NONEocr.all_pf_data_rd.l3_hit_m.hit_other_core_fwdocr.all_pf_rfo.l3_hit_f.hit_other_core_fwdOCR.ALL_PF_RFO.L3_HIT_F.HIT_OTHER_CORE_FWD  OCR.ALL_PF_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDOCR.ALL_PF_RFO.L3_HIT_M.HIT_OTHER_CORE_FWD  OCR.ALL_PF_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDocr.all_pf_rfo.l3_hit_m.hit_other_core_no_fwdocr.all_pf_rfo.l3_hit_s.any_snoopocr.all_pf_rfo.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100120OCR.ALL_PF_RFO.L3_HIT_S.HIT_OTHER_CORE_FWD  OCR.ALL_PF_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100120ocr.all_reads.l3_hit.hitm_other_coreOCR.ALL_READS.L3_HIT_F.ANY_SNOOP  OCR.ALL_READS.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0004Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_F.ANY_SNOOPCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.demand_code_rd.l3_hit_m.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080001ocr.demand_data_rd.l3_hit_f.hit_other_core_no_fwdCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_HIT.HITM_OTHER_CORE OCR.DEMAND_RFO.L3_HIT.HITM_OTHER_COREocr.demand_rfo.l3_hit_e.hit_other_core_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDocr.other.l3_hit_e.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100400ocr.pf_l1d_and_sw.l3_hit_s.no_snoop_neededCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_E.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200010ocr.pf_l2_data_rd.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100010ocr.pf_l2_rfo.l3_hit.any_snoopocr.pf_l2_rfo.l3_hit.hit_other_core_no_fwdocr.pf_l2_rfo.l3_hit.snoop_noneocr.pf_l2_rfo.l3_hit_f.no_snoop_neededCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDocr.pf_l2_rfo.l3_hit_m.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040020Counts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.PF_L3_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDocr.pf_l3_data_rd.l3_hit_s.no_snoop_neededCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_HIT.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040100offcore_response.all_data_rd.l3_hit_f.hit_other_core_no_fwdoffcore_response.all_data_rd.supplier_none.hit_other_core_no_fwdoffcore_response.all_pf_data_rd.l3_hit_e.hitm_other_coreoffcore_response.all_pf_data_rd.l3_hit_f.hit_other_core_fwdoffcore_response.all_pf_data_rd.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_S.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020490offcore_response.all_pf_rfo.l3_hit_e.snoop_noneoffcore_response.all_pf_rfo.l3_hit_f.snoop_missoffcore_response.all_reads.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_S.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDoffcore_response.all_rfo.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_RFO.SUPPLIER_NONE.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020122This event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_F.ANY_SNOOPoffcore_response.demand_code_rd.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDoffcore_response.demand_data_rd.l3_hit_m.hit_other_core_no_fwdoffcore_response.demand_data_rd.supplier_none.hit_other_core_no_fwdoffcore_response.demand_rfo.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.SNOOP_HIT_WITH_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_E.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.PMM_HIT_LOCAL_PMM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020010This event is deprecated. Refer to new event OCR.PF_L2_RFO.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.SUPPLIER_NONE.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020080This event is deprecated. Refer to new event OCR.PF_L3_RFO.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.PF_L3_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEOCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000491OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDED  OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDOCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDOCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000490OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC0007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4040007F7OCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDOCR.ALL_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDocr.demand_code_rd.l3_miss.remote_hit_forwardevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000004Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000004Counts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.DEMAND_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.NO_SNOOP_NEEDED OCR.DEMAND_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000001ocr.demand_rfo.l3_miss.any_snoopCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPCounts any other requests OCR.OTHER.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000400ocr.pf_l2_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000020Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.HIT_OTHER_CORE_FWD OCR.PF_L2_RFO.L3_MISS.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000020Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.SNOOP_NONEocr.pf_l3_data_rd.l3_miss_local_dram.hitm_other_coreocr.pf_l3_data_rd.l3_miss_local_dram.hit_other_core_no_fwdocr.pf_l3_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000100ocr.pf_l3_rfo.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.ANY_SNOOPoffcore_response.all_pf_data_rd.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDoffcore_response.all_reads.l3_miss_local_dram.no_snoop_neededoffcore_response.all_rfo.l3_miss_local_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDoffcore_response.demand_rfo.l3_miss.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEoffcore_response.other.l3_miss_remote_hop1_dram.any_snoopoffcore_response.other.l3_miss_remote_hop1_dram.hit_other_core_no_fwdoffcore_response.other.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_data_rd.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSOCR.ALL_DATA_RD.SUPPLIER_NONE.SNOOP_MISSocr.all_pf_rfo.supplier_none.no_snoop_neededCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.SUPPLIER_NONE.HITM_OTHER_COREocr.pf_l1d_and_sw.pmm_hit_local_pmm.any_snoopocr.pf_l1d_and_sw.pmm_hit_local_pmm.snoop_not_neededCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDevent=0xe3All commands for Intel Optane DC persistent memory. Unit: uncore_imc unc_m2m_imc_reads.to_pmml1d_pend_miss.l2_stallSW prefetch requests that hit L2 cacheocr.demand_rfo.l3_hit.anyCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit a cacheline in the L3 where a snoop was sentocr.hwpf_l1d_and_swpf.l3_hit.snoop_not_neededCounts hardware prefetch data reads (which bring data to L2)  that hit a cacheline in the L3 where a snoop hit in another core, data forwarding is not requiredocr.other.l3_hit.snoop_sentCounts the number of PREFETCHT1 or PREFETCHT2 instructions executedCounts the number of cycles when the optimal number of uops were delivered by the Instruction Decode Queue (IDQ) to the back-end of the pipeline when there was no back-end stalls. This event counts for one SMT thread in a given cycleevent=0xc8,period=100003,umask=0x2event=0xc9,period=100003,umask=0x2ocr.demand_data_rd.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000002Counts all branch instructions retired (Precise event)Counts cycles the Backend cluster is recovering after a miss-speculation or a Store Buffer or Load Buffer drain stallCycles when RAT does not issue Uops to RS for the threadevent=0xe,period=100003,umask=0x2Counts instruction fetch requests that miss the ITLB (Instruction TLB) and hit the STLB (Second-level TLB)exe_activity.bound_on_loadsUOPS_EXECUTED.THREAD / ( UOPS_EXECUTED.CORE_CYCLES_GE_1 / 2 )MemoryBWCounts core-originated cacheable requests to the L3 cache (Longest Latency cache). Requests include data and code reads, Reads-for-Ownership (RFOs), speculative accesses and hardware prefetches to the L1 and L2.  It does not include hardware prefetches to the L3, and may not count other types of requests to the L3This event is deprecated. Refer to new event MEM_LOAD_L3_HIT_RETIRED.XSNP_FWD  Supports address when precise (Precise event)Counts hardware and software prefetches to all cache levels that hit in the L3 or were snooped from another core's caches on the same socketocr.reads_to_core.l3_hit.snoop_hit_with_fwdocr.streaming_wr.l3_hitCounts streaming stores that hit in the L3 or were snooped from another core's caches on the same socketCycles with outstanding code read requests pending( 1000000000 * ( UNC_CHA_TOR_OCCUPANCY.IA_MISS_DRD_PMM / UNC_CHA_TOR_INSERTS.IA_MISS_DRD_PMM ) / cha_0@event\=0x0@ )Counts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were supplied by DRAMocr.demand_rfo.remote_pmmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10070Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that have any type of responseevent=0x20,umask=0x01CHA to iMC Full Line Writes Issued : Full Line Non-ISOCH. Unit: uncore_cha Lines Victimized : All Lines Victimized. Unit: uncore_cha TOR Inserts : All requests from iA Cores that Hit the LLC. Unit: uncore_cha TOR Inserts : RFOs issued by iA Cores that Missed the LLC. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_crd_prefunc_cha_tor_inserts.io_itomTOR Inserts : DRd_Prefs issued by iA Cores. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_rfo_remoteevent=0xc1,ch_mask=0x20,fc_mask=0x07,umask=0x01event=0x84,ch_mask=0x10,fc_mask=0x07,umask=0x01FAF allocation -- sent to ADQ. Unit: uncore_irp Responses to snoops of any type that hit M line in the IIO cache. Unit: uncore_irp CMS Clockticks. Unit: uncore_m2m event=0x38,umask=0x1C80uncore_uboxevent=0x35,umask=0xCD42FF04Counts the total number of L2 Cache accesses. Counts on a per core basisevent=0x24,period=200003,umask=0x2Counts the number of cycles the core is stalled due to a demand load which hit in the L2 cachemem_bound_stalls.store_buffer_fullevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0044Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedCounts L2 cache hardware prefetch data reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedocr.hwpf_l2_rfo.l3_hitThis event is deprecated. Refer to new event MEM_BOUND_STALLS.LOAD_DRAM_HITCounts the number of core cycles during which interrupts are masked (disabled)ocr.all_code_rd.local_dramCounts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by DRAMCounts streaming stores which modify a full 64 byte cacheline that have any type of responseCounts all hardware and software prefetches that have any type of responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10477ocr.uc_wr.any_responseCounts uncached memory writes that have any type of responseCounts the total number of mispredicted branch instructions retired.  All branch type instructions are accounted for.  Prediction of the branch target address enables the processor to begin executing instructions before the non-speculative execution path is known. The branch prediction unit (BPU) predicts the target address based on the instruction pointer (IP) of the branch and on the execution path through which execution reached this IP.    A branch misprediction occurs when the prediction is wrong, and results in discarding all instructions executed in the speculative path and re-fetching from the correct path (Precise event)topdown_fe_bound.ciscCounts the number of issue slots every cycle that were not delivered by the frontend due to ITLB missesRead Pending Queue Occupancy : Accumulates the occupancies of the Read Pending Queue each cycle.  This can then be used to calculate both the average occupancy (in conjunction with the number of cycles not empty) and the average latency (in conjunction with the number of allocations).  The RPQ is used to schedule reads out to the memory controller and to track the requests.  Requests allocate into the RPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the HA to the iMC. They deallocate after the CAS command has been issued to memoryevent=0x35,umask=0xC001FE01,config1=0x41833CMS ClockticksTOR Inserts : DRd_Opt_Prefs issued by iA Cores that hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Occupancy : CRDs issued by iA Cores : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsTOR Occupancy : All requests from iA Cores that Missed the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsevent=0x8,period=200003,umask=0x20Counts the number Extended Page Directory Entry misses.  The Extended Page Directory cache is used by Virtual Machine operating systems while the guest operating systems use the standard TLB cachesevent=0x4f,period=2000003,umask=0x4Counts the number of page walks completed due to instruction fetches whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 1G pages.  Includes page walks that page faultcpu_coreFraction of branches that are CALL or RET. Unit: cpu_core Fraction of cycles spent in the Operating System (OS) Kernel mode. Unit: cpu_core BR_MISP_RETIRED.ALL_BRANCHES / BACLEARS.ANYCycle cost per L2 hit. Unit: cpu_atom 100 * MEM_BOUND_STALLS.IFETCH_LLC_HIT / ( MEM_BOUND_STALLS.IFETCH )1000 * MEM_UOPS_RETIRED.ALL_LOADS / INST_RETIRED.ANYCounts the number of cycles that uops are blocked due to a load buffer full condition. Unit: cpu_atom mem_uops_retired.load_latency_gt_256l1d_pend_miss.l2_stallsRetired load instructions with L3 cache hits as data sources  Supports address when precise (Precise event). Unit: cpu_core event=0x2a,period=100003,umask=0x1,offcore_rsp=0x8003C0001Retired Instructions who experienced Instruction L1 Cache true miss (Precise event). Unit: cpu_core event=0xc6,period=100007,umask=0x1,frontend=0x600806Cycles MITE is delivering optimal number of Uops. Unit: cpu_core Cycles when uops are being delivered to IDQ while MS is busy. Unit: cpu_core Cycles while L1 cache miss demand load is outstanding. Unit: cpu_core Counts randomly selected loads when the latency from first dispatch to completion is greater than 8 cycles  Supports address when precise (Must be precise). Unit: cpu_core Counts demand data reads that were not supplied by the L3 cache. Unit: cpu_core Counts the number of machine clears that flush the pipeline and restart the machine with the use of microcode due to SMC, MEMORY_ORDERING, FP_ASSISTS, PAGE_FAULT, DISAMBIGUATION, and FPC_VIRTUAL_TRAP. Unit: cpu_atom Counts the number of issue slots every cycle that were not delivered by the frontend due to the microcode sequencer (MS). Unit: cpu_atom Counts the number of issue slots every cycle that were not delivered by the frontend due to wrong predecodes. Unit: cpu_atom Conditional branch instructions retired (Precise event). Unit: cpu_core cpu_clk_unhalted.pauseCounts number of cycles no uops were dispatched to be executed on this thread. Unit: cpu_core unc_m_vc0_requests_wrincoming write request page status is Page Hit. Unit: uncore_imc PRE command sent to DRAM for a read/write request. Unit: uncore_imc event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F803C0001ARITH.FPDIV_ACTIVEFP_ARITH_DISPATCHED.PORT_0FP_ARITH_INST_RETIRED2.128B_PACKED_HALFevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x703000001INT_VEC_RETIRED.128BITevent=0x10,umask=0x0000000002unc_m_cas_count.wr_nonpreevent=0x50,umask=0x0000000002event=0x2,umask=0x0000000001unc_upi_rxl_basic_hdr_match.ncs_opcevent=0x83,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000004event=0x84,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000004UNC_IIO_COMP_BUF_OCCUPANCY.CMPD.ALL_PARTS. Unit: uncore_iio unc_m2m_prefcam_inserts.xpt_allchevent=0x58,umask=0x0000000005: UPI - All Channels. Unit: uncore_m2m event=0x53,umask=0x0000000002event=0x54,umask=0x0000000001event=0x35,umask=0x00c001fd01TOR Inserts; All from local IO. Unit: uncore_cha TOR Inserts : IPQ. Unit: uncore_cha event=0x35,umask=0x00c000ff01event=0x35,umask=0x00C000FF05TOR Inserts : Just NonCoherent. Unit: uncore_cha TOR Inserts; CRd Pref misses from local IA. Unit: uncore_cha event=0x35,umask=0x00cc47ff01TOR Inserts;SpecItoM from Local IA. Unit: uncore_cha unc_cha_tor_occupancy.ia_hit_llcprefrfoevent=0x36TOR Occupancy : MMIO Access. Unit: uncore_cha unc_cha_tor_occupancy.isocTOR Occupancy; DRd Opt Pref hits from local IA. Unit: uncore_cha TOR Occupancy; DRd Pref from local IA. Unit: uncore_cha unc_cha_tor_inserts.ia_hit_drdpteTOR Inserts; RdCur and FsRdCur hits from local IO. Unit: uncore_cha event=0x35,umask=0x00cccffe01unc_cha_tor_occupancy.ia_miss_llcprefcodeunc_cha_tor_inserts.ia_miss_itomevent=0x35,umask=0x00CC47FE01TOR Inserts : WCiLs issued by iA Cores targeting DDR that missed the LLC - HOMed remotely. Unit: uncore_cha event=0x36,umask=0x00c8c7ff01event=0x36,umask=0x00c86f8601event=0x8bbp_snp_re_syncl2_wcb_req.zero_byte_storexi_sys_fill_latencyevent=0xc5Retired Far Control Transfersevent=0x1cf,umask=0x04Tagged IBS Ops. Number of Ops tagged by IBSThe number of fused retired branch instructions retired per cycle. The number of events logged per cycle can vary from 0 to 3fp_ret_sse_avx_ops.dp_div_flopsLS MAB allocates by type - storesL1 DTLB Reload of a page of 2M sizeevent=0x47Cycles where a dispatch group is valid but does not get dispatched due to a token stall. AGSQ Tokens unavailableall_dc_accessesall_l2_cache_accessesApproximate: Combined DRAM B/bytes of all channels on a NPS1 node (die) (may need --metric-no-group)bp_l1_tlb_fetch_hit.if4kevent=0x85,umask=0xffbp_l1_tlb_miss_l2_tlb_miss.if2mevent=0x85,umask=0x02Retired Conditional Branch Instructions MispredictedNumber of SSE Move Ops eliminated. This is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemesNumber of SSE Move Ops. This is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemesNumber of accesses to the dcache for load/store referencesDemand Data Cache Fills by Data Source. Local L2 hitSoftware Prefetch Instructions Dispatched (Speculative)Software Prefetch Data Cache Fills by Data Source. From another cache (home node remote)event=0x59,umask=0x08event=0xaa,umask=0x01de_dis_dispatch_token_stalls1.fp_sch_rsrc_stallAll Instruction Cache Accesses. Counts various IC tag related hit and miss eventsex_ret_ind_brch_instrevent=0x41,umask=0x7fAny Data Cache Fills by Data Source. From cache of different CCX in same nodeThe number of 4KB misaligned (i.e., page crossing) loadsls_tlb_flush.all_tlb_flushesCycles where a dispatch group is valid but does not get dispatched due to a Token Stall. Also counts cycles when the thread is not selected to dispatch but would have been stalled due to a Token Stall. Load Queue resource stall. Applies to all ops with load semanticsRetired near returns mispredicted. Each misprediction incurs the same penalty as a mispredicted conditional branch instructionRetired indirect branch instructionsSoftware prefetch data cache fills from either DRAM or MMIO in a different NUMA node (same or different socket)L2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L2Burst (aggressively fetch additional sequential lines into L2 cache)L2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L1Stride (fetch additional lines into L1 cache when each access is a constant distance from the previous)l2_pf_miss_l2_l3.l2_strideevent=0xad,umask=0x3fex_no_retire.emptyevent=0x120,umask=0x1ex_ret_ucode_instrevent=0x29f,umask=0x7feevent=0x29f,umask=0xbffevent=0x55e,umask=0x7feevent=0x45e,umask=0x7ffData beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 4event=0x45e,umask=0xbffevent=0x55e,umask=0xbffremote_socket_inf1_outbound_data_beats_ccm1event=0xb5f,umask=0xf3eData beats (64 bytes) for local socket outbound data from inter-socket xGMI link 4Data beats (64 bytes) for local socket outbound data from inter-socket xGMI link 6event=0x8,umask=0x08Retired packed 128-bit floating-point opsfp_ops_retired_by_width.pack_512_uops_retiredfp_ops_retired_by_type.vector_subevent=0xb,umask=0xa0event=0xb,umask=0xb0Retired 128-bit packed integer MOV opspacked_int_op_type.int256_shufflepacked_int_op_type.int256_packFloating-point dispatch faults for x87 fillsfp_disp_faults.allNumber of CAS commands sent for writesumc_data_slot_clks.rdFraction of dispatched ops that were flushed due to pipeline restarts (resyncs)Fraction of dispatch slots that remained unused because of stalls due to the memory subsysteml2_cache_accesses_from_l1_dc_missesL2 cache misses from L2 cache hardware prefetcherl2_cache_hits_from_l1_dc_missbp_l1_tlb_miss_l2_tlb_hit + bp_l1_tlb_miss_l2_tlb_miss.allMacro-ops dispatchedlocal_processor_write_data_beats_cs0 + local_processor_write_data_beats_cs1 + local_processor_write_data_beats_cs2 + local_processor_write_data_beats_cs3 + local_processor_write_data_beats_cs4 + local_processor_write_data_beats_cs5 + local_processor_write_data_beats_cs6 + local_processor_write_data_beats_cs7 + local_processor_write_data_beats_cs8 + local_processor_write_data_beats_cs9 + local_processor_write_data_beats_cs10 + local_processor_write_data_beats_cs11remote_processor_read_data_beats_cs0 + remote_processor_read_data_beats_cs1 + remote_processor_read_data_beats_cs2 + remote_processor_read_data_beats_cs3 + remote_processor_read_data_beats_cs4 + remote_processor_read_data_beats_cs5 + remote_processor_read_data_beats_cs6 + remote_processor_read_data_beats_cs7 + remote_processor_read_data_beats_cs8 + remote_processor_read_data_beats_cs9 + remote_processor_read_data_beats_cs10 + remote_processor_read_data_beats_cs11ccn read-cycles event. Unit: uncore_sys_ccn_pmu hwpmcDC_L1_DTLB_MISS_AND_L2_DTLB_MISSDC_ONE_BIT_ECC_ERRORIC_MISSFR_RETIRED_TAKEN_BRANCHES_MISPREDICTED_BY_ADDR_MISCOMPAREFR_TAKEN_HARDWARE_INTERRUPTSNB_HT_BUS0_BANDWIDTHNB_HT_BUS2_BANDWIDTHL2_STORE_BUFFERABLEL1_CACHE_PAGECOL_ALIASL2_CACHE_NEON_MEM_ACCESSEVENT_33HEVENT_51HEVENT_54HEVENT_86HEVENT_89HEVENT_9AHEVENT_ACHEVENT_BAHEVENT_EDHSW_INCREVENT_10AHEVENT_11AHEVENT_1A0HEVENT_1B2HEVENT_1D0HEVENT_1F4HEVENT_23EHEVENT_251HEVENT_271HEVENT_27FHEVENT_28FHEVENT_293HEVENT_2BDHEVENT_2F6HEVENT_30EHEVENT_31BHEVENT_34AHEVENT_364HEVENT_366HEVENT_36BHEVENT_37BHEVENT_398HEVENT_3A4HEVENT_3A7HEVENT_3E8HMEM_ACCESS_LDBUS_ACCESS_NORMALBR_RETIREDL2D_TLB_RDhnf_brd_snoops_senthnf_seq_hitxp_partial_dat_flitsbsx_rd_req_trkr_occ_cnt_ovflrnd_wrcancel_sentrni_s2_wdata_beatsclkdiv2_allocatePMON_EXCEPTVFPU_INSTR_COMPLETEDVPU_INSTR_WAIT_CYCLESDSS_INSTR_COMPLETEDDST_STREAM_0_CACHE_LINE_FETCHESLSU_STORE_QUEUE_INDEX_ALIASL2_CACHE_HITSFPU_COMPLETION_STALLL2_CACHE_DIRTY_UPDATESDVT3_DETECTEDSTWCX_FAILURESesloadmissesPPC_7450ARMV8_CORTEX_A53{"type": "initialize"%s, "version": "0x%08x", "arch": "0x%08x", "cpuid": "%s", "tsc_freq": "%jd", "sec": "%jd", "nsec": "%jd"}
%s, "oldpid": "%d", "newpid": "%d"}
IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)This category represents fraction of slots wasted due to incorrect speculations. SMT version; use when SMT is enabled and measuring per logical CPUCPU_CLK_UNHALTED.THREADBranches;Fed;PGOINST_RETIRED.ANY / ( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE) )Total number of retired Instructions, Sample with: INST_RETIRED.PREC_DIST1000 * MEM_LOAD_UOPS_RETIRED.L1_MISS / INST_RETIRED.ANYMem;CacheMisses;OffcoreCPU_CLK_UNHALTED.REF_TSC / msr@tsc@(CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_timeC2 residency percent per packageThis event counts the total number of requests from the L2 hardware prefetchersevent=0xd0,period=2000003,umask=0x81event=0xca,cmask=1,period=100003,umask=0x1eevent=0xc1,period=100003,umask=0x10event=0x79,cmask=4,period=2000003,umask=0x18event=0x79,cmask=1,period=2000003,umask=0x24This event counts the number of uops delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Counting includes uops that may bypass the IDQInstruction Decode Queue (IDQ) empty cycleshle_retired.startLoads with latency value being above 128  Spec update: BDM100, BDM35 (Must be precise)This event counts loads with latency value being above 16  Spec update: BDM100, BDM35 (Must be precise)Loads with latency value being above 32  Spec update: BDM100, BDM35 (Must be precise)This event counts loads with latency value being above four  Spec update: BDM100, BDM35 (Must be precise)rtm_retired.aborted_misc1event=0xc9,period=2000003,umask=0x20This event counts unhalted core cycles during which the thread is in rings 1, 2, or 3This is a precise version (that is, uses PEBS) of the event that counts both direct and indirect near call instructions retired (Precise event)Taken branch instructions retired. (Precise Event - PEBS) (Precise event)Not taken branch instructions retiredbr_misp_exec.taken_return_nearevent=0xc5,period=400009,umask=0x4cpu_clk_thread_unhalted.one_thread_activecpu_clk_unhalted.ref_xclkevent=0xa3,cmask=5,period=2000003,umask=0x5event=0xa3,cmask=6,period=2000003,umask=0x6Counts number of cycles nothing is executed on any execution port, while there was at least one pending demand load requestld_blocks.no_srThis event counts stall cycles caused by the store buffer (SB) overflow (excluding draining from synch). This counts cycles that the pipeline backend blocked uop delivery from the front endevent=0xa1,period=2000003,umask=0x2event=0xa1,period=2000003,umask=0x10event=0xa1,any=1,period=2000003,umask=0x20uops_executed_port.port_6uops_issued.anyUops that Resource Allocation Table (RAT) issues to Reservation Station (RS)Number of slow LEA uops being allocated. A uop is generally considered SlowLea if it has 3 sources (e.g. 2 sources + immediate) regardless if as a result of LEA instruction or notLLC prefetch misses for data reads. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox unc_h_snoop_resp.rsp_fwd_wbevent=0x85unc_p_power_state_occupancy.cores_c6Demand load Miss in all translation lookaside buffer (TLB) levels causes a page walk that completes (2M/4M)  Spec update: BDM69event=0x49,period=100003,umask=0x10This event counts cycles for an extended page table walk. The Extended Page directory cache differs from standard TLB caches by the operating system that use it. Virtual machine operating systems use the extended page directory cache, while guest operating systems use the standard TLB cachesevent=0x85,period=100003,umask=0xeNumber of DTLB page walker hits in the L1+FB  Spec update: BDM69, BDM98Number of ITLB page walker hits in the L1+FB  Spec update: BDM69, BDM98offcore_response.all_pf_code_rd.l3_hit.any_snoopoffcore_response.all_pf_data_rd.l3_hit.snoop_hitmoffcore_response.all_pf_rfo.l3_hit.snoop_hitmoffcore_response.all_rfo.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0008offcore_response.corewb.supplier_none.snoop_hit_no_fwdoffcore_response.corewb.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0002Number of SSE/AVX computational 256-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 4 calculations per elementNumber of SSE/AVX computational scalar single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT RCP FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementNumber of times HLE abort was triggered (Precise event)Counts randomly selected loads with latency value being above four  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)offcore_response.all_data_rd.l3_miss_local_dram.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000090offcore_response.all_pf_data_rd.l3_miss.snoop_noneoffcore_response.all_pf_rfo.l3_miss_local_dram.snoop_missoffcore_response.all_pf_rfo.l3_miss_local_dram.snoop_non_dramoffcore_response.all_rfo.l3_miss.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000008offcore_response.corewb.l3_miss_local_dram.snoop_non_dramoffcore_response.demand_code_rd.l3_miss.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000004offcore_response.demand_code_rd.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000004offcore_response.demand_data_rd.l3_hit.snoop_non_dramoffcore_response.demand_rfo.l3_miss.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000028000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000080Number of near branch instructions retired that were mispredicted and taken (Precise event)event=0x34,umask=0x1fevent=0x34,umask=0x16Number of entries allocated. Account for Any type: e.g. Snoop, Core aperture, etc. Unit: uncore_arb Counts all demand & prefetch data reads miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x063BC00091Counts all demand & prefetch data reads miss the L3 and clean or shared data is transferred from remote cachel1d_cache.replevent=0x28,period=200000,umask=0x4fl2_lines_in.self.anyl2_lock.self.i_statel2_rqsts.self.any.i_stateevent=0x2e,period=200000,umask=0x4fmisalign_mem_ref.bubbleevent=0x6d,period=200000,umask=0xe0ext_snoop.this_agent.anyevent=0x88,period=2000000,umask=0x10br_inst_type_retired.indevent=0x88,period=2000000,umask=0x4Micro-op reissues for any cause (At Retirement)DTLB misses due to load operationsRetired loads that miss the DTLB (precise event) (Precise event)Stores uops retired that split a cache-line (Precise event capable)  Supports address when precise (Must be precise)Counts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x04000032b7Counts requests to the uncore subsystem that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000022Counts the number of writeback transactions caused by L1 or L2 cache evictions that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of writeback transactions caused by L1 or L2 cache evictions that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.demand_code_rd.l2_miss.anyCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.demand_rfo.l2_hitevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000002offcore_response.partial_streaming_stores.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200004000offcore_response.pf_l1_data_rd.l2_miss.anyoffcore_response.pf_l1_data_rd.l2_miss.snoop_miss_or_no_snoop_neededoffcore_response.pf_l2_data_rd.l2_miss.hit_other_core_no_fwdevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000020Counts any data writes to uncacheable write combining (USWC) memory region  that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of times a BACLEAR is signaled for any reason, including, but not limited to indirect branch/call,  Jcc (Jump on Conditional Code/Jump if Condition is Met) branch, unconditional branch/call, and returnsDecode restrictions due to predicting wrong instruction lengthfetch_stall.itlb_fill_pending_cyclesRetired branch instructions (Precise event capable) (Must be precise)Retired near indirect call instructions (Precise event capable) (Must be precise)Counts retired Jcc (Jump on Conditional Code/Jump if Condition is Met) branch instructions retired, including both when the branch was taken and when it was not taken (Must be precise)Counts mispredicted near indirect CALL branch instructions retired, where the target address taken was not what the processor predicted (Must be precise)issue_slots_not_consumed.anyCounts a load blocked from using a store forward because of an address/size mismatch, only one of the loads blocked from each store will be counted (Must be precise)event=0xc2,period=2000003Counts the number of integer divide uops retired (Must be precise)Counts store uops retired that caused a DTLB miss  Supports address when precise (Must be precise)Counts bus lock and split lock requests hit the L2 cacheCounts the number of writeback transactions caused by L1 or L2 cache evictions outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand cacheable data reads of full cache lines miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredoffcore_response.full_streaming_stores.any_responseCounts reads for ownership (RFO) requests generated by L2 prefetcher hit the L2 cacheCounts any data writes to uncacheable write combining (USWC) memory region  hit the L2 cacheCounts any data writes to uncacheable write combining (USWC) memory region  hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts machine clears due to memory ordering issues.  This occurs when a snoop request happens and the machine is uncertain if memory ordering will be preserved - as another core is in the process of modifying the datadtlb_store_misses.walk_pendingRetired load uops which data sources were load uops missed L1 but hit FB due to preceding miss to the same cache line with data not ready  Supports address when precise.  Spec update: HSM30 (Precise event)Retired load uops with L2 cache hits as data sources  Supports address when precise.  Spec update: HSD76, HSD29, HSM30 (Precise event)Offcore outstanding RFO store transactions in SQ to uncore. Set Cmask=1 to count cycles  Spec update: HSD62, HSD61, HSM63Counts prefetch (that bring data to L2) data reads hit in the L3Number of X87 FP assists due to output valuesNumber of SIMD move elimination candidate uops that were eliminatedCounts cycles FE delivered 4 uops or Resource Allocation Table (RAT) was stalling FE  Spec update: HSD135( ITLB_MISSES.WALK_DURATION + DTLB_LOAD_MISSES.WALK_DURATION + DTLB_STORE_MISSES.WALK_DURATION ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )offcore_response.pf_l2_code_rd.l3_miss.any_responseevent=0xc4,period=100003,umask=0x40Misses in all TLB levels that cause a page walk of any page sizeStore misses in all DTLB levels that cause page walksStore miss in all TLB levels causes a page walk that completes. (1G)event=0xbc,period=2000003,umask=0x44offcore_response.pf_l2_code_rd.llc_miss.any_responseevent=0x28,period=200003,umask=0x8l2_lines_out.pf_dirtymem_load_uops_llc_hit_retired.xsnp_hitmRetired load uops which data sources were hits in LLC without snoops required (Precise event)Number of SSE* or AVX-128 FP Computational scalar single-precision uops issued this cycleNumber of assists associated with 256-bit AVX store operationsInstruction cache, streaming buffer and victim cache missesNumber of any page walk that had a miss in LLCReference cycles when the at least one thread on the physical core is unhalted. (counts at 100 MHz rate)Cycles per core when uops are dispatched to port 4Number of flags-merge uops allocated. Such uops adds delayunc_arb_trk_requests.evictionsLLC lookup request that access cache and found line in M-state. Unit: uncore_cbox itlb_misses.large_page_walk_completedoffcore_response.pf_l2_data_rd.llc_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x87f820004offcore_response.pf_l2_data_rd.llc_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x107fc00010(UNC_P_FREQ_GE_3000MHZ_CYCLES / UNC_P_CLOCKTICKS) * 100.l1d_blocks.bank_conflict_cyclesevent=0xbf,cmask=1,period=100003,umask=0x5dsb_fill.all_cancelValid instructions written to IQ per cycleevent=0xc5,period=100007,umask=0x2event=0x59,period=2000003,umask=0x40Cycles with at least one slow LEA uop being allocatedevent=0x4,period=100007,umask=0x4Counts Demand code reads and prefetch code read requests  that accounts for any responseoffcore_response.any_code_rd.l2_hit_this_tile_mCounts any Read request  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts any request that accounts for any responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400022offcore_response.demand_data_rd.l2_hit_far_tile_e_foffcore_response.demand_data_rd.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000002Counts Demand cacheable data writes that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080100offcore_response.partial_writes.l2_hit_this_tile_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002002000Counts L2 code HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateCounts UC code reads (valid only for Outstanding response type)  that accounts for responses which hit its own tile's L2 with data in M stateevent=0xc2,period=200003,umask=0x20Counts all instruction fetches that hit the instruction cacheoffcore_response.any_data_rd.ddr_farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101003091offcore_response.any_data_rd.mcdram_farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200070event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600022Counts demand cacheable data and L1 prefetch data reads that accounts for data responses from DRAM LocalCounts L1 data HW prefetches that accounts for data responses from DRAM LocalCounts UC code reads (valid only for Outstanding response type)  that accounts for responses from MCDRAM (local and far)Counts the store micro-ops retired that were pushed in the rehad queue because the store address buffer is fullrecycleq.st_splitsCounts the number of core cycles when allocation pipeline is stalled and is waiting for a free MEC reservation station entryL1 data cache lines allocatedl1d_cache_st.s_stateL1D hardware prefetch missesL1D hardware prefetch requestsL2 data prefetches in M stateevent=0xf1,period=100000,umask=0x4All L2 demand store RFOs that hit the cacheevent=0x2e,period=100000,umask=0x41event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F44Offcore code reads satisfied by a remote cache or remote DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFFFFOffcore requests satisfied by the LLC and HIT in a sibling coreoffcore_response.any_request.remote_cache_hitOffcore requests that HIT in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x708offcore_response.corewb.remote_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3833Offcore other requests satisfied by a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3830offcore_response.pf_data.remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1810offcore_response.pf_ifetch.llc_hit_no_other_coreOffcore prefetch code reads that HIT in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1870fp_assist.inputOffcore code reads satisfied by a remote DRAMoffcore_response.demand_data.any_llc_missevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6002offcore_response.pf_rfo.remote_dramevent=0x13,period=2000000,umask=0x2event=0xd4,period=2000000,umask=0x1arith.cycles_div_busyBACLEAR asserted with bad target addressbaclear_force_iqevent=0xe8,period=2000000,umask=0x1br_inst_exec.return_nearevent=0x89,period=20000,umask=0x7fcpu_clk_unhalted.ref_pExecution pipeline restart due to Memory ordering conflictsevent=0xc7,period=200000,umask=0x1uops_executed.port0event=0x8,period=200000,umask=0x20Retired instructions with at least 1 uncacheable load or lock  Supports address when precise (Precise event)Cycles with offcore outstanding demand rfo reads transactions in SuperQueue (SQ), queue to uncoreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0080004offcore_response.demand_rfo.l4_hit_local_l4.any_snoopoffcore_response.demand_rfo.l4_hit_local_l4.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10001C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80048000offcore_response.other.l3_hit_m.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x40048000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000408000offcore_response.other.l4_hit_local_l4.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x40408000Number of cache line split locks sent to uncoreRetired Instructions who experienced STLB (2nd level TLB) true miss (Precise event)Counts the number of cycles uops were delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Count includes uops that may 'bypass' the IDQCounts, on the per-thread basis, cycles when less than 1 uop is delivered to Resource Allocation Table (RAT). IDQ_Uops_Not_Delivered.core >= 3Counts randomly selected loads when the latency from first dispatch to completion is greater than 32 cycles  Supports address when precise (Must be precise)Cycles with at least 6 Demand Data Read requests that miss L3 cache in the superQevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFC400004offcore_response.demand_rfo.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x44000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C408000tx_mem.abort_capacityevent=0xa6,period=2000003,umask=0x4Number of retired PAUSE instructions (that do not end up with a VMExit to the VMM; TSX aborted Instructions may be counted). This event is not supported on first SKL and KBL products1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))100 * ( (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * INT_MISC.CLEAR_RESTEER_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) ) - (100 * (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * ( (ICACHE_64B.IFTAG_STALL / CPU_CLK_UNHALTED.THREAD) + (( ICACHE_16B.IFDATA_STALL + 2 * cpu@ICACHE_16B.IFDATA_STALL\,cmask\=1\,edge@ ) / CPU_CLK_UNHALTED.THREAD) + (9 * BACLEARS.ANY / CPU_CLK_UNHALTED.THREAD) ) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)))Instruction_Fetch_BW_SMT ( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * INT_MISC.CLEAR_RESTEER_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) ) * (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )) / BR_MISP_RETIRED.ALL_BRANCHESStore misses in all TLB levels causes a page walk that completes. (All page sizes)Counts the number of any STLB flush attempts (such as entire, VPID, PCID, InvPage, CR3 write, etc.)Counts the number of request that were not accepted into the L2Q because the L2Q is FULLCounts the number of request from the L2 that were not accepted into the XQCounts any rfo reads (demand & prefetch) that have any response typeCounts demand and DCU prefetch RFOs that miss L2Counts demand and DCU prefetch RFOs that are are outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts RFO requests generated by L2 prefetchers that miss L2event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680004800Loads missed DTLB (Precise event)REQUEST = DATA_INTO_CORE and RESPONSE = ANY_RESPONSECounts demand code reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts demand data writes (RFOs) that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0100offcore_response.pf_l_ifetch.any_responseoffcore_response.pf_llc_data_rd.llc_miss.dramoffcore_requests_outstanding.any.read_not_emptyoffcore_response.any_data.local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7ffREQUEST = ANY RFO and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = CORE_WB and RESPONSE = LLC_HIT_NO_OTHER_COREoffcore_response.corewb.llc_hit_other_core_hitREQUEST = DATA_IFETCH and RESPONSE = LLC_HIT_NO_OTHER_COREREQUEST = DATA_IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = DATA_IN and RESPONSE = ANY_CACHE_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff04REQUEST = DEMAND_IFETCH and RESPONSE = IO_CSR_MMIOREQUEST = DEMAND_RFO and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f80event=0xb7,period=100000,umask=0x1,offcore_rsp=0x150REQUEST = PF_DATA_RD and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = PF_RFO and RESPONSE = ANY_CACHE_DRAMREQUEST = PREFETCH and RESPONSE = REMOTE_CACHE_HITMmisalign_mem_ref.storeREQUEST = DEMAND_RFO and RESPONSE = ANY_LLC_MISSREQUEST = OTHER and RESPONSE = ANY_LLC_MISSDTLB load miss page walk cyclesdtlb_misses.large_walk_completedITLB miss page walk cyclesevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5822event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5803event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0491Counts prefetch (that bring data to L2) data reads that hit in the L3offcore_response.pf_l3_rfo.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00491Counts all prefetch data reads that miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800490offcore_response.demand_data_rd.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000002offcore_response.pf_l1d_and_sw.l3_miss.remote_hit_forwardevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00400Counts CAS (Column Access Select) regular read commands issued to DRAM on a per channel basis.  CAS commands are issued to specify the address to read or write on DRAM, and this event increments for every regular read.  This event only counts regular reads and does not includes underfill reads due to partial write requests.  This event counts whether AutoPrecharge (which closes the DRAM Page automatically after a read/write)  is enabled or notunc_m_cas_count.wr_wmmCounts transactions that looked into the multi-socket cacheline Directory state, and therefore did not send a snoop because the Directory indicated it was not neededMulti-socket cacheline Directory state lookups; Snoop Needed. Unit: uncore_cha FaST wire asserted; Horizontal. Unit: uncore_cha Lines Victimized; Lines in M state. Unit: uncore_cha Number of times that an RFO hit in S state. Unit: uncore_cha Snoop filter capacity evictions for S-state entries. Unit: uncore_cha unc_cha_snoop_resp.rspievent=0xc2,ch_mask=0x02,fc_mask=0x4,umask=0x03unc_iio_comp_buf_occupancy.cmpd.part0PCIe Completion Buffer occupancy of completions with data: Part 2unc_iio_data_req_by_cpu.peer_write.part3Peer to peer read request for 4 bytes made by IIO Part0 to an IIO target. Unit: uncore_iio Counts every peer to peer read request for 4 bytes of data made by IIO Part1 to the MMIO space of an IIO target. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busunc_iio_txn_req_by_cpu.mem_write.part0Peer to peer read request for up to a 64 byte transaction is made by a different IIO unit to IIO Part3. Unit: uncore_iio Counts every read request for up to a 64 byte transaction of data made by IIO Part0 to a unit on the main die (generally memory). In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busunc_iio_txn_req_of_cpu.mem_write.part2Peer to peer read request of up to a 64 byte transaction is made by IIO Part3 to an IIO target. Unit: uncore_iio event=0x19Occupancy of the IRP FAF queue. Unit: uncore_irp BL Ingress (from CMS) Occupancyevent=0x31,umask=0x1event=0x41ocr.all_data_rd.l3_hit.any_snoopOCR.ALL_DATA_RD.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_DATA_RD.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDOCR.ALL_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDocr.all_data_rd.l3_hit_f.hit_other_core_no_fwdOCR.ALL_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWD  OCR.ALL_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.all_data_rd.l3_hit_f.snoop_noneocr.all_data_rd.l3_hit_m.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100491OCR.ALL_PF_DATA_RD.L3_HIT.ANY_SNOOP OCR.ALL_PF_DATA_RD.L3_HIT.ANY_SNOOP OCR.ALL_PF_DATA_RD.L3_HIT.ANY_SNOOPocr.all_pf_data_rd.l3_hit_e.no_snoop_neededocr.all_pf_data_rd.l3_hit_e.snoop_noneOCR.ALL_PF_DATA_RD.L3_HIT_S.SNOOP_MISSocr.all_pf_rfo.l3_hit_m.snoop_missocr.all_reads.l3_hit_m.snoop_missocr.all_rfo.l3_hit.hit_other_core_fwdocr.all_rfo.l3_hit.snoop_hit_with_fwdOCR.ALL_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWD  OCR.ALL_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDOCR.ALL_RFO.L3_HIT_F.NO_SNOOP_NEEDED  OCR.ALL_RFO.L3_HIT_F.NO_SNOOP_NEEDEDocr.demand_code_rd.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100004Counts demand data reads OCR.DEMAND_DATA_RD.L3_HIT.SNOOP_NONECounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_M.HITM_OTHER_COREocr.demand_data_rd.l3_hit_m.no_snoop_neededocr.demand_rfo.l3_hit.no_snoop_neededocr.demand_rfo.l3_hit_e.any_snoopCounts any other requests  OCR.OTHER.L3_HIT_E.HIT_OTHER_CORE_FWDCounts any other requests  OCR.OTHER.L3_HIT_S.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0400ocr.pf_l1d_and_sw.l3_hit_f.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040400Counts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_M.ANY_SNOOPocr.pf_l2_data_rd.l3_hit.no_snoop_neededCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_HIT.SNOOP_MISSCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_S.HITM_OTHER_COREocr.pf_l2_rfo.l3_hit_s.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080080ocr.pf_l3_data_rd.l3_hit_f.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040080Counts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200100ocr.pf_l3_rfo.l3_hit_s.snoop_missoffcore_response.all_data_rd.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020491This event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_MISSoffcore_response.all_pf_data_rd.l3_hit_e.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_F.SNOOP_MISSoffcore_response.all_pf_data_rd.l3_hit_s.snoop_missoffcore_response.all_pf_data_rd.pmm_hit_local_pmm.snoop_not_neededoffcore_response.all_pf_rfo.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000207F7offcore_response.all_rfo.l3_hit_e.hit_other_core_fwdoffcore_response.all_rfo.l3_hit_f.snoop_missThis event is deprecated. Refer to new event OCR.ALL_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_RFO.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.SNOOP_NONEoffcore_response.demand_code_rd.l3_hit_f.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.SUPPLIER_NONE.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.SUPPLIER_NONE.SNOOP_NONEoffcore_response.demand_data_rd.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_F.NO_SNOOP_NEEDEDoffcore_response.pf_l1d_and_sw.l3_hit_f.snoop_missoffcore_response.pf_l1d_and_sw.l3_hit_m.any_snoopThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_S.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020400This event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.SUPPLIER_NONE.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_hit_m.hit_other_core_fwdoffcore_response.pf_l2_rfo.l3_hit_m.snoop_missoffcore_response.pf_l3_data_rd.l3_hit_e.no_snoop_neededoffcore_response.pf_l3_data_rd.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020100Intel AVX-512 computational 512-bit packed BFloat16 instructions retiredOCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORE  OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREocr.all_pf_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdOCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOP  OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPOCR.ALL_READS.L3_MISS_LOCAL_DRAM.SNOOP_NONEocr.all_reads.l3_miss_remote_hop1_dram.hit_other_core_fwdOCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000122ocr.all_rfo.l3_miss_remote_hop1_dram.no_snoop_neededCounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.NO_SNOOP_NEEDED OCR.DEMAND_CODE_RD.L3_MISS.NO_SNOOP_NEEDEDCounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDocr.demand_rfo.l3_miss_local_dram.hit_other_core_no_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000002ocr.other.l3_miss.no_snoop_neededocr.other.l3_miss.remote_hit_forwardocr.other.l3_miss_local_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000400ocr.pf_l1d_and_sw.l3_miss_remote_hop1_dram.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.ANY_SNOOP OCR.PF_L2_DATA_RD.L3_MISS.ANY_SNOOPocr.pf_l2_data_rd.l3_miss_local_dram.hit_other_core_no_fwdocr.pf_l2_rfo.l3_miss_local_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000020Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREocr.pf_l3_data_rd.l3_miss_remote_hop1_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000100offcore_response.all_pf_data_rd.l3_miss_local_dram.no_snoop_neededoffcore_response.all_pf_data_rd.l3_miss_remote_hop1_dram.snoop_noneoffcore_response.all_reads.l3_miss.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.NO_SNOOP_NEEDEDoffcore_response.demand_code_rd.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.ANY_SNOOPoffcore_response.other.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l1d_and_sw.l3_miss.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.SNOOP_NONEoffcore_response.pf_l1d_and_sw.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.pf_l2_data_rd.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDoffcore_response.pf_l3_rfo.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.all_data_rd.any_responseocr.all_pf_rfo.supplier_none.hit_other_core_fwdocr.all_rfo.any_responseCounts all demand code reads  OCR.DEMAND_CODE_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.demand_rfo.supplier_none.hit_other_core_fwdCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.pf_l2_rfo.any_responseevent=0xe7event=0xd0,period=1000003,umask=0x82Counts all retired store instructions. This event account for SW prefetch instructions and PREFETCHW instruction for stores  Supports address when precise (Precise event)Retired load instructions whose data sources were L3 hit and cross-core snoop missed in on-pkg core cache  Supports address when precise (Precise event)Counts demand instruction fetches and L1 instruction cache prefetches that hit a cacheline in the L3 where a snoop hit in another core, data forwarding is not requiredCounts demand instruction fetches and L1 instruction cache prefetches that hit a cacheline in the L3 where a snoop was sent but no other cores had the dataCounts demand data reads that hit a cacheline in the L3 where a snoop hit in another core, data forwarding is not requiredocr.hwpf_l1d_and_swpf.l3_hit.anyRetired instructions after front-end starvation of at least 1 cycle (Precise event)event=0xc6,period=100007,umask=0x1,frontend=0x501006idq.ms_cycles_anyevent=0x79,cmask=1,edge=1,period=100003,umask=0x30C9 residency percent per packageocr.hwpf_l1d_and_swpf.l3_missevent=0xc9,period=100003,umask=0x4ocr.other.dramPrecise instruction retired event with a reduced effect of PEBS shadow in IP distribution (Precise event)False dependencies due to partial compare on addressevent=0x5e,cmask=1,edge=1,inv=1,period=100003,umask=0x1event=0x8,period=100003,umask=0x4Branches;FetchBW;PGOMemoryBound;MemoryBWCounts demand data reads that resulted in a snoop that hit in another core, which did not forward the dataocr.demand_rfo.snc_cache.hit_with_fwdCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that hit in the L3 or were snooped from another core's caches on the same socketevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x830000477event=0xb7,period=100003,umask=0x1,offcore_rsp=0x94002380Counts hardware prefetches to the L3 only that missed the local socket's L1, L2, and L3 cachesCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that missed the L3 Cache and were supplied by the local socket (DRAM or PMM), whether or not in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts PMM or DRAM accesses that are controlled by the close or distant SNC Clusterocr.demand_data_rd.snc_dramCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were supplied by DRAM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeocr.reads_to_core.local_socket_dramCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by PMM attached to another socketAll DRAM write CAS commands issued. Unit: uncore_imc unc_m_pre_count.pgtRemote read requests sent to the CHA's home agent. Unit: uncore_cha unc_cha_tor_occupancy.ia_hitTOR Occupancy : RFOs issued by iA Cores that Missed the LLC. Unit: uncore_cha event=0x36,umask=0xC001FF04TOR Inserts : LLCPrefRFO issued by iA Cores. Unit: uncore_cha TOR Occupancy : DRds issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha TOR Inserts : CLFlushes issued by iA Cores. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_drd_local_pmmTOR Occupancy : DRds issued by iA Cores targeting DDR Mem that Missed the LLC. Unit: uncore_cha PCIe Completion Buffer Occupancy of completions with data : Part 7. Unit: uncore_iio unc_m2m_tag_hit.nm_rd_hit_cleanunc_m2p_cms_clockticksevent=0x24,period=200003l2_request.missmem_bound_stalls.ifetch_l2_hitCounts the number of retired split loads uops  Supports address when precise (Precise event)Counts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were supplied by the L3 cacheocr.hwpf_l2_code_rd.l3_hit.snoop_hit_with_fwdCounts streaming stores that were not supplied by the L3 cacheevent=0x63,edge=1,period=200003This event is deprecated. Refer to new event BUS_LOCK.LOCK_CYCLESCounts the number of unhalted cycles a core is blocked due to an accepted lock it issuedCounts the number of core cycles during which there are pending interrupts while interrupts are masked (disabled)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000000000000040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000477Counts the total number of issue slots that were not consumed by the backend because allocation is stalled due to a machine clear (nuke) of any kind including memory ordering and memory disambiguationLines Victimized : All Lines Victimized : Counts the number of lines that were victimized on a fill.  This can be filtered by the state that the line was inCounts snoop filter capacity evictions for entries tracking modified lines in the cores? cache.? Snoop filter capacity evictions occur when the snoop filter is full and evicts an existing entry to track a new entry.? Does not count clean evictions such as when a core?s cache replaces a tracked cacheline with a new cachelineevent=0x35,umask=0xC827FF01TOR Inserts : CRd_Prefs issued by iA Cores that hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : RFOs issued by iA Cores that Hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x35,umask=0xC877DE01TOR Inserts : WiLs issued by iA Cores that Missed LLC. Unit: uncore_cha TOR Inserts : RFOs issued by iA Cores : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interrupts: All Inserts Inbound (p2p + faf + cset)Counts the number of page walks outstanding in the page miss handler (PMH) for stores every cycle.  A page walk is outstanding from start till PMH becomes idle again (ready to serve next walk). Includes EPT-walk intervalsCounts the number of page walks outstanding for an Extended Page table walk including GTLB hits per cycleCPU_CLK_UNHALTED.COREAddress_Alias_BlocksPercentage of all uops which are ucode ops. Unit: cpu_atom MEM_BOUND_STALLS.LOAD_L2_HIT / MEM_LOAD_UOPS_RETIRED.L2_HITevent=0xd0,period=1000003,umask=0x5,ldlat=0x100L2 cache misses when fetching instructions. Unit: cpu_core Counts demand read for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that resulted in a snoop hit in another cores caches, data forwarding is required as the data is modified. Unit: cpu_core offcore_requests.data_rdevent=0x20,cmask=1,period=1000003,umask=0x8fp_arith_dispatched.port_5Retired Instructions who experienced a critical DSB miss (Precise event). Unit: cpu_core event=0xc6,period=100007,umask=0x1,frontend=0x620006Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 512 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core ld_head.other_at_retCounts the number of retired loads that are blocked because it initially appears to be store forward blocked, but subsequently is shown not to be blocked based on 4K alias check (Precise event). Unit: cpu_atom Counts the number of issue slots every cycle that were not delivered by the frontend due to other common frontend stalls not categorized. Unit: cpu_atom number of branch instructions retired that were mispredicted and taken. Non PEBS (Precise event). Unit: cpu_core TMA slots available for an unhalted logical processor. General counter - architectural event. Unit: cpu_core Uops executed on ports 5 and 11. Unit: cpu_core Cycles where at least 4 uops were executed per-thread. Unit: cpu_core Counts the number of uops to be executed per-thread each cycle. Unit: cpu_core uops_retired.cyclesunc_m_dram_page_miss_rdPage walks completed due to a demand data load to a 1G page. Unit: cpu_core Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by a cache on a remote socket where a snoop hit in another core's caches which forwarded the unmodified data to the requesting coreCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that either hit a non-modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeOFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DEMAND_RFOevent=0x2a,period=100003,umask=0x1,offcore_rsp=0xFBFF80822rs_empty.cyclesevent=0xe4,umask=0x0000000001unc_m_pre_count.wr_pch0IIO Clockticks. Unit: uncore_iio event=0x84,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000002TOR Occupancy for DRds issued by iA Cores targeting DDR Mem that Missed the LLC. Unit: uncore_cha Matches on Transmit path of a UPI Port : Non-Coherent Standard, Match Opcode. Unit: uncore_upi RxQ Occupancy - All Packets : Slot 0. Unit: uncore_upi Data requested of the CPU : Card reading from DRAM. Unit: uncore_iio event=0x84,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000001event=0xc1,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000001event=0x33,umask=0x0000000002event=0x50,umask=0x0000000010TOR Inserts; DRd from local IA. Unit: uncore_cha event=0x35,umask=0x00c827ff01TOR Occupancy; All Snoops from Remote. Unit: uncore_cha TOR Occupancy; CRd hits from local IA. Unit: uncore_cha unc_cha_tor_occupancy.evictevent=0x36,umask=0x00c8a7fd01unc_cha_tor_occupancy.io_hit_rfounc_cha_tor_occupancy.ia_miss_rfo_pref_remoteevent=0x35,umask=0x00cccffd01TOR Inserts : WCiLs issued by iA Cores targeting DDR that missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00c8168601TOR Occupancy : CRd issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00cc47fe01TOR Occupancy : UCRdFs issued by iA Cores that Missed LLC. Unit: uncore_cha TOR Occupancy : WCiLFs issued by iA Cores targeting DDR that missed the LLC - HOMed remotely. Unit: uncore_cha IC line invalidated due to L2 invalidating probe (external or LS). The number of instruction cache lines invalidated. A non-SMC event is CMC (cross modifying code), either from the other thread of the core or another coreevent=0x60,umask=0x08Miscellaneous events covered in more detail by l2_request_g2 (PMCx061)L3 cache misses. Unit: amd_l3 All L3 Miss Request Types. Ignores SliceMask and ThreadMask. Unit: amd_l3 ex_tagged_ibs_ops.ibs_tagged_ops_retevent=0x1d0event=0x107,umask=0x38Total number of fp uOps  on pipe 0All OpsThe number of serializing Ops retired. x87 bottom-executing uOps retiredThe number of software prefetches that did not fetch data outside of the processor core. Software PREFETCH instruction saw a match on an already-allocated miss request bufferevent=0x28a,umask=0x02L2 Cache Accesses from L1 Instruction Cache Misses (including prefetch)l3_missesLS MAB Allocates by Type. DC prefetcherL1 DTLB Miss. DTLB reload to a 4K page that miss the L2 TLBAll TLB Flushesevent=0xae,umask=0x20de_dis_dispatch_token_stalls1.int_sched_misc_token_stallevent=0xae,umask=0x01Counts retired Fused InstructionsDemand Data Cache Fills by Data Source. From CCX Cache in different NodeCycles where a dispatch group is valid but does not get dispatched due to a Token Stall. Also counts cycles when the thread is not selected to dispatch but would have been stalled due to a Token Stall. FP scheduler resource stall. Applies to ops that use the FP schedulerCycles where a dispatch group is valid but does not get dispatched due to a Token Stall. Also counts cycles when the thread is not selected to dispatch but would have been stalled due to a Token Stall. Taken branch buffer resource stallCycles where a dispatch group is valid but does not get dispatched due to a Token Stall. Also counts cycles when the thread is not selected to dispatch but would have been stalled due to a Token Stall. Store Queue resource stall. Applies to all ops with store semantics(xi_sys_fill_latency * 16) / xi_ccx_sdp_req1L1 Data Cache Fills: From Remote NodeL1 Data Cache Fills: From within same CCXL2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L2UpDown (fetch the next or previous line into L2 cache for all memory accesses)event=0x71,umask=0x02L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L2UpDown (fetch the next or previous line into L2 cache for all memory accesses)L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L1Region (fetch additional lines into L1 cache when the data access for a given instruction tends to be followed by a consistent pattern of other accesses within a localized region)event=0xad,umask=0x02L3 cache fill requests sourced from all data sources. Unit: amd_l3 SMIs receivedWrite data beats (64 bytes) for local processor at Coherent Station (CS) 3remote_processor_read_data_beats_cs6remote_processor_write_data_beats_cs4Write data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 0event=0x85f,umask=0xbffevent=0x41e,umask=0x7feevent=0x51e,umask=0x7felocal_socket_inf1_inbound_data_beats_ccm7local_socket_inf0_outbound_data_beats_ccm2remote_socket_inf1_inbound_data_beats_ccm7remote_socket_inf0_outbound_data_beats_ccm5fp_ops_retired_by_width.allfp_ops_retired_by_type.vector_logicalevent=0xb,umask=0x0cRetired SSE and AVX integer subtract opsRetired SSE and AVX integer shuffle ops (may include instructions not necessarily thought of as including shuffles e.g. horizontal add, dot product, and certain MOV instructions)Retired 256-bit packed floating-point add opsevent=0xc,umask=0x40fp_pack_ops_retired.allRetired 256-bit packed integer compare opsFloating-point dispatch faults of all typesNumber of CAS commands sentNumber of memory load-store operations dispatched to the load-store unitfrontend_boundL1 demand data cache fills from within the same CCXl1_demand_data_cache_fills_from_far_cacheL1 demand data cache fills from DRAM or MMIO in a different NUMA nodeMemory controller CAS command rateAttributable Level 3 cache access, readTotal read hits. Unit: hisi_sccl,l3c sys_ccn_pmu.read_cyclesIC_MICROARCHITECTURAL_RESYNC_BY_SNOOPFR_RETIRED_UOPSFR_NUMBER_OF_BREAKPOINTS_FOR_DR1EVENT_80HEVENT_8EHEVENT_97HEVENT_98HEVENT_9EHEVENT_B0HEVENT_B9HINTEGER_CORE_CLOCK_ENABLEDEVENT_102HEVENT_127HEVENT_14EHEVENT_16DHEVENT_176HEVENT_1FFHEVENT_204HEVENT_207HEVENT_236HEVENT_277HEVENT_27DHEVENT_292HEVENT_296HEVENT_2ACHEVENT_2EAHEVENT_32EHEVENT_334HEVENT_36EHEVENT_3AAHEVENT_3C0HEVENT_3C4HEVENT_3CCHEVENT_3E7HEVENT_3F5HEVENT_3FCHVFP_SPECSTALL_BACKENDdn_rxreq_bpi_dvmophnf_dir_snoops_senthni_rrt_rd_occ_cnt_ovflhni_rrt_rd_allocxp_txflit_validcxha_rdb_occclkdiv2_cycle_countclkdiv2_ranks_in_pwr_downVIU2_INSTR_WAIT_CYCLESTHRESHOLD_VEC_INSTR_QUEUE_ENTRIES_CYCLESBTIC_MISSTLBSYNC_INSTR_COMPLETEDGROUP_DISPATCHMARKED_GROUP_ISSUEDREJECT_COMPLETION_STALL_ERAT_MISSPM_EVENT_CYCLESCYCLES_MU_SCHED_STALLEDMISALIGNED_LOAD_STORE_ACCESS_TRANSLATEDINTERRUPTS_TAKENgspage-conflictmemory-controller-lo-pri-bypassbranchesdc-missesREADINTEL_IVYBRIDGE_XEONSTOPPEDLLC_MISSEScpu_clk_unhalted.threadRESOURCE_STALLS.ANYfrontend_retired.l1i_missch_maskl3_thread_mask{"type": "closelog"}
{"type": "procexit"{"type": "map_out"coreGenuineIntel-6-9AFlops;InsTypeIpArith_Scalar_SPThis event counts the number of WB requests that hit L2 cacheL2 cache hits when fetching instructions, code readsL2 cache misses when fetching instructionsThis event counts L1D writebacks that access L2 cachemem_load_uops_l3_miss_retired.local_dramThis event counts the number of offcore outstanding cacheable Core Data Read transactions in the super queue every cycle. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation). See corresponding Umask under OFFCORE_REQUESTS  Spec update: BDM76fp_arith_inst_retired.scalar_singlefp_assist.anyevent=0x58,period=1000003,umask=0x8Uops not delivered to Resource Allocation Table (RAT) per thread when backend of the machine is not stalledevent=0xc8,period=2000003,umask=0x80event=0xc9,period=2000003,umask=0x8event=0xc9,period=2000003,umask=0x1Number of times a TSX Abort was triggered due to commit but Lock Buffer not emptyevent=0x54,period=2000003,umask=0x40otherlock_cycles.split_lock_uc_lock_durationevent=0x14,period=2000003,umask=0x1br_inst_exec.nontaken_conditionalFar branch instructions retired  Spec update: BDW98Cycles while L1 cache miss demand load is outstandingNot software-prefetch load dispatches that hit FB allocated for hardware prefetchmachine_clears.cyclesevent=0xa1,period=2000003,umask=0x80Cycles at least 4 micro-op is executed from any thread on physical coreuops_executed_port.port_2event=0xc2,cmask=10,inv=1,period=2000003,umask=0x1event=0x35,umask=0x3,filter_opc=0x1c8,filter_tid=0x3eevent=0x49,period=100003,umask=0x1dtlb_store_misses.walk_durationMisses at all ITLB levels that cause page walks  Spec update: BDM69event=0x24,period=200003,umask=0xc4Counts all demand & prefetch data reads have any response typeoffcore_response.all_pf_data_rd.l3_hit.snoop_missoffcore_response.all_pf_data_rd.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020090offcore_response.all_pf_rfo.supplier_none.snoop_noneoffcore_response.demand_code_rd.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0004Counts demand data reads have any response typeoffcore_response.demand_data_rd.l3_hit.snoop_noneoffcore_response.demand_data_rd.l3_hit.snoop_not_neededoffcore_response.demand_rfo.l3_hit.snoop_hit_no_fwdoffcore_response.other.supplier_none.any_snoopoffcore_response.other.supplier_none.snoop_hit_no_fwdoffcore_response.pf_l2_data_rd.l3_hit.any_snoopoffcore_response.pf_l2_data_rd.l3_hit.snoop_noneoffcore_response.pf_l3_code_rd.any_responseoffcore_response.pf_l3_code_rd.l3_hit.snoop_missoffcore_response.pf_l3_code_rd.supplier_none.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0080offcore_response.pf_l3_data_rd.supplier_none.any_snoopoffcore_response.pf_l3_data_rd.supplier_none.snoop_noneRandomly selected loads with latency value being above 32  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000240offcore_response.all_pf_data_rd.l3_miss.snoop_hit_no_fwdoffcore_response.all_pf_rfo.l3_miss.snoop_missoffcore_response.all_pf_rfo.l3_miss.snoop_not_neededoffcore_response.demand_code_rd.l3_miss_local_dram.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000002offcore_response.pf_l3_rfo.supplier_none.snoop_non_dramoffcore_response.all_code_rd.llc_hit.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) RFOs hit in the L3Counts all requests miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00002(UNC_M_POWER_SELF_REFRESH / UNC_M_CLOCKTICKS) * 100.event=0x40,period=2000000,umask=0xa3event=0x29,period=200000,umask=0x5fL2 cache lines evictedevent=0x30,period=200000,umask=0x5fevent=0x30,period=200000,umask=0x52l2_rqsts.self.prefetch.i_stateevent=0x11,period=10000,umask=0x81event=0x11,period=10000,umask=0x1event=0xb1,period=2000000,umask=0x80x87_comp_ops_exe.any.smisalign_mem_ref.st_bubbleStreaming SIMD Extensions (SSE) PrefetchT2 instructions executedbus_trans_inval.selfbus_trans_mem.selfevent=0x6a,period=200000,umask=0xe0RFO bus transactionsbus_trans_wb.all_agentsevent=0x6,period=200000,umask=0x80event=0xc4,period=2000000,umask=0x1br_inst_type_retired.retevent=0xd1,period=200003,umask=0x2Counts memory load uops retired where the data is retrieved from the WCB (or fill buffer), indicating that the load found its data while that data was in the process of being brought into the L1 cache.  Typically a load will receive this indication when some other load or prefetch missed the L1 cache and was in the process of retrieving the cache line containing the data, but that process had not yet finished (and written the data back to the cache). For example, consider load X and Y, both referencing the same cache line that is not in the L1 cache.  If load X misses cache first, it obtains and WCB (or fill buffer) and begins the process of requesting the data.  When load Y requests the data, it will either hit the WCB, or the L1 cache, depending on exactly what time the request to Y occurs  Supports address when precise (Must be precise)mem_uops_retired.allLoad uops retired (Precise event capable)  Supports address when precise (Must be precise)Counts the number of store uops retired  Supports address when precise (Must be precise)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600003010offcore_response.any_rfo.l2_hitCounts demand cacheable data reads of full cache lines that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.full_streaming_stores.l2_miss.hit_other_core_no_fwdoffcore_response.sw_prefetch.l2_miss.hitm_other_coreevent=0xe6,period=200003,umask=0x1event=0x80,period=200003,umask=0x3ms_decoded.ms_entryCounts the number of instructions that retire execution. For instructions that consist of multiple uops, this event counts the retirement of the last uop of the instruction. The event continues counting during hardware interrupts, traps, and inside interrupt handlers.  This is an architectural performance event.  This event uses a (_P)rogrammable general purpose performance counter. *This event is Precise Event capable:  The EventingRIP field in the PEBS record is precise to the address of the instruction which caused the event.  Note: Because PEBS records can be collected only on IA32_PMC0, only one event can use the PEBS facility at a time (Must be precise)Loads blocked because address has 4k partial address false dependence (Precise event capable) (Must be precise)Uops requested but not-delivered to the back-end per cycleevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x00000132b7Counts requests to the uncore subsystem have any transaction responses from the uncore subsystemCounts bus lock and split lock requests have any transaction responses from the uncore subsystemCounts bus lock and split lock requests have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.full_streaming_stores.outstandingCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.pf_l1_data_rd.any_responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000002000Counts data cacheline reads generated by hardware L2 cache prefetcher hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of times that the machines clears due to a page fault. Covers both I-side and D-side(Loads/Stores) page faults. A page fault occurs when either page is not present, or an access violationCounts once per cycle for each page walk occurring due to a load (demand data loads or SW prefetches). Includes cycles spent traversing the Extended Page Table (EPT). Average cycles per walk can be calculated by dividing by the number of walksevent=0x48,period=2000003,umask=0x2Counts any demand and L1 HW prefetch data load requests to L2  Spec update: HSD78, HSM80All L2 requests  Spec update: HSD78, HSM80Retired store uops that miss the STLB  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)Counts all prefetch (that bring data to LLC only) data reads hit in the L3Randomly selected loads with latency value being above 32  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100400001Number of times a transactional abort was signaled due to a data conflict on a transactionally accessed addressFar branch instructions retiredNumber of far branches retiredNumber of near taken branches retired (Precise event)This event counts the number of reference cycles when the core is not in a halt state. The core enters the halt state when it is running the HLT instruction or the MWAIT instruction. This event is not affected by core frequency changes (for example, P states, TM2 transitions) but has the same incrementing frequency as the time stamp counter. This event can approximate elapsed time while the core was not in a halt stateCycles with pending L2 cache miss loads  Spec update: HSD78, HSM63, HSM80Counts total number of uops to be executed per-core each cycle  Spec update: HSD30, HSM31L3 Lookup external snoop request that access cache and found line in E or S-state. Unit: uncore_cbox page_walker_loads.itlb_memoryRetired load uop whose Data Source was: remote DRAM either Snoop not needed or Snoop Miss (RspI)  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)offcore_response.pf_l2_code_rd.llc_hit.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x083FC00091Not rejected writebacks from L1D to L2 cache lines in any statel2_rqsts.pf_missRetired store uops that split across a cacheline boundary. (Precise Event)Offcore outstanding cacheable data read transactions in SQ to uncore. Set Cmask=1 to count cyclesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00010002Number of FP Computational Uops Executed this cycle. The number of FADD, FSUB, FCOM, FMULs, integer MULsand IMULs, FDIVs, FPREMs, FSQRTS, integer DIVs, and IDIVs. This event does not distinguish an FADD used in the middle of a transcendental flow from a sCycles per thread when uops are dispatched to port 4unc_cbo_cache_lookup.sevent=0x34,umask=0x10event=0x34,umask=0x80Completed page walks in ITLB due to STLB load misses for large pagesevent=0xd3,period=100007,umask=0xcevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0080offcore_response.all_code_rd.llc_miss.remote_dramCounts demand data reads that miss the LLC  the data is found in M state in remote cache and forwarded from thereoffcore_response.demand_data_rd.llc_miss.remote_hit_forwardLLC misses - Uncacheable reads. Derived from unc_c_tor_inserts.miss_opcode.uncacheable. Unit: uncore_cbox event=0x35,umask=0x1,filter_opc=0x195llc_references.itom_writeWrite requests to home agent. Unit: uncore_ha (UNC_P_FREQ_BAND0_CYCLES / UNC_P_CLOCKTICKS) * 100.Counts the number of cycles that the uncore was running at a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band2=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Unit: uncore_pcu Counts the number of times that the uncore transitioned to a frequency greater than or equal to 1.2Ghz. Derived from unc_p_freq_band0_cycles. Unit: uncore_pcu event=0x51,period=2000003,umask=0x2Retired load uops which data sources were LLC hit and cross-core snoop missed in on-pkg core cacheRetired load uops that miss the STLB (Precise event)Speculative and retired mispredicted direct near callsevent=0xc1,period=100003,umask=0x2Resource stalls due to load or store buffers all being in useresource_stalls.ooo_rsrcCounts the number of cycles that the uncore was running at a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band2=XXX with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Unit: uncore_pcu This event counts the number of load micro-ops retiredevent=0x4,period=200003,umask=0x2event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800088000Counts Demand cacheable data write requests  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180022Counts Demand cacheable data write requests  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateCounts demand cacheable data and L1 prefetch data reads that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000080offcore_response.partial_writes.l2_hit_this_tile_sCounts L1 data HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateoffcore_response.pf_l2_code_rd.l2_hit_near_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000020offcore_response.pf_software.l2_hit_near_tile_mThis event counts the number of times that the pipeline stalled due to FP operations needing assistsCounts Demand cacheable data and L1 prefetch data read requests  that accounts for data responses from DRAM Localoffcore_response.any_pf_l2.ddr_faroffcore_response.any_read.ddrevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x01806032f7offcore_response.bus_locks.mcdramCounts demand code reads and prefetch code reads that accounts for responses from DDR (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400001event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600002event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000020offcore_response.pf_l2_rfo.mcdram_farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200020event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080801000offcore_response.uc_code_reads.ddr_farCounts the number of mispredicted branch instructions retired that were conditional jumps (Precise event)This event counts the number of instructions that retire.  For instructions that consist of multiple micro-ops, this event counts exactly once, as the last micro-op of the instruction retires.  The event continues counting while instructions retire, including during interrupt service routines caused by hardware interrupts, faults or trapsno_alloc_cycles.allevent=0xca,period=200003,umask=0x7fevent=0xca,period=200003,umask=0x90Counts the number of occurences a retired store that is a cache line split. Each split should be counted only onceuncore_edc_eclkevent=0x51,period=2000000,umask=0x2L1 data cacheable reads and writesevent=0x40,period=2000000,umask=0x1event=0x40,period=2000000,umask=0xfevent=0x41,period=2000000,umask=0x2event=0x4e,period=200000,umask=0x2l1d_prefetch.requestsL2 lines allocated in the E stateL2 prefetch missesl2_write.lock.m_stateoffcore_response.any_ifetch.any_cache_dramoffcore_response.any_ifetch.llc_hit_other_core_hitmoffcore_response.any_ifetch.remote_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7FFFoffcore_response.any_rfo.io_csr_mmioevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x408Offcore writebacks to a remote cache or remote DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3877offcore_response.data_in.remote_cache_hitoffcore_response.demand_data_rd.remote_cache_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x404event=0xb7,period=100000,umask=0x1,offcore_rsp=0x1004offcore_response.demand_rfo.llc_hit_no_other_coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x202offcore_response.demand_rfo.remote_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x780offcore_response.pf_data.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x410offcore_response.pf_ifetch.io_csr_mmioevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x120event=0xf7,period=20000,umask=0x1event=0xfd,period=200000,umask=0x40offcore_response.data_ifetch.remote_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4033event=0xb7,period=100000,umask=0x1,offcore_rsp=0xF802Offcore prefetch data reads satisfied by the local DRAMOffcore prefetch code reads that missed the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4020event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2020event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4070event=0xb8,period=100000,umask=0x2br_inst_exec.condReference cycles when thread is not halted (fixed counter)event=0x3c,period=100000,umask=0x1Total CPU cyclesInstructions retired (fixed counter)event=0xc0,period=2000000,umask=0x1resource_stalls.loadevent=0xb1,any=1,period=2000000,umask=0x80event=0xb1,any=1,period=2000000,umask=0x10ITLB miss page walksmem_inst_retired.all_storesRetired load instructions which data sources were load missed L1 but hit FB due to preceding miss to the same cache line with data not ready  Supports address when precise (Precise event)Counts retired load instructions with at least one uop that hit in the L1 data cache. This event includes all SW prefetches and lock instructions regardless of the data source  Supports address when precise (Precise event)offcore_response.demand_code_rd.l3_hit.spl_hitoffcore_response.demand_code_rd.l3_hit_m.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10001C0001offcore_response.demand_data_rd.l3_hit_e.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040001offcore_response.demand_data_rd.l3_hit_s.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x801C0002offcore_response.demand_rfo.l3_hit_m.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100108000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0408000offcore_response.other.l4_hit_local_l4.snoop_not_neededevent=0xc6,period=100007,umask=0x1,frontend=0x200206Counts the number of uops delivered to Instruction Decode Queue (IDQ) from the MITE path. Counting includes uops that may 'bypass' the IDQ. This also means that uops are not being delivered from the Decode Stream Buffer (DSB)Counts cycles during which uops initiated by Decode Stream Buffer (DSB) are being delivered to Instruction Decode Queue (IDQ) while the Microcode Sequencer (MS) is busy. Counting includes uops that may 'bypass' the IDQCycles while L3 cache miss demand load is outstandingoffcore_requests_outstanding.l3_miss_demand_data_rdoffcore_response.demand_code_rd.l3_hit_m.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C400001offcore_response.demand_data_rd.l3_miss_local_dram.spl_hitoffcore_response.demand_rfo.l3_miss_local_dram.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000108000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84008000Counts all (macro) branch instructions retired  Spec update: SKL091cpu_clk_unhalted.ring0_transinst_retired.nopNumber of uops delivered to the back-end by the LSD(Loop Stream Detector)Increments whenever there is an update to the LBR arrayevent=0xc2,cmask=1,inv=1,period=2000003,umask=0x2100 * ( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * INT_MISC.CLEAR_RESTEER_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) )100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) * ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD) / #(CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (( OFFCORE_REQUESTS_BUFFER.SQ_FULL / 2 ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) ) + ( (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( ((L1D_PEND_MISS.PENDING / ( MEM_LOAD_RETIRED.L1_MISS + MEM_LOAD_RETIRED.FB_HIT )) * cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ / CPU_CLK_UNHALTED.THREAD) / #(max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) ) Total pipeline cost of Memory Address Translation related bottlenecks (data-side TLBs)Fed;FetchBW;FrontendFed;FetchBW;Frontend_SMT(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * (DSB2MITE_SWITCHES.PENALTY_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) + ((IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD))) * (( IDQ.ALL_MITE_CYCLES_ANY_UOPS - IDQ.ALL_MITE_CYCLES_4_UOPS ) / CPU_CLK_UNHALTED.THREAD / 2) / #((IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)))DSB_Misses_CostLoads that miss the DTLB and hit the STLBCounts demand reads of partial cache lines (including UC and WC) that miss L2event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680002000event=0xca,period=200003,umask=0x3fCounts the number of cycles when no uops are allocated, the IQ is empty, and no other condition is blocking allocationI-side page-walksevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0240Counts prefetch code reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1f80400040Counts prefetch (that bring data to L2) data reads that miss the LLC and the data returned from dramoffcore_response.pf_llc_rfo.llc_miss.dramoffcore_requests.any.readoffcore_requests_outstanding.demand.read_codeevent=0x60,period=2000000,umask=0x1event=0x60,cmask=1,period=2000000,umask=0x4REQUEST = ANY IFETCH and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1ffREQUEST = CORE_WB and RESPONSE = LOCAL_CACHEREQUEST = DATA_IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = DEMAND_DATA and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff02REQUEST = PF_DATA and RESPONSE = REMOTE_CACHE_HITMoffcore_response.pf_data_rd.all_local_dram_and_remote_cache_hitREQUEST = PREFETCH and RESPONSE = IO_CSR_MMIOREQUEST = DATA_IFETCH and RESPONSE = ANY_LLC_MISSoffcore_response.data_in.any_dram_and_remote_fwdREQUEST = DEMAND_IFETCH and RESPONSE = ANY_DRAM AND REMOTE_FWDoffcore_response.pf_data.any_dram_and_remote_fwdevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf820REQUEST = PF_IFETCH and RESPONSE = OTHER_LOCAL_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5802event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2770Retired loads that hit remote socket in modified state (Precise Event)event=0xb1,cmask=1,edge=1,inv=1,period=2000000,umask=0x1fCounts all demand & prefetch data reads that have any response typeoffcore_response.pf_l1d_and_sw.l3_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0020Counts all prefetch (that bring data to LLC only) data reads that hit in the L3Counts all prefetch (that bring data to LLC only) RFOs that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.all_data_rd.l3_miss.remote_hit_forwardoffcore_response.all_pf_rfo.l3_miss.remote_hit_forwardoffcore_response.demand_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts prefetch (that bring data to L2) data reads that miss in the L3Counts all prefetch (that bring data to LLC only) data reads that miss the L3 and the data is returned from local or remote dramevent=0x28,period=200003,umask=0x7Core cycles where the core was running in a manner where Turbo may be clipped to the AVX2 turbo scheduleInstructions per FP Arithmetic AVX 512-bit instruction (lower number means higher occurrence rate). May undercount due to FMA double countingINST_RETIRED.ANY / ( FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE )1000 * L2_LINES_OUT.SILENT / INST_RETIRED.ANYcha@event\=0x36\,umask\=0x21\,config\=0x40433@ / cha@event\=0x36\,umask\=0x21\,config\=0x40433\,thresh\=1@All DRAM CAS Commands issued. Unit: uncore_imc Counts the number of entries in the Read Pending Queue (RPQ) at each cycle.  This can then be used to calculate both the average occupancy of the queue (in conjunction with the number of cycles not empty) and the average latency in the queue (in conjunction with the number of allocations).  The RPQ is used to schedule reads out to the memory controller and to track the requests.  Requests allocate into the RPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the CHA to the iMC. They deallocate from the RPQ after the CAS command has been issued to memorywrite requests from local home agent. Unit: uncore_cha unc_iio_data_req_of_cpu.mem_write.part2unc_cha_requests.invitoe_localLocal requests for exclusive ownership of a cache line  without receiving data. Unit: uncore_cha event=0x13,umask=0x01unc_cha_sf_eviction.m_stateevent=0x5c,umask=0x08PCIe Completion Buffer Inserts of completions with data: Part 1PCIe Completion Buffer occupancy of completions with data: Part 0-3event=0xd5,fc_mask=0x04,umask=0x01event=0xc0,ch_mask=0x04,fc_mask=0x07,umask=0x08Peer to peer write request of 4 bytes made to IIO Part1 by a different IIO unit. Unit: uncore_iio event=0xc1,ch_mask=0x01,fc_mask=0x07,umask=0x08Write request of up to a 64 byte transaction is made by IIO Part0 to Memory. Unit: uncore_iio Cycles when direct to core mode (which bypasses the CHA) was disabled. Unit: uncore_m2m Cycles when direct to Intel UPI was disabled. Unit: uncore_m2m unc_m2m_directory_update.i2aevent=0x37,umask=0x4unc_m2m_imc_writes.niunc_m2m_rxc_bl_insertsCycles the Rx of the Intel UPI is in L0p power mode. Unit: uncore_upi Counts incoming FLITs (FLow control unITs) which bypassed the slot0 RxQ buffer (Receive Queue) and passed directly to the Egress.  This is a latency optimization, and should generally be the common case.  If this value is less than the number of FLITs transfered, it implies that there was queueing getting onto the ring, and thus the transactions saw higher latencyunc_upi_rxl_flits.non_dataocr.all_data_rd.l3_hit.snoop_hit_with_fwdOCR.ALL_DATA_RD.L3_HIT.SNOOP_MISS OCR.ALL_DATA_RD.L3_HIT.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200491OCR.ALL_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDED  OCR.ALL_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0490ocr.all_pf_data_rd.l3_hit_f.hitm_other_coreocr.all_pf_data_rd.l3_hit_m.hit_other_core_no_fwdocr.all_pf_data_rd.l3_hit_s.any_snoopOCR.ALL_PF_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDOCR.ALL_PF_RFO.L3_HIT_E.HIT_OTHER_CORE_FWD  OCR.ALL_PF_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDocr.all_pf_rfo.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100120OCR.ALL_PF_RFO.L3_HIT_S.SNOOP_MISSocr.all_reads.l3_hit.no_snoop_neededOCR.ALL_READS.L3_HIT_F.NO_SNOOP_NEEDED  OCR.ALL_READS.L3_HIT_F.NO_SNOOP_NEEDEDOCR.ALL_READS.L3_HIT_S.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200122Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_E.HITM_OTHER_COREocr.demand_data_rd.l3_hit.hit_other_core_fwdocr.demand_rfo.l3_hit_s.hitm_other_coreCounts any other requests OCR.OTHER.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.OTHER.L3_HIT.HIT_OTHER_CORE_NO_FWDCounts any other requests  OCR.OTHER.L3_HIT_M.NO_SNOOP_NEEDEDocr.other.l3_hit_s.snoop_missCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_F.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080010ocr.pf_l2_rfo.l3_hit.hitm_other_coreocr.pf_l2_rfo.l3_hit_e.snoop_noneocr.pf_l2_rfo.l3_hit_m.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_HIT.SNOOP_NONECounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDocr.pf_l3_rfo.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080100Counts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_F.ANY_SNOOPocr.pf_l3_rfo.l3_hit_m.any_snoopocr.pf_l3_rfo.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.SNOOP_MISSoffcore_response.all_data_rd.pmm_hit_local_pmm.snoop_not_neededoffcore_response.all_pf_data_rd.l3_hit_e.hit_other_core_no_fwdoffcore_response.all_pf_data_rd.l3_hit_e.snoop_missoffcore_response.all_pf_data_rd.l3_hit_s.any_snoopThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020490offcore_response.all_pf_rfo.l3_hit_f.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400120offcore_response.all_reads.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F804007F7This event is deprecated. Refer to new event OCR.ALL_READS.SUPPLIER_NONE.SNOOP_NONEoffcore_response.all_rfo.l3_hit_e.hitm_other_coreoffcore_response.all_rfo.l3_hit_e.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020122offcore_response.demand_code_rd.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.SUPPLIER_NONE.SNOOP_MISSoffcore_response.demand_rfo.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_E.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80408000This event is deprecated. Refer to new event OCR.OTHER.PMM_HIT_LOCAL_PMM.ANY_SNOOPoffcore_response.pf_l1d_and_sw.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_data_rd.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020020This event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_E.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_S.SNOOP_MISSoffcore_response.pf_l3_rfo.l3_hit_s.snoop_missThis event is deprecated. Refer to new event OCR.PF_L3_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDLSD.UOPS / (IDQ.DSB_UOPS + LSD.UOPS + IDQ.MITE_UOPS + IDQ.MS_UOPS)( ( 64 * imc@event\=0xe3@ / 1000000000 ) / duration_time )ocr.all_data_rd.l3_miss.snoop_missOCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOP  OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000491OCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDOCR.ALL_PF_DATA_RD.L3_MISS.REMOTE_HITM OCR.ALL_PF_DATA_RD.L3_MISS.REMOTE_HITMOCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.all_pf_data_rd.l3_miss_remote_hop1_dram.snoop_missocr.all_reads.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x900007F7ocr.all_rfo.l3_miss_local_dram.hit_other_core_no_fwdOCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSocr.demand_code_rd.l3_miss_local_dram.snoop_missocr.demand_data_rd.l3_miss_local_dram.hitm_other_coreCounts any other requests OCR.OTHER.L3_MISS.HITM_OTHER_CORE OCR.OTHER.L3_MISS.HITM_OTHER_CORECounts any other requests  OCR.OTHER.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000400ocr.pf_l1d_and_sw.l3_miss_local_dram.no_snoop_neededocr.pf_l1d_and_sw.l3_miss_remote_hop1_dram.any_snoopCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDocr.pf_l2_data_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000020Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.HITM_OTHER_CORE OCR.PF_L2_RFO.L3_MISS.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000020ocr.pf_l2_rfo.l3_miss_remote_hop1_dram.any_snoopocr.pf_l2_rfo.l3_miss_remote_hop1_dram.no_snoop_neededCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.all_data_rd.l3_miss_remote_hop1_dram.any_snoopoffcore_response.all_pf_data_rd.l3_miss.hit_other_core_no_fwdoffcore_response.all_pf_data_rd.l3_miss_remote_hop1_dram.any_snoopThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.other.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.ANY_SNOOPoffcore_response.pf_l1d_and_sw.l3_miss_local_dram.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_miss_remote_hop1_dram.snoop_missoffcore_response.pf_l3_data_rd.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSOCR.ALL_DATA_RD.ANY_RESPONSE have any response typeOCR.ALL_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWD  OCR.ALL_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDOCR.ALL_PF_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDED  OCR.ALL_PF_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.all_reads.any_responseOCR.ALL_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDED OCR.ALL_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDCounts all demand code reads OCR.DEMAND_CODE_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.demand_data_rd.pmm_hit_local_pmm.snoop_noneocr.demand_data_rd.supplier_none.hitm_other_coreCounts demand data reads  OCR.DEMAND_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.demand_data_rd.supplier_none.no_snoop_neededCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.PMM_HIT_LOCAL_PMM.ANY_SNOOPCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.SUPPLIER_NONE.HITM_OTHER_CORECounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.pf_l2_rfo.supplier_none.snoop_noneCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.SUPPLIER_NONE.ANY_SNOOPocr.pf_l3_rfo.supplier_none.snoop_missunc_m_pmm_rpq_insertsWrite requests allocated in the PMM Write Pending Queue for Intel Optane DC persistent memory. Unit: uncore_imc unc_m_pmm_bandwidth.writeIntel Optane DC persistent memory bandwidth write (MB/sec). Unit: uncore_imc All commands for Intel Optane DC persistent memoryevent=0xd3,umask=0x1Number of cycles a demand request has waited due to L1D Fill Buffer (FB) unavailabilityCounts number of phases a demand request has waited due to L1D Fill Buffer (FB) unavailablability. Demand requests include cacheable/uncacheable demand load, store, lock or SW prefetch accessesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C0002ocr.demand_rfo.l3_hit.snoop_hit_no_fwdCounts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that hit a cacheline in the L3 where a snoop was not needed to satisfy the requestocr.streaming_wr.l3_hit.anyCycles where at least 1 outstanding Demand RFO request is pending.   RFOs are initiated by a core as part of a data store operation.  Demand RFO requests include RFOs, locks, and ItoM transactions.  Requests are considered outstanding from the time they miss the core's L2 cache until the transaction completion message is sent to the requestorCounts retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 256 cycles which was not interrupted by a back-end stall (Precise event)Counts retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 64 cycles which was not interrupted by a back-end stall (Precise event)idq.dsb_cycles_anyC10 residency percent per packageevent=0xc9,period=100003,umask=0x8Counts the number of times a TSX Abort was triggered due to commit but Lock Buffer not emptyevent=0xc5,period=50021,umask=0x80Counts retired mispredicted indirect (near taken) CALL instructions, including both register and memory indirect (Precise event)event=0xc0,cmask=1,inv=1,period=1000003,umask=0x1Counts end of periods where the Reservation Station (RS) was emptyNumber of TMA slots that were wasted due to incorrect speculation by branch mispredictions. This event estimates number of operations that were issued but not retired from the specualtive path as well as the out-of-order engine recovery past a branch mispredictionuops_dispatched.port_0Number of uops executed on port 61 - CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_DISTRIBUTEDCache lines that are evicted by L2 cache when triggered by an L2 cache fillevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1030000001ocr.reads_to_core.remote_cache.snoop_hit_with_fwdocr.demand_data_rd.l3_miss_localCounts demand data reads that were supplied by DRAM attached to another socketCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were supplied by PMMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F3FFC0477event=0xb7,period=100003,umask=0x1,offcore_rsp=0x70C0004772LM Tag Check : Write Hit in Near Memory Cache. Unit: uncore_imc event=0x36,umask=0xC001FE01TOR Inserts : CRd_Prefs issued by iA Cores that hit the LLC. Unit: uncore_cha event=0x35,umask=0xC887FD01TOR Inserts : CRDs issued by iA Cores. Unit: uncore_cha unc_cha_tor_occupancy.ia_crdTOR Inserts : DRds issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha unc_cha_tor_inserts.io_hit_itomcachenearNumber requests PCIe makes of the main die : All. Unit: uncore_iio event=0x84,ch_mask=0x20,fc_mask=0x07,umask=0x01unc_iio_txn_req_of_cpu.mem_read.part7PCIe Completion Buffer Inserts of completions with data: Part 0-7. Unit: uncore_iio Tag Hit : Clean NearMem Read Hit. Unit: uncore_m2m uncore_m2pciemem_bound_stalls.loadmem_bound_stalls.load_llc_hitCounts the total number of load uops retired  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x201F803C0000event=0xe6,period=200003,umask=0x2Counts all code reads that were not supplied by the L3 cacheocr.full_streaming_wr.l3_miss_localevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000477c0_stalls.load_dram_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000000010000ocr.reads_to_core.outstandingCounts the number of unhalted core clock cycles. (Fixed event)Counts the number of issue slots every cycle that were not consumed by the backend due to fast nukes such as memory ordering and memory disambiguation machine clearsevent=0x71,period=1000003,umask=0x8dCounts the number of uops issued by the front end every cycle. When 4-uops are requested and only 2-uops are delivered, the event counts 2.  Uops_issued correlates to the number of ROB entries.  If uop takes 2 ROB slots it counts as 2 uops_issuedTOR Inserts : RFO_Prefs issued by iA Cores that Hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsunc_cha_tor_inserts.ia_miss_wcilfTOR Inserts : WCiLF issued by iA Cores that Missed the LLC. Unit: uncore_cha Data requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 7Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 1Data requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 5Number Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 7Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Counts the number of page walks completed due to store DTLB misses to a 1G pageevent=0x4f,period=2000003,umask=0x1Counts the number of first level TLB misses but second level hits due to an instruction fetch that did not start a page walk. Account for all pages sizes. Will results in a DTLB write from STLBFraction of branches that are taken conditionals. Unit: cpu_core cpu_atomPercent of instruction miss cost that hit in the L2. Unit: cpu_atom Inst_Miss_Cost_L3Hit_PercentCounts the number of cycles that uops are blocked due to an RSV full condition. Unit: cpu_atom mem_uops_retired.load_latency_gt_512Counts the number of stores uops retired. Counts with or without PEBS enabled (Precise event). Unit: cpu_atom Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 128 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core Counts the number of cycles that the head (oldest load) of the load buffer is stalled due to any number of reasons, including an L1 miss, WCB full, pagewalk, store address block or store data block, on a load that retires. Unit: cpu_atom Counts streaming stores that have any type of response. Unit: cpu_core Counts the number of machines clears due to memory renaming. Unit: cpu_atom event=0x75,period=200003,umask=0x2Counts the number of issue slots every cycle that were not consumed by the backend due to IEC or FPC RAT stalls, which can be due to FIQ or IEC reservation stalls in which the integer, floating point or SIMD scheduler is not able to accept uops. Unit: cpu_atom Counts the total number of issue slots every cycle that were not consumed by the backend due to frontend stalls. Unit: cpu_atom int_vec_retired.shufflesevent=0x3,period=100003,umask=0x4Number of clocks. Unit: uncore_imc MEM_STORE_RETIRED.L2_HITfp_arith_inst_retired2.complex_scalar_halfevent=0xcf,period=100003,umask=0x2FP_ARITH_INST_RETIRED2.COMPLEX_SCALAR_HALFevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x10004event=0x2a,period=100003,umask=0x1,offcore_rsp=0x708000004event=0x2a,period=100003,umask=0x1,offcore_rsp=0x104004477Number of uops dispatch to execution ports 2, 3 and 10unc_m_pmm_wpq_occupancy.all_sch1DRAM Precharge commands. : Precharge due to (?). Unit: uncore_imc DRAM RD_CAS and WR_CAS Commands. Unit: uncore_imc unc_m_pre_count.wr_pch1event=0x35,umask=0x00c817fe01TOR Inserts for ItoM from local IO. Unit: uncore_cha event=0x36,umask=0x00c8177e01RxQ Flit Buffer Allocations : Slot 1. Unit: uncore_upi event=0xc0,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000001event=0x18,umask=0x03event=0x20,umask=0x04unc_m2m_tracker_occupancy.ch1unc_m2m_prefcam_demand_drops.ch1_upiCache and Snoop Filter Lookups; Snoop Requests from a Remote Socket. Unit: uncore_cha unc_cha_tor_inserts.loc_ioevent=0x35,umask=0x00c827fe01TOR Occupancy; DRd hits from local IA. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_llcprefrfoevent=0x36,umask=0x0000000010unc_cha_tor_occupancy.rrqunc_cha_tor_occupancy.hitTOR Occupancy : Just ISOC. Unit: uncore_cha TOR Occupancy; RFO Pref hits from local IA. Unit: uncore_cha event=0x36,umask=0x00c8f3fe04unc_cha_tor_inserts.ia_miss_remote_wcilf_ddrevent=0x35,umask=0x00C86F8601event=0x35,umask=0x00C86F0601TOR Occupancy : WCiLFs issued by iA Cores targeting PMM that missed the LLC - HOMed locally. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_wcill2_request_g2.group1event=0x64,umask=0x01L2 prefetcher hits in L3. Counts all L2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit the L3Retired Taken Branch Instructions MispredictedSSE instructions (SSE, SSE2, SSE3, SSSE3, SSE4A, SSE41, SSE42, AVX)Total number multi-pipe uOps assigned to pipe 3fpu_pipe_assignment.dual0event=0,umask=0x08event=0x4,umask=0x04The number of serializing Ops retired. x87 control word mispredict traps due to mispredictions in RC or PC, or changes in mask bitsevent=0x5,umask=0x04L1 DTLB Reload of a page of 32K sizeevent=0x46,umask=0x03OC Mode Switch. IC to OC mode switchl3_cacheMicro-ops Retirednps1_die_to_dramMultiply-add FLOPS. Multiply-add counts as 2 FLOPS. This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15fp_ret_sse_avx_ops.div_flopsFloating Point Dispatch Faults. x87 fill faultls_st_commit_cancel2.st_commit_cancel_wcb_fullCycles where a dispatch group is valid but does not get dispatched due to a token stall. FP scheduler resource stall. Applies to ops that use the FP schedulerevent=0xae,umask=0x02event=0xccThe number of retired conditional branch instructions that were not correctly predicted because of a branch direction mismatchls_dmnd_fills_from_sys.mem_io_remotels_dmnd_fills_from_sys.ext_cache_localHardware Prefetch Data Cache Fills by Data Source. From Local L2 to the coreCycles where a dispatch group is valid but does not get dispatched due to a token stall. Insufficient Retire Queue tokens availableRetired unconditional indirect branch instructions mispredictedevent=0x5a,umask=0xdfL2 cache requests: data cache shared readsevent=0x60,umask=0xffCore to L2 cache requests (not including L2 prefetch) with status: data cache request miss in L2event=0x70,umask=0x20L2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache cache of all typesl3_lookup_state.l3_hitevent=0xac,umask=0x01l3_xi_sampled_latency.ext_farL3 cache fill requests sourced from another CCX's cache when the address was in the same NUMA node. Unit: amd_l3 ex_no_retire.thread_not_selectedWrite data beats (64 bytes) for local processor at Coherent Station (CS) 11event=0x15f,umask=0xbferemote_processor_read_data_beats_cs10remote_socket_upstream_read_beats_iom3local_socket_inf0_outbound_data_beats_ccm0event=0x49e,umask=0xbfeevent=0x51e,umask=0xbfeevent=0x59f,umask=0xbfeData beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 4Data beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 6event=0x3,umask=0x1fRetired x87 bottom-executing ops. Bottom-executing ops wait for all older ops to retire before executingevent=0xa,umask=0x03Retired vector floating-point logical opsfp_ops_retired_by_type.allRetired 128-bit packed floating-point compare opsevent=0xc,umask=0x0dfp_pack_ops_retired.fp256_addevent=0xc,umask=0x60Retired 128-bit packed integer SHA opsRetired 256-bit packed integer shift opsInstruction fetches that miss in the L1 ITLB but hit in the L2 ITLBInstruction fetches that miss in both the L1 and L2 ITLBs (page-table walks are requested) for 1G pagesFraction of dispatch slots that remained unused because of a bandwidth bottleneck in the frontend (such as decode or op cache fetch bandwidth)frontend_bound_bandwidthretiring * d_ratio(ex_ret_ucode_ops, ex_ret_ops)local_socket_inbound_data_to_cpuremote_socket_inf0_outbound_data_beats_ccm0 + remote_socket_inf1_outbound_data_beats_ccm0 + remote_socket_inf0_outbound_data_beats_ccm1 + remote_socket_inf1_outbound_data_beats_ccm1 + remote_socket_inf0_outbound_data_beats_ccm2 + remote_socket_inf1_outbound_data_beats_ccm2 + remote_socket_inf0_outbound_data_beats_ccm3 + remote_socket_inf1_outbound_data_beats_ccm3 + remote_socket_inf0_outbound_data_beats_ccm4 + remote_socket_inf1_outbound_data_beats_ccm4 + remote_socket_inf0_outbound_data_beats_ccm5 + remote_socket_inf1_outbound_data_beats_ccm5 + remote_socket_inf0_outbound_data_beats_ccm6 + remote_socket_inf1_outbound_data_beats_ccm6 + remote_socket_inf0_outbound_data_beats_ccm7 + remote_socket_inf1_outbound_data_beats_ccm7d_ratio(umc_cas_cmd.rd, umc_cas_cmd.all)(umc_cas_cmd.all * 64) / 1e6 / duration_timeDDRC write commands. Unit: hisi_sccl,ddrc uncore_imc_free_runningRET_STACK_MISPREDICTEVENT_12HEVENT_14HEVENT_23HEVENT_39HEVENT_61HEVENT_6BHEVENT_79HEVENT_A1HEVENT_D1HEVENT_E9HJAZELLE_BACKWARD_BRANCHISSUE_IS_EMPTYMAIN_EXECUTION_UNIT_PIPECID_WRITE_RETIREDMEMORY_ERRORST_RETIREDEVENT_147HEVENT_1B1HEVENT_1B6HEVENT_1B9HEVENT_203HEVENT_217HEVENT_218HEVENT_229HEVENT_250HEVENT_2A1HEVENT_2A3HEVENT_2C6HEVENT_2C8HEVENT_2FAHEVENT_312HEVENT_33FHEVENT_36AHEVENT_37AHEVENT_389HEVENT_3A3HEVENT_3E3HEVENT_3F0HBR_RETURN_SPECL2D_CACHE_ALLOCATEL3_CACHE_RDhnf_sfbi_brd_snp_senthnf_stash_data_pullhni_arready_no_arvalidrni_wrcancel_sentcxra_req_pcrd_stalls_lnk2cxra_ext_dat_stallclkdiv2_bk_fsm_trackerclk_requestMTVSCR_INSTR_COMPLETEDMTVRSAVE_INSTR_COMPLETEDLOAD_MISS_ALIASL1_DATA_TOUCH_HITL1_DATA_STORE_HITLSU_LMQ_INDEX_ALIASDISPATCHES_TO_FPR_ISSUE_QUEUEL1_DATA_TOUCH_MISSSWITCHES_BETWEEN_PRIV_USERVR_ISSUE_QUEUE_DISPATCHESL2_DATA_CACHE_MISSESPREFETCH_ENGINE_COLLISION_VS_STOREGCT_EMPTY_BY_SRQ_FULLMARKED_STORE_WITH_INTRFXU0_BUSY_FXU1_IDLEDATA_MMU_TLB4K_RELOADSBIU_MASTER_DATA_SIDE_CASTOUT_REQUESTSDVT7_DETECTEDFPU_FPSCR_FULL_STALLSYSTEMARMV7_CORTEX_A17DELETEDpmc_pmu_event_get_by_idxmetric_name: %s
LLC-MISS-RHITMGenuineIntel-6-5FGenuineIntel-6-57v2NO_NMI_WATCHDOGInstructions Per Cycle across hyper-threads (per physical core)Instructions per Branch (lower number means higher occurrence rate)DSB_Coverage64 * L1D.REPLACEMENT / 1000000000 / duration_timeMem;CacheMissesSocket actual clocks when any core is active on that socketevent=0x48,period=2000003,umask=0x1l2_rqsts.all_demand_data_rdevent=0x24,period=200003,umask=0x42event=0xd2,period=100003,umask=0x8Retired load uops which data sources were hits in L3 without snoops required. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM100 (Precise event)event=0xd0,period=100007,umask=0x21Offcore outstanding cacheable Core Data Read transactions in SuperQueue (SQ), queue to uncore  Spec update: BDM76event=0xc7,period=2000004,umask=0x3cevent=0xc7,period=2000003,umask=0x2This event counts x87 floating point (FP) micro-code assist (invalid operation, denormal operand, SNaN operand) when the input value (one of the source operands to an FP instruction) is invalididq.all_mite_cycles_4_uopsUops delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) pathidq.ms_dsb_cyclesidq.ms_dsb_uopsidq_uops_not_delivered.cycles_le_3_uop_deliv.coreNumber of times HLE abort was triggered (PEBS) (Precise event)Number of times a disallowed operation caused an HLE abortevent=0xc8,period=2000003,umask=0x1machine_clears.memory_orderingCounts the number of machine clears due to memory order conflictscpl_cycles.ring123event=0x63,period=2000003,umask=0x1number of near branch instructions retired that were mispredicted and taken. (Precise Event - PEBS) (Precise event)This is a precise version (that is, uses PEBS) of the event that counts mispredicted return instructions retiredcycle_activity.stalls_l1d_misscycle_activity.stalls_l1d_pendingThis event counts FP operations retired. For X87 FP operations that have no exceptions counting also includes flows that have several X87, or flows that use X87 uops in the exception handlingCycles when Resource Allocation Table (RAT) external stall is sent to Instruction Decode Queue (IDQ) for the threadevent=0xa8,period=2000003,umask=0x1This event counts both thread-specific (TS) and all-thread (AT) nukesCycles when Reservation Station (RS) is empty for the threadCycles per thread when uops are executed in port 4event=0xb1,period=2000003,umask=0x2uops_executed.cycles_ge_2_uops_execevent=0xb1,cmask=2,period=2000003,umask=0x1uops_executed_port.port_6_coreuops_executed_port.port_7uops_issued.slow_leaRetirement slots used. (Precise Event - PEBS) (Precise event)This event counts cycles without actually retired uopsunc_c_clockticksuncore memoryuncore_imcevent=0x8,period=100003,umask=0x1virtual memorydtlb_store_misses.stlb_hitStore misses that miss the  DTLB and hit the STLB (2M)Store miss in all TLB levels causes a page walk that completes. (4K)  Spec update: BDM69itlb_misses.stlb_hit_2mevent=0xbc,period=2000003,umask=0x18bdw metricsRetired load uops which data sources were L3 and cross-core snoop hits in on-pkg core cache  Supports address when precise.  Spec update: BDM100 (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020091offcore_response.all_pf_rfo.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0004offcore_response.pf_l2_code_rd.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020040offcore_response.pf_l2_code_rd.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020010Counts all prefetch (that bring data to L2) RFOsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0020offcore_response.pf_l3_code_rd.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0100offcore_response.all_data_rd.l3_hit.snoop_non_dramoffcore_response.all_rfo.l3_miss_local_dram.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0040offcore_response.pf_l2_code_rd.l3_miss_local_dram.snoop_hit_no_fwdoffcore_response.pf_l2_code_rd.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000010offcore_response.pf_l3_data_rd.l3_miss.snoop_noneoffcore_response.pf_l3_data_rd.l3_miss.snoop_not_neededoffcore_response.pf_l3_rfo.l3_miss_local_dram.snoop_hit_no_fwdevent=0x3c,any=1,period=100003,umask=0x1unc_cbo_cache_lookup.any_esL3 Lookup write request that access cache and found line in E or S-state. Unit: uncore_cbox Each cycle count number of all Core outgoing valid entries. Such entry is defined as valid from it's allocation till first of IDI0 or DRS0 messages is sent out. Accounts for Coherent and non-coherent traffic. Unit: uncore_arb Counts all demand & prefetch data reads hit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedevent=0x14l2_ifetch.self.i_statel2_lines_in.self.prefetchl2_st.self.mesisimd_inst_retired.packed_singleSIMD packed arithmetic micro-ops retiredevent=0xb3,period=2000000,umask=0x90event=0x7a,period=200000,umask=0x0All bus transactionsbus_trans_any.selfevent=0xc6,period=2000000,umask=0x1event=0x77,period=200000,umask=0x22snoop_stall_drv.all_agentsNumber of thermal tripsGood store forwardsevent=0xc,period=200000,umask=0x1Requests rejected by the L2Qevent=0xd1,period=200003,umask=0x1Load uops retired that missed L2 (Precise event capable)  Supports address when precise (Must be precise)Counts load uops retired where the data requested spans a 64 byte cache line boundary  Supports address when precise (Must be precise)offcore_response.any_pf_data_rd.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000022event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040008Counts the number of writeback transactions caused by L1 or L2 cache evictions that true miss for the L2 cache with a snoop miss in the other processor moduleoffcore_response.demand_data_rd.l2_miss.anyCounts demand cacheable data reads of full cache lines that are outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040002Counts demand reads for ownership (RFO) requests generated by a write to full data cache line that are outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.full_streaming_stores.l2_miss.hitm_other_coreCounts data cache line reads generated by hardware L1 data cache prefetcher that hit the L2 cacheoffcore_response.pf_l2_rfo.l2_miss.hitm_other_coreCycles the FP divide unit is busyCounts when a memory load of a uop spans a page boundary (a split) is retired (Must be precise)Cycles code-fetch stalled due to an outstanding ITLB missbr_misp_retired.non_return_indbr_misp_retired.returnCounts the number of core cycles while the core is not in a halt state.  The core enters the halt state when it is running the HLT instruction. In mobile systems the core frequency may change from time to time. For this reason this event may have a changing ratio with regards to time.  This event uses fixed counter 1.  You cannot collect a PEBs record for this eventThis event used to measure front-end inefficiencies. I.e. when front-end of the machine is not delivering uops to the back-end and the back-end has is not stalled. This event can be used to identify if the machine is truly front-end bound.  When this event occurs, it is an indication that the front-end of the machine is operating at less than its theoretical peak performance. Background: We can think of the processor pipeline as being divided into 2 broader parts: Front-end and Back-end. Front-end is responsible for fetching the instruction, decoding into uops in machine understandable format and putting them into a uop queue to be consumed by back end. The back-end then takes these uops, allocates the required resources.  When all resources are ready, uops are executed. If the back-end is not ready to accept uops from the front-end, then we do not want to count these as front-end bottlenecks.  However, whenever we have bottlenecks in the back-end, we will have allocation unit stalls and eventually forcing the front-end to wait until the back-end is ready to receive more uops. This event counts only when back-end is requesting more uops and front-end is not able to provide them. When 3 uops are requested and no uops are delivered, the event counts 3. When 3 are requested, and only 1 is delivered, the event counts 2. When only 2 are delivered, the event counts 1. Alternatively stated, the event will not count if 3 uops are delivered, or if the back end is stalled and not requesting any uops at all.  Counts indicate missed opportunities for the front-end to deliver a uop to the back end. Some examples of conditions that cause front-end efficiencies are: ICache misses, ITLB misses, and decoder restrictions that limit the front-end bandwidth. Known Issues: Some uops require multiple allocation slots.  These uops will not be charged as a front end 'not delivered' opportunity, and will be regarded as a back end problem. For example, the INC instruction has one uop that requires 2 issue slots.  A stream of INC instructions will not count as UOPS_NOT_DELIVERED, even though only one instruction can be issued per clock.  The low uop issue rate for a stream of INC instructions is considered to be a back end issueDuration of I-side pagewalks in cyclesCounts data reads generated by L1 or L2 prefetchers outstanding, per cycle, from the time of the L2 miss to when any response is receivedoffcore_response.any_read.outstandingCounts requests to the uncore subsystem hit the L2 cacheCounts reads for ownership (RFO) requests (demand & prefetch) true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of writeback transactions caused by L1 or L2 cache evictions true miss for the L2 cache with a snoop miss in the other processor moduleCounts data cache line reads generated by hardware L1 data cache prefetcher hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xc3,period=20003,umask=0x20ept.walk_pendingAny MLC or L3 HW prefetch accessing L2, including rejectsDemand Data Read requests sent to uncore  Spec update: HSD78, HSM80Cycles with any input/output SSE* or FP assistsicache.ifetch_stallUops not delivered to Resource Allocation Table (RAT) per thread when backend of the machine is not stalled  Spec update: HSD135Cycles with less than 3 uops delivered by the front end  Spec update: HSD135event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00091Counts all demand code reads miss the L3 and the data is returned from local dramCycles in which the L1D and L2 are locked, due to a UC lock or split lockThis event counts the number of thread cycles while the thread is not in a halt state. The thread enters the halt state when it is running the HLT instruction. The core frequency may change from time to time due to power or thermal throttlingCycles with pending memory loadsCycles where at least 1 uop was executed per-thread  Spec update: HSD144, HSD30, HSM31event=0xc2,any=1,cmask=1,inv=1,period=2000003,umask=0x1An external snoop hits a non-modified line in some processor core. Unit: uncore_cbox page_walker_loads.ept_dtlb_memoryevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0091Counts demand & prefetch RFOs that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresDecode Stream Buffer (DSB)-to-MITE switchesdsb_fill.exceed_dsb_lines( 1 * ( FP_COMP_OPS_EXE.SSE_SCALAR_SINGLE + FP_COMP_OPS_EXE.SSE_SCALAR_DOUBLE ) + 2 * FP_COMP_OPS_EXE.SSE_PACKED_DOUBLE + 4 * ( FP_COMP_OPS_EXE.SSE_PACKED_SINGLE + SIMD_FP_256.PACKED_DOUBLE ) + 8 * SIMD_FP_256.PACKED_SINGLE ) / CPU_CLK_UNHALTED.THREADevent=0xcd,period=2000003,umask=0x2offcore_response.all_code_rd.llc_miss.dramevent=0xbe,period=100003,umask=0x1Cycles with pending memory loads. Set AnyThread to count per coreevent=0x85,period=100003,umask=0x80Counts demand & prefetch data reads that hit in the LLC and sibling core snoop returned a clean responseoffcore_response.demand_code_rd.llc_miss.remote_dramoffcore_response.demand_code_rd.llc_miss.remote_hitmCounts prefetch (that bring data to L2) data reads that miss the LLC  and the data returned from remote & local dramllc_references.pcie_partial_readevent=0xdCounts the number of cycles that the uncore was running at a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band0=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Unit: uncore_pcu event=0xcCounts the number of cycles that the uncore was running at a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band3=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Unit: uncore_pcu event=0x7event=0x51,period=2000003,umask=0x8Cycles when dispatched loads are cancelled due to L1D bank conflicts with other load portsData from remote DRAM either Snoop not needed or Snoop Miss (RspI)offcore_requests_outstanding.demand_data_rd_c6event=0x9c,cmask=4,inv=1,period=2000003,umask=0x1event=0xb7,period=100003,umask=0x1,offcore_rsp=0x187FC20077event=0x5e,cmask=1,edge=1,inv=1,period=2000003,umask=0x1Counts the number of cycles that the uncore was running at a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band0=XXX with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Unit: uncore_pcu l2_requests.referenceCounts Demand code reads and prefetch code read requests  that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateCounts Demand code reads and prefetch code read requests  that accounts for responses which hit its own tile's L2 with data in F stateCounts Demand cacheable data and L1 prefetch data read requests  that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400070offcore_response.any_pf_l2.outstandingoffcore_response.any_read.l2_hit_near_tile_mCounts any Read request  that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts Demand cacheable data write requests  that accounts for responses which hit its own tile's L2 with data in F stateCounts Bus locks and split lock requests that accounts for responses which hit its own tile's L2 with data in F stateCounts demand cacheable data and L1 prefetch data reads that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts demand cacheable data and L1 prefetch data reads that accounts for responses which hit its own tile's L2 with data in E stateCounts Demand cacheable data writes that accounts for any responseoffcore_response.partial_writes.l2_hit_far_tileoffcore_response.partial_writes.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000100Counts L1 data HW prefetches that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400020event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010200This event counts the number of scalar SSE, AVX, AVX2, AVX-512 micro-ops retired (floating point, integer and store) except for loads (memory-to-register mov-type micro ops), division, sqrtoffcore_response.any_code_rd.ddr_nearCounts any Read request  that accounts for data responses from MCDRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200400offcore_response.demand_data_rd.ddrCounts demand cacheable data and L1 prefetch data reads that accounts for responses from DDR (local and far)Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses from MCDRAM (local and far)offcore_response.partial_reads.mcdram_nearoffcore_response.pf_l2_rfo.ddr_nearCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for data responses from MCDRAM Localoffcore_response.pf_software.ddr_nearCounts the number of near CALL branch instructions retired (Precise event)Counts the number of branch instructions retired that were near indirect CALL or near indirect JMP (Precise event)This event counts cycles when the divider is busy. More specifically cycles when the divide unit is unable to accept a new divide uop because it is busy processing a previously dispatched uop. The cycles will be counted irrespective of whether or not another divide uop is waiting to enter the divide unit (from the RS). This event counts integer divides, x87 divides, divss, divsd, sqrtss, sqrtsd event and does not count vector dividesevent=0x3,period=200003,umask=0x80recycleq.lockevent=0x63,period=2000000,umask=0x2l1d_cache_ld.s_stateL1D prefetch load lock accepted in fill bufferevent=0xf2,period=100000,umask=0x4event=0xf2,period=100000,umask=0x8event=0x24,period=200000,umask=0x80L2 demand store RFOs in S stateevent=0xb,period=5,umask=0x10,ldlat=0x4000Memory instructions retired above 4096 clocks (Precise Event)Load instructions retired with a data source of local DRAM or locally homed remote hitm (Precise Event)offcore_requests_sq_fulloffcore_response.any_data.remote_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7FFAll offcore RFO requestsoffcore_response.corewb.remote_cache_hitAll offcore code or data read requestsOffcore code or data read requests that HIT in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF03Offcore demand data requests satisfied by a remote cacheoffcore_response.demand_data_rd.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4701event=0xb7,period=100000,umask=0x1,offcore_rsp=0x1801offcore_response.demand_rfo.remote_cache_hitmAll offcore other requestsOffcore other requests that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF10offcore_response.pf_data_rd.llc_hit_no_other_coreoffcore_response.pf_ifetch.llc_hit_other_core_hitmOffcore prefetch code reads satisfied by the LLCOffcore prefetch code reads that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1020event=0xb7,period=100000,umask=0x1,offcore_rsp=0x270offcore_response.prefetch.local_cache_dramLoads delayed with at-Retirement block codetwo_uop_insts_decodedOffcore demand code reads satisfied by any DRAMevent=0x80,period=2000000,umask=0x1L1I Instruction fetchessnoop_response.hitmUnconditional branches executedevent=0x89,period=2000,umask=0x10lsd.inactiveuops_decoded.ms_cycles_activeuops_executed.core_stall_count_no_port5event=0xb1,period=2000000,umask=0x2uops_executed.port234_coreUops issuedevent=0xc2,period=2000000,umask=0x4uop_unfusionDTLB second level hitITLB missoffcore_response.demand_data_rd.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1001C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000001rtm_retired.aborted_memDirect and indirect near call instructions retired  Spec update: SKL091 (Precise event)Mispredicted direct and indirect near call instructions retired (Precise event)Number of times a microcode assist is invoked by HW other than FP-assist. Examples include AD (page Access Dirty) and AVX* related assistsCounts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 0Memory_BandwidthMem;MemoryLat;Offcore_SMT100 * (( BR_INST_RETIRED.CONDITIONAL + 3 * BR_INST_RETIRED.NEAR_CALL + (BR_INST_RETIRED.NEAR_TAKEN - ( BR_INST_RETIRED.CONDITIONAL - BR_INST_RETIRED.NOT_TAKEN ) - 2 * BR_INST_RETIRED.NEAR_CALL) ) / (4 * CPU_CLK_UNHALTED.THREAD))100 * (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * ( (ICACHE_64B.IFTAG_STALL / CPU_CLK_UNHALTED.THREAD) + (( ICACHE_16B.IFDATA_STALL + 2 * cpu@ICACHE_16B.IFDATA_STALL\,cmask\=1\,edge@ ) / CPU_CLK_UNHALTED.THREAD) + (9 * BACLEARS.ANY / CPU_CLK_UNHALTED.THREAD) ) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD))INST_RETIRED.ANY / MEM_INST_RETIRED.ALL_LOADSAverage number of Uops issued by front-end when it issued somethingFraction of branches that are CALL or RET1000 * MEM_LOAD_RETIRED.L1_MISS / INST_RETIRED.ANYCode miss in all TLB levels causes a page walk that completes. (1G)Counts any code reads (demand & prefetch) that hit in the other module where modified copies were found in other core's L1 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000044Counts data cacheline reads generated by L2 prefetchers that miss L2 with a snoop miss responseThis event counts the number of load uops reissued from Rehabqrehabq.st_splitsCounts the number of times entered into a ucode flow in the FEC.  Includes inserted flows due to front-end detected faults or assists.  Speculative countThis event counts the number of times that pipeline was cleared due to memory ordering issuesCounts the number of reference cycles while the core is not in a halt state. The core enters the halt state when it is running the HLT instruction. This event is a component in many key event ratios.  The core frequency may change from time. This event is not affected by core frequency changes but counts as if the core is running at the maximum frequency all the time.  Divide this event count by core frequency to determine the elapsed time while the core was not in halt state.  Divide this event count by core frequency to determine the elapsed time while the core was not in halt state.  This event is architecturally defined and is a designated fixed counter.  CPU_CLK_UNHALTED.CORE and CPU_CLK_UNHALTED.CORE_P use the core frequency which may change from time to time.  CPU_CLK_UNHALTE.REF_TSC and CPU_CLK_UNHALTED.REF are not affected by core frequency changes but counts as if the core is running at the maximum frequency all the time.  The fixed events are CPU_CLK_UNHALTED.CORE and CPU_CLK_UNHALTED.REF_TSC and the programmable events are CPU_CLK_UNHALTED.CORE_P and CPU_CLK_UNHALTED.REFMSROM micro-ops retiredRetired load uops with L2 cache hits as data sources. (Precise Event - PEBS) (Precise event)This event counts retired load uops that hit in the last-level (L3) cache without snoops required. (Precise Event - PEBS) (Precise event)Counts prefetch code reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts prefetch code reads that hit in the LLC and the snoops sent to sibling cores return clean responseCounts prefetch (that bring data to L2) code reads that hit in the LLC and the snoops sent to sibling cores return clean responseoffcore_response.pf_llc_code_rd.llc_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0200offcore_response.pf_llc_code_rd.llc_hit.no_snoop_neededevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff44offcore_response.demand_data_rd.all_local_dram_and_remote_cache_hitREQUEST = DEMAND_DATA_RD and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DEMAND_DATA_RD and RESPONSE = ANY_CACHE_DRAMREQUEST = DEMAND_RFO and RESPONSE = REMOTE_CACHE_HITMREQUEST = PF_DATA_RD and RESPONSE = ANY_CACHE_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3077REQUEST = DEMAND_IFETCH and RESPONSE = ANY_LLC_MISSREQUEST = DEMAND_RFO and RESPONSE = OTHER_LOCAL_DRAMoffcore_response.pf_data_rd.other_local_dramREQUEST = PF_DATA_RD and RESPONSE = REMOTE_DRAMevent=0xb4,period=100000,umask=0x1Snoop invalidate requestssnoopq_requests_outstanding.code_not_emptyevent=0xc5,period=20000,umask=0x1Mispredicted conditional retired branches (Precise Event)Offcore data reads, RFOs, and prefetches satisfied by the local DRAMCounts all prefetch data reads that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts demand data reads that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0001Counts all demand data writes (RFOs) that hit in the L3Counts L1 data cache hardware prefetch requests and software prefetch requests that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all prefetch (that bring data to LLC only) RFOs that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedfp_arith_inst_retired.512b_packed_doubleevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00122offcore_response.demand_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdCounts prefetch (that bring data to L2) data reads that miss the L3 and the modified data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00010offcore_response.pf_l2_rfo.l3_miss.any_snoopL2_Evictions_Silent_PKICORE_POWER.LVL0_TURBO_LICENSE / CPU_CLK_UNHALTED.THREAD( UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART0 + UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART1 + UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART2 + UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART3 ) * 4 / 1000000000 / duration_timeevent=0x35,umask=0x21,config1=0x41833event=0x50,umask=0x0CUPI interconnect send bandwidth for payload. Derived from unc_upi_txl_flits.all_data. Unit: uncore_upi PCI Express bandwidth reading at IIO, part 1. Unit: uncore_iio unc_iio_data_req_of_cpu.mem_write.part1PCI Express bandwidth writing at IIO, part 1. Unit: uncore_iio unc_cha_dir_update.torevent=0x54,umask=0x02Counts the total number of requests coming from a remote socket for exclusive ownership of a cache line without receiving data (INVITOE) to the CHAPCIe Completion Buffer occupancy of completions with data: Part 0-3. Unit: uncore_iio unc_iio_data_req_by_cpu.mem_read.part0unc_iio_data_req_by_cpu.mem_write.part3Write request of 4 bytes made to IIO Part3 by the CPU. Unit: uncore_iio unc_iio_data_req_by_cpu.peer_write.part0Peer to peer read request for 4 bytes made by IIO Part1 to an IIO target. Unit: uncore_iio Peer to peer write request of up to a 64 byte transaction is made to IIO Part0 by a different IIO unit. Unit: uncore_iio event=0x84,ch_mask=0x02,fc_mask=0x07,umask=0x04unc_iio_txn_req_of_cpu.mem_write.part0Peer to peer read request of up to a 64 byte transaction is made by IIO Part1 to an IIO target. Unit: uncore_iio Peer to peer read request of up to a 64 byte transaction is made by IIO Part2 to an IIO target. Unit: uncore_iio event=0x18event=0x11,umask=0x8Number of reads in which direct to Intel UPI transactions were overridden. Unit: uncore_m2m Counts when the M2M (Mesh to Memory) looks into the multi-socket cacheline Directory state , and found the cacheline marked in the I (Invalid) state indicating the cacheline is not stored in another socket, and so there is no need to snoop the other sockets for the latest data.  The data may be stored in any state in the local socketevent=0x38,umask=0x80FLITs that bypassed the TxL Buffer. Unit: uncore_upi mem_load_l3_miss_retired.remote_pmmocr.all_data_rd.l3_hit.hit_other_core_fwdOCR.ALL_DATA_RD.L3_HIT.SNOOP_NONE OCR.ALL_DATA_RD.L3_HIT.SNOOP_NONEocr.all_data_rd.l3_hit_e.any_snoopocr.all_data_rd.l3_hit_e.no_snoop_neededOCR.ALL_DATA_RD.L3_HIT_F.SNOOP_NONEocr.all_data_rd.l3_hit_m.hitm_other_coreOCR.ALL_DATA_RD.L3_HIT_S.HITM_OTHER_CORE  OCR.ALL_DATA_RD.L3_HIT_S.HITM_OTHER_COREOCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_MISS OCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200490OCR.ALL_PF_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDED  OCR.ALL_PF_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDOCR.ALL_PF_RFO.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_PF_RFO.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_PF_RFO.L3_HIT.NO_SNOOP_NEEDEDocr.all_pf_rfo.l3_hit_m.hitm_other_coreOCR.ALL_PF_RFO.L3_HIT_M.HITM_OTHER_CORE  OCR.ALL_PF_RFO.L3_HIT_M.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C07F7ocr.all_reads.l3_hit_e.hit_other_core_fwdocr.all_rfo.l3_hit.any_snoopocr.all_rfo.l3_hit_e.snoop_missocr.all_rfo.l3_hit_f.snoop_noneocr.all_rfo.l3_hit_s.hitm_other_coreocr.all_rfo.l3_hit_s.snoop_noneocr.demand_code_rd.l3_hit.any_snoopocr.demand_code_rd.l3_hit.no_snoop_neededocr.demand_data_rd.l3_hit_f.any_snoopCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDocr.demand_data_rd.l3_hit_f.snoop_noneocr.demand_data_rd.l3_hit_s.any_snoopCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_E.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_M.ANY_SNOOPocr.demand_rfo.l3_hit_m.snoop_missocr.other.l3_hit_e.hit_other_core_fwdocr.other.l3_hit_f.snoop_noneocr.other.l3_hit_m.hitm_other_coreocr.other.l3_hit_s.hitm_other_coreCounts any other requests  OCR.OTHER.L3_HIT_S.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200400Counts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_F.HITM_OTHER_COREocr.pf_l2_data_rd.l3_hit_f.hit_other_core_fwdCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.PF_L2_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDocr.pf_l2_rfo.l3_hit.no_snoop_neededCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_E.NO_SNOOP_NEEDEDocr.pf_l2_rfo.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200080Counts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.PF_L3_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDocr.pf_l3_rfo.l3_hit_e.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200100ocr.pf_l3_rfo.l3_hit_f.no_snoop_neededCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_M.NO_SNOOP_NEEDEDoffcore_response.all_data_rd.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020491offcore_response.all_pf_data_rd.l3_hit_e.any_snoopoffcore_response.all_pf_data_rd.l3_hit_e.snoop_noneoffcore_response.all_pf_data_rd.l3_hit_m.any_snoopThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_M.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDoffcore_response.all_reads.l3_hit.snoop_missThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_S.HITM_OTHER_COREoffcore_response.all_reads.supplier_none.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F800207F7offcore_response.all_rfo.l3_hit_e.hit_other_core_no_fwdoffcore_response.all_rfo.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_F.HITM_OTHER_COREoffcore_response.all_rfo.l3_hit_m.snoop_missoffcore_response.demand_code_rd.pmm_hit_local_pmm.snoop_not_neededoffcore_response.demand_code_rd.supplier_none.hitm_other_coreoffcore_response.demand_code_rd.supplier_none.hit_other_core_fwdoffcore_response.demand_data_rd.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_F.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDoffcore_response.demand_rfo.l3_hit_e.no_snoop_neededoffcore_response.demand_rfo.l3_hit_f.no_snoop_neededoffcore_response.demand_rfo.l3_hit_s.hitm_other_coreoffcore_response.demand_rfo.l3_hit_s.no_snoop_neededoffcore_response.demand_rfo.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.DEMAND_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_F.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_S.SNOOP_NONEoffcore_response.pf_l1d_and_sw.l3_hit_e.hit_other_core_no_fwdoffcore_response.pf_l1d_and_sw.l3_hit_s.hit_other_core_fwdoffcore_response.pf_l2_data_rd.l3_hit_m.no_snoop_neededoffcore_response.pf_l2_data_rd.l3_hit_s.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_F.HITM_OTHER_COREoffcore_response.pf_l2_rfo.l3_hit_f.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_M.HITM_OTHER_COREoffcore_response.pf_l3_data_rd.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREoffcore_response.pf_l3_data_rd.supplier_none.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_rfo.l3_hit_e.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDLSD_Coverageevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000491OCR.ALL_DATA_RD.L3_MISS.REMOTE_HIT_FORWARD OCR.ALL_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDOCR.ALL_DATA_RD.L3_MISS.SNOOP_MISS OCR.ALL_DATA_RD.L3_MISS.SNOOP_MISSOCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORE  OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREOCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDOCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDOCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORE  OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREOCR.ALL_PF_DATA_RD.L3_MISS.ANY_SNOOP OCR.ALL_PF_DATA_RD.L3_MISS.ANY_SNOOP OCR.ALL_PF_DATA_RD.L3_MISS.ANY_SNOOPocr.all_pf_data_rd.l3_miss.snoop_noneocr.all_pf_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000490ocr.all_pf_rfo.l3_miss.snoop_noneocr.all_pf_rfo.l3_miss_remote_hop1_dram.no_snoop_neededOCR.ALL_READS.L3_MISS.HITM_OTHER_CORE OCR.ALL_READS.L3_MISS.HITM_OTHER_CORE OCR.ALL_READS.L3_MISS.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8040007F7ocr.all_rfo.l3_miss.hitm_other_coreOCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORE  OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORECounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDocr.demand_data_rd.l3_miss_remote_hop1_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000002Counts any other requests OCR.OTHER.L3_MISS.REMOTE_HIT_FORWARDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604008000ocr.other.l3_miss_remote_hop1_dram.snoop_noneocr.pf_l1d_and_sw.l3_miss.snoop_missocr.pf_l1d_and_sw.l3_miss_local_dram.snoop_noneocr.pf_l1d_and_sw.l3_miss_remote_hop1_dram.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000010Counts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.pf_l2_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.pf_l3_data_rd.l3_miss_local_dram.any_snoopocr.pf_l3_data_rd.l3_miss_local_dram.snoop_missCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000080Counts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.SNOOP_MISSoffcore_response.all_data_rd.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.all_pf_data_rd.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.demand_data_rd.l3_miss.hit_other_core_fwdoffcore_response.other.l3_miss.remote_hitmThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l2_data_rd.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEoffcore_response.pf_l3_rfo.l3_miss.hitm_other_coreOCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOP OCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPocr.all_pf_data_rd.any_responseocr.all_pf_data_rd.supplier_none.hitm_other_coreOCR.ALL_PF_RFO.SUPPLIER_NONE.SNOOP_NONEOCR.ALL_READS.ANY_RESPONSE have any response typeocr.all_reads.supplier_none.snoop_missOCR.ALL_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDED  OCR.ALL_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.pf_l1d_and_sw.any_responseocr.pf_l1d_and_sw.supplier_none.snoop_noneevent=0xd3,umask=0x2UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART0 + UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART1 + UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART2 + UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART3l1d_pend_miss.fb_full_periodsevent=0x48,period=1000003,umask=0x1Counts the number of demand Data Read requests initiated by load instructions that hit L2 cacheevent=0x24,period=200003,umask=0x28Counts hardware prefetch data reads (which bring data to L2)  that hit a cacheline in the L3 where a snoop was sent or notocr.hwpf_l2_data_rd.l3_hit.snoop_hit_no_fwdDSB-to-MITE switch true penalty cyclesidq.dsb_cycles_okCounts the number of cycles uops were delivered to the Instruction Decode Queue (IDQ) from the MITE (legacy decode pipeline) path. During these cycles uops are not being delivered from the Decode Stream Buffer (DSB)event=0xc8,period=100003,umask=0x1event=0x54,period=100003,umask=0x1event=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000400Counts all the retired branch instructions that were mispredicted by the processor. A branch misprediction occurs when the processor incorrectly predicts the destination of the branch.  When the misprediction is discovered at execution, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct path (Precise event)br_misp_retired.cond_takenCounts all miss-predicted indirect branch instructions retired (excluding RETs. TSX aborts is considered indirect branch) (Precise event)Counts the number of available slots for an unhalted logical processor. The event increments by machine-width of the narrowest pipeline as employed by the Top-down Microarchitecture Analysis method. The count is distributed among unhalted logical processors (hyper-threads) who share the same physical coreCounts the number of page walks outstanding for a demand load in the PMH (Page Miss Handler) each cycleCounts demand instruction fetches and L1 instruction cache prefetches that hit a modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x808000001ocr.reads_to_core.remote_cache.snoop_fwdCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by a cache on a remote socket where a snoop hit in another core's caches which forwarded the unmodified data to the requesting coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F3FC00002ocr.itom.l3_miss_localevent=0xef,period=1000003,umask=0x40event=0xb7,period=100003,umask=0x1,offcore_rsp=0x73C000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x708000001DRAM Precharge commands. : Precharge due to page table. Unit: uncore_imc unc_m_pre_count.allDRAM Precharge commands. Unit: uncore_imc Local write requests that miss the SF/LLC and remote write requests sent to the CHA's home agent. Unit: uncore_cha TOR Occupancy : All requests from iA Cores that Hit the LLC. Unit: uncore_cha TOR Occupancy : All requests from iA Cores that Missed the LLC. Unit: uncore_cha event=0x36,umask=0xC001FD04event=0x35,umask=0xC88FFE01unc_cha_tor_inserts.ia_miss_drd_pref_remoteTOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_partial_streaming_wrevent=0x84,ch_mask=0x02,fc_mask=0x07,umask=0x80event=0x83,ch_mask=0x10,fc_mask=0x07,umask=0x80unc_iio_clockticks_freerunevent=0xd5,fc_mask=0x04,umask=0x20PCIe Completion Buffer Occupancy of completions with data : Part 3. Unit: uncore_iio Misc Events - Set 1 : Lost Forward. Unit: uncore_irp event=0x2d,umask=0x02Clockticks of the mesh to memory (M2M). Unit: uncore_m2m l2_request.allCounts the number of cycles the core is stalled due to an instruction cache or TLB miss which hit in DRAM or MMIO (Non-DRAM)mem_bound_stalls.ifetch_llc_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0044event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0001Counts uncached memory reads that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that were not supplied by the L3 cacheocr.uc_rd.l3_missbus_lock.self_locksevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000000000000044ocr.demand_data_and_l1pf_rd.any_responseThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.ANY_RESPONSEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800000010000Counts the number of retired loads that are blocked because it initially appears to be store forward blocked, but subsequently is shown not to be blocked based on 4K alias check (Precise event)Counts the number of issue slots every cycle that were not consumed by the backend due to certain allocation restrictionstopdown_fe_bound.allCounts the number of issue slots every cycle that were not delivered by the frontend due to the microcode sequencer (MS)TOR Inserts : DRd_Opt_Prefs issued by iA Cores that missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : All requests from IO Devices : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : PCIRdCurs issued by IO Devices : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsData requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 1Data requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 3Number Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Number Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 5FAF RF fullCounts the number of Extended Page Directory Entry missesInstructions per FP Arithmetic AVX* 256-bit instruction (lower number means higher occurrence rate). Unit: cpu_core L2 cache misses per kilo instruction for all request types (including speculative). Unit: cpu_core CPU_CLK_UNHALTED.CORE / INST_RETIRED.ANYPercentage of total non-speculative loads with a address aliasing block. Unit: cpu_atom Estimated_Pause_CostCounts the number of cycles the core is stalled due to a demand load which hit in the L2 cache. Unit: cpu_atom Counts the number of cycles the core is stalled due to a demand load which hit in the LLC or other core with HITE/F/M. Unit: cpu_atom mem_scheduler_block.allCounts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 8 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom L2 code requests. Unit: cpu_core Demand Data Read requests. Unit: cpu_core All L2 requests.[This event is alias to L2_REQUEST.ALL]. Unit: cpu_core RFO requests that hit L2 cache. Unit: cpu_core All retired store instructions  Supports address when precise (Precise event). Unit: cpu_core Demand and prefetch data reads. Unit: cpu_core Stalls caused by changing prefix length of the instruction. Unit: cpu_core Counts the number of cycles that the head (oldest load) of the load buffer is stalled due to a store address match when load subsequently retires. Unit: cpu_atom event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3FBFC00001Counts the number of issue slots every cycle that were not consumed by the backend due to scoreboards from the instruction queue (IQ), jump execution unit (JEU), or microcode sequencer (MS). Unit: cpu_atom Not taken branch instructions retired (Precise event). Unit: cpu_core Core cycles the allocator was stalled due to recovery from earlier clear event for this thread. Unit: cpu_core event=0xad,period=1000003,umask=0x10event=0xe7,period=1000003,umask=0xcevent=0xe7,period=1000003,umask=0x80int_vec_retired.vnni_128TMA slots wasted due to incorrect speculation by branch mispredictions. Unit: cpu_core event=0xb2,period=2000003,umask=0x10incoming write request page status is Page Empty. Unit: uncore_imc Counts the number of page walks completed due to store DTLB misses to any page size. Unit: cpu_atom All L2 requests.[This event is alias to L2_RQSTS.REFERENCES]event=0x2a,period=100003,umask=0x1,offcore_rsp=0x830000001event=0x2a,period=100003,umask=0x1,offcore_rsp=0x808000002Number of all Scalar Half-Precision FP arithmetic instructions(1) retired - regular and complexMEMORY_ACTIVITY.STALLS_L3_MISSevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x94000800event=0x2a,period=100003,umask=0x1,offcore_rsp=0x104000001AMX_OPS_RETIRED.INT8This event is deprecated. Refer to new event ARITH.IDIV_ACTIVEMiss-predicted near indirect branch instructions retired (excluding returns) (Precise event)INST_RETIRED.REP_ITERATIONunc_m_pmm_rpq_occupancy.all_sch0unc_m_pmm_wpq_occupancy.all_sch0event=0x3,umask=0x0000000010unc_m_cas_count.pch0event=0x84,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000002unc_upi_rxl_occupancy.slot0event=0xc1,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000001event=0xc1,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000004event=0xc2,ch_mask=0x02,fc_mask=0x07,umask=0x0000000004event=0x35,umask=0x00c001fe04event=0x35,umask=0x00cc43fe04event=0x35,umask=0x00C001FFC8event=0x35,umask=0x00c807ff01event=0x36,umask=0x00c001fe01TOR Occupancy; Misses from local IO. Unit: uncore_cha event=0x36,umask=0x00c803fe04event=0x36,umask=0x0000000002event=0x37,umask=0x0000000002TOR Inserts; LLCPrefCode hits from local IA. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_llcprefcodeevent=0x36,umask=0x00cccfff01event=0x35,umask=0x00c8170a01event=0x35,umask=0x00c8170601unc_cha_tor_inserts.io_clflushevent=0x36,umask=0x00c8968601unc_cha_tor_occupancy.ia_miss_crd_localTOR Occupancy : CLFlushOpts issued by iA Cores. Unit: uncore_cha event=0x36,umask=0x00cc27ff01unc_cha_tor_occupancy.ia_miss_ucrdfL1 BTB CorrectionAll L2 Cache Requests (Breakdown 1 - Common). Data cache reads (including hardware and software prefetch)All L2 Cache Requests (Breakdown 2 - Rare). Instruction cache read sized non-cacheableCore to L2 cacheable request access status (not including L2 Prefetch). Instruction cache request miss in L2The number of uOps retired. This includes all processor activity (instructions, exceptions, interrupts, microcode assists, etc.). The number of events logged per cycle can vary from 0 to 4event=0xc3Retired Conditional Branch Instructionsevent=0xd3fp_retx87_fp_ops.allDouble precision multiply-add FLOPS. Multiply-add counts as 2 FLOPSThis is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Double precision divide/square root FLOPSfp_num_mov_elim_scal_op.sse_mov_ops_elimevent=0x45,umask=0xffls_tablewalker.dsidels_pref_instr_disp.prefetch_ntaSoftware Prefetch Instructions (3DNow PREFETCHW instruction) DispatchedSoftware Prefetch Instructions Dispatched. Prefetch, Prefetch_T0_T1_T2ls_inef_sw_pref.mab_mch_cntevent=0xaf,umask=0x08Cycles where a dispatch group is valid but does not get dispatched due to a token stall. ALSQ 3_0 Tokens unavailable1core clocksNon-forwardable conflict; used to reduce STLI's via software. All reasons. Store To Load Interlock (STLI) are loads that were unable to complete because of a possible match with an older store, and the older store could not do STLF for some reasonNumber of SMIs receivedevent=0x2dAll L1 DTLB Misses or ReloadsL1 DTLB Miss. DTLB reload to a 1G page that miss in the L2 TLBL1 DTLB Miss. DTLB reload to a 2M page that miss in the L2 TLBls_hw_pf_dc_fill.ls_mabresp_rmt_dramde_dis_uop_queue_empty_di0Cycles where a dispatch group is valid but does not get dispatched due to a token stall. Store queue resource stall. Applies to all ops with store semanticsevent=0x18e,umask=0x18Software Prefetch Data Cache Fills by Data Source. From Local L2 to the coreHardware Prefetch Data Cache Fills by Data Source. From L3 or different L2 in same CCXCount of Allocated Mabsde_dis_dispatch_token_stalls1.int_phy_reg_file_rsrc_stallde_dis_dispatch_token_stalls2.int_sch0_token_stallCycles where a dispatch group is valid but does not get dispatched due to a token stall. No tokens for Integer Scheduler Queue 0 availableall_data_cache_accessesL1 Data Cache Fills: From External CCX CacheMiss Address Buffer (MAB) entries allocated by a Load-Store (LS) pipe for all types of allocationsAny data cache fills from either DRAM or MMIO in any NUMA node (same or different socket)ls_any_fills_from_sys.allevent=0x59,umask=0x80event=0x70,umask=0x02event=0x72,umask=0x04l2_pf_miss_l2_l3.allevent=0x15f,umask=0x7felocal_processor_write_data_beats_cs8Read data beats (64 bytes) for remote processor at Coherent Station (CS) 1remote_processor_read_data_beats_cs3remote_processor_read_data_beats_cs7Read data beats (64 bytes) for remote processor at Coherent Station (CS) 10remote_processor_write_data_beats_cs3event=0x81f,umask=0x7feRead data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 0event=0x85f,umask=0x7ffevent=0x85f,umask=0xbfeData beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 7event=0x51f,umask=0x7ffremote_socket_inf0_inbound_data_beats_ccm6Data beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 7remote_socket_inf0_outbound_data_beats_ccm0remote_socket_inf1_outbound_data_beats_ccm0Retired scalar floating-point add opsRetired scalar floating-point multiply-accumulate opsRetired scalar floating-point compare opsevent=0xa,umask=0x0eRetired vector floating-point subtract opsRetired MMX integer multiply opssse_avx_ops_retired.mmx_macsse_avx_ops_retired.sse_avx_addRetired SSE and AVX integer CLM opsRetired SSE and AVX integer shift opsfp_pack_ops_retired.fp128_mulRetired 256-bit packed floating-point multiply-accumulate opsevent=0xc,umask=0x80Retired 256-bit packed integer multiply-accumulate opsumc_act_cmd.wr4kB misaligned (page crossing) loadsNumber of cycles dispatch is stalled for integer scheduler queue 1 tokensPipelineL2;backend_bound_groupretiring_microcodeInstruction cache miss ratio for all fetches. An instruction cache miss will not be counted by this metric if it is an OC hitl1_dcacheL1 data cache fills from within the same CCXL1 instruction TLB missesdram_read_data_for_remote_processor3.0517578125e-5MiBmemory_controllerUNC_CBO_HYPHEN. Unit: uncore_cbox FR_RETIRED_TAKEN_BRANCHESFR_DISPATCH_STALL_WHEN_FAR_XFER_OR_RESYNC_BRANCH_PENDINGITLB_REFILLEXC_EXECUTEDAXI_WRITEMEM_UNALIGNED_ACCESS_REPLAYEVENT_24HEVENT_2CHEVENT_5FHEVENT_69HEVENT_75HEVENT_7AHEVENT_B6HEVENT_E3HEVENT_131HEVENT_17CHEVENT_182HEVENT_1A8HEVENT_1C9HEVENT_1D5HEVENT_230HEVENT_23BHEVENT_245HEVENT_26FHEVENT_270HEVENT_273HEVENT_2A7HEVENT_2BBHEVENT_2C1HEVENT_2CBHEVENT_2DAHEVENT_2DCHEVENT_332HEVENT_33BHEVENT_371HEVENT_378HEVENT_387HEVENT_3AFHEVENT_3C2HEVENT_3C8HEVENT_3D0HEVENT_3D7HL1D_CACHE_LDEXC_TRAP_OTHERRC_ST_SPEChnf_cache_fillhnf_intv_dirtyhni_nonpcie_serializationcxha_snppcrd_lnk0_stallTRUE_BRANCH_TARGET_HITSL1_DATA_SNOOP_HITSL1_DATA_LOAD_HITFP_STORE_INSTR_COMPLETED_IN_LSULSU_INDEXED_ALIAS_STALLLWARX_INSTR_COMPLETEDL1_DATA_CACHE_RELOADSSYNC_INSTR_COMPLETEDDST_STREAM_3_CACHE_LINE_FETCHESEXTERNAL_SNOOP_RETRYGROUP_MARKED_IDUPM_EVENT_TRANSITIONSLOAD_GUARDED_MISS_CYCLESPART2_MISALIGNED_CACHE_ACCESS_CYCLESL2MMU_MISSESSTWCX_SUCCESSESmultiply-pipe-excluding-junk-opssharedmemory-controller-hi-pri-bypassQUALIFIERK7BRANCH-INSTRUCTION-RETIREDLONGEST_LAT_CACHE.REFERENCEbr_inst_retired.all_branchesregex '%s' failed to compile, ignoring
%s, "tid": "%d"}
GenuineIntel-6-3Cv24v131 - ( (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) + (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD)) + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) )UPIUpTBILPBranch instructions per taken branch1000 * L2_RQSTS.MISS / INST_RETIRED.ANYGFLOPs1 - CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / ( CPU_CLK_UNHALTED.REF_XCLK_ANY / 2 ) if #SMT_on else 0Cycles Per Instruction for the Operating System (OS) Kernel modeAverage latency of data read request to external memory (in nanoseconds). Accounts for demand loads and L1/L2 prefetchesThis event counts duration of L1D miss outstanding in cyclesl2_rqsts.rfo_hitRFO requests that hit L2 cacheRFO requests that miss L2 cacheevent=0xf0,period=200003,umask=0x8l2_trans.rfoRetired load uops which data sources were L3 and cross-core snoop hits in on-pkg core cache. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM100 (Precise event)This is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were misses in the mid-level (L2) cache. Counting excludes unknown and UC data source  Supports address when precise (Precise event)This is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were data hits in the last-level (L3) cache without snoops required  Supports address when precise.  Spec update: BDM100 (Precise event)Miss in last-level (L3) cache. Excludes Unknown data-source. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM100, BDE70 (Precise event)event=0x60,period=2000003,umask=0x2Number of SSE/AVX computational double precision floating-point instructions retired. Applies to SSE* and AVX*scalar, double and single precision floating-point: ADD SUB MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per element.  ?event=0xc7,period=2000003,umask=0x3event=0xc7,period=2000005,umask=0x2aNumber of SIMD Move Elimination candidate uops that were not eliminatedevent=0x80,period=200003,umask=0x2This event counts the number of cycles  uops were  delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Counting includes uops that may bypass the IDQevent=0xc8,period=2000003,umask=0x10hle_retired.aborted_misc5br_inst_exec.all_branchesThis event counts taken speculative and retired macro-conditional branch instructions excluding calls and indirect branchesThis event counts taken speculative and retired indirect branches that have a return mnemonicevent=0xc5,period=400009,umask=0x20cpu_clk_thread_unhalted.ref_xclk_anyevent=0xa3,cmask=1,period=2000003,umask=0x1inst_retired.prec_distevent=0xc0,period=2000003,umask=0x1event=0x7,period=100003,umask=0x1load_hit_pre.sw_pfevent=0xc1,period=100003,umask=0x40Cycles per thread when uops are executed in port 3event=0xe,period=2000003,umask=0x40This is a precise version (that is, uses PEBS) of the event that counts all actually retired uops. Counting increments by two for micro-fused uops, and by one for macro-fused and other uops. Maximal increment value for one cycle is eight  Supports address when precise (Precise event)Number of cycles using always true condition (uops_ret < 16) applied to non PEBS uops retired eventevent=0x35,umask=0x3,filter_opc=0x18f,filter_nc=1event=0x35,umask=0x1,filter_opc=0x181L2 demand and L2 prefetch code references to LLC. Derived from unc_c_tor_inserts.opcode. Unit: uncore_cbox event=0x35,umask=0x1,filter_opc=0x18cevent=0x1,umask=0xCunc_p_prochot_external_cyclesevent=0x85,period=100003,umask=0x60Core misses that miss the  DTLB and hit the STLB (4K)event=0xbc,period=2000003,umask=0x24event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020091offcore_response.all_data_rd.supplier_none.snoop_missoffcore_response.all_pf_code_rd.l3_hit.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010040offcore_response.pf_l2_data_rd.supplier_none.snoop_hitmoffcore_response.pf_l3_code_rd.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020200Number of SSE/AVX computational scalar double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementoffcore_response.all_data_rd.l3_miss_local_dram.snoop_noneoffcore_response.corewb.l3_miss.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000004offcore_response.demand_data_rd.supplier_none.snoop_non_dramoffcore_response.other.l3_miss.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C008000offcore_response.pf_l2_rfo.l3_miss_local_dram.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000100This event counts taken branch instructions retired (Precise event)unc_cbo_cache_lookup.any_mesiL3 Lookup write request that access cache and found line in E or S-stateunc_arb_trk_requests.drd_directNumber of Writes allocated - any write transactions: full/partials writes and evictions( ITLB_MISSES.WALK_DURATION + DTLB_LOAD_MISSES.WALK_DURATION + DTLB_STORE_MISSES.WALK_DURATION + 7 * ( DTLB_STORE_MISSES.WALK_COMPLETED + DTLB_LOAD_MISSES.WALK_COMPLETED + ITLB_MISSES.WALK_COMPLETED ) ) / ( 2 * CPU_CLK_UNHALTED.THREAD )event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00100l1d_cache.ldevent=0x28,period=200000,umask=0x44event=0x28,period=200000,umask=0x48l2_ld.self.demand.m_stateevent=0x24,period=200000,umask=0x70event=0x30,period=200000,umask=0x74L2 store requestsevent=0x2a,period=200000,umask=0x4fevent=0x2a,period=200000,umask=0x48Retired computational Streaming SIMD Extensions 2 (SSE2) scalar-double instructionsevent=0xc7,period=2000000,umask=0x10event=0xb1,period=2000000,umask=0x0Icache missuops.ms_cyclesevent=0x5,period=200000,umask=0x97event=0x5,period=200000,umask=0x9misalign_mem_ref.st_splitprefetch.software_prefetch.arevent=0x68,period=200000,umask=0x40cycles_div_busyevent=0xdc,period=2000000,umask=0x2All store forwardsevent=0x82,period=200000,umask=0x2event=0xcb,period=200000,umask=0x4page_walks.i_side_cyclesCounts data reads (demand & prefetch) that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_pf_data_rd.l2_hitCounts data reads generated by L1 or L2 prefetchers that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts the number of writeback transactions caused by L1 or L2 cache evictions that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000004Counts demand cacheable data reads of full cache lines that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.demand_data_rd.outstandingCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that hit the L2 cacheCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that miss the L2 cacheCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredStore uops that split a page (Precise event capable) (Must be precise)Retired near call instructions (Precise event capable) (Must be precise)Counts near indirect CALL branch instructions retired (Must be precise)event=0xc4,period=200003,umask=0xfdStore uops retired that missed the DTLB (Precise event capable)  Supports address when precise (Must be precise)offcore_response.any_read.any_responseCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts any data writes to uncacheable write combining (USWC) memory region  outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache lines requests by software prefetch instructions true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000001000Counts data cache lines requests by software prefetch instructions outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0x49,period=2000003,umask=0x2Number of instruction fetches that hit the L2 cacheNumber of instruction fetches that missed the L2 cacheThis event counts requests originating from the core that reference a cache line in the last level cacheNumber of SIMD FP assists due to output valuesNumber of front end re-steers due to BPU mispredictionIncrement each cycle. # of uops delivered to IDQ from DSB path. Set Cmask = 1 to count cyclesoffcore_response.all_reads.l3_miss.any_responseCounts the number of conditional branch instructions retired (Precise event)Reference cycles when the thread is unhalted. (counts at 100 MHz rate)unc_cbo_xsnp_response.miss_externalCycles with at least one request outstanding is waiting for data return from memory controller. Account for coherent and non-coherent requests initiated by IA Cores, Processor Graphics Unit, or LLC. Unit: uncore_arb ITLB misses that hit STLB (2M)This event counts cycles when the  page miss handler (PMH) is servicing page walks caused by ITLB missesevent=0xbc,period=2000003,umask=0x82offcore_response.demand_data_rd.llc_miss.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00080event=0x24,period=200003,umask=0x1Counts all demand & prefetch prefetch RFOsCounts number of X87 uops executedoffcore_response.demand_code_rd.llc_miss.dramuops_dispatched_port.port_1_coreCycles per thread when uops are dispatched to port 5A snoop misses in some processor core. Unit: uncore_cbox LLC lookup request that access cache and found line in I-stateevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0010event=0x36,umask=0x2Aunc_p_freq_band3_cyclesfreq_ge_3000mhz_cycles %Number of GSSE-256 Computational FP single precision uops issued this cycleThis event counts the number of instruction cache, streaming buffer and victim cache misses. Counting includes unchacheable accessesHardware Prefetch requests that miss the L1D cache. This accounts for both L1 streamer and IP-based (IPP) HW prefetchers. A request is being counted each time it access the cache & miss it, including if a block is applicable or if hit the Fill Buffer for This event counts the number of the divide operations executedMultiply packed/scalar single precision uops allocatedThis event counts the number of cycles with at least one slow LEA uop being allocated. A uop is generally considered as slow LEA if it has three sources (for example, two sources and immediate) regardless of whether it is a result of LEA instruction or not. Examples of the slow LEA uop are or uops with base, index, and offset source operands using base and index reqisters, where base is EBR/RBP/R13, using RIP relative or 16-bit addressing modes. See the Intel? 64 and IA-32 Architectures Optimization Reference Manual for more details about slow LEA instructionsCycles with either free list is emptyevent=0x5b,period=2000003,umask=0xfThis event counts the number of retirement slots used each cycle.  There are potentially 4 slots that can be used each cycle - meaning, 4 micro-ops or 4 instructions could retire each cycle.  This event is used in determining the 'Retiring' category of the Top-Down pipeline slots characterization (Precise event)Counts the number of load micro-ops retired that caused micro TLB missCounts the matrix events specified by MSR_OFFCORE_RESPxoffcore_response.any_code_rd.l2_hit_near_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000044Counts Demand cacheable data and L1 prefetch data read requests  that accounts for responses which hit its own tile's L2 with data in F stateoffcore_response.any_pf_l2.l2_hit_far_tileCounts any request that accounts for responses which hit its own tile's L2 with data in E stateoffcore_response.any_rfo.l2_hit_far_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400400offcore_response.partial_writes.l2_hit_near_tile_moffcore_response.pf_l1_data_rd.l2_hit_near_tile_moffcore_response.pf_l2_rfo.l2_hit_near_tile_moffcore_response.pf_l2_rfo.l2_hit_this_tile_eCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000020uops_retired.scalar_simdCounts the number of times the MSROM starts a flow of uopsCounts Demand code reads and prefetch code read requests  that accounts for responses from DDR (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180603091offcore_response.any_pf_l2.mcdramCounts Demand cacheable data write requests  that accounts for data responses from DRAM Localoffcore_response.demand_code_rd.mcdram_faroffcore_response.demand_rfo.ddr_farCounts L2 code HW prefetches that accounts for responses from DDR (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400040offcore_response.pf_l2_rfo.mcdramevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000200Counts the number of mispredicted near indirect CALL branch instructions retired (Precise event)Counts the number of mispredicted branch instructions retired that were near indirect CALL or near indirect JMP (Precise event)recycleq.any_stevent=0x3,umask=0x01event=0x51,period=2000000,umask=0x1event=0x40,period=2000000,umask=0x4l1d_cache_lock.m_stateL1 data cache stores in M statel2_data_rqsts.anyL2 data demand loads in E stateL2 data demand loads in S stateevent=0xf2,period=100000,umask=0xfL2 instruction fetch missesl2_transactions.rfomem_inst_retired.latency_above_threshold_4event=0xcb,period=200000,umask=0x2Offcore requests blocked due to Super Queue fullevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8FFoffcore_response.any_rfo.remote_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4708offcore_response.corewb.remote_cache_dramOffcore writebacks that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x277Offcore request = all data, response = any locationevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4703event=0xb7,period=100000,umask=0x1,offcore_rsp=0x230Offcore prefetch data requests satisfied by a remote cache or remote DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4740event=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF20event=0xb7,period=100000,umask=0x1,offcore_rsp=0x820offcore_response.prefetch.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F70Offcore prefetch requests that HIT in a remote cacheevent=0x10,period=2000000,umask=0x40fp_mmx_trans.to_fp128 bit SIMD integer pack operationsevent=0xa6,period=2000000,umask=0x1event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4011offcore_response.corewb.remote_dramOffcore demand RFO requests satisfied by a remote DRAMOffcore prefetch data requests that missed the LLCoffcore_response.pf_data.local_dramOffcore prefetch RFO requests satisfied by a remote DRAMLoads dispatched that bypass the MOBpartial_address_aliasarith.divevent=0x88,period=20000,umask=0x4event=0x88,period=200000,umask=0x7Instructions that must be decoded by decoder 0resource_stalls.storeCycles Uops were issued on either threadevent=0xc2,period=2000000,umask=0x1Counts L1D data line replacements including opportunistic replacements, and replacements that require stall-for-replace or block-for-replaceNumber of times a request needed a FB (Fill Buffer) entry but there was no entry available for it. A request includes cacheable/uncacheable demands that are load, store or SW prefetch instructionsCounts the number of lines that are silently dropped by L2 cache when triggered by an L2 cache fill. These lines are typically in Shared or Exclusive state. A non-threaded eventCounts the number of demand Data Read requests (including requests from L1D hardware prefetchers). These loads may hit or miss L2 cache. Only non rejected loads are countedAll retired memory instructions  Supports address when precise (Precise event)Retired load instructions that split across a cacheline boundary  Supports address when precise (Precise event)Retired load instructions that miss the STLB  Supports address when precise (Precise event)mem_load_l3_hit_retired.xsnp_hitCounts the demand RFO (read for ownership) requests including regular RFOs, locks, ItoMoffcore_response.demand_code_rd.l3_hit_e.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC01C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040001Counts all demand data writes (RFOs)have any response typeoffcore_response.demand_rfo.l3_hit_e.snoop_missoffcore_response.demand_rfo.l3_hit_e.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0040002offcore_response.demand_rfo.supplier_none.snoop_not_neededoffcore_response.other.l3_hit_m.snoop_not_neededsw_prefetch_access.ntaCounts once for most SIMD 128-bit packed computational single precision floating-point instruction retired. Counts twice for DPP and FM(N)ADD/SUB instructions retiredCounts once for most SIMD scalar computational double precision floating-point instructions retired. Counts twice for DPP and FM(N)ADD/SUB instructions retiredCounts cycles with any input and output SSE or x87 FP assist. If an input and output assist are detected on the same cycle the event increments by 1Cycles where a code fetch is stalled due to L1 instruction cache tag missCounts randomly selected loads when the latency from first dispatch to completion is greater than 64 cycles  Supports address when precise (Must be precise)offcore_requests.l3_miss_demand_data_rdCycles with at least 1 Demand Data Read requests who miss L3 cache in the superQoffcore_response.demand_code_rd.l3_miss.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC4000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20001C8000offcore_response.other.l3_miss.snoop_non_dramNumber of times a transactional abort was signaled due to a data capacity limitation for transactional reads or writesevent=0x9,period=2000003,umask=0x1event=0xc5,period=400009,umask=0x2event=0x3c,cmask=1,edge=1,period=100007Cycles total of 2 uops are executed on all ports and Reservation Station was not emptyCycles total of 3 uops are executed on all ports and Reservation Station was not emptyCounts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 4Counts the number of Blend Uops issued by the Resource Allocation Table (RAT) to the reservation station (RS) in order to preserve upper bits of vector registers. Starting with the Skylake microarchitecture, these Blend uops are needed since every Intel SSE instruction executed in Dirty Upper State needs to preserve bits 128-255 of the destination register. For more information, refer to Mixing Intel AVX and Intel SSE Code section of the Optimization GuideBad;BadSpec;BrMispredicts_SMTarb@event\=0x80\,umask\=0x2@ / arb@event\=0x80\,umask\=0x2\,cmask\=1@Each cycle count number of all Core outgoing valid entries. Such entry is defined as valid from its allocation till first of IDI0 or DRS0 messages is sent out. Accounts for Coherent and non-coherent traffic. Unit: uncore_arb Page walk completed due to a demand data store to a 1G pagePage walk completed due to a demand data store to a 2M/4M pageCode miss in all TLB levels causes a page walk that completes. (All page sizes)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000044Counts any request that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts demand and DCU prefetch instruction cacheline that miss L2rehabq.any_strehabq.sta_fullThis event counts the number of instructions that retire.  For instructions that consist of multiple micro-ops, this event counts exactly once, as the last micro-op of the instruction retires.  The event continues counting while instructions retire, including during interrupt service routines caused by hardware interrupts, faults or traps.  Background: Modern microprocessors employ extensive pipelining and speculative techniques.  Since sometimes an instruction is started but never completed, the notion of "retirement" is introduced.  A retired instruction is one that commits its states. Or stated differently, an instruction might be abandoned at some point. No instruction is truly finished until it retires.  This counter measures the number of completed instructions.  The fixed event is INST_RETIRED.ANY and the programmable event is INST_RETIRED.ANY_PCounts the number of cycles the Alloc pipeline is stalled when any one of the RSs (IEC, FPC and MEC) is full. This event is a superset of all the individual RS stall event countsRetired load uops with unknown information as data source in cache serviced the load. (Precise Event - PEBS) (Precise event)Retired store uops that miss the STLB. (Precise Event - PEBS) (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0244offcore_response.all_pf_rfo.llc_hit.no_snoop_neededCounts prefetch RFOs that hit in the LLC and the snoops sent to sibling cores return clean responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0122offcore_response.all_rfo.llc_hit.snoop_missCounts all prefetch (that bring data to L2) RFOs that hit in the LLCCounts prefetch (that bring data to LLC only) RFOs that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all prefetch RFOs that miss the LLC  and the data returned from dramREQUEST = DEMAND_IFETCH and RESPONSE = LLC_MISS_LOCAL and SNOOP = DRAMevent=0xb0,period=100000,umask=0x10REQUEST = ANY_DATA read and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = CORE_WB and RESPONSE = ANY_CACHE_DRAMREQUEST = CORE_WB and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = DATA_IN and RESPONSE = LLC_HIT_NO_OTHER_COREREQUEST = DEMAND_DATA and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f04offcore_response.other.all_local_dram_and_remote_cache_hitREQUEST = OTHER and RESPONSE = ANY_LOCATIONREQUEST = PF_DATA_RD and RESPONSE = LLC_HIT_NO_OTHER_COREREQUEST = PF_DATA_RD and RESPONSE = LLC_HIT_OTHER_CORE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff20REQUEST = PF_IFETCH and RESPONSE = LOCAL_CACHEevent=0xf4,period=2000000,umask=0x4offcore_response.any_request.any_dram_and_remote_fwdREQUEST = ANY_REQUEST and RESPONSE = OTHER_LOCAL_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x20ffoffcore_response.data_ifetch.other_local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3033REQUEST = DEMAND_DATA_RD and RESPONSE = ANY_DRAM AND REMOTE_FWDevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3002REQUEST = OTHER and RESPONSE = REMOTE_DRAMREQUEST = PREFETCH and RESPONSE = ANY_DRAM AND REMOTE_FWDsnoopq_requests.codeevent=0x8,period=200000,umask=0x80Offcore data reads, RFOs, and prefetches satisfied by the IO, CSR, MMIO unitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF50Retired load instructions which data sources missed L3 but serviced from local dram  Supports address when precise (Precise event)mem_load_l3_miss_retired.remote_fwdoffcore_response.all_pf_data_rd.l3_hit.hitm_other_coreCounts prefetch RFOs that have any response typeCounts prefetch RFOs that hit in the L3offcore_response.pf_l3_data_rd.l3_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000120offcore_response.all_pf_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdoffcore_response.all_rfo.l3_miss.any_snoopoffcore_response.demand_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdoffcore_response.demand_rfo.l3_miss.snoop_miss_or_no_fwdoffcore_response.pf_l1d_and_sw.l3_miss.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000080offcore_response.pf_l3_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdCounts all prefetch (that bring data to LLC only) RFOs that miss the L3 and the data is returned from remote dram( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE) ) / ( 2 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ) )Average latency of data read request to external DRAM memory [in nanoseconds]. Accounts for demand loads and L1/L2 data-read prefetchesunc_m_rpq_insertsevent=0x83,ch_mask=0x08,fc_mask=0x07,umask=0x04Counts only multi-socket cacheline Directory state updates due to memory writes issued from the TOR pipe which are the result of remote transaction hitting the SF/LLC and returning data Core2Core. This does not include memory write requests which are for I (Invalid) or E (Exclusive) cachelinesunc_cha_misc.rfo_hit_sPCIe Completion Buffer Inserts of completions with data: Part 2. Unit: uncore_iio event=0xd5,fc_mask=0x04,umask=0x08Counts every peer to peer write request of 4 bytes of data made to the MMIO space of a card on IIO Part2 by a different IIO unit. Does not include requests made by the same IIO unit. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busCounts every peer to peer read request for 4 bytes of data made by IIO Part0 to the MMIO space of an IIO target. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busevent=0xc1,ch_mask=0x02,fc_mask=0x07,umask=0x01Counts reads in which direct to core transactions (which would have bypassed the CHA) were overriddenCounts reads in which direct to Intel Ultra Path Interconnect (UPI) transactions (which would have bypassed the CHA) were overriddenunc_m2m_imc_reads.normalevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100491ocr.all_pf_data_rd.l3_hit_e.any_snoopocr.all_pf_data_rd.l3_hit_f.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200490ocr.all_pf_data_rd.l3_hit_s.no_snoop_neededOCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_FWDOCR.ALL_PF_RFO.L3_HIT.SNOOP_HIT_WITH_FWDocr.all_pf_rfo.l3_hit_f.snoop_missOCR.ALL_PF_RFO.L3_HIT_S.HITM_OTHER_CORE  OCR.ALL_PF_RFO.L3_HIT_S.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100120ocr.all_reads.l3_hit.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C07F7ocr.all_reads.l3_hit_f.no_snoop_neededOCR.ALL_RFO.L3_HIT.SNOOP_NONE OCR.ALL_RFO.L3_HIT.SNOOP_NONEOCR.ALL_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWD  OCR.ALL_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDocr.all_rfo.l3_hit_s.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100122ocr.all_rfo.l3_hit_s.hit_other_core_no_fwdocr.all_rfo.l3_hit_s.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080004Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_S.ANY_SNOOPCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C8000ocr.other.l3_hit.no_snoop_neededocr.other.l3_hit_s.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080400ocr.pf_l1d_and_sw.l3_hit_s.any_snoopocr.pf_l1d_and_sw.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0010Counts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDocr.pf_l2_data_rd.l3_hit_m.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080080Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDocr.pf_l3_rfo.l3_hit.any_snoopCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_HIT.HIT_OTHER_CORE_FWD OCR.PF_L3_RFO.L3_HIT.HIT_OTHER_CORE_FWDoffcore_response.all_data_rd.l3_hit_f.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_M.ANY_SNOOPoffcore_response.all_data_rd.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400491This event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_NONEoffcore_response.all_pf_data_rd.l3_hit_e.hit_other_core_fwdoffcore_response.all_pf_data_rd.l3_hit_m.hitm_other_coreoffcore_response.all_pf_rfo.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_S.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400120This event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_E.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_M.NO_SNOOP_NEEDEDoffcore_response.all_reads.l3_hit_m.snoop_missThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_M.SNOOP_MISSoffcore_response.all_reads.l3_hit_s.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_data_rd.l3_hit_s.hitm_other_coreoffcore_response.demand_rfo.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_F.NO_SNOOP_NEEDEDoffcore_response.demand_rfo.l3_hit_f.snoop_missThis event is deprecated. Refer to new event OCR.DEMAND_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_RFO.SUPPLIER_NONE.SNOOP_MISSoffcore_response.other.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_S.HIT_OTHER_CORE_FWDoffcore_response.other.supplier_none.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.pf_l1d_and_sw.l3_hit_f.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_F.SNOOP_NONEoffcore_response.pf_l1d_and_sw.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020400offcore_response.pf_l2_data_rd.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_E.HITM_OTHER_COREoffcore_response.pf_l2_rfo.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_F.SNOOP_MISSoffcore_response.pf_l3_data_rd.l3_hit_m.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400080This event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000490OCR.ALL_PF_DATA_RD.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_PF_DATA_RD.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_PF_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDOCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDOCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDocr.all_pf_rfo.l3_miss.remote_hit_forwardOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDED  OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDOCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDED  OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDOCR.ALL_READS.L3_MISS.ANY_SNOOP OCR.ALL_READS.L3_MISS.ANY_SNOOP OCR.ALL_READS.L3_MISS.ANY_SNOOPOCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C0007F7ocr.all_rfo.l3_miss_local_dram.hitm_other_coreCounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.HITM_OTHER_CORE OCR.DEMAND_CODE_RD.L3_MISS.HITM_OTHER_COREocr.demand_code_rd.l3_miss.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000001Counts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.DEMAND_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC08000Counts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.ANY_SNOOP OCR.PF_L1D_AND_SW.L3_MISS.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000400ocr.pf_l1d_and_sw.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPocr.pf_l1d_and_sw.l3_miss_remote_hop1_dram.snoop_missocr.pf_l2_data_rd.l3_miss.hit_other_core_fwdocr.pf_l2_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDocr.pf_l2_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000080This event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.SNOOP_MISSoffcore_response.all_reads.l3_miss.any_snoopThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.REMOTE_HITMoffcore_response.all_rfo.l3_miss_remote_hop1_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDoffcore_response.demand_rfo.l3_miss_remote_hop1_dram.snoop_missoffcore_response.other.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.SNOOP_MISSoffcore_response.pf_l1d_and_sw.l3_miss_remote_hop1_dram.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.pf_l2_rfo.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.pf_l3_data_rd.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.pf_l3_rfo.l3_miss_remote_hop1_dram.hitm_other_coreoffcore_response.pf_l3_rfo.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.pf_l3_rfo.l3_miss_remote_hop1_dram.snoop_missOCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HITM_OTHER_CORE  OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREOCR.ALL_PF_RFO.SUPPLIER_NONE.HITM_OTHER_CORE  OCR.ALL_PF_RFO.SUPPLIER_NONE.HITM_OTHER_COREocr.all_pf_rfo.supplier_none.snoop_noneocr.all_rfo.supplier_none.hit_other_core_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.SUPPLIER_NONE.ANY_SNOOPocr.demand_rfo.supplier_none.hit_other_core_no_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.other.supplier_none.no_snoop_neededCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.SUPPLIER_NONE.HITM_OTHER_CORECounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.pf_l3_data_rd.pmm_hit_local_pmm.snoop_noneocr.pf_l3_data_rd.supplier_none.snoop_missocr.pf_l3_rfo.pmm_hit_local_pmm.any_snoopCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDevent=0x48,cmask=1,edge=1,period=1000003,umask=0x2event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1E003C0002Counts hardware prefetch RFOs (which bring data to L2) that hit a cacheline in the L3 where a snoop was sent or not1 - ( (BR_INST_RETIRED.COND_NTAKEN / BR_INST_RETIRED.ALL_BRANCHES) + (BR_INST_RETIRED.COND_TAKEN / BR_INST_RETIRED.ALL_BRANCHES) + (( BR_INST_RETIRED.NEAR_CALL + BR_INST_RETIRED.NEAR_RETURN ) / BR_INST_RETIRED.ALL_BRANCHES) + ((BR_INST_RETIRED.NEAR_TAKEN - BR_INST_RETIRED.COND_TAKEN - 2 * BR_INST_RETIRED.NEAR_CALL) / BR_INST_RETIRED.ALL_BRANCHES) )event=0x60,cmask=1,period=1000003,umask=0x10Counts the number of times RTM commit succeededCounts the number of times a class of instructions that may cause a transactional abort was executed inside a transactional regionCounts taken branch instructions retired (Precise event)event=0xc5,period=50021Counts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to ports 5 and 9MemoryBound;MemoryLatevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1008000002C1_Core_Residencyevent=0xef,period=1000003,umask=0x1event=0xb7,period=100003,umask=0x1,offcore_rsp=0x708000002ocr.reads_to_core.snc_dramCounts retired mispredicted indirect (near taken) calls, including both register and memory indirect (Precise event)All DRAM CAS commands issued. Unit: uncore_imc DRAM Clockticks. Unit: uncore_imc unc_m_wpq_occupancy_pch1PMM Read Pending Queue Occupancy. Unit: uncore_imc PMM Commands : Reads - RPQ. Unit: uncore_imc event=0x36,umask=0xC001FD01unc_cha_cms_clockticksCMS Clockticks. Unit: uncore_cha event=0x35,umask=0xC897FF01event=0x36,umask=0xC816FE01unc_cha_tor_inserts.ia_miss_drd_remoteTOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices that hit the LLC. Unit: uncore_cha TOR Inserts; WCiLF misses from local IA. Unit: uncore_cha event=0x83,ch_mask=0x10,fc_mask=0x07,umask=0x01event=0x83,ch_mask=0x80,fc_mask=0x07,umask=0x04unc_iio_txn_req_by_cpu.mem_write.part5unc_iio_txn_req_by_cpu.mem_write.part7event=0x1f,umask=0x10unc_i_coherent_ops.wbmtoievent=0x11,umask=0x08unc_i_snoop_resp.all_hit_mCounts the number of load uops retired that hit in the L3 cache (Precise event)ocr.corewb_m.l3_hitCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedocr.hwpf_l2_data_rd.l3_hitCounts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were supplied by the L3 cacheCounts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedocr.uc_rd.l3_hit.snoop_hitmCounts uncached memory reads that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1001003C0000Counts the number of floating point operations retired that required microcode assistCounts the number of times a decode restriction reduces the decode throughput due to wrong instruction length predictionCounts the number of machine clears due to memory ordering caused by a snoop from an external agent. Does not count internally generated machine clears such as those due to memory disambiguationCounts uncached memory reads that were not supplied by the L3 cacheocr.partial_streaming_wr.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10470Counts the total number of issue slots every cycle that were not consumed by the backend due to backend stallsevent=0x71,period=1000003,umask=0x80Counts the number of times the LLC was accessed - this includes code, data, prefetches and hints coming from L2.  This has numerous filters available.  Note the non-standard filtering equation.  This event will count requests that lookup the cache multiple times with multiple increments.  One must ALWAYS set umask bit 0 and select a state or states to match.  Otherwise, the event will count nothing.   CHAFilter0[24:21,17] bits correspond to [FMESI] state. Read transactionsunc_cha_tor_occupancy.ia_drd_optData requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Number Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 7Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 7Counts the number of first level TLB misses but second level hits due to loads that did not start a page walk. Account for all pages sizes. Will result in a DTLB write from STLBevent=0x49,period=2000003,umask=0x80Instructions per FP Arithmetic AVX/SSE 128-bit instruction (lower number means higher occurrence rate). Unit: cpu_core Average data fill bandwidth to the L1 data cache [GB / sec]. Unit: cpu_core INST_RETIRED.ANY / CPU_CLK_UNHALTED.CORE100 * MEM_BOUND_STALLS.IFETCH_L2_HIT / ( MEM_BOUND_STALLS.IFETCH )Inst_Miss_Cost_L2Hit_PercentNumber of cycles a demand request has waited due to L1D Fill Buffer (FB) unavailability. Unit: cpu_core Read requests with true-miss in L2 cache.[This event is alias to L2_REQUEST.MISS]. Unit: cpu_core event=0x20,period=1000003,umask=0x8event=0x40,period=100003,umask=0x8arith.fpdiv_activefp_arith_dispatched.port_0Retired instructions after front-end starvation of at least 1 cycle (Precise event). Unit: cpu_core icache_data.stallsCounts the number of instructions retired. (Fixed event) (Precise event). Unit: cpu_atom Mispredicted indirect CALL retired (Precise event). Unit: cpu_core Core cycles when the thread is not in halt state. Unit: cpu_core event=0xc0,period=2000003,umask=0x10int_vec_retired.add_128Cycles when at least one PMH is busy with a page walk for a demand load. Unit: cpu_core Page walks completed due to a demand data store to a 4K page. Unit: cpu_core Instruction fetch requests that miss the ITLB and hit the STLB. Unit: cpu_core Code miss in all TLB levels causes a page walk that completes. (All page sizes). Unit: cpu_core Counts read requests of any type with true-miss in the L2 cache. True-miss excludes L2 misses that were merged with ongoing L2 misses.[This event is alias to L2_RQSTS.MISS]Retired load instructions with local Intel Optane DC persistent memory as the data source where the data request missed all caches  Supports address when preciseOFFCORE_REQUESTS.ALL_REQUESTSfp_arith_inst_retired2.256b_packed_halfevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x70C004477event=0x2a,period=100003,umask=0x1,offcore_rsp=0x703004477Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by PMM attached to another socketevent=0xec,period=2000003,umask=0x10INT_VEC_RETIRED.256BITNumber of uops dispatch to execution ports 4 and 9unc_m_cas_count.rd_pre_regValid Flits Sent : Slot 1. Unit: uncore_upi unc_upi_txl_flits.llctrlunc_upi_rxl_flits.dataevent=0x32,umask=0x0000000002unc_upi_txl_occupancyevent=0x42event=0x84,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000004event=0xc1,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000004unc_m2m_prefcam_demand_drops.ch0_upiunc_m2m_prefcam_demand_merge.upi_allchevent=0x50,umask=0x000000000cWrite requests made into the CHA. Unit: uncore_cha event=0x35,umask=0x0000000020event=0x35,umask=0x0000000040TOR Inserts : Just Remote Targets. Unit: uncore_cha TOR Occupancy; Misses from Local IA. Unit: uncore_cha TOR Occupancy : All from Local iA. Unit: uncore_cha event=0x36,umask=0x00C000FF05TOR Occupancy : Just Hits. Unit: uncore_cha unc_cha_tor_occupancy.ia_hit_rfo_prefevent=0x36,umask=0x00cc43ff04event=0x36,umask=0x00c88fff01unc_cha_tor_occupancy.ia_miss_rfo_remoteTOR Inserts; RdCur and FsRdCur misses from local IO. Unit: uncore_cha unc_cha_tor_occupancy.ia_llcprefcodeunc_cha_tor_inserts.ia_miss_drd_pref_ddrTOR Inserts : ItoMs issued by iA Cores that Hit LLC. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_remote_wcil_pmmunc_cha_tor_occupancy.ia_miss_drd_local_pmmunc_cha_tor_occupancy.ia_miss_drd_pref_local_ddrTOR Occupancy : CRd_Prefs issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha TOR Occupancy : WCiLF issued by iA Cores that Missed the LLC. Unit: uncore_cha TOR Occupancy : WCiLFs issued by iA Cores targeting DDR that missed the LLC - HOMed locally. Unit: uncore_cha TOR Occupancy : WCiLs issued by iA Cores targeting PMM that missed the LLC - HOMed remotely. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_remote_wcil_ddrunc_cha_tor_occupancy.io_clflushevent=0x91Decoder Overrides Existing Branch Prediction (speculative)ic_fetch_stall.ic_stall_dq_emptyInstruction Pipe Stall. IC pipe was stalled during this clock cycle (including IC to OC fetches) due to back-pressureevent=0x8c,umask=0x02The number of ITLB reload requestsAll L2 Cache Requests (Breakdown 1 - Common). Instruction cache readsevent=0x61,umask=0x40Total cycles spent waiting for L2 fills to complete from L3 or memory, divided by four. Event counts are for both threads. To calculate average latency, the number of fills from both threads must be usedevent=0x63,umask=0x01Core to L2 cacheable request access status (not including L2 Prefetch). Instruction cache hit modifiable line in L2Core to L2 cacheable request access status (not including L2 Prefetch). Instruction cache requests in L2L2 prefetch hit in L2. Use l2_cache_hits_from_l2_hwpf insteadevent=0x1cf,umask=0x02Unit: amd_df fp_retx87_fp_ops.div_sqr_r_opsDivide and square root Opsevent=0x3,umask=0xffx87 control word mispredict traps due to mispredictions in RC or PC, or changes in mask bitsSSE control word mispredict traps due to mispredictions in RC, FTZ or DAZ, or changes in mask bitsevent=0x5,umask=0x01ls_mab_alloc.loadsls_l1_d_tlb_miss.tlb_reload_1g_l2_missbp_l1_tlb_miss_l2_tlb_miss.if1gTotal number uOps assigned to pipe 2Divide/square root FLOPS. This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15event=0x24,umask=0x02ls_smi_rxls_refills_from_sys.ls_mabresp_lcl_l2L1 DTLB Miss. DTLB reload to a 1G page that hit in the L2 TLBevent=0x4b,umask=0xffCycles where a dispatch group is valid but does not get dispatched due to a token stall. Integer Scheduler miscellaneous resource stallDispatch of a single op that performs a memory load. Counts the number of operations dispatched to the LS unit. Unit Masks ADDedls_mab_alloc.all_allocationsls_mab_alloc.hardware_prefetcher_allocationsevent=0x41,umask=0x40ls_dmnd_fills_from_sys.mem_io_localls_dmnd_fills_from_sys.int_cacheDemand Data Cache Fills by Data Source. From L3 or different L2 in same CCXls_sw_pf_dc_fills.int_cacheRetired taken branch instructions (all types of architectural control flow changes, including exceptions and interrupts)Retired indirect branch instructions mispredicted (only EX mispredicts). Each misprediction incurs the same penalty as a mispredicted conditional branch instructionex_ret_uncond_brnch_instr_mispredRetired unconditional branch instructionsevent=0x43,umask=0x80Any data cache fills from L3 cache or different L2 cache in the same CCXSoftware prefetch instructions dispatched (speculative) of type PrefetchT0 (move data to all cache levels), T1 (move data to all cache levels except L1) and T2 (move data to all cache levels except L1 and L2)Software prefetches that did not fetch data outside of the processor core as the PREFETCH instruction saw a match on an already allocated Miss Address Buffer (MAB)ls_sw_pf_dc_fills.dram_io_farSoftware prefetch data cache fills from all types of data sourcesls_hw_pf_dc_fills.dram_io_farCore to L2 cache requests (not including L2 prefetch) with status: instruction cache hit non-modifiable line in L2l2_pf_hit_l2.l2_next_lineInstruction cache missesL3 cache fill requests sourced from DRAM in a different NUMA node. Unit: amd_l3 Core cycles not in haltlocal_processor_read_data_beats_cs1Read data beats (64 bytes) for local processor at Coherent Station (CS) 5event=0x9f,umask=0xbffWrite data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 2event=0x89f,umask=0xbfelocal_socket_inf1_outbound_data_beats_ccm1Data beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 3local_socket_inf1_outbound_data_beats_ccm4event=0x41f,umask=0xbfeData beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 3event=0xcdf,umask=0xf3efp_ops_retired_by_type.scalar_divfp_ops_retired_by_type.vector_mulRetired vector floating-point multiply-accumulate opsfp_ops_retired_by_type.vector_cvtRetired vector floating-point convert opsfp_ops_retired_by_type.vector_otherevent=0xb,umask=0x0aevent=0xb,umask=0x20sse_avx_ops_retired.sse_avx_aesevent=0xb,umask=0xe0Retired 128-bit packed floating-point logical opsevent=0xd,umask=0xc0packed_int_op_type.int256_logicalOps dispatched from loop bufferNumber of cycles dispatch is stalled for Load queue token6 * ls_not_halted_cycd_ratio(bad_speculation * resyncs_or_nc_redirects, ex_ret_brn_misp + resyncs_or_nc_redirects)Fraction of dispatch slots used by microcode ops that retiredAll L2 cache misses(l3_xi_sampled_latency.all * 10) / l3_xi_sampled_latency_requests.alll1_demand_data_cache_fills_from_local_l2local_processor_read_data_beats_cs0 + local_processor_read_data_beats_cs1 + local_processor_read_data_beats_cs2 + local_processor_read_data_beats_cs3 + local_processor_read_data_beats_cs4 + local_processor_read_data_beats_cs5 + local_processor_read_data_beats_cs6 + local_processor_read_data_beats_cs7 + local_processor_read_data_beats_cs8 + local_processor_read_data_beats_cs9 + local_processor_read_data_beats_cs10 + local_processor_read_data_beats_cs11remote_socket_upstream_read_beats_iom0 + remote_socket_upstream_read_beats_iom1 + remote_socket_upstream_read_beats_iom2 + remote_socket_upstream_read_beats_iom3remote_socket_inf0_inbound_data_beats_ccm0 + remote_socket_inf1_inbound_data_beats_ccm0 + remote_socket_inf0_inbound_data_beats_ccm1 + remote_socket_inf1_inbound_data_beats_ccm1 + remote_socket_inf0_inbound_data_beats_ccm2 + remote_socket_inf1_inbound_data_beats_ccm2 + remote_socket_inf0_inbound_data_beats_ccm3 + remote_socket_inf1_inbound_data_beats_ccm3 + remote_socket_inf0_inbound_data_beats_ccm4 + remote_socket_inf1_inbound_data_beats_ccm4 + remote_socket_inf0_inbound_data_beats_ccm5 + remote_socket_inf1_inbound_data_beats_ccm5 + remote_socket_inf0_inbound_data_beats_ccm6 + remote_socket_inf1_inbound_data_beats_ccm6 + remote_socket_inf0_inbound_data_beats_ccm7 + remote_socket_inf1_inbound_data_beats_ccm7event=0x34LS_MICROARCHITECTURAL_RESYNC_BY_SNOOPDC_MICROARCHITECTURAL_EARLY_CANCELFR_INTERRUPTS_MASKED_CYCLESFR_DISPATCH_STALL_WHEN_REORDER_BUFFER_IS_FULLFR_DISPATCH_STALL_WHEN_RESERVATION_STATIONS_ARE_FULLFR_DISPATCH_STALL_WHEN_FPU_IS_FULLFR_NUMBER_OF_BREAKPOINTS_FOR_DR2L1_ICACHE_REFILLL2_CACHE_REFILLL2_CACHE_MISSL1_ICACHE_HASH_MISSEVENT_2FHEVENT_3CHEVENT_40HEVENT_6DHEVENT_7DHPLE_FIFO_OVERFLOWEVENT_130HEVENT_166HEVENT_16FHEVENT_17BHEVENT_18DHEVENT_1C1HEVENT_1E8HEVENT_1EFHEVENT_216HEVENT_21DHEVENT_22EHEVENT_256HEVENT_272HEVENT_28CHEVENT_294HEVENT_29AHEVENT_2A5HEVENT_2FCHEVENT_321HEVENT_338HEVENT_352HEVENT_355HEVENT_374HEVENT_39DHEVENT_3B3HEVENT_3F7Hdn_rxreq_dvmopsbsx_txdat_stallrni_txreq_flits_totalrni_rdb_unordcxra_req_pcrd_stalls_lnk0cxla_rx_tlp_link0IU2_INSTR_COMPLETEDLSU_TOUCH_ALIAS_VS_CSQL3_TOUCH_HITSBUS_RETRYPREFETCH_ENGINE_COLLISION_VS_LOAD_STORE_INSTR_FETCHBUS_LOWCRIT_INPUT_INTR_PENDING_LATENCY_CYCLESSTORE_RETRIESL2_CACHE_DATA_HITSFPU_DOUBLE_PUMPFPU_DIVIDE_CYCLESFPU_PIPE_SYNC_STALLSstore-pipe-excluding-junk-opsaccesseslow-op-pos-0sse-and-x87-microtrapsRUNNINGps->ps_len == 0fc_mask{"type": "procfork"GenuineIntel-6-3Dv8GenuineIntel-6-85GenuineIntel-6-2CGenuineIntel-6-7DBackend_Bound_SMTCPIExecute_per_IssueBpTkBranchC6 residency percent per packageL2 cache lines filling L2L2 prefetch requests that hit L2 cacheL2 prefetch requests that miss L2 cacheevent=0xf0,period=200003,umask=0x80L2 cache accesses when fetching instructionsevent=0xf0,period=200003,umask=0x1event=0xf0,period=200003,umask=0x20event=0xd3,period=100007,umask=0x20event=0xd1,period=50021,umask=0x10event=0xb2,period=2000003,umask=0x1This event counts any input SSE* FP assist - invalid operation, denormal operand, dividing by zero, SNaN operand. Counting includes only cases involving penalties that required micro-code assist interventionfp_assist.x87_inputevent=0x79,cmask=1,period=2000003,umask=0x10Counts cycles FE delivered 4 uops or Resource Allocation Table (RAT) was stalling FECycles with less than 2 uops delivered by the front endevent=0xc8,period=2000003,umask=0x8Number of times HLE commit succeededmisalign_mem_ref.storesNumber of times RTM commit succeededNumber of times a TSX line had a cache conflictevent=0x5c,period=2000003,umask=0x1All (macro) branch instructions retired. (Precise Event - PEBS)  Spec update: BDW98 (Must be precise)event=0xc4,period=400009,umask=0x10Taken speculative and retired mispredicted indirect branches with return mnemonicCycles while memory subsystem has an outstanding loadload_hit_pre.hw_pfNot software-prefetch load dispatches that hit FB allocated for software prefetchNumber of integer Move Elimination candidate uops that were eliminatedevent=0xa2,period=2000003,umask=0x10This event counts ROB full stall cycles. This counts cycles that the pipeline backend blocked uop delivery from the front endevent=0x5e,period=2000003,umask=0x1Cycles per thread when uops are executed in port 1uops_executed.core_cycles_ge_3Cycles per core when uops are exectuted in port 5read requests to local home agent. Unit: uncore_ha unc_m_pre_count.page_missPre-charges due to page misses. Unit: uncore_imc Pre-charge for reads. Unit: uncore_imc This event counts store misses in all DTLB levels that cause page walks of any page size (4K/2M/4M/1G)  Spec update: BDM69event=0x49,period=100003,umask=0x40Operations that miss the first ITLB level but hit the second and do not cause any page walksCounts the number of demand Data Read requests, initiated by load instructions, that hit L2 cacheevent=0x24,period=200003,umask=0xd0This event counts line-splitted load uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K)  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0091offcore_response.all_data_rd.l3_hit.snoop_hit_no_fwdoffcore_response.all_data_rd.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0120offcore_response.all_rfo.l3_hit.snoop_noneoffcore_response.demand_code_rd.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020040offcore_response.pf_l2_data_rd.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0100offcore_response.pf_l3_rfo.l3_hit.snoop_not_neededoffcore_response.pf_l3_rfo.supplier_none.snoop_hit_no_fwdoffcore_response.pf_l3_rfo.supplier_none.snoop_not_neededCounts randomly selected loads with latency value being above 16  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)Counts randomly selected loads with latency value being above 32  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)offcore_response.all_pf_code_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000090offcore_response.all_rfo.l3_miss_local_dram.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000122offcore_response.pf_l2_data_rd.l3_miss.snoop_hit_no_fwdoffcore_response.pf_l3_data_rd.l3_miss_local_dram.snoop_missevent=0x3c,period=100003,umask=0x2unc_cbo_cache_lookup.any_mL3 Lookup read request that access cache and found line in any MESI-stateunc_clock.socketCounts all data/code/rfo reads (demand & prefetch) miss in the L3Counts all data/code/rfo reads (demand & prefetch) miss the L3 and the modified data is transferred from remote cacheCounts all data/code/rfo reads (demand & prefetch) miss the L3 and clean or shared data is transferred from remote cacheQPI clock ticks. Unit: uncore_qpi (UNC_M_POWER_CRITICAL_THROTTLE_CYCLES / UNC_M_CLOCKTICKS) * 100.event=0x29,period=200000,umask=0x58event=0x26,period=200000,umask=0x70l2_lines_out.self.demandl2_reject_busq.self.demand.e_statel2_st.self.e_stateLoad splits (At Retirement)Bus cycles while processor receives databus_drdy_clocks.this_agentHIT signal assertedevent=0xc4,period=200000,umask=0x8Retired branch instructions that were mispredicted takenevent=0xc4,period=2000000,umask=0xcdiv.arevent=0xc,period=2000000,umask=0x1event=0xc,period=200000,umask=0x2event=0xd1,period=200003,umask=0x40Counts memory uops retired where the data requested spans a 64 byte cache line boundary  Supports address when precise (Must be precise)Counts data reads (demand & prefetch) that hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x00000432b7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x02000032b7Counts requests to the uncore subsystem that true miss for the L2 cache with a snoop miss in the other processor moduleCounts reads for ownership (RFO) requests (demand & prefetch) that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.corewb.l2_miss.snoop_miss_or_no_snoop_neededevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000004offcore_response.demand_data_rd.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000001Counts demand cacheable data reads of full cache lines that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.demand_rfo.l2_miss.hit_other_core_no_fwdCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000002offcore_response.partial_streaming_stores.l2_miss.anyCounts data cache line reads generated by hardware L1 data cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts data cacheline reads generated by hardware L2 cache prefetcher that true miss for the L2 cache with a snoop miss in the other processor moduleCounts reads for ownership (RFO) requests generated by L2 prefetcher that hit the L2 cacheCounts the number of core cycles during which interrupts are masked (disabled). Increments by 1 each core cycle that EFLAGS.IF is 0, regardless of whether interrupts are pending or notCycles pending interrupts are maskedevent=0xc4,period=200003,umask=0xbfbr_inst_retired.taken_jccevent=0xc5,period=200003,umask=0xebLoads blocked due to store data not ready (Precise event capable) (Must be precise)Counts load uops retired that caused a DTLB miss  Supports address when precise (Must be precise)Counts data reads generated by L1 or L2 prefetchers hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data reads generated by L1 or L2 prefetchers miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache lines requests by software prefetch instructions have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache lines requests by software prefetch instructions hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts once per cycle for each page walk occurring due to a demand data store. Includes cycles spent traversing the Extended Page Table (EPT). Average cycles per walk can be calculated by dividing by the number of walksPage walk completed due to an instruction fetch in a 4K pageRetired load uops which data sources were L3 hit and cross-core snoop missed in on-pkg core cache  Supports address when precise.  Spec update: HSD29, HSD25, HSM26, HSM30 (Precise event)event=0xd1,period=100003,umask=0x20Retired store uops that split across a cacheline boundary  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)Data read requests sent to uncore (demand and prefetch)hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all demand data writes (RFOs) hit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedNumber of transitions from AVX-256 to legacy SSE when penalty applicable  Spec update: HSD56, HSM57Counts cycles MITE is delivered four uops. Set Cmask = 4Counts demand data reads miss in the L3Counts the number of near return instructions retired (Precise event)Execution stalls due to L1 data cache miss loads. Set Cmask=0CHCycles at least 2 micro-op is executed from any thread on physical core  Spec update: HSD30, HSM31This events counts the cycles where at least two uop were executed. It is counted per thread  Spec update: HSD144, HSD30, HSM31Cycles per core when uops are executed in port 5L3 Lookup write request that access cache and found line in I-state. Unit: uncore_cbox event=0x22,umask=0x21Code miss in all TLB levels causes a page walk that completes. (4K)event=0xbd,period=100003,umask=0x20offcore_response.demand_code_rd.llc_hit.hitm_other_coreoffcore_response.demand_code_rd.llc_miss.local_dramevent=0x27,period=200003,umask=0x8event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0004offcore_response.all_data_rd.llc_miss.dramNumber of requests allocated in Coherency Trackerunc_cbo_xsnp_response.invalevent=0x34,umask=0x40Cycles PMH is busy with this walkoffcore_response.pf_llc_data_rd.llc_hit.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads that miss in the LLCCounts prefetch (that bring data to L2) data reads that miss the LLC  the data is found in M state in remote cache and forwarded from thereevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x87f820010PCIe allocating writes that hit in LLC (DDIO hits). Derived from unc_c_tor_inserts.opcode.ddio_hit. Unit: uncore_cbox Occupancy for all LLC misses that are addressed to local memory. Unit: uncore_cbox unc_q_rxl0p_power_cycles(UNC_Q_RxL0P_POWER_CYCLES / UNC_Q_CLOCKTICKS) * 100.unc_p_freq_band2_cyclesCounts the number of times that the uncore transitioned a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band0=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Derived from unc_p_freq_band0_cycles. Unit: uncore_pcu freq_max_current_cycles %l1d.allocated_in_mInstructions retired. (Precise Event - PEBS) (Must be precise)Number of cases where any load ends up with a valid block-code written to the load buffer (including blocks due to Memory Order Buffer (MOB), Data Cache Unit (DCU), TLB, but load has no DCU miss)other_assists.itlb_miss_retiredOccupancy counter for all LLC misses; we divide this by UNC_C_CLOCKTICKS to get average Q depth. Unit: uncore_cbox tor_occupancy.miss_all %event=0x8,period=100003,umask=0x2This event counts the number of load micro-ops retired that miss in L1 Data cache. Note that prefetch misses will not be countedevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800403091Counts Demand cacheable data and L1 prefetch data read requests  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004003091offcore_response.any_pf_l2.l2_hit_near_tile_mCounts any request that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateoffcore_response.bus_locks.l2_hit_far_tileCounts Bus locks and split lock requests that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeoffcore_response.demand_code_rd.l2_hit_far_tile_mCounts demand code reads and prefetch code reads that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateCounts demand code reads and prefetch code reads that accounts for responses which hit its own tile's L2 with data in E stateCounts demand cacheable data and L1 prefetch data reads that accounts for responses which hit its own tile's L2 with data in M stateCounts demand cacheable data and L1 prefetch data reads that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400002Counts Demand cacheable data writes that accounts for responses which hit its own tile's L2 with data in F stateoffcore_response.pf_l2_code_rd.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800181000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080803091event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181808000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080808000offcore_response.demand_code_rd.mcdram_nearCounts Demand cacheable data writes that accounts for responses from DDR (local and far)offcore_response.partial_reads.ddr_nearevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101002000offcore_response.pf_l2_rfo.ddrevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800020no_alloc_cycles.rob_fulll1d_wb_l2.mesil1d_wb_l2.m_statel2_data_rqsts.demand.i_stateL2 load missesL2 instruction fetch transactionsevent=0xf0,period=200000,umask=0x1event=0xb,period=20,umask=0x10,ldlat=0x1000event=0xcb,period=40000,umask=0x8offcore_response.any_data.llc_hit_other_core_hitmoffcore_response.any_ifetch.any_locationoffcore_response.any_request.any_locationoffcore_response.any_rfo.llc_hit_no_other_coreOffcore writebacks to the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1808event=0xb7,period=100000,umask=0x1,offcore_rsp=0x8004offcore_response.demand_ifetch.local_cache_dramoffcore_response.demand_rfo.llc_hit_other_core_hitmOffcore prefetch data requests satisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F20offcore_response.pf_rfo.local_cacheOffcore prefetch requests satisfied by the LLC or local DRAMoffcore_response.prefetch.remote_cache_hitAll Floating Point to and from MMX transitionsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4044offcore_response.data_in.remote_dramOffcore demand data reads satisfied by a remote DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4004Offcore prefetch data reads satisfied by a remote DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2040offcore_response.prefetch.remote_draml1i.missesevent=0x13,period=2000000,umask=0x4Mispredicted indirect non call branches executedMispredicted taken branches executedevent=0xc0,cmask=16,inv=1,period=2000000,umask=0x1Cycles no uops were delivered by the LSDUops executed on port 5uops_issued.cycles_all_threadsFused Uops issuedevent=0xc2,cmask=1,inv=1,period=2000000,umask=0x1Counts all retired memory instructions - loads and stores  Supports address when precise (Precise event)mem_load_retired.fb_hitevent=0xd1,period=100007,umask=0x40Offcore outstanding Code Reads transactions in the SuperQueue (SQ), queue to uncore, every cycleevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080001offcore_response.demand_data_rd.l3_hit_s.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100002offcore_response.other.l4_hit_local_l4.snoop_hit_no_fwdCounts once for most SIMD 256-bit packed double computational precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsRetired instructions that are fetched after an interval where the front-end delivered no uops for a period of at least 1 cycle which was not interrupted by a back-end stall (Must be precise)frontend_retired.latency_ge_16frontend_retired.latency_ge_2frontend_retired.latency_ge_512event=0xb7,period=100003,umask=0x1,offcore_rsp=0x44000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x203C408000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000408000Counts the number of hardware interruptions received by the processorNot taken branch instructions retired  Spec update: SKL091Cycles where the Store Buffer was full and no outstanding loaduops_executed.x87Load miss in all TLB levels causes a page walk that completes. (All page sizes)event=0x49,period=2000003,umask=0x10event=0x85,cmask=1,period=100003,umask=0x10Counts any code reads (demand & prefetch) that miss L2 with a snoop miss responseoffcore_response.any_rfo.l2_miss.snoop_missoffcore_response.corewb.l2_miss.no_snoop_neededoffcore_response.demand_code_rd.l2_miss.snoop_missCounts demand and DCU prefetch data read that miss L2Any reissued load uopsevent=0x86,period=200003,umask=0x3fCounts the number of cycles and allocation pipeline is stalled and is waiting for a free MEC reservation station entry.  The cycles should be appropriately counted in case of the cracked ops e.g. In case of a cracked load-op, the load portion is sent to Mevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0240Counts all prefetch data reads that hit in the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0004Counts prefetch (that bring data to LLC only) code reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts prefetch (that bring data to LLC only) code reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10080offcore_response.pf_l2_code_rd.llc_miss.dramoffcore_requests.anyOffcore demand RFO requestsOutstanding offcore readsevent=0x60,cmask=1,period=2000000,umask=0x1event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f11REQUEST = ANY_REQUEST and RESPONSE = REMOTE_CACHE_HITMoffcore_response.data_in.local_dram_and_remote_cache_hitREQUEST = DATA_IN and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DEMAND_DATA and RESPONSE = ANY_CACHE_DRAMREQUEST = OTHER and RESPONSE = IO_CSR_MMIOevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f50REQUEST = DATA_IFETCH and RESPONSE = ANY_DRAM AND REMOTE_FWDoffcore_response.pf_ifetch.any_dram_and_remote_fwdevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2780offcore_response.all_data_rd.l3_hit.snoop_hit_with_fwdCounts all prefetch data reads that have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0100Counts all demand & prefetch data reads that miss the L3 and the data is returned from local dramCounts all demand code reads that miss the L3 and clean or shared data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800010Fraction of Core cycles where the core was running with power-delivery for license level 1.  This includes high current AVX 256-bit instructions as well as low current AVX 512-bit instructions. SMT version; use when SMT is enabled and measuring per logical CPUunc_m_cas_count.allDRAM Underfill Read CAS Commands issued. Unit: uncore_imc event=0x4,umask=0x4DRAM CAS (Column Address Strobe) Commands.; DRAM WR_CAS (w/ and w/out auto-pre) in Write Major Mode. Unit: uncore_imc unc_cha_tor_inserts.ia_missunc_cha_requests.readsMulti-socket cacheline Directory state updates; Directory Updated memory write from the HA pipe. Unit: uncore_cha unc_cha_sf_eviction.e_stateCounts when a a transaction with the opcode type RspIFwd Snoop Response was received which indicates a remote caching agent forwarded the data and the requesting agent is able to acquire the data in E (Exclusive) or M (modified) states.  This is commonly returned with RFO (the Read for Ownership issued before a write) transactions.  The snoop could have either been to a cacheline in the M,E,F (Modified, Exclusive or Forward)  statesClockticks of the IIO Traffic Controller. Unit: uncore_iio PCIe Completion Buffer occupancy of completions with data: Part 0unc_iio_data_req_by_cpu.peer_read.part1unc_iio_data_req_by_cpu.peer_read.part3unc_iio_txn_req_by_cpu.mem_read.part2unc_iio_txn_req_by_cpu.mem_read.part3Counts every write request of up to a 64 byte transaction of data made by IIO Part1 to a unit on the main die (generally memory). In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busCounts every write request of up to a 64 byte transaction of data made by IIO Part3 to a unit on the main die (generally memory). In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busPeer to peer write request of up to a 64 byte transaction is made by IIO Part3 to an IIO target. Unit: uncore_iio unc_i_faf_insertsevent=0x27Messages sent direct to the Intel UPI. Unit: uncore_m2m unc_m2m_directory_lookup.state_aCounts when the M2M (Mesh to Memory) looks into the multi-socket cacheline Directory state, and found the cacheline marked in the A (SnoopAll) state, indicating the cacheline is stored in another socket in any state, and we must snoop the other sockets to make sure we get the latest data.  The data may be stored in any state in the local socketCounts when the M2M (Mesh to Memory) updates the multi-socket cacheline Directory state from from A (SnoopAll) to I (Invalid)AD Egress (to CMS) Occupancy. Unit: uncore_m2m event=0x16,umask=0x03Counts clockticks of the fixed frequency clock controlling the Intel Ultra Path Interconnect (UPI).  This clock runs at1/8th the 'GT/s' speed of the UPI link.  For example, a  9.6GT/s  link will have a fixed Frequency of 1.2 GhzCounts cycles when the Intel Ultra Path Interconnect (UPI) is in L1 power mode.  L1 is a mode that totally shuts down the UPI link.  Link power states are per link and per direction, so for example the Tx direction could be in one state while Rx was in another, this event only coutns when both links are shutdownCounts cycles when the transmit side (Tx) of the Intel Ultra Path Interconnect(UPI) is in L0p power mode. L0p is a mode where we disable 60% of the UPI lanes, decreasing our bandwidth in order to save powerevent=0x2,umask=0x27event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100491OCR.ALL_PF_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDocr.all_pf_data_rd.l3_hit_m.hitm_other_coreOCR.ALL_PF_DATA_RD.L3_HIT_S.HITM_OTHER_CORE  OCR.ALL_PF_DATA_RD.L3_HIT_S.HITM_OTHER_COREOCR.ALL_PF_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDOCR.ALL_READS.L3_HIT_F.HIT_OTHER_CORE_NO_FWD  OCR.ALL_READS.L3_HIT_F.HIT_OTHER_CORE_NO_FWDOCR.ALL_READS.L3_HIT_M.HITM_OTHER_CORE  OCR.ALL_READS.L3_HIT_M.HITM_OTHER_COREOCR.ALL_RFO.L3_HIT_E.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200004Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDocr.demand_data_rd.l3_hit_s.hit_other_core_fwdocr.demand_data_rd.l3_hit_s.snoop_missCounts any other requests OCR.OTHER.L3_HIT.SNOOP_NONECounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_HIT.HITM_OTHER_CORE OCR.PF_L1D_AND_SW.L3_HIT.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0400Counts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.pf_l1d_and_sw.l3_hit_f.no_snoop_neededCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_M.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040010ocr.pf_l2_data_rd.l3_hit_s.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080020Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_M.HITM_OTHER_CORECounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_E.ANY_SNOOPocr.pf_l3_data_rd.l3_hit_f.any_snoopCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_F.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040080ocr.pf_l3_data_rd.l3_hit_s.hitm_other_coreocr.pf_l3_rfo.l3_hit_e.hitm_other_coreCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.ANY_SNOOPoffcore_response.all_data_rd.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_F.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.all_pf_data_rd.l3_hit_f.snoop_noneoffcore_response.all_pf_data_rd.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.ANY_SNOOPoffcore_response.all_pf_rfo.l3_hit_m.hit_other_core_fwdoffcore_response.all_pf_rfo.l3_hit_m.snoop_missThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.SNOOP_NONEoffcore_response.all_reads.l3_hit_e.snoop_missoffcore_response.all_reads.l3_hit_s.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_F.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400001This event is deprecated. Refer to new event OCR.DEMAND_RFO.ANY_RESPONSEoffcore_response.demand_rfo.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.SNOOP_MISSoffcore_response.other.l3_hit_f.snoop_missoffcore_response.other.pmm_hit_local_pmm.any_snoopoffcore_response.pf_l1d_and_sw.l3_hit_s.any_snoopoffcore_response.pf_l2_data_rd.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDoffcore_response.pf_l3_data_rd.l3_hit_e.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.SNOOP_NONENumber of SSE/AVX computational 512-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 16 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP14 RSQRT14 SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementocr.all_data_rd.l3_miss_local_dram.hitm_other_coreocr.all_data_rd.l3_miss_local_dram.no_snoop_neededocr.all_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdocr.all_pf_data_rd.l3_miss.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000490ocr.all_pf_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.all_pf_data_rd.l3_miss_remote_hop1_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1040007F7ocr.all_reads.l3_miss_local_dram.snoop_miss_or_no_fwdocr.all_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDocr.demand_data_rd.l3_miss.any_snoopocr.demand_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000001Counts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.HIT_OTHER_CORE_FWD OCR.DEMAND_RFO.L3_MISS.HIT_OTHER_CORE_FWDocr.demand_rfo.l3_miss_local_dram.no_snoop_neededCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.REMOTE_HIT_FORWARDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000080Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000100ocr.pf_l3_rfo.l3_miss_local_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.HITM_OTHER_COREoffcore_response.other.l3_miss.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDocr.all_reads.supplier_none.any_snoopocr.pf_l2_data_rd.supplier_none.hit_other_core_no_fwdocr.pf_l2_rfo.pmm_hit_local_pmm.snoop_noneocr.pf_l2_rfo.supplier_none.snoop_missunc_m_pmm_bandwidth.totalunc_m_pmm_wpq_occupancy.allunc_m2m_imc_writes.to_pmmClean line underfill read hits to Near Memory(DRAM cache) in Memory Mode. Unit: uncore_m2m Number of cycles a demand request has waited due to L1D due to lack of L2 resourcesCounts demand requests that miss L2 cacheNumber of completed demand load requests that missed the L1, but hit the FB(fill buffer), because a preceding miss to the same cacheline initiated the line to be brought into L1, but data is not yet ready in L1  Supports address when precise (Precise event)Counts the number of off-core outstanding read-for-ownership (RFO) store transactions every cycle. An RFO transaction is considered to be in the Off-core outstanding state between L2 cache miss and transaction completionCounts number of SSE/AVX computational 128-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 2 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementCounts retired Instructions who experienced Instruction L1 Cache true miss (Precise event)event=0xc6,period=100007,umask=0x1,frontend=0x510006TOPDOWN.SLOTS / ( TOPDOWN.SLOTS / 2 ) if #SMT_on else 1Other_BranchesNumber of machine clears due to memory ordering conflictsNumber of times an instruction execution caused the transactional nest count supported to be exceededevent=0x54,period=100003,umask=0x8br_inst_retired.condMispredicted non-taken conditional branch instructions retired (Precise event)number of branch instructions retired that were mispredicted and taken. Non PEBS (Precise event)Counts the number of reference cycles when the core is not in a halt state. The core enters the halt state when it is running the HLT instruction or the MWAIT instruction. This event is not affected by core frequency changes (for example, P states, TM2 transitions) but has the same incrementing frequency as the time stamp counter. This event can approximate elapsed time while the core was not in a halt state. This event has a constant ratio with the CPU_CLK_UNHALTED.REF_XCLK event. It is counted on a dedicated fixed counter, leaving the eight programmable counters available for other events. Note: On all current platforms this event stops counting during 'throttling (TM)' states duty off periods the processor is 'halted'.  The counter update is done at a lower clock rate then the core clock the overflow status bit for this counter may appear 'sticky'.  After the counter has overflowed and software clears the overflow status bit and resets the counter to less than MAX. The reset value to the counter is not clocked immediately so the overflow status bit will flip 'high (1)' and generate another PMI (if enabled) after which the reset value gets clocked into the counter. Therefore, software will get the interrupt, read the overflow status bit '1 for bit 34 while the counter value is less than MAX. Software should ignore this caseevent=0xa3,cmask=4,period=1000003,umask=0x4Number of all retired NOP instructions (Precise event)misc_retired.pause_instuops_dispatched.port_5event=0xe,cmask=1,inv=1,period=1000003,umask=0x1Pipeline;PortsUtilFlops;HPCocr.demand_data_rd.snc_cache.hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1030000477event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080800event=0x60,period=1000003,umask=0x2UOPS_RETIRED.SLOTS / BR_INST_RETIRED.NEAR_TAKENCounts demand data reads that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallyevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x84002380ocr.other.l3_miss_localocr.reads_to_core.l3_missFor every cycle, increments by the number of demand data read requests pending that are known to have missed the L3 cache.  Note that this does not capture all elapsed cycles while requests are outstanding - only cycles from when the requests were known to have missed the L3 cacheCounts responses to snoops indicating the line may be kept on this core in the (S)hared state, after the data is forwarded back to the requestor, initially the data was found in the cache in the (FS) Forward or Shared state.  A single snoop response from the core counts on all hyperthreads of the coreCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by DRAMocr.demand_rfo.snc_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x12380Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those DRAM accesses that are controlled by the close SNC Clusterevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004004772LM Tag Check : Hit in Near Memory Cache. Unit: uncore_imc event=0x2,umask=0x10unc_cha_tor_inserts.ia_hit_crdevent=0x35,umask=0xCCC7FD01TOR Inserts : LLCPrefRFO issued by iA Cores that hit the LLC. Unit: uncore_cha event=0x35,umask=0xCC43FD04TOR Inserts : PCIRdCurs issued by IO Devices that missed the LLC. Unit: uncore_cha unc_iio_data_req_by_cpu.mem_write.part6unc_iio_data_req_by_cpu.mem_write.part7event=0xc0,ch_mask=0x40,fc_mask=0x07,umask=0x04unc_iio_data_req_of_cpu.cmpd.part4event=0xc2,ch_mask=0x01,fc_mask=0x04,umask=0x03unc_i_irp_all.inbound_insertsunc_i_faf_transactionsTag Hit : Dirty NearMem Read Hit. Unit: uncore_m2m unc_m3upi_clockticksCycles in L0p. Unit: uncore_upi unc_cha_tor_inserts.io_itomcachenear_remotel2_request.hitCounts the number of cycles the core is stalled due to an instruction cache or TLB miss which hit in the L2 cacheCounts modified writebacks from L1 cache and L2 cache that were supplied by the L3 cacheocr.all_code_rd.l3_miss_localevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000001ocr.hwpf_l2_rfo.l3_miss_localocr.all_code_rd.dramCounts L2 cache hardware prefetch RFOs (written to the L2 cache only) that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)Counts the number of far branch instructions retired, includes far jump, far call and return, and interrupt call and return (Precise event)Counts the total number of instructions retired. (Fixed event) (Precise event)Counts the number of machine clears due to a page fault.  Counts both I-Side and D-Side (Loads/Stores) page faults.  A page fault occurs when either the page is not present, or an access violation occursDRAM Activate Count : All Activates : Counts the number of DRAM Activate commands sent on this channel.  Activate commands are issued to open up a page on the DRAM devices so that it can be read or written to with a CAS.  One can calculate the number of Page Misses by subtracting the number of Page Miss precharges from the number of ActivatesCounts snoop filter capacity evictions for entries tracking shared lines in the cores? cache.? Snoop filter capacity evictions occur when the snoop filter is full and evicts an existing entry to track a new entry.? Does not count clean evictions such as when a core?s cache replaces a tracked cacheline with a new cachelineevent=0x35,umask=0xC827FE01unc_cha_tor_occupancy.ia_drd_opt_prefunc_cha_tor_occupancy.ia_hit_drd_optData requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 3Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 7itlb.fillsehl metricsCounts the total number of machine clears including memory ordering, memory disambiguation, self-modifying code, page faults and floating point assistadl metrics( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING ) / ( 4 * CPU_CLK_UNHALTED.DISTRIBUTED )Average Frequency Utilization relative nominal frequency. Unit: cpu_core Percentage of total non-speculative loads with a store forward or unknown store address block. Unit: cpu_atom Branch_Mispredict_to_Unknown_Branch_RatioInst_Miss_Cost_DRAMHit_Percentmem_scheduler_block.rsvL2 cache lines filling L2. Unit: cpu_core SW prefetch requests that hit L2 cache. Unit: cpu_core SW prefetch requests that miss L2 cache. Unit: cpu_core mem_store_retired.l2_hitCounts number of SSE/AVX computational 256-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. Unit: cpu_core Counts number of SSE/AVX computational scalar double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. Unit: cpu_core Retired Instructions who experienced iTLB true miss (Precise event). Unit: cpu_core event=0xc6,period=100007,umask=0x1,frontend=0x600106Retired Instructions who experienced STLB (2nd level TLB) true miss (Precise event). Unit: cpu_core Counts streaming stores that have any type of response. Unit: cpu_atom assists.page_faultevent=0xc1,period=1000003,umask=0x8Counts the total number of mispredicted branch instructions retired for all branch types (Precise event). Unit: cpu_atom Cycles when divide unit is busy executing divide or square root operations. Unit: cpu_core arith.int_divider_activeevent=0xec,period=2000003,umask=0x40Number of machine clears (nukes) of any type. Unit: cpu_core Uops executed on port 1. Unit: cpu_core Incoming read prefetch request from IA. Unit: uncore_imc unc_m_dram_page_empty_wrunc_mc1_rdcas_count_freerunevent=0x11,cmask=1,period=100003,umask=0x10event=0x2a,period=100003,umask=0x1,offcore_rsp=0x1008004477event=0xcf,period=100003,umask=0x3event=0xce,period=1000003,umask=0x2Mispredicted indirect CALL retired (Precise event)M2P Clockticks. Unit: uncore_m2pcie event=0x1,ch_mask=0x0000Write Requests from a unit on this socket. Unit: uncore_cha unc_cha_requests.invitoeTOR Occupancy for DRd misses from local IA targeting remote memory. Unit: uncore_cha TOR Inserts for DRds issued by IA Cores targeting DDR Mem that Missed the LLC. Unit: uncore_cha unc_upi_rxl_flits.slot1event=0x3,umask=0x0000000097event=0x3,umask=0x0000000047Matches on Transmit path of a UPI Port : Non-Coherent Bypass. Unit: uncore_upi event=0x5,umask=0x000000000eevent=0x12,umask=0x0000000001event=0x84,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000004event=0xc1,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000004event=0x21,umask=0x0320unc_cha_osb.local_invitoeevent=0x35,umask=0x00c001ff08event=0x35,umask=0x00c80ffd01TOR Inserts : Just Hits. Unit: uncore_cha unc_cha_tor_inserts.mmioevent=0x35,umask=0x00c8a7fe01event=0x35,umask=0x00c887fe01event=0x35,umask=0x00c8c7ff01event=0x35,umask=0x00c8d7ff01TOR Occupancy : IRQ - Non iA. Unit: uncore_cha TOR Occupancy : WBQ. Unit: uncore_cha TOR Occupancy : All from Local iA and IO. Unit: uncore_cha TOR Occupancy : All Snoops from Remote. Unit: uncore_cha TOR Occupancy : Just Misses. Unit: uncore_cha event=0x36,umask=0x00c827fd01event=0x36,umask=0x00c897fe01TOR Occupancy; DRd Pref misses from local IA. Unit: uncore_cha All LLC lines in M state that are victimized on a fill. Unit: uncore_cha event=0x35,umask=0x00c8f3fe04unc_cha_tor_inserts.ia_hit_llcprefcodeTOR Inserts; LLCPrefData misses from local IA. Unit: uncore_cha event=0x36,umask=0x00ccd7fe01unc_cha_tor_inserts.ia_miss_crd_localevent=0x35,umask=0x00C8668601TOR Inserts : WCiLs issued by iA Cores targeting PMM that missed the LLC. Unit: uncore_cha TOR Occupancy : DRds issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha TOR Occupancy : DRd_Prefs issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_drd_pref_ddrevent=0x36,umask=0x00c86fff01event=0x36,umask=0x00c86ffe01ic_cache_inval.l2_invalidating_probeevent=0x60,umask=0x40l2_request_g1.group2event=0x61,umask=0x80l2_request_g2.ic_rd_sizedevent=0x61,umask=0x10ex_ret_brn_tkn_mispevent=0xcb,umask=0x01event=0x887,umask=0x02dram_channel_data_controller_6Total number multi-pipe uOps assigned to pipe 0fpu_pipe_assignment.total0event=0x4,umask=0x02event=0x25,umask=0x01Number of STLF hitsevent=0x46,umask=0x04The number of software prefetches that did not fetch data outside of the processor core. Software PREFETCH instruction saw a DC hitde_dis_dispatch_token_stalls0.alsq3_token_stallbranch_misprediction_ratiol2_cache_req_stat.ic_dc_miss_in_l2 + l2_pf_miss_l2_hit_l3 + l2_pf_miss_l2_l3all_l2_cache_missesevent=0x85,umask=0x07event=0x85,umask=0x01event=0xae,umask=0x40SSE/AVX bottom-executing ops retired. The number of serializing Ops retiredThe number of retired CLFLUSH instructions. This is a non-speculative eventAny Data Cache Fills by Data Source. From L3 or different L2 in same CCXls_any_fills_from_sys.lcl_l2Software Prefetch Instructions Dispatched (Speculative). PrefetchW instruction. See docAPM3 PREFETCHWSoftware Prefetch Data Cache Fills by Data Source. From CCX Cache in different Nodede_dis_cops_from_decoder.disp_op_type.any_fp_dispatchMiss Address Buffer (MAB) entries allocated by a Load-Store (LS) pipe for hardware prefetcher allocationsls_dmnd_fills_from_sys.local_ccxls_dmnd_fills_from_sys.dram_io_farls_any_fills_from_sys.dram_io_allls_sw_pf_dc_fills.dram_io_nearHardware prefetch data cache fills from L3 cache or different L2 cache in the same CCXHardware prefetch data cache fills from cache of another CCX when the address was in the same NUMA nodel2_cache_req_stat.allevent=0x71,umask=0x01event=0x4,umask=0xfeevent=0xac,umask=0x04L3 cache fill requests sourced from another CCX's cache when the address was in a different NUMA node. Unit: amd_l3 Retired instructionsCycles with no retire due  to the lack of valid ops in the retire queue (may be caused by front-end bottlenecks or pipeline redirects)event=0xdf,umask=0x7feevent=0x21f,umask=0xbferemote_processor_write_data_beats_cs0event=0x5f,umask=0xbffWrite data beats (64 bytes) for remote processor at Coherent Station (CS) 11local_socket_upstream_read_beats_iom1event=0x89f,umask=0x7felocal_socket_upstream_write_beats_iom2remote_socket_upstream_write_beats_iom3event=0x45f,umask=0x7felocal_socket_inf0_outbound_data_beats_ccm4event=0x41f,umask=0x7ffremote_socket_inf1_inbound_data_beats_ccm0event=0x4de,umask=0xbfffp_ops_retired_by_type.scalar_mulevent=0xa,umask=0x08event=0xa,umask=0xe0Retired MMX integer subtract opsevent=0xb,umask=0x40sse_avx_ops_retired.sse_avx_shiftRetired SSE and AVX integer pack opsevent=0xc,umask=0x04fp_pack_ops_retired.fp128_logicalevent=0xc,umask=0x0ffp_pack_ops_retired.fp256_macRetired 256-bit packed floating-point blend opspacked_int_op_type.int128_subpacked_int_op_type.int256_cmpRetired 256-bit packed integer ops of other typesumc_cas_cmd.wrNumber of memory load operations dispatched to the load-store unitbp_l1_tlb_fetch_hit.allInstruction fetches that hit in the L1 ITLB for all page sizesbackend_boundl2_pf_hit_l2.all + l2_pf_miss_l2_hit_l3.all + l2_pf_miss_l2_l3.allL3 cache access, readTotal cache missesDC_REFILL_FROM_L2FR_RETIRED_X86_INSTRUCTIONSPC_BRANCH_MIS_PREDL1_ICACHE_ACCESSEVENT_44HEVENT_67HEVENT_99HEVENT_B5HEVENT_BDHEVENT_CFHEVENT_D5HEVENT_E2HEVENT_FBHEXTERNAL_INTERRUPTPLE_FIFO_FLUSHEVENT_10CHEVENT_124HEVENT_167HEVENT_17FHEVENT_1ADHEVENT_1C0HEVENT_232HEVENT_255HEVENT_284HEVENT_288HEVENT_2E9HEVENT_302HEVENT_304HEVENT_316HEVENT_32CHEVENT_360HEVENT_36DHEVENT_37EHEVENT_383HEVENT_386HEVENT_3D6HEVENT_3E5HL1D_CACHE_REFILL_STBUS_ACCESS_PERIPHdn_rxreq_retriedhnf_cmp_adq_fullrni_s2_rdata_beatsrni_wrt_occ_ovflrni_rrt_allocclk_upload_stallTLB_BIT_TRANSITIONSLS_LM_COMPLETEDL1_INSTR_CACHE_ACCESSESVEC_LOAD_INSTR_COMPLETEDCYCLES_ONE_INSTR_DISPATCHEDVTQ_STREAM_CANCELED_PREMATURELYVTQ_LINE_FETCH_MISSBRANCH_FLUSHESL2_EXTERNAL_INTERVENTIONSBUS_TAS_FOR_WRITESSNOOP_REQUESTSdc-fillvictim-from-l2write-to-read-turnaroundSOFT-CLOCK.HARDDMC620_PMU_CINTEL_ALDERLAKEINTEL_ATOM_TREMONT%s:	%s
anyGenuineIntel-6-46Frontend_Bound_SMTTopdownL1_SMTThis category represents fraction of slots wasted due to incorrect speculationsThis category represents fraction of slots utilized by useful work i.e. issued uops that eventually get retired. Ideally; all pipeline slots would be attributed to the Retiring category.  Retiring of 100% would indicate the maximum Pipeline_Width throughput was achieved.  Maximizing Retiring typically increases the Instructions-per-cycle (see IPC metric). Note that a high Retiring value does not necessary mean there is no room for more performance.  For example; Heavy-operations or Microcode Assists are categorized under Retiring. They often indicate suboptimal performance and can often be optimized or avoidedUOPS_RETIRED.RETIRE_SLOTS / INST_RETIRED.ANY( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE ) / CPU_CLK_UNHALTED.THREAD( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE) ) / ( 2 * CPU_CLK_UNHALTED.THREAD )IpArith_Scalar_DPAverage data fill bandwidth to the L2 cache [GB / sec]64 * L2_LINES_IN.ALL / 1000000000 / duration_timeL2 cache misses per kilo instruction for all demand loads  (including speculative)Mem;MemoryLat;SoCIpFarBranchl1d_pend_miss.pending_cyclesevent=0xf1,period=100003,umask=0x4l2_rqsts.l2_pf_hitevent=0xf0,period=200003,umask=0x40This event counts Read for Ownership (RFO) requests that access L2 cachemem_load_uops_l3_hit_retired.xsnp_hitoffcore_requests_outstanding.cycles_with_demand_data_rdfp_arith_inst_retired.256b_packed_doubleNumber of SSE/AVX computational 256-bit packed double precision floating-point instructions retired.  Each count represents 4 computations. Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementNumber of SSE/AVX computational packed floating-point instructions retired. Applies to SSE* and AVX*, packed, double and single precision floating-point: ADD SUB MUL DIV MIN MAX RSQRT RCP SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementNumber of SSE/AVX computational scalar single precision floating-point instructions retired.  Each count represents 1 computation. Applies to SSE* and AVX* scalar single precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP RSQRT SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementThis event counts cycles with any input and output SSE or x87 FP assist. If an input and output assist are detected on the same cycle the event increments by 1This event counts the number of SSE* floating point (FP) micro-code assist (numeric overflow/underflow) when the output value (destination register) is invalid. Counting covers only cases involving penalties that require micro-code assist interventionevent=0xc1,period=100003,umask=0x8idq_uops_not_delivered.cycles_le_1_uop_deliv.coreNumber of times an HLE execution aborted due to HLE-unfriendly instructionsmem_trans_retired.load_latency_gt_4event=0xcd,period=100003,umask=0x1,ldlat=0x4Number of times an RTM abort was attributed to a Memory condition (See TSX_Memory event for additional details)Counts the number of times an HLE XACQUIRE instruction was executed inside an RTM transactional regionNumber of times a TSX Abort was triggered due to a non-release/commit store to locktx_mem.hle_elision_buffer_fullThis event counts cycles in which the L1 and L2 are locked due to a UC lock or split lock. A lock is asserted in case of locked memory access, due to noncacheable memory, locked operation that spans two cache lines, or a page walk from the noncacheable page table. L1D and L2 locks have a very high performance penalty and it is highly recommended to avoid such accessbr_inst_exec.all_direct_near_callevent=0x88,period=200003,umask=0xd0br_inst_exec.taken_indirect_near_returnevent=0xc4,period=400009,umask=0x4br_misp_exec.all_indirect_jump_non_call_retevent=0xc5,period=400009cpu_clk_unhalted.one_thread_activecpu_clk_unhalted.thread_pint_misc.rat_stall_cyclesint_misc.recovery_cyclesmachine_clears.maskmovSelf-modifying code (SMC) detectedNumber of times any microcode assist is invoked by HW upon uop writebackresource_stalls.robrob_misc_events.lbr_insertsCount cases of saving new LBRrs_events.empty_endThis event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 0event=0xb1,cmask=1,period=2000003,umask=0x2uops_executed.cycles_ge_4_uops_execllc_misses.mmio_readLLC misses for PCIe read current. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox (UNC_P_PROCHOT_EXTERNAL_CYCLES / UNC_P_CLOCKTICKS) * 100.Load misses that miss the  DTLB and hit the STLB (4K)event=0x49,period=100003,umask=0x60dtlb_store_misses.stlb_hit_2mpage_walker_loads.itlb_l2 ( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * (BR_MISP_RETIRED.ALL_BRANCHES * (12 * ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY ) / CPU_CLK_UNHALTED.THREAD) / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY )) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) ) * (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )) / BR_MISP_RETIRED.ALL_BRANCHESThis event counts retired load uops which data sources were hits in the nearest-level (L1) cache.
Note: Only two data-sources of L1/FB are applicable for AVX-256bit  even though the corresponding AVX load could be serviced by a deeper level in the memory hierarchy. Data source is reported for the Low-half load. This event also counts SW prefetches independent of the actual data source  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020120offcore_response.demand_code_rd.l3_hit.any_snoopoffcore_response.demand_code_rd.supplier_none.snoop_noneCounts all demand data writes (RFOs) have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100028000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020010offcore_response.pf_l2_rfo.any_responseoffcore_response.pf_l3_data_rd.l3_hit.snoop_hitmCounts all prefetch (that bring data to LLC only) RFOsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000090offcore_response.demand_code_rd.l3_miss_local_dram.any_snoopoffcore_response.demand_code_rd.l3_miss_local_dram.snoop_missoffcore_response.demand_rfo.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084008000offcore_response.pf_l2_code_rd.l3_miss_local_dram.snoop_missoffcore_response.pf_l2_data_rd.l3_miss_local_dram.snoop_missoffcore_response.pf_l2_rfo.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020080Return instructions retired (Precise event)L3 Lookup any request that access cache and found line in I-stateevent=0x34,umask=0x86unc_cbo_cache_lookup.write_esunc_arb_trk_occupancy.drd_directevent=0x81,umask=0x01event=0x84,umask=0x01offcore_response.demand_rfo.llc_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00091l2_data_rqsts.self.m_statel2_ld.self.demand.i_statesimd_instr_retiredSIMD micro-ops executed (excluding stores)prefetch.prefetcht2busq_empty.selfevent=0x7b,period=200000,umask=0x0event=0x6e,period=200000,umask=0xe0Instruction-fetch bus transactionsevent=0x6a,period=200000,umask=0x40br_inst_retired.pred_not_takenevent=0x88,period=2000000,umask=0x1Instructions retiredMicro-op reissues for any causestore_forwards.gooduops_retired.stallsevent=0x82,period=200000,umask=0x4Requests rejected by the XQCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that hit the L2 cacheCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_read.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000008000Counts requests to the uncore subsystem that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts requests to the uncore subsystem that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000002000offcore_response.pf_l2_data_rd.l2_miss.anyevent=0xcd,period=200003,umask=0x2Counts the number of times the prediction (from the predecode cache) for instruction length is incorrecthw_interrupts.pending_and_maskedCounts hardware interrupts received by the processorRetired far branch instructions (Precise event capable) (Must be precise)Retired near relative call instructions (Precise event capable) (Must be precise)Counts loads that block because their address modulo 4K matches a pending store (Must be precise)event=0x3,period=200003,umask=0x1Counts data read, code read, and read for ownership (RFO) requests (demand & prefetch) hit the L2 cacheCounts reads for ownership (RFO) requests (demand & prefetch) true miss for the L2 cache with a snoop miss in the other processor moduleCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes have any transaction responses from the uncore subsystemoffcore_response.streaming_stores.l2_miss.hitm_other_coreCounts data cache lines requests by software prefetch instructions hit the L2 cacheCycles the code-fetch stalls and an ITLB miss is outstandingRetired load uops with L1 cache hits as data sources  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)All retired load uops  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)Offcore outstanding cacheable Core Data Read transactions in SuperQueue (SQ), queue to uncore  Spec update: HSD62, HSD61, HSM63avx_insts.allevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00244FP operations retired. X87 FP operations that have no exceptions: Counts also flows that have several X87 or flows that use X87 uops in the exception handlingCycles which a uop is dispatched on port 1 in this threadunc_cbo_cache_lookup.extsnp_mEach cycle count number of valid entries in Coherency Tracker queue from allocation till deallocation. Aperture requests (snoops) appear as NC decoded internally and become coherent (snoop L3, access memory). Unit: uncore_arb event=0xbc,period=2000003,umask=0x41event=0xbc,period=2000003,umask=0x84event=0xf2,period=100003,umask=0x1event=0x24,period=200003,umask=0x3l2_store_lock_rqsts.allRetired load uops which data sources were LLC and cross-core snoop hits in on-pkg core cache (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0001Counts all data/code/rfo reads (demand & prefetch) that miss the LLC  and the data returned from dramLoads blocked by overlapping with store buffer that cannot be forwardedCycles which a Uop is dispatched on port 5event=0x22,umask=0x01unc_cbo_xsnp_response.hitDemand load cycles page miss handler (PMH) is busy with this walkoffcore_response.all_pf_data_rd.llc_hit.no_snoop_neededCounts all data/code/rfo reads (demand & prefetch) that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803c8000Counts prefetch (that bring data to L2) data reads that hit in the LLC and the snoops sent to sibling cores return clean responseoffcore_response.pf_l2_data_rd.llc_miss.local_dramLLC prefetch misses for RFO. Derived from unc_c_tor_inserts.miss_opcode.rfo_prefetch. Unit: uncore_cbox Partial PCIe reads. Derived from unc_c_tor_inserts.opcode.pcie_partial. Unit: uncore_cbox event=0x35,umask=0x1,filter_opc=0x1e5unc_p_freq_band1_cyclesL1D data cache lines in M state evicted due to replacementoffcore_response.all_demand_mlc_pref_reads.llc_miss.local_drambr_misp_retired.takenevent=0x7,period=100003,umask=0x8Performance sensitive flags-merging uops added by Sandy Bridge u-archresource_stalls.lbresource_stalls.lb_sbevent=0x5b,period=2000003,umask=0x40Memory controller clock ticks. Used to get percentages of memory controller cycles events. Unit: uncore_imc event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080044Counts Demand code reads and prefetch code read requests  that accounts for responses which hit its own tile's L2 with data in M stateCounts any request that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateCounts Demand cacheable data write requests  that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateoffcore_response.bus_locks.l2_hit_near_tile_e_foffcore_response.demand_data_rd.l2_hit_this_tile_fCounts demand cacheable data and L1 prefetch data reads that accounts for responses which hit its own tile's L2 with data in S stateoffcore_response.demand_rfo.l2_hit_near_tile_e_foffcore_response.partial_reads.l2_hit_far_tileCounts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses which hit its own tile's L2 with data in F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080040event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000040Counts L2 code HW prefetches that accounts for responses which hit its own tile's L2 with data in E stateCounts any Prefetch requests that accounts for data responses from MCDRAM Localoffcore_response.bus_locks.ddrCounts Bus locks and split lock requests that accounts for data responses from DRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400004offcore_response.demand_rfo.mcdramevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x2000020080event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800100event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080202000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800040Counts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for data responses from MCDRAM Far or Other tile L2 hit farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101001000offcore_response.uc_code_reads.ddrCounts the number of mispredicted branch instructions retired (Precise event)This event counts the number of core cycles when no uops are allocated, the instruction queue is empty and the alloc pipe is stalled waiting for instructions to be fetchedddr bandwidth write (CPU traffic only) (MB/sec). Unit: uncore_imc unc_e_wpq_insertsevent=0x42,period=2000000,umask=0x2l2_data_rqsts.demand.e_stateevent=0xf1,period=100000,umask=0x7L2 RFO hitsL2 RFO missesl2_write.lock.hitL2 demand store RFOs in I state (misses)Memory instructions retired above 64 clocks (Precise Event)Retired loads that hit valid versions in the LLC cache (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x111offcore_response.any_request.io_csr_mmioevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2FFevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8022offcore_response.any_rfo.llc_hit_other_core_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x722event=0xb7,period=100000,umask=0x1,offcore_rsp=0x3822event=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF08Offcore data reads, RFO's and prefetches statisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8003Offcore demand data reads satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x204offcore_response.other.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8080Offcore other requests satisfied by the LLCoffcore_response.pf_data.remote_cache_dramoffcore_response.pf_data_rd.remote_cache_hitmoffcore_response.pf_ifetch.local_cache_dramoffcore_response.pf_rfo.llc_hit_other_core_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3820Offcore prefetch RFO requests that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4770fp_mmx_trans.anysimd_int_64.packsimd_int_64.unpackInstructions decodedMacro-fused instructions decodedevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF822offcore_response.corewb.any_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2008Offcore data reads, RFO's and prefetches statisfied by the local DRAMoffcore_response.pf_data_rd.local_dramoffcore_response.pf_rfo.any_dramOffcore prefetch requests satisfied by any DRAMLarge ITLB hitsq_full_stall_cyclesevent=0x88,period=20000,umask=0x20Retired conditional branch instructions (Precise Event)br_misp_exec.return_nearevent=0x3c,cmask=2,inv=1,period=2000000,umask=0x0Instructions retired (Programmable counter and Precise Event) (Precise event)event=0xc3,period=20000,umask=0x1rat_stalls.rob_read_portSIMD Packed-Double Uops retired (Precise Event)SIMD Packed-Single Uops retired (Precise Event)Cycles Uops executed on ports 0-4 (core count)Cycles no Uops issued on any port (core count)event=0xb1,period=2000000,umask=0x1Retirement slots used (Precise Event)itlb_misses.anyevent=0xd0,period=2000003,umask=0x83event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2001C0004offcore_response.demand_code_rd.l3_hit_m.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040004offcore_response.demand_code_rd.l3_hit_s.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0400004offcore_response.demand_rfo.l3_hit_s.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400088000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100088000offcore_response.other.l3_hit_m.snoop_noneNumber of PREFETCHT0 instructions executedsw_prefetch_access.t1_t2Counts once for most SIMD 128-bit packed computational double precision floating-point instructions retired. Counts twice for DPP and FM(N)ADD/SUB instructions retiredRetired instructions that are fetched after an interval where the front-end delivered no uops for a period of 64 cycles which was not interrupted by a back-end stall (Precise event)Counts the number of cycles uops were delivered to the Instruction Decode Queue (IDQ) from the MITE (legacy decode pipeline) path. Counting includes uops that may 'bypass' the IDQ. During these cycles uops are not being delivered from the Decode Stream Buffer (DSB)Counts randomly selected loads when the latency from first dispatch to completion is greater than 256 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000100001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC4000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFC400002This event counts conditional branch instructions retired  Spec update: SKL091 (Precise event)Counts both taken and not taken retired mispredicted direct and indirect near calls, including both register and memory indirect (Precise event)Counts all not software-prefetch load dispatches that hit the fill buffer (FB) allocated for the software prefetch. It can also be incremented by some lock instructions. So it should only be used with profiling so that the locks can be excluded by ASM (Assembly File) inspection of the nearby instructionsCycles where the pipeline is stalled due to serializing operationsCounts the number of x87 uops executedFraction of branches that are unconditional (direct or indirect) jumpsCounts completed page walks  (2M/4M sizes) caused by demand data loads. This implies address translations missed in the DTLB and further levels of TLB. The page walk can end with or without a faultCounts 1 per cycle for each PMH that is busy with a page walk for a store. EPT page walk duration are excluded in Skylake microarchitectureThis event counts requests originating from the core that references a cache line in the L2 cacheThis event counts the number of load ops retired that miss in L1 Data cache. Note that prefetch misses will not be countedCounts DCU hardware prefetcher data read that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.pf_l1_data_rd.l2_miss.snoop_missREL_CALL counts the number of near relative CALL branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)This event counts the number of store uops retired. (Precise Event - PEBS) (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0004Counts all demand rfo's Counts prefetch (that bring data to L2) RFOs that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0100offcore_response.pf_data_rd.llc_miss_local.dramoffcore_response.pf_ifetch.llc_miss_local.dramThis event counts the number of cycles with at least one slow LEA uop being allocated. A uop is generally considered as slow LEA if it has three sources (for example, two sources and immediate) regardless of whether it is a result of LEA instruction or not. Examples of the slow LEA uop are or uops with base, index, and offset source operands using base and index reqisters, where base is EBR/RBP/R13, using RIP relative or 16-bit addressing modes. See the Intel 64 and IA-32 Architectures Optimization Reference Manual for more details about slow LEA instructionsevent=0x60,period=2000000,umask=0x4event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7fffevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2ffREQUEST = DATA_IN and RESPONSE = LOCAL_CACHEREQUEST = DEMAND_DATA_RD and RESPONSE = LLC_HIT_NO_OTHER_COREevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f10event=0xb7,period=100000,umask=0x1,offcore_rsp=0xf803offcore_response.demand_data_rd.other_local_dramREQUEST = PF_DATA and RESPONSE = ANY_DRAM AND REMOTE_FWDREQUEST = PF_DATA and RESPONSE = ANY_LLC_MISSevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10120Counts all demand & prefetch RFOs that hit in the L3Counts all demand & prefetch RFOs that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts all demand data writes (RFOs) that have any response typeNumber of SSE/AVX computational 512-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT14 RCP14 FM(N)ADD/SUB. FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00491Counts demand data reads that miss the L3 and clean or shared data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00080offcore_response.pf_l3_rfo.l3_miss.any_snoopoffcore_response.pf_l3_rfo.l3_miss.remote_hit_forwardread requests to memory controller. Unit: uncore_imc event=0x35,umask=0x21,config1=0x40e33read requests from remote home agent. Unit: uncore_cha unc_cha_requests.writesevent=0x83,ch_mask=0x04,fc_mask=0x07,umask=0x04LLC_MISSES.PCIE_WRITEevent=0x53,umask=0x01unc_cha_sf_eviction.s_stateunc_iio_clockticksPCIe Completion Buffer Inserts of completions with data: Part 1. Unit: uncore_iio PCIe Completion Buffer occupancy of completions with data: Part 1. Unit: uncore_iio unc_iio_txn_req_by_cpu.mem_write.part2unc_iio_txn_req_of_cpu.mem_read.part2Counts every peer to peer read request of up to a 64 byte transaction made by IIO Part2 to the MMIO space of an IIO target. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busNumber of reads in which direct to core transaction were overridden. Unit: uncore_m2m Multi-socket cacheline Directory lookup (cacheline found in S state). Unit: uncore_m2m Multi-socket cacheline Directory update from A to S. Unit: uncore_m2m Counts when the M2M (Mesh to Memory) issues reads to the iMC (Memory Controller)unc_m2m_txc_ad_occupancyCycles Intel UPI is in L1 power mode (shutdown). Unit: uncore_upi FLITs received which bypassed the Slot0 Receive Buffer. Unit: uncore_upi unc_upi_rxl_bypassed.slot1event=0x3,umask=0x97Shows legal flit time (hides impact of L0p and L0c).; Count Data Flits (which consume all slots), but how much to count is based on Slot0-2 mask, so count can be 0-3 depending on which slots are enabled for counting.Counts when the Intel Ultra Path Interconnect(UPI) transmits an idle FLIT(80 bit FLow control unITs).  Every UPI cycle must be sending either data FLITs, protocol/credit FLITs or idle FLITsocr.all_data_rd.l3_hit.no_snoop_neededOCR.ALL_DATA_RD.L3_HIT_E.HITM_OTHER_CORE  OCR.ALL_DATA_RD.L3_HIT_E.HITM_OTHER_COREOCR.ALL_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWD  OCR.ALL_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200491OCR.ALL_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWD  OCR.ALL_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDocr.all_data_rd.l3_hit_s.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F800807F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4000807F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8002007F7ocr.all_reads.l3_hit_m.hit_other_core_fwdOCR.ALL_READS.L3_HIT_S.HIT_OTHER_CORE_FWD  OCR.ALL_READS.L3_HIT_S.HIT_OTHER_CORE_FWDocr.all_rfo.l3_hit_e.hitm_other_coreocr.demand_data_rd.l3_hit.snoop_missocr.demand_data_rd.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_M.NO_SNOOP_NEEDEDocr.other.l3_hit.snoop_noneocr.other.l3_hit_m.hit_other_core_fwdocr.other.l3_hit_s.any_snoopCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_HIT.ANY_SNOOP OCR.PF_L1D_AND_SW.L3_HIT.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040400ocr.pf_l2_data_rd.l3_hit_e.hit_other_core_fwdocr.pf_l2_data_rd.l3_hit_f.hitm_other_coreCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDocr.pf_l2_rfo.l3_hit.snoop_hit_with_fwdocr.pf_l2_rfo.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200020ocr.pf_l2_rfo.l3_hit_f.hit_other_core_no_fwdocr.pf_l2_rfo.l3_hit_f.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100080ocr.pf_l3_rfo.l3_hit_e.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200100ocr.pf_l3_rfo.l3_hit_s.hit_other_core_no_fwdocr.pf_l3_rfo.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_hit_f.any_snoopoffcore_response.all_pf_rfo.l3_hit_f.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_F.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_hit_s.any_snoopThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDoffcore_response.all_pf_rfo.supplier_none.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_F.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_F.SNOOP_NONEoffcore_response.all_reads.l3_hit_s.any_snoopThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_S.SNOOP_MISSoffcore_response.all_reads.supplier_none.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10000207F7This event is deprecated. Refer to new event OCR.ALL_READS.SUPPLIER_NONE.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_F.SNOOP_MISSoffcore_response.all_rfo.supplier_none.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_E.SNOOP_NONEoffcore_response.demand_code_rd.l3_hit_f.no_snoop_neededoffcore_response.demand_code_rd.l3_hit_s.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_E.SNOOP_NONEoffcore_response.demand_data_rd.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.ANY_SNOOPoffcore_response.other.l3_hit_f.no_snoop_neededoffcore_response.other.pmm_hit_local_pmm.snoop_not_neededThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_F.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_M.HITM_OTHER_COREoffcore_response.pf_l1d_and_sw.pmm_hit_local_pmm.snoop_noneoffcore_response.pf_l1d_and_sw.supplier_none.hitm_other_coreoffcore_response.pf_l1d_and_sw.supplier_none.hit_other_core_no_fwdoffcore_response.pf_l2_data_rd.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.pf_l2_rfo.l3_hit_e.snoop_missThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_S.HITM_OTHER_COREoffcore_response.pf_l2_rfo.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_RFO.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_data_rd.l3_hit_f.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_F.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.SUPPLIER_NONE.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020100OCR.ALL_DATA_RD.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_DATA_RD.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000491ocr.all_pf_data_rd.l3_miss.remote_hitmOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISSOCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEocr.all_reads.l3_miss.no_snoop_neededocr.all_rfo.l3_miss.snoop_missocr.all_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000004ocr.demand_code_rd.l3_miss_remote_hop1_dram.no_snoop_neededocr.demand_data_rd.l3_miss.hit_other_core_no_fwdocr.demand_data_rd.l3_miss_local_dram.hit_other_core_fwdocr.demand_data_rd.l3_miss_local_dram.snoop_noneCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDocr.other.l3_miss.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804008000ocr.other.l3_miss_local_dram.hit_other_core_no_fwdocr.other.l3_miss_remote_hop1_dram.any_snoopCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDocr.pf_l2_data_rd.l3_miss.snoop_missCounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.SNOOP_MISSocr.pf_l3_data_rd.l3_miss.hitm_other_coreocr.pf_l3_data_rd.l3_miss.no_snoop_neededCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPocr.pf_l3_rfo.l3_miss.remote_hit_forwardCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.SNOOP_MISSocr.pf_l3_rfo.l3_miss_local_dram.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDocr.pf_l3_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdocr.pf_l3_rfo.l3_miss_remote_hop1_dram.snoop_missoffcore_response.all_pf_data_rd.l3_miss_local_dram.hit_other_core_fwdoffcore_response.all_pf_rfo.l3_miss.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_FWDoffcore_response.all_reads.l3_miss_remote_hop1_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_code_rd.l3_miss_remote_hop1_dram.any_snoopoffcore_response.demand_code_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdoffcore_response.demand_rfo.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_data_rd.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdoffcore_response.pf_l3_data_rd.l3_miss_remote_hop1_dram.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.SNOOP_MISSOCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOP OCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPOCR.ALL_READS.SUPPLIER_NONE.SNOOP_MISSCounts demand data reads  OCR.DEMAND_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests have any response typeCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDunc_m_pmm_cmd1.wrunc_m_tagchk.miss_dirtyTag Check; Dirtyevent=0x48,period=1000003,umask=0x2Counts number of cycles a demand request has waited due to L1D Fill Buffer (FB) unavailablability. Demand requests include cacheable/uncacheable demand load, store, lock or SW prefetch accessesCounts number of L1D misses that are outstanding in each cycle, that is each cycle the number of Fill Buffers (FB) outstanding required by Demand Reads. FB either is held by demand loads, or it is held by non-demand loads and gets hit at least once by demand. The valid outstanding interval is defined until the FB deallocation by one of the following ways: from FB allocation, if FB is allocated by demand from the demand Hit FB, if it is allocated by hardware or software prefetch. Note: In the L1D, a Demand Read contains cacheable or noncacheable demand loads, including ones causing cache-line splits and reads due to page walks resulted from any request typeModified cache lines that are evicted by L2 cache when triggered by an L2 cache fillocr.demand_code_rd.l3_hit.snoop_not_neededCounts demand data reads that hit a cacheline in the L3 where a snoop hit in another cores caches, data forwarding is required as the data is modifiedCounts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that hit a cacheline in the L3 where a snoop was sent or notocr.hwpf_l2_data_rd.l3_hit.snoop_hitmevent=0x60,cmask=1,period=1000003,umask=0x8For every cycle, increments by the number of outstanding demand data read requests pendingFor every cycle, increments by the number of outstanding demand data read requests pending.   Requests are considered outstanding from the time they miss the core's L2 cache until the transaction completion message is sent to the requestorsq_misc.sq_fullCycles the queue waiting for offcore responses is fullevent=0xc7,period=100003,umask=0x10event=0xab,cmask=1,edge=1,period=100003,umask=0x2Decode Stream Buffer (DSB)-to-MITE transitions countDecode Stream Buffer (DSB) is a Uop-cache that holds translations of previously fetched instructions that were decoded by the legacy x86 decode pipeline (MITE). This event counts fetch penalty cycles when a transition occurs from DSB to MITECounts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularityevent=0x79,period=100003,umask=0x301 - CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_DISTRIBUTED if #SMT_on else 0event=0xc8,period=100003,umask=0x80Cycles where at least one demand data read request known to have missed the L3 cache is pendingCounts Unfriendly TSX abort triggered by a nest count that is too deepSpeculatively counts the number of TSX aborts due to a data capacity limitation for transactional readsevent=0x54,period=100003,umask=0x2Speculatively counts the number of Transactional Synchronization Extensions (TSX) aborts due to a data capacity limitation for transactional writesocr.hwpf_l2_rfo.any_responseMispredicted indirect CALL instructions retired (Precise event)Counts the cycles when optimal number of uops is delivered by the LSD (Loop-stream detector)TMA slots wasted due to incorrect speculation by branch mispredictionsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x18300004771000000000 * ( UNC_CHA_TOR_OCCUPANCY.IA_MISS_DRD / UNC_CHA_TOR_INSERTS.IA_MISS_DRD ) / ( cha_0@event\=0x0@ / duration_time )Counts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were not supplied by the local socket's L1, L2, or L3 caches and were supplied by the local socketcore_snoop_response.s_fwd_feCounts demand data reads that were supplied by DRAM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by PMM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those PMM accesses that are controlled by the close SNC Cluster2LM Tag Check : Read Hit in Near Memory Cache. Unit: uncore_imc unc_m_tagchk.nm_wr_hitunc_cha_tor_inserts.ioevent=0x36,umask=0xC817FE01unc_cha_tor_occupancy.ia_miss_drd_localTOR Inserts : DRds issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x35,umask=0xC886FE01event=0x35,umask=0xc86ffe01event=0xc0,ch_mask=0x10,fc_mask=0x07,umask=0x01event=0x83,ch_mask=0x40,fc_mask=0x07,umask=0x01PCIe Completion Buffer Inserts of completions with data: Part 5. Unit: uncore_iio PCIe Completion Buffer Inserts of completions with data: Part 6. Unit: uncore_iio Page walks completed due to a demand data store to a 1G pageCounts the number of L2 Cache accesses that miss the L2 and get BBL reject  short and long rejects (includes those counted in L2_reject_XQ.any). Counts on a per core basisCounts the number of cacheable memory requests that access the LLC. Counts on a per core basismem_bound_stalls.ifetchCounts the number of load uops retired that performed one or more locks  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0044Counts all code reads that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedCounts streaming stores that were supplied by the L3 cacheocr.uc_rd.l3_hitCounts the number of cycles the floating point divider is busy.  Does not imply a stall waiting for the dividerCounts the number of instruction cache missesocr.demand_data_and_l1pf_rd.dramocr.demand_data_and_l1pf_rd.outstandingCounts the total number of instructions that retired. For instructions that consist of multiple uops, this event counts the retirement of the last uop of the instruction. This event continues counting during hardware interrupts, traps, and inside interrupt handlers. This event uses fixed counter 0 (Precise event)Counts the total number of instructions that retired. For instructions that consist of multiple uops, this event counts the retirement of the last uop of the instruction. This event continues counting during hardware interrupts, traps, and inside interrupt handlers. This event uses a programmable general purpose performance counter (Precise event)topdown_fe_bound.itlbCounts the number of uops that are from complex flows issued by the micro-sequencer (MS) (Precise event)unc_cha_tor_inserts.ia_hit_drd_optTOR Inserts : UCRdFs issued by iA Cores that Missed LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x35,umask=0xC86FFE01Free running counter that increments for integrated IO (IIO) traffic controller clockticksData requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 5event=0x8,period=200003,umask=0x80Counts the number of page walks completed due to load DTLB misses to any page sizeTotal number of retired InstructionsInstructions per Load (lower number means higher occurrence rate). Unit: cpu_core Counts the numer of issue slots  that result in retirement slots. Unit: cpu_atom 5 * CPU_CLK_UNHALTED.COREInstructions per Far Branch. Unit: cpu_atom Branch_Mispredict_RatioDemand Data Read requests that hit L2 cache. Unit: cpu_core assists.sse_avx_mixevent=0x5,period=1000003,umask=0x84Counts randomly selected loads when the latency from first dispatch to completion is greater than 16 cycles  Supports address when precise (Must be precise). Unit: cpu_core Counts randomly selected loads when the latency from first dispatch to completion is greater than 512 cycles  Supports address when precise (Must be precise). Unit: cpu_core core_power.license_3topdown_bad_speculation.nukeCounts the number of issue slots every cycle that were not delivered by the frontend due to ITLB misses. Unit: cpu_atom cpu_clk_unhalted.pause_instCycles stalled due to no store buffers available. (not including draining form sync). Unit: cpu_core Incoming VC1 write request. Unit: uncore_imc incoming read request page status is Page Empty. Unit: uncore_imc ACT command for a write request sent to DRAM. Unit: uncore_imc event=0x49,period=2000003,umask=0xePage walks completed due to a demand data store to a 2M/4M page. Unit: cpu_core Read requests with true-miss in L2 cache.[This event is alias to L2_RQSTS.MISS]Retired load instructions which data sources missed L3 but serviced from local DRAM  Supports address when preciseevent=0xcf,period=100003,umask=0x8event=0xcf,period=100003,umask=0x1event=0x2a,period=100003,umask=0x1,offcore_rsp=0x84000800event=0x2a,period=100003,umask=0x1,offcore_rsp=0x73C000004Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM or PMM attached to another socketcpu_clk_unhalted.c01TMA slots wasted due to incorrect speculationsevent=0x5,umask=0x00000000c1event=0x20,umask=0x0000000002event=0x5,umask=0x00000000e0event=0xe0,umask=0x0000000004unc_m_pmm_rpq_occupancy.no_gnt_sch1unc_m_pre_count.ufill_pch0M2M Clockticks. Unit: uncore_m2m M3UPI Clockticks. Unit: uncore_m3upi CHA Clockticks. Unit: uncore_cha event=0x35,umask=0x00c897fe01unc_upi_txl_flits.prothdrunc_upi_rxl_flits.llcrdevent=0x3,umask=0x0000000027unc_upi_rxl_inserts.slot2Read request for 4 bytes made by IIO Part2 to Memory. Unit: uncore_iio event=0xc1,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000001Number Transactions requested by the CPU : Core reading from Cards MMIO space. Unit: uncore_iio unc_cha_tor_inserts.snps_from_remevent=0x35,umask=0x0000000008event=0x35,umask=0x00c817ff01event=0x35,umask=0x00C88FFF01TOR Inserts;CLFlushOpt from Local IA. Unit: uncore_cha TOR Occupancy; LLCPrefRFO hits from local IA. Unit: uncore_cha TOR Occupancy; Hits from local IO. Unit: uncore_cha event=0x36,umask=0x0000000004TOR Occupancy : RRQ. Unit: uncore_cha unc_cha_tor_occupancy.loc_iounc_cha_tor_occupancy.match_opcunc_cha_tor_occupancy.premorph_opcevent=0x37,umask=0x0000000001event=0x37,umask=0x0000000004event=0x35,umask=0xcc3fff01unc_cha_tor_inserts.ia_miss_drd_pref_remote_pmmevent=0x35,umask=0x00C80EFE01event=0x35,umask=0x00C867FE01unc_cha_tor_inserts.ia_miss_local_wcilf_ddrunc_cha_tor_occupancy.ia_miss_drd_local_ddrevent=0x36,umask=0x00c8d7ff01unc_cha_tor_occupancy.ia_wcilfTOR Occupancy : WCiLs issued by iA Cores targeting PMM that missed the LLC - HOMed locally. Unit: uncore_cha bp_de_redirectl2_request_g2.ic_rd_sized_ncl2_cache_req_stat.ls_rd_blk_l_hit_sCore to L2 cacheable request access status (not including L2 Prefetch). Instruction cache hit clean line in L2ex_ret_condTagged IBS Ops. Number of times an op could not be tagged by IBS because of a previous tagged op that has not retiredfpu_pipe_assignment.dualfpu_pipe_assignment.dual2Total number of fp uOps on pipe 3event=0x5,umask=0x02event=0x29,umask=0x02Total Page Table Walks on I-sideevent=0xaf,umask=0x20Cycles where a dispatch group is valid but does not get dispatched due to a token stall. ALSQ 1 Tokens unavailableL2 Cache Misses from L1 Data Cache Missesl2_pf_miss_l2_hit_l3 + l2_pf_miss_l2_l3l2_dtlb_missesThe number of instruction fetches that miss in both the L1 and L2 TLBs. Instruction fetches to a 1GB pageNumber of Ops that are candidates for optimization (have Z-bit either set or pass). This is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemesNumber of retired CLFLUSH instructionsls_refills_from_sys.ls_mabresp_lcl_dramls_sw_pf_dc_fill.ls_mabresp_lcl_dramevent=0x5a,umask=0x08Hardware Prefetch Data Cache Fills by Data Source. From another cache (home node local)Cycles where a dispatch group is valid but does not get dispatched due to a token stall. Floating point register file resource stall. Applies to all FP ops that have a destination registerevent=0x18e,umask=0x1fInstruction Cache Hit. Counts various IC tag related hit and miss eventsAny Data Cache Fills by Data Source. From Local L2 to the corels_sw_pf_dc_fills.ext_cache_remoteHardware Prefetch Data Cache Fills by Data Source. From cache of different CCX in same nodels_hw_pf_dc_fills.int_cacheCycles where a dispatch group is valid but does not get dispatched due to a Token Stall. Also counts cycles when the thread is not selected to dispatch but would have been stalled due to a Token Stall. FP Flush recovery stallls_sw_pf_dc_fills.allL2 cache requests: data cache storesevent=0x71,umask=0x80Retired CPUID instructionsevent=0x1c1Ops tagged by IBSOps tagged by IBS that retiredRead data beats (64 bytes) for local processor at Coherent Station (CS) 3Read data beats (64 bytes) for local processor at Coherent Station (CS) 4event=0x19f,umask=0x7ffevent=0x19f,umask=0xbferemote_processor_read_data_beats_cs9remote_processor_write_data_beats_cs6Write data beats (64 bytes) for remote processor at Coherent Station (CS) 8remote_socket_upstream_write_beats_iom2local_socket_inf0_outbound_data_beats_ccm1remote_socket_inf0_inbound_data_beats_ccm0event=0x4df,umask=0xbfefp_ops_retired_by_type.scalar_subfp_ops_retired_by_type.scalar_cvtevent=0xb,umask=0x90event=0xb,umask=0xffRetired 128-bit packed floating-point multiply-accumulate opsevent=0xc,umask=0x10fp_pack_ops_retired.fp256_otherevent=0xd,umask=0x0bNumber of clocks used by the data bus for readsls_tlb_flush.alll2_request_g1.all_no_prefetch + l2_pf_hit_l2.all + l2_pf_miss_l2_hit_l3.all + l2_pf_miss_l2_l3.allL1 data cache fills from a different NUMA nodel1_demand_data_cache_fills_from_near_memoryDRAM read data for remote processorRemote socket upstream DMA read dataremote_socket_upstream_dma_write_dataEstimated combined memory bandwidthUNC_CBO_TWO_HYPHddr write-cycles event. Unit: uncore_sys_ddr_pmu uncore_sys_ccn_pmu,FR_RETIRED_FASTPATH_DOUBLE_OP_INSTRUCTIONSFR_NUMBER_OF_BREAKPOINTS_FOR_DR3L2_CACHE_WBCYCLES_NONIDLE_NEON_INTEVENT_29HEVENT_34HEVENT_3BHEVENT_4BHEVENT_68HEVENT_DBHEVENT_E4HEVENT_F5HEVENT_FDHSTREX_FAILEDISSUE_DNOT_DISPATCH_ANY_INSTRDMBEVENT_101HEVENT_10EHEVENT_12EHEVENT_139HEVENT_1B8HEVENT_1C5HEVENT_1C8HEVENT_1E0HEVENT_1FDHEVENT_201HEVENT_2B2HEVENT_2B6HEVENT_2B7HEVENT_301HEVENT_36CHEVENT_37FHEVENT_3B9HBUS_ACCESS_SHAREDREMOTE_ACCESSdn_rxreq_trk_fullhnf_pocq_retryhnf_qos_pocqhnf_snp_fwdedrnd_s0_wdata_beatsrnd_wrt_alloccxha_snppcrd_lnk1_stallcxra_req_pcrd_stalls_lnk1cxla_avg_rx_tlp_sz_ccix_msgcxla_rx_tlp_buffer_fullclkdiv2_arbitrateEIEIO_INSTR_COMPLETEDVTQ_SUSPENDS_DUE_TO_CTX_CHANGELSU_RA_LATCH_STALLFALL_THROUGH_BRANCHES_PROCESSEDFIRST_SPECULATIVE_BRANCH_BUFFER_RESOLVED_CORRECTLYTHIRD_SPECULATION_BUFFER_ACTIVEGPR_ISSUE_QUEUE_ENTRIES_OVER_THRESHOLDFP_STORE_DOUBLE_COMPLETES_IN_LSUBUS_RETRY_DUE_TO_COLLISIONREJECT_COMPLETION_STALLTOTAL_ALLOCATED_TO_DLFBDATA_MMU_BUSYCRIT_INPUT_INTR_LATENCY_CYCLESSTALLS_NO_CAQ_OR_COBtlb-reloadnopINTEL_ATOM_GOLDMONTINTEL_ICELAKE_XEONPPC_E500PPC_POWER8SCex_ret_instr{"type": "procexec"{"type": "callchain"Cor;PipelineFP_Arith_Utilization_SMTBranches;Fed;InsType( cpu@ITLB_MISSES.WALK_DURATION\,cmask\=1@ + cpu@DTLB_LOAD_MISSES.WALK_DURATION\,cmask\=1@ + cpu@DTLB_STORE_MISSES.WALK_DURATION\,cmask\=1@ + 7 * ( DTLB_STORE_MISSES.WALK_COMPLETED + DTLB_LOAD_MISSES.WALK_COMPLETED + ITLB_MISSES.WALK_COMPLETED ) ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )SMT_2T_UtilizationC6_Core_ResidencyThis event counts duration of L1D miss outstanding, that is each cycle number of Fill Buffers (FB) outstanding required by Demand Reads. FB either is held by demand loads, or it is held by non-demand loads and gets hit at least once by demand. The valid outstanding interval is defined until the FB deallocation by one of the following ways: from FB allocation, if FB is allocated by demand; from the demand Hit FB, if it is allocated by hardware or software prefetch.
Note: In the L1D, a Demand Read contains cacheable or noncacheable demand loads, including ones causing cache-line splits and reads due to page walks resulted from any request typeCycles with L1D load Misses outstanding from any thread on physical coreThis event counts the number of demand Data Read requests that hit L2 cache. Only not rejected loads are countedevent=0x24,period=200003,umask=0x22This event counts the number of L2 cache accesses when fetching instructionsevent=0xd2,period=20011,umask=0x2This is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were L3 hit and a cross-core snoop hit in the on-pkg core cache  Supports address when precise.  Spec update: BDM100 (Precise event)mem_load_uops_l3_miss_retired.remote_dramRetired load uops which data sources were load uops missed L1 but hit FB due to preceding miss to the same cache line with data not ready. (Precise Event - PEBS)  Supports address when precise (Precise event)event=0xd1,period=100003,umask=0x2event=0xd0,period=100003,umask=0x12event=0x60,cmask=1,period=2000003,umask=0x8This event counts cycles during which the demand fetch waits for data (wfdM104H) from L2 or iSB (opportunistic hit)idq.all_dsb_cycles_4_uopsThis event counts cycles during which uops are being delivered to Instruction Decode Queue (IDQ) while the Microcode Sequenser (MS) is busy. Counting includes uops that may bypass the IDQ. Uops maybe initiated by Decode Stream Buffer (DSB) or MITEevent=0x79,period=2000003,umask=0x20idq.ms_uopsmem_trans_retired.load_latency_gt_256event=0x5,period=2000003,umask=0x1Speculative cache line split load uops dispatched to L1 cachertm_retired.aborted_misc2Counts the number of times a XBEGIN instruction was executed inside an HLE transactional regiontx_mem.abort_hle_store_to_elided_lockarith.fpu_div_activeevent=0x88,period=200003,umask=0x82br_inst_retired.all_branches_pebsDirect and indirect macro near call instructions retired (captured in ring 3). (Precise Event - PEBS) (Precise event)Core cycles when at least one thread on the physical core is not in halt statecycle_activity.cycles_l1d_pendingcycle_activity.stalls_l2_missmachine_clears.countother_assists.any_wb_assistCycles stalled due to re-order buffer fullevent=0xe,period=2000003,umask=0x1uncore cacheMMIO writes. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox LLC prefetch misses for code reads. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox llc_references.streaming_partialwrite requests to home agent. Unit: uncore_ha unc_m_power_channel_ppduncore powerprochot_external_cycles %Load misses in all DTLB levels that cause page walks  Spec update: BDM69This event counts load misses in all DTLB levels that cause a completed page walk (4K page size). The page walk can end with or without a fault  Spec update: BDM69event=0x49,period=100003,umask=0x4itlb_misses.walk_durationSTLB flush attemptsThis event counts retired load uops which data sources were data hits in the last-level (L3) cache without snoops required  Supports address when precise.  Spec update: BDM100 (Precise event)offcore_response.all_pf_data_rd.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010122offcore_response.corewb.l3_hit.snoop_hit_no_fwdoffcore_response.demand_data_rd.l3_hit.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400028000Counts prefetch (that bring data to L2) data reads have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020200offcore_response.pf_l3_data_rd.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020080Number of SSE/AVX computational 256-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 8 calculations per elementevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000120offcore_response.all_pf_rfo.l3_miss_local_dram.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000120offcore_response.all_pf_rfo.l3_miss_local_dram.snoop_not_neededoffcore_response.all_rfo.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000122offcore_response.demand_code_rd.l3_miss_local_dram.snoop_hitmoffcore_response.demand_code_rd.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000040offcore_response.pf_l2_code_rd.l3_miss.snoop_noneoffcore_response.pf_l2_data_rd.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000080Mispredicted conditional branch instructions retired (Precise event)L3 Lookup any request that access cache and found line in M-state. Unit: uncore_cbox event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0604000244offcore_response.all_reads.llc_miss.local_dramevent=0,umask=0x2event=0x2c,period=200000,umask=0x44event=0x2c,period=200000,umask=0x41event=0x28,period=200000,umask=0x42event=0x29,period=200000,umask=0x42l2_ld.self.prefetch.m_stateevent=0x30,period=200000,umask=0x48event=0x2e,period=200000,umask=0x42event=0x2e,period=200000,umask=0x5fSIMD packed shift micro-ops executedx87_comp_ops_exe.any.arx87_comp_ops_exe.fxch.sDecode stall due to PFB emptyInstruction fetchesevent=0x7,period=2000000,umask=0x10event=0x7,period=200000,umask=0x86Memory bus transactionshw_int_rcvOnly taken macro conditional branch instructionsbr_inst_type_retired.dir_callMultiply operations retiredDuration of I-Side page walksevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400003091Counts data reads (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000043010Counts data reads generated by L1 or L2 prefetchers that hit the L2 cacheCounts data reads generated by L1 or L2 prefetchers that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_request.l2_miss.hit_other_core_no_fwdevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040022offcore_response.any_rfo.l2_miss.hitm_other_coreCounts reads for ownership (RFO) requests (demand & prefetch) that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of writeback transactions caused by L1 or L2 cache evictions that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that are outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000800Load uops that split a page (Precise event capable) (Must be precise)Counts the number of issue slots per core cycle that were not consumed by the backend due to either a full resource  in the backend (RESOURCE_FULL) or due to the processor recovering from some event (RECOVERY)Unfilled issue slots per cycle to recoverld_blocks.all_blockuops_not_delivered.anyCounts uops which retired (Must be precise)event=0x5,period=200003,umask=0x1event=0xb7,period=100007,umask=0x1,offcore_rsp=0x40000032b7Counts requests to the uncore subsystem hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts bus lock and split lock requests true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000004Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts any data writes to uncacheable write combining (USWC) memory region  true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts any data writes to uncacheable write combining (USWC) memory region  outstanding, per cycle, from the time of the L2 miss to when any response is receivedoffcore_response.sw_prefetch.outstandingDemand requests that miss L2 cache  Spec update: HSD78, HSM80Retired load uops with L3 cache hits as data sources  Supports address when precise.  Spec update: HSD74, HSD29, HSD25, HSM26, HSM30 (Precise event)offcore_response.all_reads.l3_hit.hitm_other_coreoffcore_response.demand_rfo.l3_hit.hitm_other_coreIncrement each cycle # of uops delivered to IDQ when MS_busy by MITE. Set Cmask = 1 to count cyclesmiss the L3 and the data is returned from local dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00080Cycles with pending memory loads. Set Cmask=2 to count cycleCycles where at least 4 uops were executed per-thread  Spec update: HSD144, HSD30, HSM31A cross-core snoop resulted from L3 Eviction which hits a modified line in some processor core. Unit: uncore_cbox A cross-core snoop resulted from L3 Eviction which hits a non-modified line in some processor coreCompleted page walks due to demand load misses that caused 4K page walks in any TLB levelsThis event counts store operations from a 2M page that miss the first DTLB level but hit the second and do not cause page walksevent=0xae,period=100003,umask=0x1Counts the number of ITLB flushes, includes 4k/2M/4M pagesCompleted page walks in ITLB of any page sizeNumber of DTLB page walker loads that hit in the L1+FBNumber of ITLB page walker hits in the L1+FBNumber of ITLB page walker hits in Memory  Spec update: HSD25mem_load_uops_llc_hit_retired.xsnp_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0091( 1 * ( FP_COMP_OPS_EXE.SSE_SCALAR_SINGLE + FP_COMP_OPS_EXE.SSE_SCALAR_DOUBLE ) + 2 * FP_COMP_OPS_EXE.SSE_PACKED_DOUBLE + 4 * ( FP_COMP_OPS_EXE.SSE_PACKED_SINGLE + SIMD_FP_256.PACKED_DOUBLE ) + 8 * SIMD_FP_256.PACKED_SINGLE ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )event=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400091Number of self-modifying-code machine clears detectedunc_cbo_cache_lookup.ievent=0x34,umask=0x08unc_cbo_cache_lookup.read_filterevent=0xd3,period=100007,umask=0x3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c03f7offcore_response.other.lru_hintsoffcore_response.pf_llc_data_rd.llc_hit.hitm_other_coreCounts all data/code/rfo reads (demand & prefetch) that miss the LLC  and the data returned from local dramCounts all data/code/rfo reads (demand & prefetch) that miss the LLC  and the data forwarded from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc20004Number of occurrences waiting for the checkpoints in Resource Allocation Table (RAT) to be recovered after Nuke due to all other cases except JEClear (e.g. whenever a ucode assist is needed like SSE exception, memory disambiguation, etc.)unc_m_act_count.rdevent=0xb,filter_band0=12Retired load uops which data sources were hits in LLC without snoops requiredThis event counts line-splitted load uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K) (Precise event)This event counts line-splitted store uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K) (Precise event)event=0x17,period=2000003,umask=0x1resource_stalls.mem_rsevent=0xa2,period=2000003,umask=0xf0resource_stalls2.ooo_rsrcuops_dispatched.coreunc_m_rpq_occupancyevent=0x4,period=200003,umask=0x1mem_uops_retired.l2_hit_loadsCounts Demand cacheable data and L1 prefetch data read requests  that accounts for responses which hit its own tile's L2 with data in E stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000070event=0xb7,period=100007,umask=0x1,offcore_rsp=0x08004032f7offcore_response.any_request.l2_hit_far_tileCounts demand code reads and prefetch code reads that accounts for responses which hit its own tile's L2 with data in S stateoffcore_response.partial_reads.l2_hit_far_tile_mCounts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateoffcore_response.pf_l1_data_rd.l2_hit_this_tile_sevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008002000offcore_response.pf_l2_rfo.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400200offcore_response.uc_code_reads.l2_hit_this_tile_eCounts UC code reads (valid only for Outstanding response type)  that accounts for responses which hit its own tile's L2 with data in F stateCounts the number of vector SSE, AVX, AVX2, AVX-512 micro-ops retired. More specifically, it counts packed SSE, AVX, AVX2, AVX-512 micro-ops (both floating point and integer) except for loads (memory-to-register mov-type micro-ops), packed byte and word multipliesevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200044Counts Demand cacheable data write requests  that accounts for data responses from MCDRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000400offcore_response.demand_code_rd.ddrevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200004Counts Demand cacheable data writes that accounts for responses from MCDRAM (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100402000Counts L2 code HW prefetches that accounts for data responses from DRAM Localoffcore_response.pf_l2_code_rd.mcdram_nearoffcore_response.pf_software.ddr_farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180601000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800200Counts UC code reads (valid only for Outstanding response type)  that accounts for data responses from DRAM FarCounts the number of near relative CALL branch instructions retired (Precise event)Fixed Counter: Counts the number of instructions retiredno_alloc_cycles.rat_stallCounts any retired load that was pushed into the recycle queue for any reasonl1d_all_ref.cacheableL1 data cache readsl1d_cache_st.e_statel1d_prefetch.triggersl1d_wb_l2.i_stateevent=0x26,period=200000,umask=0xf0l2_lines_in.e_stateL2 Load transactionsL2 demand lock RFOs in E stateAll demand L2 lock RFOs that hit the cachel2_write.lock.s_stateMemory instructions retired above 16384 clocks (Precise Event)mem_inst_retired.latency_above_threshold_8192Retired loads that hit the L2 cache (Precise Event)event=0xf,period=10000,umask=0x20offcore_response.any_data.any_locationOffcore code reads satisfied by the LLC  and HITM in a sibling coreoffcore_response.any_request.remote_cache_hitmoffcore_response.data_ifetch.any_locationevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4777event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F33offcore_response.data_in.llc_hit_other_core_hitmOffcore data reads, RFO's and prefetches that HITM in a remote cacheOffcore demand data requests satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x101offcore_response.other.local_cache_dramoffcore_response.pf_ifetch.any_locationOffcore prefetch requests satisfied by the LLC  and HITM in a sibling coreevent=0x10,period=2000000,umask=0x80Offcore data reads satisfied by a remote DRAMoffcore_response.any_ifetch.local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x20FFOffcore writebacks to any DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2077offcore_response.data_in.local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF804offcore_response.demand_rfo.any_dramI/O transactionsMispredicted branches executedbr_misp_exec.directAny Instruction Length Decoder stall cyclesevent=0x20,period=2000000,umask=0x1event=0xb1,any=1,cmask=1,inv=1,period=2000000,umask=0x1fuops_executed.port1Uops executed on port 3 (core count)uops_executed.port5event=0xe,cmask=1,inv=1,period=2000000,umask=0x1itlb_flushmem_load_retired.l1_missmem_load_retired.l3_missoffcore_response.demand_code_rd.l3_hit_m.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4001C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100001offcore_response.demand_data_rd.l4_hit_local_l4.snoop_noneoffcore_response.demand_rfo.l3_hit_s.snoop_missoffcore_response.demand_rfo.supplier_none.snoop_noneoffcore_response.other.l4_hit_local_l4.snoop_hitmsw_prefetch_access.t0Counts randomly selected loads when the latency from first dispatch to completion is greater than 32 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)Counts randomly selected loads when the latency from first dispatch to completion is greater than 512 cycles  Supports address when precise (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x44000004offcore_response.demand_data_rd.l3_miss.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000048000offcore_response.other.l3_hit_s.snoop_non_drammemory_disambiguation.history_resetAll (macro) branch instructions retired  Spec update: SKL091 (Must be precise)Cycles total of 3 uops are executed on all ports and Reservation Station (RS) was not emptyCounts the number of instructions retired from execution. For instructions that consist of multiple micro-ops, Counts the retirement of the last micro-op of the instruction. Counting continues during hardware interrupts, traps, and inside interrupt handlers. Notes: INST_RETIRED.ANY is counted by a designated fixed counter, leaving the four (eight when Hyperthreading is disabled) programmable counters available for other events. INST_RETIRED.ANY_P is counted by a programmable counter and it is an architectural performance event. Counting: Faulting executions of GETSEC/VM entry/VM Exit/MWait will not count as retired instructionsCounts cycles during which the reservation station (RS) is empty for the thread.; Note: In ST-mode, not active thread should drive 0. This is usually caused by severely costly branch mispredictions, or allocator/FE issues100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) * ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD) / #(CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (OFFCORE_REQUESTS_BUFFER.SQ_FULL / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) ) + ( (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( ((L1D_PEND_MISS.PENDING / ( MEM_LOAD_RETIRED.L1_MISS + MEM_LOAD_RETIRED.FB_HIT )) * cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ / CPU_CLK_UNHALTED.THREAD) / #(max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) ) DSBmiss;Fed_SMTCounts demand data loads that caused a page walk of any page size (4K/2M/4M/1G). This implies it missed in all TLB levels, but the walk need not have completedevent=0x8,cmask=1,period=100003,umask=0x10Counts 1 per cycle for each PMH that is busy with a EPT (Extended Page Table) walk for any request typeCycles when at least one PMH is busy with a page walk for code (instruction fetch) request. EPT page walk duration are excluded in SkylakeLoads missed UTLBCounts any data read (demand & prefetch) that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts demand and DCU prefetch data read that miss L2 with a snoop miss responserehabq.ld_block_std_notreadyThis event counts the number of retire loads that experienced cache line boundary splits (Precise event)This event counts all instruction fetches that miss the Instruction cache or produce memory requests. This includes uncacheable fetches. An instruction fetch miss is counted only once and not once for every cycle it is outstandingRetired load uops that split across a cacheline boundary. (Precise Event - PEBS) (Precise event)This event counts line-splitted store uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K). (Precise Event - PEBS) (Precise event)offcore_response.all_pf_code_rd.llc_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0120Counts data/code/rfo reads (demand & prefetch) that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts demand & prefetch RFOs that hit in the LLC and the snoops sent to sibling cores return clean responseCounts demand data reads that hit in the LLC and the snoops sent to sibling cores return clean responseoffcore_response.pf_llc_code_rd.llc_hit.snoop_missoffcore_response.all_rfo.llc_miss.dramThis event counts the number of cycles spent executing performance-sensitive flags-merging uops. For example, shift CL (merge_arith_flags). For more details, See the Intel 64 and IA-32 Architectures Optimization Reference Manualsnb metricsOffcore read requestsevent=0xb0,period=100000,umask=0x2REQUEST = ANY IFETCH and RESPONSE = IO_CSR_MMIOREQUEST = ANY IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x80ffREQUEST = DEMAND_DATA and RESPONSE = REMOTE_CACHE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff01REQUEST = DEMAND_IFETCH and RESPONSE = LLC_HIT_NO_OTHER_COREoffcore_response.pf_data.local_dram_and_remote_cache_hitREQUEST = PF_RFO and RESPONSE = REMOTE_CACHE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf833Loads that partially overlap an earlier storeOffcore data reads, RFOs, and prefetches satisfied by the LLC and not found in a sibling coreOffcore data reads, RFOs, and prefetches satisfied by the LLC  and HITM in a sibling coreCounts all prefetch data reads that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts all demand code reads that have any response typeCounts all demand code reads that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts prefetch (that bring data to L2) data reads that have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800004offcore_response.demand_rfo.l3_miss.remote_hit_forwardCounts L1 data cache hardware prefetch requests and software prefetch requests that miss the L3 and the modified data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00020offcore_response.pf_l3_rfo.l3_miss.remote_hitmidi_misc.wb_upgrade( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * ( FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE ) + 16 * FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )( UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART0 + UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART1 + UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART2 + UNC_IIO_DATA_REQ_OF_CPU.MEM_READ.PART3 ) * 4 / 1000000000 / duration_timecha_0@event\=0x0@Counts the number of read requests allocated into the Read Pending Queue (RPQ).  This queue is used to schedule reads out to the memory controller and to track the requests.  Requests allocate into the RPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the CHA to the iMC.  The requests deallocate after the read CAS command has been issued to DRAM.  This event counts both Isochronous and non-Isochronous requests which were issued to the RPQevent=0x35,umask=0x21,config1=0x40041e33event=0x50,umask=0x01Counts the number of transactions that trigger a configurable number of cross snoops.  Cores are snooped if the transaction looks up the cache and determines that it is necessary based on the operation type and what CoreValid bits are set.  For example, if 2 CV bits are set on a data read, the cores must have the data in S state so it is not necessary to snoop them.  However, if only 1 CV bit is set the core my have modified the data.  If the transaction was an RFO, it would need to invalidate the lines.  This event can be filtered based on who triggered the initial snoop(s)Counts only multi-socket cacheline Directory state updates memory writes issued from the HA pipe. This does not include memory write requests which are for I (Invalid) or E (Exclusive) cachelinesMulti-socket cacheline Directory state updates; Directory Updated memory write from TOR pipe. Unit: uncore_cha Counts when a a transaction with the opcode type RspSFwd Snoop Response was received which indicates a remote caching agent forwarded the data but held on to its current copy.  This is common for data and code reads that hit in a remote socket in E (Exclusive) or F (Forward) stateRead request for 4 bytes made by the CPU to IIO Part2. Unit: uncore_iio Write request of 4 bytes made to IIO Part1 by the CPU. Unit: uncore_iio unc_iio_data_req_by_cpu.mem_write.part2Peer to peer read request for 4 bytes made by a different IIO unit to IIO Part1. Unit: uncore_iio unc_iio_data_req_of_cpu.peer_write.part1Counts every peer to peer write request of 4 bytes of data made by IIO Part2 to the MMIO space of an IIO target. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busunc_iio_txn_req_by_cpu.peer_read.part0Counts every read request for up to a 64 byte transaction of data made by IIO Part2 to a unit on the main die (generally memory). In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busunc_iio_txn_req_of_cpu.peer_write.part1event=0x84,ch_mask=0x04,fc_mask=0x07,umask=0x02Inbound write (fast path) requests received by the IRP. Unit: uncore_irp Messages sent direct to core (bypassing the CHA). Unit: uncore_m2m unc_m2m_direct2core_txn_overrideunc_m2m_directory_lookup.anyAD Egress (to CMS) AllocationsCounts valid data FLITs  (80 bit FLow control unITs: 64bits of data) received from any of the 3 Intel Ultra Path Interconnect (UPI) Receive Queue slots on this UPI unitCounts protocol header and credit FLITs  (80 bit FLow control unITs) received from any of the 3 UPI slots on this UPI unitCounts retired load instructions with local Intel Optane DC persistent memory as the data source and the data request missed L3 (AppDirect or Memory Mode) and DRAM cache(Memory Mode). Precise event  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0491OCR.ALL_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWD  OCR.ALL_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDocr.all_pf_rfo.l3_hit.snoop_noneocr.all_pf_rfo.l3_hit_e.snoop_noneOCR.ALL_PF_RFO.L3_HIT_E.SNOOP_NONEocr.all_pf_rfo.l3_hit_f.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040120OCR.ALL_READS.L3_HIT.SNOOP_HIT_WITH_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1002007F7ocr.all_reads.l3_hit_m.hitm_other_coreOCR.ALL_READS.L3_HIT_S.ANY_SNOOP  OCR.ALL_READS.L3_HIT_S.ANY_SNOOPocr.all_rfo.l3_hit_e.hit_other_core_fwdocr.all_rfo.l3_hit_e.snoop_noneOCR.ALL_RFO.L3_HIT_F.SNOOP_MISSOCR.ALL_RFO.L3_HIT_F.SNOOP_NONEocr.all_rfo.l3_hit_m.hitm_other_coreocr.all_rfo.l3_hit_m.snoop_noneOCR.ALL_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWD  OCR.ALL_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDCounts all demand code reads OCR.DEMAND_CODE_RD.L3_HIT.HITM_OTHER_CORE OCR.DEMAND_CODE_RD.L3_HIT.HITM_OTHER_COREocr.demand_code_rd.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080004ocr.demand_code_rd.l3_hit_m.snoop_missCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_S.HITM_OTHER_CORECounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_M.ANY_SNOOPCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.DEMAND_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDocr.other.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C8000ocr.pf_l1d_and_sw.l3_hit_m.hit_other_core_no_fwdCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_M.NO_SNOOP_NEEDEDCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_HIT.NO_SNOOP_NEEDED OCR.PF_L2_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDocr.pf_l2_data_rd.l3_hit_m.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_S.NO_SNOOP_NEEDEDocr.pf_l3_data_rd.l3_hit.no_snoop_neededocr.pf_l3_data_rd.l3_hit.snoop_noneocr.pf_l3_data_rd.l3_hit_m.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100080ocr.pf_l3_rfo.l3_hit_f.any_snoopocr.pf_l3_rfo.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDoffcore_response.all_data_rd.l3_hit_e.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPoffcore_response.all_pf_data_rd.supplier_none.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.demand_code_rd.l3_hit_e.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_M.ANY_SNOOPoffcore_response.demand_rfo.l3_hit_m.hit_other_core_no_fwdoffcore_response.demand_rfo.l3_hit_s.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020002offcore_response.pf_l1d_and_sw.l3_hit_e.snoop_noneoffcore_response.pf_l1d_and_sw.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.ANY_RESPONSEoffcore_response.pf_l2_data_rd.pmm_hit_local_pmm.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020020This event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDoffcore_response.pf_l3_data_rd.pmm_hit_local_pmm.snoop_noneoffcore_response.pf_l3_rfo.l3_hit_e.hitm_other_coreoffcore_response.pf_l3_rfo.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400100100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) * ( ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD) / #( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (OFFCORE_REQUESTS_BUFFER.SQ_FULL / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) ) + ( (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( ((L1D_PEND_MISS.PENDING / ( MEM_LOAD_RETIRED.L1_MISS + MEM_LOAD_RETIRED.FB_HIT )) * cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ / CPU_CLK_UNHALTED.THREAD) / #(max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) ) ( 1000000000 * ( imc@event\=0xe0\,umask\=0x1@ / imc@event\=0xe3@ ) / imc_0@event\=0x0@ )( ( 64 * imc@event\=0xe7@ / 1000000000 ) / duration_time )ocr.all_data_rd.l3_miss.any_snoopOCR.ALL_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDocr.all_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000120OCR.ALL_PF_RFO.L3_MISS.REMOTE_HIT_FORWARD OCR.ALL_PF_RFO.L3_MISS.REMOTE_HIT_FORWARDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000120OCR.ALL_READS.L3_MISS_LOCAL_DRAM.SNOOP_MISSocr.all_rfo.l3_miss.remote_hit_forwardOCR.ALL_RFO.L3_MISS.SNOOP_MISS OCR.ALL_RFO.L3_MISS.SNOOP_MISSocr.all_rfo.l3_miss_remote_hop1_dram.hitm_other_coreOCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONECounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDocr.demand_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreocr.demand_rfo.l3_miss.no_snoop_neededocr.demand_rfo.l3_miss_local_dram.snoop_noneocr.demand_rfo.l3_miss_remote_hop1_dram.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.REMOTE_HITMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000080ocr.pf_l3_rfo.l3_miss_local_dram.any_snoopThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDoffcore_response.all_data_rd.l3_miss_local_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.all_reads.l3_miss.hitm_other_coreoffcore_response.all_reads.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.SNOOP_MISSoffcore_response.all_rfo.l3_miss.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.SNOOP_NONEoffcore_response.demand_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.ANY_SNOOPoffcore_response.pf_l2_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.pf_l2_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdoffcore_response.pf_l2_rfo.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDoffcore_response.pf_l3_rfo.l3_miss.no_snoop_neededocr.all_data_rd.supplier_none.hitm_other_coreCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.pf_l2_data_rd.supplier_none.hit_other_core_fwdocr.pf_l3_data_rd.supplier_none.hit_other_core_fwdocr.pf_l3_rfo.any_responseAll Reads - RPQ or Ufillevent=0xea,umask=0x8unc_cha_tor_inserts.ia_miss_drdevent=0x35,umask=0x21,config1=0x40433M2M Reads Issued to iMC; All, regardless of priorityevent=0x2c,umask=0x04Counts the number of cache lines replaced in L1 data cacheCounts number of cycles a demand request has waited due to L1D due to lack of L2 resources. Demand requests include cacheable/uncacheable demand load, store, lock or SW prefetch accessesRetired load instructions whose data sources were L3 and cross-core snoop hits in on-pkg core cache  Supports address when precise (Precise event)Counts hardware prefetch data reads (which bring data to L2)  that hit a cacheline in the L3 where a snoop was not needed to satisfy the requestCycles where at least 1 outstanding data read request is pending.  Data read requests include cacheable demand reads and L2 prefetches, but do not include RFOs, code reads or prefetches to the L3.  Reads due to page walks resulting from any request type will also be counted.  Requests are considered outstanding from the time they miss the core's L2 cache until the transaction completion message is sent to the requestorCounts number of SSE/AVX computational scalar double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementevent=0x79,cmask=5,period=2000003,umask=0x8event=0x9c,cmask=5,period=1000003,umask=0x1Counts the number of times an RTM execution aborted due to various memory events (e.g. read/write capacity and conflicts)ocr.other.local_dramevent=0x3c,period=2000003,umask=0x8This event counts cycles without actually retired instructionsint_misc.uop_droppingCounts the number of demand load dispatches that hit L1D fill buffer (FB) allocated for software prefetchNumber of available slots for an unhalted logical processor. The event increments by machine-width of the narrowest pipeline as employed by the Top-down Microarchitecture Analysis method (TMA). The count is distributed among unhalted logical processors (hyper-threads) who share the same physical core. Software can use this event as the denominator for the top-level metrics of the TMA method. This architectural event is counted on a designated fixed counter (Fixed Counter 3)event=0xa6,cmask=5,period=2000003,umask=0x21Counts demand data reads that resulted in a snoop hit a modified line in another core's caches which forwarded the dataocr.demand_data_rd.snc_cache.hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84400001ocr.hwpf_l1d_and_swpf.l3_miss_localevent=0xef,period=1000003,umask=0x20Counts demand data reads that were supplied by DRAMocr.hwpf_l2.any_responseCounts hardware prefetches to the L3 only that have any type of responseCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were not supplied by the local socket's L1, L2, or L3 caches and were supplied by a remote socketunc_m_wpq_occupancy_pch0event=0xe4,umask=0x01unc_cha_tor_inserts.ia_hit_drd_prefunc_cha_tor_inserts.ia_rfounc_cha_tor_inserts.ia_miss_drd_localevent=0x35,umask=0xCCD7FF01Number Transactions requested of the CPU : Card reading from DRAM. Unit: uncore_iio event=0x84,ch_mask=0x01,fc_mask=0x07,umask=0x80Number Transactions requested by the CPU : Core reading from Card's MMIO space. Unit: uncore_iio unc_iio_data_req_by_cpu.mem_read.part7event=0x83,ch_mask=0x20,fc_mask=0x07,umask=0x04Free running counter that increments for IIO clocktick. Unit: uncore_iio Clockticks of the IO coherency tracker (IRP). Unit: uncore_irp M2M Writes Issued to iMC : PMM - All Channels. Unit: uncore_m2m TOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices to locally HOMed memory. Unit: uncore_cha Counts the number of load uops retired  Supports address when precise (Precise event)This event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_HIT.SNOOP_HIT_NO_FWDCounts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by the L3 cache where no snoop was needed to satisfy the requestevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x801F803C0000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0040Counts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedCounts modified writebacks from L1 cache and L2 cache that were not supplied by the L3 cacheocr.partial_streaming_wr.l3_miss_localevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x202184000000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000044ocr.hwpf_l2_code_rd.any_responseCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)Counts modified writeBacks from L2 cache that miss the L3 cache that have any type of responseCounts the total number of instructions in which the instruction pointer (IP) of the processor is resteered due to a branch instruction and the branch instruction successfully retires.  All branch type instructions are accounted for (Precise event)Counts the total number of BTCLEARS which occurs when the Branch Target Buffer (BTB) predicts a taken branchevent=0x74,period=1000003,umask=0x8TOR Inserts : UCRdFs issued by iA Cores that Missed LLC. Unit: uncore_cha TOR Inserts : ItoMs issued by IO Devices that missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : PCIRdCurs issued by IO Devices that missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Occupancy : DRd_Opt_Prefs issued by iA Cores : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsTOR Occupancy : DRd_Opt issued by iA Cores that missed the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsNumber Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 5Counts the number of page walks completed due to instruction fetches whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to any page size.  Includes page walks that page faultCounts the number of retired loads that are blocked due to a first level TLB miss (Precise event)Per-Logical Processor actual clocks when the Logical Processor is active. Unit: cpu_core Instructions per FP Arithmetic instruction (lower number means higher occurrence rate). Unit: cpu_core Actual Average Latency for L1 data-cache miss demand load instructions (in core cycles). Unit: cpu_core L1 cache true misses per kilo instruction for retired demand loads. Unit: cpu_core L2 cache true misses per kilo instruction for retired demand loads. Unit: cpu_core Counts the number of cycles the core is stalled due to an instruction cache or tlb miss which hit in the last level cache or other core with HITE/F/M. Unit: cpu_atom event=0x4,period=20003,umask=0x4event=0xc6,period=100007,umask=0x1,frontend=0x604006Counts the number of cycles that the head (oldest load) of the load buffer is stalled due to a core bound stall including a store address match, a DTLB miss or a page walk that detains the load from retiring. Unit: cpu_atom event=0x76,period=1000003,umask=0x1Cycles where at least 1 uop was executed per-thread. Unit: cpu_core Counts the number of x87 uops dispatched. Unit: cpu_core ACT command sent to DRAM. Unit: uncore_imc event=0x2a,period=100003,umask=0x1,offcore_rsp=0x1008000001event=0x2a,period=100003,umask=0x1,offcore_rsp=0x10003C4477FP_ARITH_INST_RETIRED2.SCALARevent=0xcf,period=100003,umask=0x1cevent=0xc6,period=100007,umask=0x1,frontend=0x8event=0x2a,period=100003,umask=0x1,offcore_rsp=0x94002380event=0x2a,period=100003,umask=0x1,offcore_rsp=0x70CC04477event=0xa5,period=1000003,umask=0x7Core clocks when the thread is in the C0.2 light-weight faster wakeup time but less power saving optimized stateINT_VEC_RETIRED.MUL_256UPI Clockticks. Unit: uncore_upi event=0x2,umask=0x0000000004Valid Flits Sent : LLCRD Not Empty. Unit: uncore_upi event=0x2,umask=0x0000000040Valid Flits Received : Protocol Header. Unit: uncore_upi event=0x4,umask=0x000000000eevent=0x83,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000004event=0xc1,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000004event=0x83,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000080unc_m2m_tracker_inserts.ch1event=0x32,umask=0x0000000204event=0x58,umask=0x0000000001event=0x35,umask=0x0000000002unc_cha_tor_inserts.irq_non_iaunc_cha_tor_inserts.prq_non_iosfevent=0x35,umask=0x00c806fe01TOR Inserts RFO misses from local IA. Unit: uncore_cha event=0x36,umask=0x00c80ffe01unc_cha_tor_occupancy.io_miss_itomTOR Occupancy : IPQ. Unit: uncore_cha event=0x36,umask=0x00c897fd01event=0x36,umask=0x00c806fe01event=0x35,umask=0x00cccfff01event=0x35,umask=0x00C8968601unc_cha_tor_inserts.ia_itomcachenearevent=0x35,umask=0x00C8678A01TOR Inserts : WCiLs issued by iA Cores. Unit: uncore_cha event=0x35,umask=0x00C86E8601event=0x35,umask=0x00C8C3FF04event=0x36,umask=0x00c8168a01event=0x36,umask=0x00c88efe01TOR Occupancy : CRd_Prefs issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha event=0x36,umask=0x00cc47ff01unc_cha_tor_occupancy.ia_miss_wcilfTOR Occupancy : WCiLs issued by iA Cores targeting DDR that missed the LLC - HOMed remotely. Unit: uncore_cha The number of instruction fetches that miss in the L1 ITLB but hit in the L2 ITLBevent=0x60,umask=0x80event=0x64,umask=0x09Core to L2 cacheable request access status (not including L2 Prefetch). Instruction cache request hit in L2 and Data cache request hit in L2 (all types)l2_pf_hit_l2Div Op CountTotal number multi-pipe uOps assigned to all pipesevent=0x3,umask=0x20fp_ret_sse_avx_ops.sp_mult_add_flopsls_l1_d_tlb_miss.tlb_reload_4k_l2_missls_l1_d_tlb_miss.tlb_reload_1g_l2_hitevent=0xaf,umask=0x10de_dis_dispatch_token_stalls0.alsq2_token_stalll2_cache_accesses_from_ic_missesl1_itlb_missesevent=0x94,umask=0x02Multiply FLOPS. This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15Demand Data Cache Fills by Data Source. DRAM or IO from different dieL1 DTLB Miss. DTLB reload coalesced page missSoftware Prefetch Instructions Dispatched (Speculative). See docAPM3 PREFETCHWCore to L2 cacheable request access status (not including L2 Prefetch). Instruction cache hit non-modifiable line in L2ls_any_fills_from_sys.int_cacheSoftware Prefetch Data Cache Fills by Data Source. From DRAM or IO connected in different NodeAll TLB Flushes. Requires unit mask 0xFF to engage event for counting. Use all_tlbs_flushed insteadAny Integer dispatch. Types of Oops Dispatched from Decodermacro_ops_retiredInstruction decoder corrects the predicted target and resteers the branch predictorRetired far control transfers (far call/jump/return, IRET, SYSCALL and SYSRET, plus exceptions and interrupts). Far control transfers are not subject to branch predictionevent=0x59,umask=0xdfls_hw_pf_dc_fills.dram_io_nearHardware prefetch data cache fills from extension memoryL2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L1Stride (fetch additional lines into L1 cache when each access is a constant distance from the previous)l2_pf_miss_l2_hit_l3.l2_up_downevent=0xac,umask=0x3fevent=0xd6,umask=0xa2Read data beats (64 bytes) for local processor at Coherent Station (CS) 9Write data beats (64 bytes) for local processor at Coherent Station (CS) 6remote_processor_read_data_beats_cs1event=0x11f,umask=0xbfeRead data beats (64 bytes) for remote processor at Coherent Station (CS) 11event=0x89f,umask=0x7fflocal_socket_upstream_write_beats_iom3Data beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 0Data beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 3Data beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 6event=0x45f,umask=0x7ffremote_socket_inf0_inbound_data_beats_ccm2event=0x51e,umask=0xbffremote_socket_inf1_outbound_data_beats_ccm5local_socket_outbound_data_beats_link4event=0x8,umask=0x01event=0x8,umask=0x04event=0xa,umask=0x05Retired scalar floating-point blend opsRetired vector floating-point ops of all typesfp_pack_ops_retired.fp256_blendevent=0xc,umask=0xb0Retired 256-bit packed floating-point logical opsRetired 256-bit packed floating-point ops of all typespacked_int_op_type.int128_packRetired 128-bit packed integer ops of all typesRetired 256-bit packed integer subtract opspacked_int_op_type.int256_mulevent=0xd,umask=0x30Retired 256-bit packed integer MOV opspacked_int_op_type.allNumber of CAS commands sent for readsde_src_op_disp.allevent=0x1a0,umask=0x1eFraction of dispatch slots that remained unused because of a latency bottleneck in the frontend (such as instruction cache or TLB misses)l2_pf_miss_l2_hit_l3.all + l2_pf_miss_l2_l3.allL1 demand data cache fills from another CCX cache in a different NUMA nodeRatio of memory controller CAS commands for writes(umc_cas_cmd.rd * 64) / 1e6 / duration_timeMemory controller PRECHARGE command rateDC_COPYBACKBUS_CYCLESBRANCH_DIR_MISPREDICTEVENT_41HEVENT_6CHEVENT_9BHEVENT_115HEVENT_125HEVENT_161HEVENT_174HEVENT_1CBHEVENT_1CCHEVENT_1D2HEVENT_1D4HEVENT_1DDHEVENT_1E7HEVENT_2C0HEVENT_2C2HEVENT_2E2HEVENT_2FBHEVENT_31AHEVENT_324HEVENT_341HEVENT_348HEVENT_3B4HEVENT_3D9HEVENT_3F4HUNALIGNED_ST_SPECSTREX_PASS_SPECDP_SPECEXC_SMCdn_rxreq_trk_occupancyhnf_sfbi_dir_snp_sentrnd_s1_wdata_beatsrni_s1_wdata_beatscxra_chi_rsp_upload_stallscxla_tx_tlp_link1clkdiv2_highhigh_qos_depthclkdiv2_refreshVIU2_INSTR_COMPLETEDGPR_ISSUE_QUEUE_DISPATCHESLSU_LOAD_HIT_LINE_ALIAS_VS_CSQ0FP_LOAD_DOUBLE_COMPLETED_IN_LSULSU_MISALIGNED_LOAD_FINISHTLBIE_INSTR_COMPLETEDDTLB_HW_SEARCH_CYCLESMARKED_STORE_COMPLETEDSTORE_TRANSLATE_WHEN_QUEUE_FULL_CYCLESL2_CACHE_CLEAN_REDUNDANT_UPDATESIAC1S_DETECTEDpiggybackunhalted-cyclesINTERRUPTINTEL_CORE2EXTREMEpmclog_get_eventLLC-MISSES"0x%016jx", v4GenuineIntel-6-3ARetiring_SMTSLOTSINST_RETIRED.ANY / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )Bad;BadSpec;BrMispredictsINST_RETIRED.ANY / BR_INST_RETIRED.ALL_BRANCHESAverage CPU UtilizationOSCycles with L1D load Misses outstandingThis event counts the number of L2 cache lines in the Exclusive state filling the L2. Counting does not cover rejectsThis event counts the number of L2 cache lines in the Shared state filling the L2. Counting does not cover rejectsl2_rqsts.all_rfoThis event counts the number of requests from the L2 hardware prefetchers that hit L2 cache. L3 prefetch new typesThis event counts the number of cycles when the L1D is locked. It is a superset of the 0x1 mask (BUS_LOCK_CLOCKS.BUS_LOCK_DURATION)This event counts the number of offcore outstanding demand rfo Reads transactions in the super queue every cycle. The Offcore outstanding state of the transaction lasts from the L2 miss until the sending transaction completion to requestor (SQ deallocation). See the corresponding Umask under OFFCORE_REQUESTS  Spec update: BDM76event=0x79,cmask=4,period=2000003,umask=0x24This event counts loads with latency value being above 32  Spec update: BDM100, BDM35 (Must be precise)This event counts loads with latency value being above eight  Spec update: BDM100, BDM35 (Must be precise)Number of times RTM abort was triggered (PEBS) (Precise event)event=0xc9,period=2000003,umask=0x40event=0xc9,period=2000003,umask=0x80rtm_retired.commitNumber of times a TSX Abort was triggered due to an evicted line caused by a transaction overflowevent=0x5c,cmask=1,edge=1,period=100007,umask=0x1event=0x88,period=200003,umask=0xc2This event counts both taken and not taken speculative and retired indirect branches that have a return mnemonicThis event counts taken speculative and retired indirect branches excluding calls and return branchesbr_inst_retired.near_call_r3This event counts both taken and not taken speculative and retired mispredicted macro conditional branch instructionsThis event counts the number of core cycles while the thread is not in a halt state. The thread enters the halt state when it is running the HLT instruction. This event is a component in many key event ratios. The core frequency may change from time to time due to transitions associated with Enhanced Intel SpeedStep Technology or TM2. For this reason this event may have a changing ratio with regards to time. When the core frequency is constant, this event can approximate elapsed time while the core was not in the halt state. It is counted on a dedicated fixed counter, leaving the four (eight when Hyperthreading is disabled) programmable counters available for other eventsint_misc.recovery_cycles_anyNumber of Uops delivered by the LSDevent=0xa2,period=2000003,umask=0x1Cycles stalled due to no store buffers available. (not including draining form sync)This event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 5uops_dispatched_port.port_6Cycles where at least 1 uop was executed per-threadevent=0xa1,any=1,period=2000003,umask=0x1llc_misses.pcie_writePCIe write misses (full cache line). Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox uncore_haunc_h_snoop_resp.rspsfwdfreq_trans_cycles %dtlb_store_misses.walk_completeditlb_misses.walk_completed_2m_4mevent=0x85,period=100003,umask=0x4 ( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * (BR_MISP_RETIRED.ALL_BRANCHES * (12 * ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY ) / CPU_CLK_UNHALTED.THREAD) / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY )) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) ) * (4 * CPU_CLK_UNHALTED.THREAD) / BR_MISP_RETIRED.ALL_BRANCHESevent=0x24,period=200003,umask=0xc1Retired load uops which data sources were hits in L3 without snoops required  Supports address when precise.  Spec update: BDM100 (Precise event)This event counts line-splitted store uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K)  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020240offcore_response.all_pf_data_rd.supplier_none.snoop_hitmCounts prefetch RFOs have any response typeoffcore_response.corewb.supplier_none.snoop_missCounts any other requests have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C8000offcore_response.other.l3_hit.snoop_noneoffcore_response.other.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020010offcore_response.pf_l2_rfo.l3_hit.snoop_hit_no_fwdoffcore_response.pf_l2_rfo.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000091offcore_response.all_pf_rfo.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020120offcore_response.all_rfo.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000080offcore_response.pf_l3_data_rd.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000100offcore_response.pf_l3_rfo.l3_miss_local_dram.snoop_not_neededDirect and indirect near call instructions retired (Precise event)event=0x22,umask=0x41A cross-core snoop initiated by this Cbox due to processor core memory request which misses in some processor core. Unit: uncore_cbox event=0x22,umask=0x81L3 Lookup any request that access cache and found line in M-stateunc_cbo_cache_lookup.read_ievent=0x34,umask=0x18event=0x81,umask=0x02offcore_response.all_requests.llc_hit.any_responseCounts prefetch (that bring data to LLC only) code reads hit in the L3Counts all demand & prefetch code reads miss the L3 and the data is returned from local draml2_data_rqsts.self.s_stateL2 cache readsevent=0x29,period=200000,umask=0x41event=0x24,period=200000,umask=0x50l2_reject_busq.self.any.mesievent=0x30,period=200000,umask=0x42l2_rqsts.self.demand.i_stateRetired Streaming SIMD Extensions (SSE) packed-single instructionsevent=0xc7,period=2000000,umask=0x2event=0xb3,period=2000000,umask=0x10simd_uop_type_exec.shift.smacro_insts.cisc_decodedbus_bnr_drv.all_agentsevent=0x64,period=200000,umask=0x40bus_trans_p.all_agentsNumber of Enhanced Intel SpeedStep(R) Technology (EIST) transitionsevent=0xe0,period=2000000,umask=0x1br_inst_retired.mispred_takenDivide operations executedevent=0x86,period=200003,umask=0x2mem_uops_retired.splitoffcore_response.any_data_rd.l2_miss.hitm_other_coreoffcore_response.demand_data_rd.l2_hitCounts data cache line reads generated by hardware L1 data cache prefetcher that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache line reads generated by hardware L1 data cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts data cache line reads generated by hardware L1 data cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040020event=0xc5,period=200003,umask=0xf7Counts core cycles the integer divide unit is busyevent=0x3,period=200003,umask=0x8Counts the number of times the machine was unable to find a translation in the Instruction Translation Lookaside Buffer (ITLB) for a linear address of an instruction fetch.  It counts when new translation are filled into the ITLB.  The event is speculative in nature, but will not count translations (page walks) that are begun and not finished, or translations that are finished but not filled into the ITLBCounts data reads (demand & prefetch) miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredoffcore_response.any_pf_data_rd.outstandingevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010022Counts reads for ownership (RFO) requests (demand & prefetch) have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache hit the L2 cacheCounts demand reads for ownership (RFO) requests generated by a write to full data cache line outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts data cacheline reads generated by hardware L2 cache prefetcher hit the L2 cachePage walks outstanding due to a demand load every cycleevent=0xbd,period=20003,umask=0x20Number of times a request needed a FB entry but there was no entry available for it. That is the FB unavailability was dominant reason for blocking the request. A request includes cacheable/uncacheable demands that is load, store or SW prefetch. HWP are eDirty L2 cache lines evicted by demandDemand data read requests that missed L2, no rejects  Spec update: HSD78, HSM80All requests that miss L2 cache  Spec update: HSD78, HSM80All requests that missed L2  Spec update: HSD78, HSM80Demand data read requests that access L2 cacheMiss in last-level (L3) cache. Excludes Unknown data-source  Supports address when precise.  Spec update: HSD74, HSD29, HSD25, HSM26, HSM30 (Precise event)Retired load uops that split across a cacheline boundary  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)offcore_response.all_reads.l3_hit.hit_other_core_no_fwdmiss in the L3event=0xc4,period=100003,umask=0x8This event counts cycles where the decoder is stalled on an instruction with a length changing prefix (LCP)This event counts loads that followed a store to the same address, where the data could not be forwarded inside the pipeline from the store to the load.  The most common reason why store forwarding would be blocked is when a load's address range overlaps with a preceding smaller uncompleted store. The penalty for blocked store forwarding is that the load must wait for the store to write its value to the cache before it can be issuedCycles per core when uops are executed in port 0uops_issued.core_stall_cyclesCycles when PMH is busy with page walkspage_walker_loads.ept_itlb_memoryl2_lines_out.pf_cleanClean L2 cache lines evicted by the MLC prefetcherMiss in last-level (L3) cache. Excludes Unknown data-source (Precise event)Offcore outstanding cacheable Core Data Read transactions in SuperQueue (SQ), queue to uncoreoffcore_requests_outstanding.cycles_with_demand_code_rdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0244event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0002( ( 1 * ( FP_COMP_OPS_EXE.SSE_SCALAR_SINGLE + FP_COMP_OPS_EXE.SSE_SCALAR_DOUBLE ) + 2 * FP_COMP_OPS_EXE.SSE_PACKED_DOUBLE + 4 * ( FP_COMP_OPS_EXE.SSE_PACKED_SINGLE + SIMD_FP_256.PACKED_DOUBLE ) + 8 * SIMD_FP_256.PACKED_SINGLE ) / 1000000000 ) / duration_timeLoads with latency value being above 16 (Must be precise)mem_trans_retired.precise_storeuops_dispatched_port.port_5_coreevent=0x22,umask=0x08Filter on cross-core snoops initiated by this Cbox due to external snoop request. Unit: uncore_cbox Filter on cross-core snoops initiated by this Cbox due to external snoop requestMisses in all TLB levels that caused page walk completed of any size by demand loadsmem_load_uops_llc_miss_retired.remote_fwdoffcore_response.all_pf_data_rd.llc_hit.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c03f7Counts prefetch (that bring data to LLC only) data reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwarded(UNC_P_FREQ_BAND1_CYCLES / UNC_P_CLOCKTICKS) * 100.RFOs that hit cache lines in E stateThis event counts cycles during which the microcode sequencer assisted the front-end in delivering uops.  Microcode assists are used for complex instructions or scenarios that can't be handled by the standard decoder.  Using other instructions, if possible, will usually improve performance.  See the Intel? 64 and IA-32 Architectures Optimization Reference Manual for more informationThis event counts the number of memory ordering Machine Clears detected. Memory Ordering Machine Clears can result from memory disambiguation, external snoops, or cross SMT-HW-thread snoop (stores) hitting load buffers.  Machine clears can have a significant performance impact if they are happening frequentlyThis event counts all remote cache-to-cache transfers (includes HITM and HIT-Forward) for all demand and L2 prefetches. LLC prefetches are excludedTaken speculative and retired mispredicted direct near callsevent=0x59,period=2000003,umask=0x20event=0x1offcore_response.any_data_rd.l2_hit_near_tile_e_fCounts Demand cacheable data and L1 prefetch data read requests  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008003091event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180070Counts any Prefetch requests that accounts for responses which hit its own tile's L2 with data in E stateoffcore_response.any_read.l2_hit_near_tile_e_fCounts any Read request  that accounts for responses which hit its own tile's L2 with data in F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002008000Counts Demand cacheable data write requests  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts Demand cacheable data write requests  that accounts for responses which hit its own tile's L2 with data in E stateCounts Bus locks and split lock requests that accounts for any responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000400Counts demand code reads and prefetch code reads that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000002Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses which hit its own tile's L2 with data in F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800402000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080040offcore_response.pf_l2_rfo.l2_hit_near_tileoffcore_response.pf_software.l2_hit_near_tile_e_foffcore_response.pf_software.l2_hit_this_tile_sCounts Software Prefetches that accounts for responses which hit its own tile's L2 with data in S stateoffcore_response.uc_code_reads.l2_hit_far_tile_mCounts UC code reads (valid only for Outstanding response type)  that accounts for responses which hit its own tile's L2 with data in S stateCounts Demand code reads and prefetch code read requests  that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts any Prefetch requests that accounts for responses from MCDRAM (local and far)Counts any Read request  that accounts for data responses from DRAM LocalCounts Demand cacheable data write requests  that accounts for data responses from DRAM Farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400022Counts Bus locks and split lock requests that accounts for responses from DDR (local and far)Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses from any NON_DRAM system address. This includes MMIO transactionsoffcore_response.pf_l1_data_rd.ddrevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181802000Counts UC code reads (valid only for Outstanding response type)  that accounts for data responses from DRAM LocalCounts the number of branch instructions retired that were conditional jumps and predicted taken (Precise event)br_misp_retired.callCounts the number of core cycles when no micro-ops are allocated and a RATstall (caused by reservation station full) is assertedcache_lock_cycles.l1devent=0x42,period=2000000,umask=0x1event=0x28,period=100000,umask=0x1event=0x26,period=200000,umask=0x1L2 data prefetches in E stateL2 modified lines evicted by a prefetch requestevent=0xf0,period=200000,umask=0x80event=0xf0,period=200000,umask=0x4All demand L2 lock RFOsL2 demand lock RFOs in S stateevent=0xb,period=10000,umask=0x10,ldlat=0x10event=0xb,period=2000000,umask=0x2mem_load_retired.llc_missevent=0xf,period=20000,umask=0x8offcore_requests.l1d_writebackevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F11Offcore data reads satisfied by the LLC and not found in a sibling coreOffcore data reads satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3811offcore_response.any_data.remote_cache_hitmOffcore code reads satisfied by the IO, CSR, MMIO unitoffcore_response.any_rfo.any_cache_dramoffcore_response.corewb.io_csr_mmiooffcore_response.corewb.llc_hit_other_core_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1877event=0xb7,period=100000,umask=0x1,offcore_rsp=0x233offcore_response.demand_data.llc_hit_other_core_hitoffcore_response.demand_data_rd.remote_cache_dramoffcore_response.demand_rfo.local_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1002Offcore demand RFO requests that HIT in a remote cacheoffcore_response.pf_data.any_locationevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8040offcore_response.pf_rfo.llc_hit_no_other_coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3870128 bit SIMD integer multiply operationsSIMD integer 64 bit pack operationsSIMD integer 64 bit logical operationssimd_int_64.packed_shiftevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2044event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4022offcore_response.data_ifetch.any_llc_missoffcore_response.data_in.any_dramOffcore data reads, RFO's and prefetches statisfied by the remote DRAMOffcore demand data reads that missed the LLCoffcore_response.demand_ifetch.any_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6040event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4040Offcore prefetch code reads satisfied by a remote DRAMES segment renameslarge_itlb.hitThread responded HITE to snoopDivide Operations executedevent=0xc4,period=20000,umask=0x2br_misp_exec.direct_near_callbr_misp_exec.near_callsStall cycles due to BPU MRU bypassLoad buffer stall cyclesRetired loads that miss the DTLB (Precise Event)Offcore data reads, RFO's and prefetches that HIT in a remote cache event=0xf2,period=200003,umask=0x2Counts core-originated cacheable requests to the  L3 cache (Longest Latency cache). Requests include data and code reads, Reads-for-Ownership (RFOs), speculative accesses and hardware prefetches from L1 and L2.  It does not include all accesses to the L3  Spec update: SKL057Retired load instructions which data sources were hits in L3 without snoops required  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080004offcore_response.demand_code_rd.l3_hit_s.snoop_noneoffcore_response.demand_code_rd.l4_hit_local_l4.snoop_noneoffcore_response.demand_data_rd.l3_hit_m.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40100002offcore_response.demand_rfo.l4_hit_local_l4.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40088000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100048000event=0x32,period=2000003,umask=0x1sw_prefetch_access.prefetchwThis event counts the number of the Decode Stream Buffer (DSB)-to-MITE switches including all misses because of missing Decode Stream Buffer (DSB) cache and u-arch forced misses.
Note: Invoking MITE requires two or three cycles delayRetired Instructions who experienced Instruction L2 Cache true miss (Precise event)frontend_retired.latency_ge_2_bubbles_ge_1Counts retired instructions that are delivered to the back-end after the front-end had at least 1 bubble-slot for a period of 2 cycles. A bubble-slot is an empty issue-pipeline slot while there was no RAT stall (Precise event)frontend_retired.latency_ge_32offcore_response.demand_rfo.l3_miss.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C408000offcore_response.other.l4_hit_local_l4.snoop_non_dramrtm_retired.aborted_unfriendlyReturn instructions retired  Spec update: SKL091 (Precise event)Counts when the Current Privilege Level (CPL) transitions from ring 1, 2 or 3 to ring 0 (Kernel)Counts the number of core cycles while the thread is not in a halt state. The thread enters the halt state when it is running the HLT instruction. This event is a component in many key event ratios. The core frequency may change from time to time due to transitions associated with Enhanced Intel SpeedStep Technology or TM2. For this reason this event may have a changing ratio with regards to time. When the core frequency is constant, this event can approximate elapsed time while the core was not in the halt state. It is counted on a dedicated fixed counter, leaving the four (eight when Hyperthreading is disabled) programmable counters available for other eventsevent=0x59,period=2000003,umask=0x1Counts the number of macro-fused uops retired. (non precise)Total pipeline cost of Branch Misprediction related bottlenecks100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) * ( ( (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) / ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (min( 9 * cpu@DTLB_LOAD_MISSES.STLB_HIT\,cmask\=1@ + DTLB_LOAD_MISSES.WALK_ACTIVE , max( CYCLE_ACTIVITY.CYCLES_MEM_ANY - CYCLE_ACTIVITY.CYCLES_L1D_MISS , 0 ) ) / CPU_CLK_UNHALTED.THREAD) / (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) ) + ( (EXE_ACTIVITY.BOUND_ON_STORES / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (( 9 * cpu@DTLB_STORE_MISSES.STLB_HIT\,cmask\=1@ + DTLB_STORE_MISSES.WALK_ACTIVE ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )) / #(EXE_ACTIVITY.BOUND_ON_STORES / CPU_CLK_UNHALTED.THREAD) ) ) BigFoot;Fed;Frontend;IcMiss;MemoryTLBCond_TKThis event counts the number of load ops retired that hit in the L2 (Precise event)Counts any data read (demand & prefetch) that hit in the other module where modified copies were found in other core's L1 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000010Counts RFO requests generated by L2 prefetchers that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.pf_l2_rfo.l2_miss.snoop_missThis event counts the cases where a forward was technically possible, but did not occur because the store data was not available at the right timeCounts the number of RETURN baclearsevent=0xca,period=200003,umask=0x50Counts the number of cycles when no uops are allocated and the ROB is full (less than 2 entries available)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0120offcore_response.data_in.any_responseCounts demand code reads that hit in the LLC and the snoops sent to sibling cores return clean responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0020Counts prefetch (that bring data to LLC only) RFOs that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.pf_l_data_rd.any_responseCounts all prefetch data reads that miss the LLC  and the data returned from dramoffcore_response.all_pf_rfo.llc_miss.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400002offcore_response.pf_llc_code_rd.llc_miss.dramoffcore_requests.any.rfooffcore_response.any_ifetch.local_dram_and_remote_cache_hitREQUEST = ANY RFO and RESPONSE = LLC_HIT_NO_OTHER_COREevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x208offcore_response.demand_ifetch.local_dram_and_remote_cache_hitREQUEST = DEMAND_RFO and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x450REQUEST = PREFETCH and RESPONSE = ANY_CACHE_DRAMMisaligned store referencesREQUEST = PREFETCH and RESPONSE = REMOTE_DRAMOffcore data reads, RFOs, and prefetches that HIT in a remote cacheoffcore_response.all_data_rd.l3_hit.no_snoop_neededoffcore_response.pf_l1d_and_sw.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0400offcore_response.pf_l2_data_rd.l3_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0010Counts all prefetch (that bring data to L2) RFOs that have any response typeNumber of SSE/AVX computational 512-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 16 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT14 RCP14 FM(N)ADD/SUB. FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000491offcore_response.all_pf_rfo.l3_miss.any_snoopCounts prefetch RFOs that miss the L3 and clean or shared data is transferred from remote cacheCounts demand data reads that miss the L3 and the data is returned from local or remote dramoffcore_response.pf_l2_data_rd.l3_miss.any_snoopCounts all prefetch (that bring data to LLC only) RFOs that miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800100skx metrics100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) * ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , OFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DATA_RD ) / CPU_CLK_UNHALTED.THREAD - (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD)) / #(CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (( (20.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) - (3.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) ) * MEM_LOAD_RETIRED.L3_HIT * (1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) / 2) / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) + ( (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD)) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) )INST_RETIRED.ANY / ( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE) )( ( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * ( FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE ) + 16 * FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE ) / 1000000000 ) / duration_timeunc_m_cas_count.rd_underfillwrite requests from home agent. Unit: uncore_cha Core Cross Snoops Issued; Multiple Eviction. Unit: uncore_cha unc_cha_dir_lookup.snpPCIe Completion Buffer Inserts of completions with data: Part 0event=0xc0,ch_mask=0x01,fc_mask=0x07,umask=0x01event=0xc0,ch_mask=0x04,fc_mask=0x07,umask=0x02event=0x83,ch_mask=0x01,fc_mask=0x07,umask=0x02Counts every read request for up to a 64 byte transaction of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part0. In the general case, part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busunc_iio_txn_req_of_cpu.mem_read.part0unc_iio_txn_req_of_cpu.mem_read.part3event=0x84,ch_mask=0x04,fc_mask=0x07,umask=0x08Counts cycles when the ability to send messages direct to the Intel Ultra Path Interconnect (bypassing the CHA) was disabledCounts when the M2M (Mesh to Memory) updates the multi-socket cacheline Directory state from from I (Invalid) to A (SnoopAll)unc_m2m_prefcam_insertsunc_m2m_rxc_ad_occupancyevent=0x31,umask=0x4Retired load instructions with remote Intel Optane DC persistent memory as the data source where the data request missed all caches. Precise event  Supports address when precise (Precise event)event=0xd1,period=100003,umask=0x80ocr.all_data_rd.l3_hit.snoop_noneocr.all_pf_data_rd.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080490OCR.ALL_PF_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDED  OCR.ALL_PF_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDocr.all_pf_data_rd.l3_hit_s.snoop_noneOCR.ALL_PF_RFO.L3_HIT_E.NO_SNOOP_NEEDED  OCR.ALL_PF_RFO.L3_HIT_E.NO_SNOOP_NEEDEDocr.all_pf_rfo.l3_hit_s.hit_other_core_no_fwdOCR.ALL_READS.L3_HIT_E.NO_SNOOP_NEEDED  OCR.ALL_READS.L3_HIT_E.NO_SNOOP_NEEDEDOCR.ALL_READS.L3_HIT_F.SNOOP_MISSocr.all_reads.l3_hit_m.hit_other_core_no_fwdocr.demand_code_rd.l3_hit.snoop_noneCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200001Counts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_HIT.SNOOP_MISSCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.demand_rfo.l3_hit_f.no_snoop_neededCounts any other requests OCR.OTHER.L3_HIT.ANY_SNOOP OCR.OTHER.L3_HIT.ANY_SNOOPCounts any other requests  OCR.OTHER.L3_HIT_E.ANY_SNOOPCounts any other requests  OCR.OTHER.L3_HIT_E.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000208000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200400ocr.pf_l1d_and_sw.l3_hit_m.hitm_other_coreocr.pf_l1d_and_sw.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0010ocr.pf_l2_data_rd.l3_hit_e.any_snoopCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040100ocr.pf_l3_rfo.l3_hit_s.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDoffcore_response.all_data_rd.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.all_pf_rfo.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.SUPPLIER_NONE.ANY_SNOOPoffcore_response.all_reads.l3_hit_e.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_READS.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.SUPPLIER_NONE.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.all_rfo.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.other.l3_hit.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.OTHER.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_F.HITM_OTHER_COREoffcore_response.pf_l1d_and_sw.l3_hit_f.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l1d_and_sw.supplier_none.no_snoop_neededoffcore_response.pf_l2_data_rd.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.SUPPLIER_NONE.ANY_SNOOPoffcore_response.pf_l2_data_rd.supplier_none.hit_other_core_no_fwdoffcore_response.pf_l2_data_rd.supplier_none.no_snoop_neededoffcore_response.pf_l2_rfo.l3_hit_s.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_M.HITM_OTHER_COREoffcore_response.pf_l3_data_rd.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEoffcore_response.pf_l3_data_rd.pmm_hit_local_pmm.snoop_not_neededThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.ANY_SNOOPoffcore_response.pf_l3_rfo.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_M.ANY_SNOOPFed;LSDPMM_Read_BWMem;MemoryBW;SoC;ServerPMM_Write_BWOCR.ALL_DATA_RD.L3_MISS.REMOTE_HITM OCR.ALL_DATA_RD.L3_MISS.REMOTE_HITMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000490ocr.all_pf_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000120ocr.all_rfo.l3_miss_local_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000122ocr.demand_code_rd.l3_miss.snoop_missocr.demand_code_rd.l3_miss_remote_hop1_dram.hitm_other_coreocr.demand_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000001ocr.demand_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededocr.demand_rfo.l3_miss.hitm_other_coreocr.demand_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdocr.other.l3_miss_local_dram.snoop_missCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.NO_SNOOP_NEEDED OCR.PF_L2_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000010Counts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.ANY_SNOOP OCR.PF_L2_RFO.L3_MISS.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000080Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000100ocr.pf_l3_rfo.l3_miss_local_dram.snoop_missCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.all_pf_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.all_pf_rfo.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.demand_code_rd.l3_miss_local_dram.hitm_other_coreoffcore_response.demand_data_rd.l3_miss_local_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSoffcore_response.pf_l2_rfo.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDOCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONE OCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEOCR.ALL_READS.SUPPLIER_NONE.SNOOP_NONEocr.demand_rfo.supplier_none.any_snoopocr.demand_rfo.supplier_none.snoop_noneocr.other.pmm_hit_local_pmm.snoop_noneCounts any other requests  OCR.OTHER.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.other.supplier_none.snoop_noneCounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPCounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.pf_l3_rfo.supplier_none.hit_other_core_no_fwdSW prefetch requests that miss L2 cacheCounts retired load instructions with locked access  Supports address when precise (Precise event)ocr.demand_code_rd.l3_hit.anyocr.hwpf_l2_data_rd.l3_hit.snoop_not_neededocr.hwpf_l2_rfo.l3_hit.snoop_sentCycles DSB is delivering optimal number of UopsCounts the number of uops delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) pathUops delivered to IDQ while MS is busyCORE_POWER.LVL2_TURBO_LICENSE / CPU_CLK_UNHALTED.DISTRIBUTED(cstate_pkg@c9\-residency@ / msr@tsc@) * 100event=0xc8,period=100003,umask=0x20Counts the number of times we entered an HLE region. Does not count nested transactionsevent=0x5d,period=100003,umask=0x4Counts the number of times a TSX Abort was triggered due to attempting an unsupported alignment from Lock BufferCounts demand instruction fetches and L1 instruction cache prefetches that have any type of responseCounts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that have any type of responseCounts far branch instructions retired (Precise event)Counts core crystal clock cycles when the thread is unhaltedevent=0x87,period=500009,umask=0x1Counts number of retired PAUSE instructions. This event is not supported on first SKL and KBL productsevent=0xa2,period=100003,umask=0x8event=0xa2,period=100003,umask=0x2event=0xa4,period=10000003,umask=0x8Counts cycles when at least 4 micro-ops are executed from any thread on physical coreNumber of page walks outstanding for an outstanding code request in the PMH each cycleocr.demand_code_rd.snc_cache.hitmCounts demand data reads that either hit a non-modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0477event=0x60,cmask=1,period=1000003,umask=0x2UNC_CHA_TOR_INSERTS.IO_PCIRDCUR * 64 / 1000000000 / duration_timeCounts hardware prefetches to the L3 only that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallyCounts streaming stores that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallyevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x700C00477event=0xb7,period=100003,umask=0x1,offcore_rsp=0x700800477event=0x4,umask=0x30event=0x10,umask=0x02Half clockticks for IMC. Unit: uncore_imc event=0x1,umask=0x0BTOR Inserts : DRds issued by iA Cores that Hit the LLC. Unit: uncore_cha event=0x35,umask=0xC001FE04unc_cha_tor_occupancy.ioTOR Inserts : DRd_Prefs issued by iA Cores that Hit the LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_rfounc_cha_tor_inserts.ia_miss_drd_local_ddrevent=0x85,ch_mask=0xFF,fc_mask=0x07,umask=0x01event=0x84,ch_mask=0x40,fc_mask=0x07,umask=0x01event=0x84,ch_mask=0x40,fc_mask=0x07,umask=0x80Valid Flits Received : Null FLITs received from any slot. Unit: uncore_upi Counts the total number of L2 Cache Accesses, includes hits, misses, rejects  front door requests for CRd/DRd/RFO/ItoM/L2 Prefetches only.  Counts on a per core basisl2_request.rejectsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x401F803C0000Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by the L3 cache where a snoop was sent but the snoop missedCounts uncached memory reads that were supplied by the L3 cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010003C0000Counts the number of floating point operations retired that required microcode assist, which is not a reflection of the number of FP operations, instructions or uopsCounts the number of floating point divide uops retired (x87 and SSE, including x87 sqrt) (Precise event)Counts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were not supplied by the L3 cacheocr.uc_wr.l3_missbus_lock.cycles_self_blockCounts the number of reference cycles that the core is not in a halt state. The core enters the halt state when it is running the HLT instruction. This event is not affected by core frequency changes and increments at a fixed frequency that is also used for the Time Stamp Counter (TSC). This event uses fixed counter 2topdown_bad_speculation.fastnuketopdown_be_bound.reorder_bufferevent=0x71,period=1000003,umask=0x2Counts the number of issue slots every cycle that were not delivered by the frontend due to decode stallsCounts the number of issue slots every cycle that were not delivered by the frontend due to frontend bandwidth restrictions due to decode, predecode, cisc, and other limitationsevent=0x35,umask=0xC001FE01,config1=0x41a33TOR Inserts : DRd_Opt issued by iA Cores that missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x35,umask=0xC867FE01TOR Inserts : WiLs issued by iA Cores that Missed LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Occupancy : DRd_Opts issued by iA Cores. Unit: uncore_cha TOR Occupancy : DRd_Opt_Prefs issued by iA Cores. Unit: uncore_cha Data requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Number requests PCIe makes of the main die : All : Counts full PCIe requests before they're broken into a series of cache-line size requests as measured by DATA_REQ_OF_CPU and TXN_REQ_OF_CPUNumber Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6event=0x49,period=200003,umask=0xeevent=0x4f,period=2000003,umask=0x2Counts the number of memory ordering machine clears triggered by a snoop from an external agent. Does not count internally generated machine clears such as those due to disambiguationsCounts the total number of issue slots that were not consumed by the backend because allocation is stalled due to a mispredicted jump or a machine clear. Only issue slots wasted due to fast nukes such as memory ordering nukes are counted. Other nukes are not accounted for. Counts all issue slots blocked during this recovery window including relevant microcode flows and while uops are not yet available in the instruction queue (IQ). Also includes the issue slots that were consumed by the backend but were thrown away because they were younger than the mispredict or machine clearInstructions per Floating Point (FP) Operation (lower number means higher occurrence rate). Unit: cpu_core L3 cache true misses per kilo instruction for retired demand loads. Unit: cpu_core event=0xd0,period=1000003,umask=0x5,ldlat=0x10Counts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 32 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom mem_uops_retired.load_latency_gt_64RFO requests to L2 cache. Unit: cpu_core event=0x79,period=1000003,umask=0x20event=0x9c,cmask=6,period=1000003,umask=0x1assists.hardwareCounts the total number of consumed retirement slots (Precise event). Unit: cpu_atom Number of instructions retired. Fixed Counter - architectural event (Precise event). Unit: cpu_core integer ADD, SUB, SAD 256-bit vector instructions. Unit: cpu_core int_vec_retired.mul_256Self-modifying code (SMC) detected. Unit: cpu_core topdown.memory_bound_slotsCycles at least 3 micro-op is executed from any thread on physical core. Unit: cpu_core Cycles at least 4 micro-op is executed from any thread on physical core. Unit: cpu_core Cycles where at least 2 uops were executed per-thread. Unit: cpu_core event=0x1aevent=0x22ld_head.dtlb_miss_at_retStores that miss the DTLB and hit the STLB. Unit: cpu_core event=0xd3,period=1000003,umask=0x4MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM  Supports address when preciseevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F803C0004event=0x2a,period=100003,umask=0x1,offcore_rsp=0x80080800fp_arith_inst_retired2.128b_packed_halfFP_ARITH_INST_RETIRED2.256B_PACKED_HALFFP_ARITH_INST_RETIRED2.VECTORevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x104000002amx_ops_retired.bf16event=0xce,period=1000003,umask=0x1integer ADD, SUB, SAD 128-bit vector instructionsinteger ADD, SUB, SAD 256-bit vector instructionsevent=0x83,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000001Valid Flits Sent : Slot 0. Unit: uncore_upi event=0x3,umask=0x000000000fMatches on Receive path of a UPI Port : Non-Coherent Bypass. Unit: uncore_upi event=0x5,umask=0x000000010fevent=0x30,umask=0x0000000004Read request for 4 bytes made by IIO Part1 to Memory. Unit: uncore_iio event=0xc1,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000004event=0x20,umask=0x08unc_cha_osb.local_readTOR Inserts; ItoM misses from local IO. Unit: uncore_cha unc_cha_tor_inserts.rrqTOR Occupancy; CRd misses from local IA. Unit: uncore_cha unc_cha_tor_occupancy.prqunc_cha_tor_occupancy.loc_iaunc_cha_tor_occupancy.remote_tgtTOR Occupancy; CRd Pref hits from local IA. Unit: uncore_cha TOR Occupancy; CRd from local IA. Unit: uncore_cha event=0x36,umask=0x00c886fe01unc_cha_tor_occupancy.ia_hit_llcprefcodeunc_cha_tor_occupancy.ia_miss_llcprefdataTOR Inserts : DRd_Prefs issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_wcilf_ddrunc_cha_tor_occupancy.ia_miss_drd_remote_pmmevent=0x36,umask=0x00c8678a01event=0x36,umask=0x00c8670a01unc_cha_tor_occupancy.io_hit_itomcachenearL2 BTB CorrectionThe number of instruction fetches that hit in the L1 ITLBevent=0x87,umask=0x04l2_cache_req_stat.ic_dc_hit_in_l2event=0x9a,umask=0x3fex_ret_brn_farRetired Branch Resyncsex_ret_mmx_fp_instr.mmx_instrdram_channel_data_controller_1The number of operations (uOps) and dual-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number multi-pipe uOps assigned to pipe 3event=0,umask=0x20fpu_pipe_assignment.totalThe number of x87 floating-point Ops that have retired. The number of events logged per cycle can vary from 0 to 8. Add/subtract Opsevent=0x3,umask=0x08fp_num_mov_elim_scal_op.optimizedThis is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemes. Number of Ops that are candidates for optimization (have Z-bit either set or pass)event=0x4,umask=0x01Counts the number of operations dispatched to the LS unit. Unit Masks ADDed. Load-op-Storesls_mab_alloc.dc_prefetcherevent=0x4b,umask=0x04Cycles where a dispatch group is valid but does not get dispatched due to a token stall. ALU tokens total unavailablerecommendedL1 Instruction Cache (32B) Fetch Miss Ratioevent=0x78,umask=0xdfevent=0xe,umask=0x04ls_locks.spec_lock_lo_specCycles where the Micro-Op Queue is emptyThe number of retired branch instructions, that were mispredictedMultiply FLOPs. This is a retire-based event. The number of retired SSE/AVX FLOPs. The number of events logged per cycle can vary from 0 to 64. This event requires the use of the MergeEvent since it can count above 15 events per cycle. See 2.1.17.3 [Large Increment per Cycle Events]. It does not provide a useful count without the use of the MergeEventRetired lock instructions. Comparable to legacy bus lockevent=0x41,umask=0x3fLoad Store Allocations. Counts when a LS pipe allocates a MAB entryevent=0x44,umask=0x08Op Cache (64B) Fetch Miss RatioDemand data cache fills from either DRAM or MMIO in a different NUMA node (same or different socket)Any data cache fills from either DRAM or MMIO in the same NUMA nodeL2 cache requests of common types not including prefetchesevent=0x71,umask=0x08Average sampled latency when data is sourced from another CCX's cache when the address was in the same NUMA node. Unit: amd_l3 local_processor_read_data_beats_cs2event=0xdf,umask=0x7ffremote_processor_read_data_beats_cs4remote_processor_write_data_beats_cs2Write data beats (64 bytes) for remote processor at Coherent Station (CS) 7local_socket_upstream_write_beats_iom1event=0x4de,umask=0x7felocal_socket_inf0_outbound_data_beats_ccm3remote_socket_inf1_inbound_data_beats_ccm6Data beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 6Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 0event=0xa,umask=0x0fevent=0xa,umask=0x50event=0xb,umask=0x0dsse_avx_ops_retired.sse_avx_clmRetired 256-bit packed floating-point square root opsRetired 128-bit packed integer AES opsRetired 128-bit packed integer pack opspacked_int_op_type.int128_otherFloating-point dispatch faults for YMM fillsFloating-point dispatch faults for YMM spillsumc_mem_clkNumber of ACTIVATE commands sent for readsd_ratio(bad_speculation * ex_ret_brn_misp, ex_ret_brn_misp + resyncs_or_nc_redirects)All data cache accesses0x01IC_REFILL_FROM_L2NB_MEMORY_CONTROLLER_PAGE_TABLE_OVERFLOWEVENT_0DHEVENT_2AHEVENT_3AHEVENT_47HEVENT_4DHEVENT_58HEVENT_A5HEVENT_AEHEVENT_D7HEVENT_DCHEVENT_E5HPC_WRITE_RETIREDEVENT_105HEVENT_110HEVENT_143HEVENT_14CHEVENT_177HEVENT_196HEVENT_1D1HEVENT_20FHEVENT_231HEVENT_234HEVENT_257HEVENT_267HEVENT_27BHEVENT_2B4HEVENT_30DHEVENT_32AHEVENT_32BHEVENT_335HEVENT_343HEVENT_350HEVENT_38AHEVENT_38EHEVENT_390HEVENT_3AEHEVENT_3B8HEVENT_3DDHEXC_TRAP_FIQSTALL_FRONTENDL1I_TLBL1D_CACHE_REFILL_INNERL1D_CACHE_REFILL_OUTERhnf_seq_fullrnd_txreq_flits_replayedcxra_snp_sink_buf_occDST_INSTR_DISPATCHEDCYCLES_WAITING_FROM_L1_INSTR_CACHE_MISSCANCELLED_L1_INSTR_CACHE_MISSESVT2_FETCHESBRANCH_UNIT_STALL_ON_CTR_DEPENDENCYL2_LOAD_HITSINTERVENTIONBUS_WRITES_NOT_RETRIEDTHRESHOLD_TIMEOUTCR_MARKED_INSTR_FINISHCACHE_INHIBITED_ACCESS_TRANSLATEDINSTR_MMU_VSP_RELOADSSTASH_L1_HITSCOHERENT_LOOKUP_MISS_DUE_TO_VALID_BUT_INCOHERENT_MATCHEShw_interrupts.received{"type": "userdata"%s, "pid": "%d", "start": "0x%016jx", "pathname": "%s"}
GenuineIntel-6-27v1GenuineIntel-6-7EGenuineIntel-6-6CTopdownL1Backend_BoundIPCRet;Summary4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )SLOTS_SMTCore actual clocks when any Logical Processor is active on the Physical CoreInstructions per Store (lower number means higher occurrence rate)INST_RETIRED.ANY / FP_ARITH_INST_RETIRED.SCALAR_SINGLEInstructionsL2MPKI_LoadAverage number of parallel data read requests to external memory. Accounts for demand loads and L1/L2 prefetchesevent=0xf1,period=100003,umask=0x7This event counts L2 writebacks that access L2 cacheevent=0xd3,period=100007,umask=0x1All retired load uops. (Precise Event - PEBS)  Supports address when precise (Precise event)Retired store uops that miss the STLB. (Precise Event - PEBS)  Supports address when precise (Precise event)Cycles when offcore outstanding cacheable Core Data Read transactions are present in SuperQueue (SQ), queue to uncore  Spec update: BDM76offcore_requests_outstanding.demand_data_rdSplit locks in SQevent=0xc7,period=2000003,umask=0x20idq.dsb_uopsevent=0x79,cmask=1,edge=1,period=2000003,umask=0x10event=0x9c,period=2000003,umask=0x1idq_uops_not_delivered.cycles_0_uops_deliv.coreCycles per thread when 4 or more uops are not delivered to Resource Allocation Table (RAT) when backend of the machine is not stalledNumber of times an HLE execution aborted due to uncommon conditionsLoads with latency value being above 512  Spec update: BDM100, BDM35 (Must be precise)event=0xcd,period=50021,umask=0x1,ldlat=0x8rtm_retired.aborted_misc4Number of times we entered an RTM region
 does not count nested transactionsevent=0x54,period=2000003,umask=0x10Speculative and retired  branchesevent=0xc4,period=100007,umask=0x8br_misp_exec.taken_indirect_near_callevent=0xc5,period=400009,umask=0x1event=0xc5,period=100007,umask=0x8Reference cycles when the thread is unhalted (counts at 100 MHz rate)event=0x3c,any=1,period=2000003,umask=0x1event=0x3c,any=1,period=2000003Counts number of cycles the CPU has at least one pending  demand* load request missing the L2 cacheThis event counts self-modifying code (SMC) detected, which causes a machine clearevent=0xa1,period=2000003,umask=0x40uops_executed.core_cycles_noneevent=0xb1,cmask=4,period=2000003,umask=0x1uops_executed_port.port_1_coreevent=0x21,umask=0x20Cycles Memory is in self refresh power mode. Unit: uncore_imc freq_max_power_cycles %This event counts store misses in all DTLB levels that cause a completed page walk (1G  page size). The page walk can end with or without a fault  Spec update: BDM69This event counts retired load uops which data sources were L3 Hit and a cross-core snoop missed in the on-pkg core cache  Supports address when precise.  Spec update: BDM100 (Precise event)Data from local DRAM either Snoop not needed or Snoop Miss (RspI)  Supports address when precise.  Spec update: BDE70, BDM100 (Precise event)Retired load uops that miss the STLB  Supports address when precise (Precise event)This event counts load uops with true STLB miss retired to the architected path. True STLB miss is an uop triggering page walk that gets completed without blocks, and later gets retired. This page walk can end up with or without a fault  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020004Counts all demand data writes (RFOs)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200028000offcore_response.pf_l2_data_rd.l3_hit.snoop_missoffcore_response.pf_l2_rfo.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0200Counts all prefetch (that bring data to LLC only) data readsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020100Counts randomly selected loads with latency value being above eight  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000091offcore_response.all_pf_data_rd.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000122offcore_response.all_rfo.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000040offcore_response.pf_l2_code_rd.l3_miss_local_dram.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000100This 48-bit fixed counter counts the UCLK cycles. Unit: uncore_ncu offcore_response.all_reads.llc_hit.hitm_other_coreCounts all demand & prefetch data reads miss the L3 and the data is returned from remote dramoffcore_response.all_reads.llc_miss.remote_hit_forwardCycles the L2 transfers data to the corel2_ifetch.self.mesievent=0x2d,period=200000,umask=0x4fl2_reject_busq.self.demand.m_stateL2 cache requestsbus_trans_burst.all_agentsbus_trans_ifetch.selfIO bus transactionsevent=0x67,period=200000,umask=0x40dispatch_blocked.anyevent=0x7e,period=200000,umask=0xe0bogus_brbr_missp_type_retired.cond_takenevent=0x12,period=2000000,umask=0x81resource_stalls.div_busystore_forwards.anyevent=0x8,period=200000,umask=0x9event=0x2e,period=200003,umask=0x41Loads retired that came from DRAM (Precise event capable)  Supports address when precise (Must be precise)offcore_response.any_request.l2_hitCounts bus lock and split lock requests that have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000008Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that hit the L2 cacheCounts demand data partial reads, including data in uncacheable (UC) or uncacheable write combining (USWC) memory types that miss the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000020Counts reads for ownership (RFO) requests generated by L2 prefetcher that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts any data writes to uncacheable write combining (USWC) memory region  that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xe6,period=200003,umask=0x8Counts core cycles during which there are pending interrupts, but interrupts are masked (EFLAGS.IF = 0)Counts far branch instructions retired.  This includes far jump, far call and return, and Interrupt call and return (Must be precise)Counts mispredicted retired Jcc (Jump on Conditional Code/Jump if Condition is Met) branch instructions retired that were supposed to be taken but the processor predicted that it would not be taken (Must be precise)Counts the number of instructions that retire execution. For instructions that consist of multiple uops, this event counts the retirement of the last uop of the instruction. The counter continues counting during hardware interrupts, traps, and inside interrupt handlers.  This event uses fixed counter 0.  You cannot collect a PEBs record for this eventevent=0xca,period=200003,umask=0x1Loads blocked (Precise event capable) (Must be precise)event=0xc3,period=200003Uops retired (Precise event capable) (Must be precise)Counts data reads generated by L1 or L2 prefetchers outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000008000Counts reads for ownership (RFO) requests (demand & prefetch) have any transaction responses from the uncore subsystemCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache true miss for the L2 cache with a snoop miss in the other processor moduleCounts data cacheline reads generated by hardware L2 cache prefetcher have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests generated by L2 prefetcher outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000004800event=0xc3,period=20003,umask=0x8Page walk completed due to a demand load to a 2M or 4M pageitlb_misses.walk_completed_1gbCounts page walks completed due to instruction fetches whose address translations missed in the TLB and were mapped to 1GB pages.  The page walks can end with or without a page faultevent=0x85,period=2000003,umask=0x4Counts page walks completed due to instruction fetches whose address translations missed in the TLB and were mapped to 4K pages.  The page walks can end with or without a page faultl2_lines_out.demand_dirtyoffcore_response.pf_l3_rfo.l3_hit.any_responseCycles per thread when 3 or more uops are not delivered to Resource Allocation Table (RAT) when backend of the machine is not stalled  Spec update: HSD135Randomly selected loads with latency value being above 256  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)offcore_response.all_code_rd.l3_miss.local_dramCounts all prefetch (that bring data to L2) RFOs miss in the L3Cycles with pending L1 cache miss loadsunc_cbo_cache_lookup.extsnp_iMiss in all TLB levels causes a page walk of any page size (4K/2M/4M/1G)event=0x49,period=100003,umask=0x80Completed page walks due to misses in ITLB 2M/4M page entriesNumber of DTLB page walker hits in the L3 + XSNP  Spec update: HSD25page_walker_loads.ept_itlb_l1offcore_response.demand_data_rd.llc_hit.hitm_other_corel2_store_lock_rqsts.missCore-originated cacheable demand requests missed LLCmem_load_uops_llc_hit_retired.xsnp_noneRetired store uops that miss the STLB. (Precise Event)Counts demand & prefetch data reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00010004Counts requests where the address of an atomic lock instruction spans a cache line boundary or the lock instruction is executed on uncacheable addressNumber of SSE* or AVX-128 FP Computational packed double-precision uops issued this cycleNumber of SSE* or AVX-128 FP Computational packed single-precision uops issued this cycleevent=0xc1,period=100003,umask=0x20event=0x11,period=2000003,umask=0x21 / ( ((FP_COMP_OPS_EXE.SSE_SCALAR_SINGLE + FP_COMP_OPS_EXE.SSE_SCALAR_DOUBLE) / UOPS_EXECUTED.THREAD) + ((FP_COMP_OPS_EXE.SSE_PACKED_DOUBLE + FP_COMP_OPS_EXE.SSE_PACKED_SINGLE + SIMD_FP_256.PACKED_SINGLE + SIMD_FP_256.PACKED_DOUBLE) / UOPS_EXECUTED.THREAD) )Loads with latency value being above 128 (Must be precise)Loads with latency value being above 32 (Must be precise)Cycles which a Uop is dispatched on port 2event=0xa1,any=1,period=2000003,umask=0xcRetired uops (Precise event)unc_cbo_cache_lookup.write_filterCycle PMH is busy with a walk due to demand loadsCounts all data/code/rfo reads (demand & prefetch) that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts miscellaneous accesses that include port i/o, MMIO and uncacheable memory accessesoffcore_response.pf_llc_data_rd.llc_hit.snoop_missCounts all demand & prefetch code reads that miss the LLC  and the data returned from remote dramevent=0xb(UNC_P_FREQ_BAND2_CYCLES / UNC_P_CLOCKTICKS) * 100.event=0xd,edge=1dtlb_load_misses.demand_ld_walk_completedRetired load uops which data sources were LLC and cross-core snoop hits in on-pkg core cacheNumber of occurences waiting for the checkpoints in Resource Allocation Table (RAT) to be recovered after Nuke due to all other cases except JEClear (e.g. whenever a ucode assist is needed like SSE exception, memory disambiguation, etc...)This event counts the number of Uops issued by the front-end of the pipeilne to the back-endevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010070offcore_response.any_pf_l2.l2_hit_this_tile_eCounts any Read request  that accounts for responses which hit its own tile's L2 with data in S stateoffcore_response.any_request.l2_hit_this_tile_mCounts Demand cacheable data writes that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeoffcore_response.demand_rfo.l2_hit_this_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080080offcore_response.partial_reads.l2_hit_this_tile_fCounts Software Prefetches that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateCounts UC code reads (valid only for Outstanding response type)  that accounts for any responseCounts all instruction fetches, including uncacheable fetchesCounts the number of times the machine clears due to memory ordering hazardsevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000100Counts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses from MCDRAM (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400020offcore_response.uc_code_reads.mcdram_nearevent=0xc5,period=200003,umask=0xbfCounts the number of core cycles when no micro-ops are allocated and the alloc pipe is stalled waiting for a mispredicted branch to retirerecycleq.sta_fullevent=0xcb,period=200003,umask=0x1event=0x5,edge=1,period=100003,umask=0x3l1d_cache_lock.e_stateL1 data cache load lock hitsL2 lines allocated in the S stateevent=0x24,period=200000,umask=0x30L2 instruction fetchesMemory instructions retired above 16 clocks (Precise Event)mem_inst_retired.latency_above_threshold_32768mem_inst_retired.latency_above_threshold_4096Instructions retired which contains a load (Precise Event)Retired loads that miss the LLC cache (Precise Event)Load instructions retired that HIT modified data in sibling core (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x144Offcore code reads satisfied by the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1844Offcore requests that HITM in a remote cacheOffcore RFO requests satisfied by the LLC and HIT in a sibling coreoffcore_response.any_rfo.local_cacheAll offcore writebacksoffcore_response.data_ifetch.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8077Offcore code or data read requests satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x777offcore_response.data_ifetch.remote_cache_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F03event=0xb7,period=100000,umask=0x1,offcore_rsp=0x403offcore_response.demand_data.local_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF01offcore_response.demand_ifetch.any_cache_dramOffcore demand code reads satisfied by the LLC and not found in a sibling coreoffcore_response.demand_rfo.any_locationevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F80Offcore prefetch RFO requests satisfied by the LLC and HIT in a sibling coreOffcore prefetch RFO requests that HIT in a remote cacheOffcore prefetch requests satisfied by any cache or DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x770X87 Floating point assists for invalid output value (Precise Event)fp_comp_ops_exe.sse_fp_packedOffcore requests satisfied by a remote DRAMOffcore RFO requests satisfied by the local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6077event=0xb7,period=100000,umask=0x1,offcore_rsp=0xF803offcore_response.demand_data_rd.any_dramoffcore_response.demand_ifetch.any_llc_missOffcore prefetch data requests satisfied by any DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2010event=0x89,period=20000,umask=0x1Mispredicted non call branches executedCycles when thread is not halted (fixed counter)Instruction Queue full stall cyclesload_hit_preevent=0xa2,period=2000000,umask=0x4SIMD Scalar-Double Uops retired (Precise Event)uops_executed.port2_coreevent=0xae,period=2000000,umask=0x1event=0xcb,period=200000,umask=0x80Counts the number of demand Data Read requests that miss L2 cache. Only not rejected loads are countedmem_inst_retired.split_loadsCounts retired store instructions that split across a cacheline boundary  Supports address when precise (Precise event)mem_load_l3_hit_retired.xsnp_missCounts both cacheable and non-cacheable code read requestsoffcore_response.demand_code_rd.l3_hit_e.spl_hitoffcore_response.demand_code_rd.l4_hit_local_l4.snoop_hitmoffcore_response.demand_code_rd.l4_hit_local_l4.snoop_not_neededoffcore_response.demand_rfo.l3_hit_e.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x401C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000088000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40108000event=0xe6,period=100003,umask=0x1frontend_retired.itlb_missRetired instructions that are fetched after an interval where the front-end delivered no uops for a period of 256 cycles which was not interrupted by a back-end stall (Precise event)Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 4 cycles which was not interrupted by a back-end stall (Precise event)Cycles where a code fetch is stalled due to L1 instruction cache misscycle_activity.stalls_l3_missCounts randomly selected loads when the latency from first dispatch to completion is greater than 256 cycles  Supports address when precise (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000100002Instruction decoders utilized in a cycleCounts resource-related stall cyclesCounts cycles during which no uops were dispatched from the Reservation Station (RS) per threadskl metricsBigFoot;Fed;Frontend;IcMiss;MemoryTLB_SMTAverage per-core data access bandwidth to the L3 cache [GB / sec]1000 * MEM_LOAD_RETIRED.L3_MISS / INST_RETIRED.ANYCounts completed page walks  (all page sizes) caused by demand data loads. This implies it missed in the DTLB and further levels of TLB. The page walk can end with or without a faultCounts 1 per cycle for each PMH that is busy with a page walk for a load. EPT page walk duration are excluded in Skylake microarchitectureCounts any code reads (demand & prefetch) that have any response typeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000044Counts any request that hit in the other module where modified copies were found in other core's L1 cacheCounts any rfo reads (demand & prefetch) that hit in the other module where modified copies were found in other core's L1 cacheoffcore_response.pf_l2_code_rd.l2_miss.anyALL_BRANCHES counts the number of any branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)RETURN counts the number of near RET branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)Retired load uops which data sources were hits in LLC without snoops required. (Precise Event - PEBS) (Precise event)This event counts all data requests (demand/prefetch data reads and demand data writes (RFOs) that miss the LLC  where the data is returned from local DRAMCounts all prefetch (that bring data to L2) RFOs that miss the LLC  and the data returned from dramNumber of any page walk that had a miss in LLC. Does not necessary cause a SUSPENDMispredicted taken branch instructions retired. (Precise Event - PEBS) (Precise event)event=0xb0,period=100000,umask=0x80offcore_response.any_data.all_local_dram_and_remote_cache_hitREQUEST = ANY_DATA read and RESPONSE = IO_CSR_MMIOREQUEST = ANY IFETCH and RESPONSE = REMOTE_CACHE_HITMREQUEST = ANY_REQUEST and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = ANY RFO and RESPONSE = LOCAL_CACHEREQUEST = DEMAND_RFO and RESPONSE = LLC_HIT_NO_OTHER_COREevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5010REQUEST = PF_RFO and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DATA_IN and RESPONSE = OTHER_LOCAL_DRAMoffcore_response.demand_data.other_local_dramevent=0x49,period=200000,umask=0x20Counts all prefetch data reads that hit in the L3offcore_response.all_rfo.l3_hit.no_snoop_neededoffcore_response.all_rfo.l3_hit.snoop_hit_with_fwdCounts demand data reads that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all demand & prefetch data reads that miss in the L3Counts prefetch RFOs that miss the L3 and the data is returned from local dramCounts all demand code reads that miss the L3 and the data is returned from local or remote dramoffcore_response.pf_l3_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdCounts all prefetch (that bring data to LLC only) RFOs that miss the L3 and the data is returned from local dramCORE_POWER.LVL0_TURBO_LICENSE / 2 / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )Read Pending Queue Occupancy. Unit: uncore_imc MMIO reads. Derived from unc_cha_tor_inserts.ia_miss. Unit: uncore_cha PCI Express bandwidth writing at IIO, part 0. Unit: uncore_iio event=0x83,ch_mask=0x08,fc_mask=0x07,umask=0x01event=0x59,umask=0x01CHA to iMC Full Line Writes Issued; Full Line Non-ISOCH. Unit: uncore_cha event=0x3d,umask=0x04unc_iio_comp_buf_inserts.cmpd.part1unc_iio_data_req_of_cpu.peer_write.part3Read request for up to a 64 byte transaction is  made by IIO Part1 to Memory. Unit: uncore_iio event=0x84,ch_mask=0x01,fc_mask=0x07,umask=0x02Inbound read requests to coherent memory, received by the IRP and inserted into the Fire and Forget queue (FAF), a queue used for processing inbound reads in the IRPunc_i_faf_occupancyevent=0x23unc_m2m_direct2upi_txn_overrideCounts when a read message that was sent direct to the Intel Ultra Path Interconnect (bypassing the CHA) was overriddenMulti-socket cacheline Directory lookup (cacheline found in I state). Unit: uncore_m2m unc_m2m_directory_lookup.state_sevent=0x2d,umask=0x4event=0x12,umask=0x2OCR.ALL_DATA_RD.L3_HIT_E.ANY_SNOOP  OCR.ALL_DATA_RD.L3_HIT_E.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080491ocr.all_pf_data_rd.l3_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080490OCR.ALL_PF_DATA_RD.L3_HIT_M.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080120ocr.all_pf_rfo.l3_hit_m.any_snoopocr.all_pf_rfo.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040120OCR.ALL_READS.L3_HIT_E.HIT_OTHER_CORE_NO_FWD  OCR.ALL_READS.L3_HIT_E.HIT_OTHER_CORE_NO_FWDocr.all_reads.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4000407F7OCR.ALL_READS.L3_HIT_M.SNOOP_MISSocr.all_rfo.l3_hit_f.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040122Counts all demand code reads OCR.DEMAND_CODE_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.DEMAND_CODE_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDocr.demand_code_rd.l3_hit_e.any_snoopocr.demand_code_rd.l3_hit_f.hit_other_core_no_fwdCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_S.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200001ocr.demand_rfo.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_M.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80208000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800208000ocr.pf_l1d_and_sw.l3_hit_f.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200400ocr.pf_l2_data_rd.l3_hit_e.snoop_noneocr.pf_l2_data_rd.l3_hit_f.snoop_noneocr.pf_l2_data_rd.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_HIT.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080020ocr.pf_l2_rfo.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040080Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDocr.pf_l3_data_rd.l3_hit_s.snoop_noneCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_HIT.HITM_OTHER_CORE OCR.PF_L3_RFO.L3_HIT.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100100offcore_response.all_pf_data_rd.l3_hit_f.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_S.SNOOP_NONEoffcore_response.all_pf_rfo.pmm_hit_local_pmm.snoop_not_neededThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_M.HIT_OTHER_CORE_FWDoffcore_response.all_rfo.supplier_none.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_data_rd.l3_hit_s.hit_other_core_no_fwdoffcore_response.demand_rfo.l3_hit_f.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDoffcore_response.demand_rfo.supplier_none.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_E.SNOOP_NONEoffcore_response.other.l3_hit_f.hit_other_core_no_fwdoffcore_response.other.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDoffcore_response.pf_l2_data_rd.l3_hit_e.snoop_noneoffcore_response.pf_l2_data_rd.l3_hit_m.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_hit_e.hit_other_core_fwdoffcore_response.pf_l2_rfo.l3_hit_f.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_M.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400080This event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_F.SNOOP_NONEoffcore_response.pf_l3_rfo.l3_hit_m.hitm_other_coreoffcore_response.pf_l3_rfo.supplier_none.hit_other_core_no_fwdocr.all_data_rd.l3_miss.hitm_other_coreocr.all_data_rd.l3_miss_local_dram.snoop_noneocr.all_data_rd.l3_miss_remote_hop1_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000491OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000490OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSOCR.ALL_PF_RFO.L3_MISS.SNOOP_MISS OCR.ALL_PF_RFO.L3_MISS.SNOOP_MISSOCR.ALL_READS.L3_MISS_LOCAL_DRAM.ANY_SNOOP  OCR.ALL_READS.L3_MISS_LOCAL_DRAM.ANY_SNOOPocr.all_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREocr.demand_rfo.l3_miss_remote_hop1_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810008000Counts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.SNOOP_NONEocr.pf_l1d_and_sw.l3_miss_local_dram.snoop_missCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDocr.pf_l2_rfo.l3_miss.hitm_other_coreocr.pf_l2_rfo.l3_miss.remote_hitmocr.pf_l2_rfo.l3_miss_local_dram.snoop_missocr.pf_l3_data_rd.l3_miss.any_snoopocr.pf_l3_data_rd.l3_miss.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.NO_SNOOP_NEEDEDoffcore_response.all_reads.l3_miss.snoop_missoffcore_response.all_reads.l3_miss_remote_hop1_dram.hitm_other_coreoffcore_response.all_rfo.l3_miss.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.SNOOP_MISSoffcore_response.demand_data_rd.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.SNOOP_NONEoffcore_response.demand_data_rd.l3_miss_remote_hop1_dram.any_snoopThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l1d_and_sw.l3_miss.hit_other_core_no_fwdoffcore_response.pf_l1d_and_sw.l3_miss_local_dram.snoop_missThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEoffcore_response.pf_l2_data_rd.l3_miss.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.pf_l3_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.pf_l3_rfo.l3_miss.hit_other_core_fwdOCR.ALL_DATA_RD.SUPPLIER_NONE.ANY_SNOOP  OCR.ALL_DATA_RD.SUPPLIER_NONE.ANY_SNOOPOCR.ALL_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWD  OCR.ALL_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDOCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDED OCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDOCR.ALL_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWD  OCR.ALL_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDOCR.ALL_RFO.SUPPLIER_NONE.SNOOP_MISSocr.demand_rfo.supplier_none.snoop_missCounts any other requests OCR.OTHER.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.other.supplier_none.hitm_other_coreocr.pf_l2_rfo.supplier_none.any_snoopocr.pf_l3_data_rd.any_responseCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREIntel Optane DC persistent memory bandwidth total (MB/sec). Unit: uncore_imc event=0xea,umask=0x2Tag Check; HitRetired load instructions whose data sources were hits in L3 without snoops required  Supports address when precise (Precise event)ocr.other.l3_hit.snoop_hit_no_fwdCounts miscellaneous requests, such as I/O and un-cacheable accesses that hit a cacheline in the L3 where a snoop hit in another core, data forwarding is not requiredCounts instruction fetch tag lookups that hit in the instruction cache (L1I). Counts at 64-byte cache-line granularity. Accounts for both cacheable and uncacheable accessesBR_INST_RETIRED.COND_TAKEN / BR_INST_RETIRED.ALL_BRANCHESCounts the number of times RTM abort was triggeredevent=0xc9,period=100003,umask=0x40event=0x54,period=100003,umask=0x10All branch instructions retired (Precise event)event=0xc5,period=50021,umask=0x20Counts core cycles when the Resource allocator was stalled due to recovery from an earlier branch misprediction or machine clear eventtopdown.slots_pCycles when at least one PMH is busy with a page walk for a demand load( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING ) / ( 2 * CORE_CLKS )Retired instructions with at least one load to uncacheable memory-type, or at least one cache-line split locked access (Bus Lock)  Supports address when precise (Precise event)Counts demand data reads that were supplied by a cache on a remote socket where a snoop hit in another core's caches which forwarded the unmodified data to the requesting coreCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit a modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C27F0event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84408000Counts miscellaneous requests, such as I/O and un-cacheable accesses that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallyocr.prefetches.l3_miss_localevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x70CC00477Counts responses to snoops indicating the line will now be (I)nvalidated in this core's caches without being forwarded back to the requestor. The line was in Forward, Shared or Exclusive (FSE) state in this cores caches.  A single snoop response from the core counts on all hyperthreads of the coreocr.demand_data_rd.local_pmmCounts demand data reads that were supplied by PMM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAMevent=0x20,umask=0x02PMM Read Queue Inserts. Unit: uncore_imc event=0x35,umask=0xC001FD01event=0x35,umask=0xC807FE01TOR Inserts : RFO_Prefs issued by iA Cores that Missed the LLC. Unit: uncore_cha event=0x35,umask=0xC80FFF01event=0x35,umask=0xC896FE01event=0x35,umask=0xC806FE01unc_cha_tor_inserts.ia_miss_rfo_pref_localunc_iio_data_req_of_cpu.cmpd.part5event=0xc1,ch_mask=0x40,fc_mask=0x07,umask=0x01event=0xc2,ch_mask=0x10,fc_mask=0x04,umask=0x03Clockticks of the power control unit (PCU). Unit: uncore_pcu Counts the number of first level data cacheline (dirty) evictions caused by misses, stores, and prefetches.  Does not count evictions or dirty writebacks caused by snoops.  Does not count a replacement unless a (dirty) line was written backCounts the number of L2 Cache accesses that miss the L2 and get rejected. Counts on a per core basisocr.l1wb_m.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0800Counts the number of BACLEARS due to an indirect branchbaclears.uncondevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x802184000000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2002184000000This event is deprecated. Refer to new event BUS_LOCK.SELF_LOCKSbus_lock.cycles_other_blockocr.demand_data_and_l1pf_rd.local_dramocr.demand_data_rd.outstandingocr.demand_rfo.outstandingevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000000010000Counts the number of reference cycles that the core is not in a halt state. The core enters the halt state when it is running the HLT instruction.  This event is not affected by core frequency changes and increments at a fixed frequency that is also used for the Time Stamp Counter (TSC). This event uses a programmable general purpose performance countercycles_div_busy.anyevent=0x3,period=1000003,umask=0x4event=0x3,period=1000003,umask=0x2topdown_fe_bound.branch_detectCounts the number of x87 uops retired, includes those in MS flows (Precise event)event=0x35,umask=0xC8A7FF01TOR Inserts : All requests from IO Devices that hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsunc_cha_tor_occupancy.ia_hit_drd_opt_prefNumber Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 7Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 1Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Number Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Number Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 3Counts the number of page walks completed due to loads (including SW prefetches) whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to any page size. Includes page walks that page faultevent=0x3,period=1000003,umask=0x8Average per-core data access bandwidth to the L3 cache [GB / sec]. Unit: cpu_core Instructions per Far Branch ( Far Branches apply upon transition from application to operating system, handling interrupts, exceptions) [lower number means higher occurrence rate]. Unit: cpu_core Counts the total number of issue slots  that were not consumed by the backend due to backend stalls.  Note that UOPS must be available for consumption in order for this event to count.  If a uop is not available (IQ is empty), this event will not count.  All of these subevents count backend stalls, in slots, due to a resource limitation.   These are not cycle based events and therefore can not be precisely added or subtracted from the Backend_Bound subevents which are cycle based.  These subevents are supplementary to Backend_Bound and can be used to analyze results from a resource perspective at allocationAverage Frequency Utilization relative nominal frequency. Unit: cpu_atom Cycles_per_Demand_Load_DRAM_HitMemLoadPKIRetired load instructions missed L1 cache as data sources  Supports address when precise (Precise event). Unit: cpu_core Retired load instructions missed L3 cache as data sources  Supports address when precise (Precise event). Unit: cpu_core event=0xc6,period=100007,umask=0x1,frontend=0x17Counts the number of issue slots every cycle that were not consumed by the backend due to memory reservation stalls in which a scheduler is not able to accept uops. Unit: cpu_atom Taken branch instructions retired (Precise event). Unit: cpu_core int_vec_retired.128bitevent=0xe7,period=1000003,umask=0x20Cycles optimal number of Uops delivered by the LSD, but did not come from the decoder. Unit: cpu_core topdown.bad_spec_slotsThis event is deprecated. Refer to new event UOPS_RETIRED.STALLS. Unit: cpu_core unc_m_dram_thermal_hotACT command for a read request sent to DRAM. Unit: uncore_imc Each cycle counts number of all outgoing valid entries in ReqTrk. Such entry is defined as valid from its allocation in ReqTrk till deallocation. Accounts for Coherent and non-coherent traffic. Unit: uncore_arb event=0x26,period=200003,umask=0x2ASSISTS.SSE_AVX_MIXFP_ARITH_DISPATCHED.PORT_1event=0x2a,period=100003,umask=0x1,offcore_rsp=0x84002380Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those DRAM accesses that are controlled by the close SNC Clusterocr.write_estimate.memoryCounts cycles during which the reservation station (RS) is empty for this logical processorCycles no uop executed while RS was not empty, the SB was not full and there was no outstanding loadUOPS_DECODED.DEC0_UOPSevent=0x3,umask=0x0000000001unc_m_pre_count.pgt_pch0event=0x5,umask=0x0000000080Read and Write Requests; Writes Remote. Unit: uncore_cha Valid Flits Sent : Slot 2. Unit: uncore_upi event=0x2,umask=0x0000000010event=0x2,umask=0x0000000097Valid Flits Received : Data. Unit: uncore_upi event=0x32,umask=0x0000000004FAF occupancy. Unit: uncore_irp event=0x83,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000080TOR Inserts; All Snoops from Remote. Unit: uncore_cha TOR Inserts; All from Local IA. Unit: uncore_cha event=0x35,umask=0x00C001FF08unc_cha_tor_inserts.io_rfoevent=0x35,umask=0x00c887ff01event=0x35,umask=0x00c8a7ff01unc_cha_tor_occupancy.io_miss_rfounc_cha_tor_occupancy.irq_non_iaunc_cha_tor_occupancy.prq_non_iosfevent=0x36,umask=0x00C000FF01unc_cha_tor_occupancy.rem_snpsevent=0x36,umask=0x00C001FFC8unc_cha_tor_occupancy.ia_hit_drd_prefTOR Occupancy; CRd Pref misses from local IA. Unit: uncore_cha event=0x36,umask=0x00c803fd04TOR Occupancy; DRd Opt from local IA. Unit: uncore_cha event=0x36,umask=0x00c8077e01event=0x35,umask=0x00CC47FD01event=0x35,umask=0x00C867FF01event=0x35,umask=0x00C8668A01TOR Inserts : WCiLFs issued by iA Cores targeting DDR that missed the LLC. Unit: uncore_cha event=0x35,umask=0x00C8670601unc_cha_tor_inserts.ia_wcilevent=0x36,umask=0x00cc47fd01Indirect Branch Prediction for potential multi-target branch (speculative)ic_fw32l2_cache_req_stat.ls_rd_blk_l_hit_xevent=0x64,umask=0x02l3_lookup_state.all_l3_req_typsevent=0x4,umask=0xffex_ret_copsex_ret_brn_resyncMMX instructionsex_div_busyRemote Link Controller Outbound Packet Types: Data (32B): Remote Link Controller 1The number of operations (uOps) and dual-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number multi-pipe uOps assigned to all pipesfp_sched_emptyfp_retired_ser_ops.sse_bot_retls_tablewalker.isideCycles where a dispatch group is valid but does not get dispatched due to a token stall. RETIRE Tokens unavailabledata_fabricSoftware Prefetch Instructions Dispatched (Speculative). PrefetchNTA instruction. See docAPM3 PREFETCHlevelevent=0x59,umask=0x02event=0xae,umask=0x80de_dis_dispatch_token_stalls1.fp_reg_file_rsrc_stalld_ratio(l2_cache_req_stat.ic_access_in_l2, bp_l1_tlb_fetch_hit + bp_l1_tlb_miss_l2_hit + bp_l1_tlb_miss_l2_tlb_miss)ls_sw_pf_dc_fills.lcl_l2Cycles where a dispatch group is valid but does not get dispatched due to a token stall. No tokens for Integer Scheduler Queue 2 availableop_cache_fetch_miss_ratiod_ratio(ic_tag_hit_miss.instruction_cache_miss, ic_tag_hit_miss.all_instruction_cache_accesses)ls_pref_instr_disp.allevent=0x4b,umask=0x07Software prefetch data cache fills from cache of another CCX in a different NUMA nodels_hw_pf_dc_fills.alternate_memoriesevent=0x70,umask=0x04l2_pf_hit_l2.l1_strideevent=0x72,umask=0x08l3_lookup_state.all_coherent_accesses_to_l3Average sampled latency when data is sourced from another CCX's cache when the address was in a different NUMA node. Unit: amd_l3 event=0x21f,umask=0x7ffWrite data beats (64 bytes) for local processor at Coherent Station (CS) 8event=0x1df,umask=0xbfeevent=0xdf,umask=0xbffevent=0x25f,umask=0xbffremote_socket_upstream_write_beats_iom0Data beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 7Data beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 4remote_socket_inf1_inbound_data_beats_ccm1Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 4remote_socket_inf1_outbound_data_beats_ccm6event=0xbdf,umask=0xf3efp_ret_x87_fp_ops.mul_opsevent=0xa,umask=0xf0event=0xb,umask=0x01Retired SSE and AVX integer multiply opsevent=0xb,umask=0x80fp_pack_ops_retired.fp256_sqrtpacked_int_op_type.int128_shiftevent=0x94,umask=0x07In each cycle counts ops unable to dispatch because the dispatch cycle was granted to the other SMT threadbad_speculation_mispredictsbackend_bound * (1 - d_ratio(ex_no_retire.load_not_complete, ex_no_retire.not_complete))all_l1_data_cache_fillsdram_write_data_for_remote_processorumc_activate_cmd_rateFP_DISPATCHED_FPU_OPSDC_DISPATCHED_PREFETCH_INSTRUCTIONSFR_RETIRED_BRANCHESL1_DCACHE_NEON_CACHEABLEPRED_BRANCH_PRED_TAKENCYCLES_STALLED_NEON_MRCEVENT_60HEVENT_82HEVENT_A0HPLE_REQUEST_PROGRAMMEDL1I_CACHE_REFILLBR_PREDEVENT_10BHEVENT_153HEVENT_162HEVENT_171HEVENT_175HEVENT_187HEVENT_18AHEVENT_192HEVENT_1AEHEVENT_1BCHEVENT_1E6HEVENT_1FCHEVENT_23DHEVENT_286HEVENT_2A6HEVENT_2ADHEVENT_2EDHEVENT_2F5HEVENT_320HEVENT_337HEVENT_353HEVENT_367HEVENT_377HEVENT_38CHEVENT_3C3HEVENT_3F1HUNALIGNED_LDST_SPECBR_IMMED_SPEChnf_sf_hithni_pcie_serializationrnd_rdb_replaycxra_req_chain_avg_lencxla_rx_tlp_link1clkdiv2_read_depthBRANCH_LINK_STAC_PREDICTEDITLB_HW_TABLE_SEARCH_CYCLESTOUCH_ALIASLSU_LMQ_FULL_STALLFP_DENORMALIZED_RESULTVTQ_RESUMES_DUE_TO_CTX_CHANGELSU_COMPLETES_FP_STORE_SINGLESTWCX_INSTR_COMPLETEDL1_DATA_LOAD_MISS_CYCLESBUS_RETRY_DUE_TO_L1_RETRYMARKED_GROUP_COMPLETEDBTB_HITS_PSEUDO_HITSCYCLES_SU2_SCHED_STALLEDPART2_MISALIGNED_CACHE_ACCESSPMC2_OVERFLOWL2_CACHE_ACCESSESic-fillnonpostwrszbytebuffer-releasetscPPC970AMD_K7GENERICTSname: %s
mem_load_l3_miss_retired.remote_hitm{"type": "pmcallocate"{"type": "proccsw"{"type": "map_in"TmaL1_SMTINST_RETIRED.ANY / BR_INST_RETIRED.NEAR_CALL1000 * MEM_LOAD_UOPS_RETIRED.L2_MISS / INST_RETIRED.ANY1000 * ( L2_RQSTS.REFERENCES - L2_RQSTS.MISS ) / INST_RETIRED.ANYL3MPKI( cpu@ITLB_MISSES.WALK_DURATION\,cmask\=1@ + cpu@DTLB_LOAD_MISSES.WALK_DURATION\,cmask\=1@ + cpu@DTLB_STORE_MISSES.WALK_DURATION\,cmask\=1@ + 7 * ( DTLB_STORE_MISSES.WALK_COMPLETED + DTLB_LOAD_MISSES.WALK_COMPLETED + ITLB_MISSES.WALK_COMPLETED ) ) / CPU_CLK_UNHALTED.THREADC6 residency percent per coreC7_Core_Residencyl2_trans.all_pfmem_uops_retired.stlb_miss_loadsCacheable and noncachaeble code read requestsevent=0xb0,period=100003,umask=0x1fp_arith_inst_retired.128b_packed_doubleNumber of SSE/AVX computational 128-bit packed single precision floating-point instructions retired.  Each count represents 4 computations. Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP RSQRT SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementevent=0xca,period=100003,umask=0x8Cycles with less than 3 uops delivered by the front endhle_retired.aborted_misc1Number of times we entered an HLE region
 does not count nested transactionsThis event counts loads with latency value being above 256  Spec update: BDM100, BDM35 (Must be precise)tx_exec.misc1Unfriendly TSX abort triggered by  a vzeroupper instructionevent=0x54,period=2000003,umask=0x4Unhalted core cycles when the thread is in ring 0Cycles when divider is busy executing divide operationsbr_inst_exec.taken_conditionalTaken speculative and retired macro-conditional branch instructions excluding calls and indirectsThis is a precise version of BR_INST_RETIRED.ALL_BRANCHES that counts all (macro) branch instructions retired  Spec update: BDW98 (Must be precise)br_inst_retired.near_returnThis event counts taken speculative and retired mispredicted macro conditional branch instructionsThis event counts the number of mispredicted ret instructions retired.(Precise Event)event=0x3c,period=2000003,umask=0x1event=0,period=2000003,umask=0x3event=0x87,period=2000003,umask=0x1This event counts false dependencies in MOB when the partial comparison upon loose net check and dependency was resolved by the Enhanced Loose net mechanism. This may not result in high performance penalties. Loose net checks can fail when loads and stores are 4k aliasedevent=0x58,period=1000003,umask=0x1Cycles at least 2 micro-op is executed from any thread on physical coreCycles at least 3 micro-op is executed from any thread on physical coreThis event counts cycles during which the Resource Allocation Table (RAT) does not issue any Uops to the reservation station (RS) for the current threadMMIO reads. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox event=0x35,umask=0x3,filter_opc=0x190llc_misses.pcie_readllc_references.streaming_fullunc_h_requests.reads_localevent=0x21,umask=0x8power_state_occupancy.cores_c0 %Counts the number of cycles that we are in external PROCHOT mode.  This mode is triggered when a sensor off the die determines that something off-die (like DRAM) is too hot and must throttle to avoid damaging the chip. Unit: uncore_pcu event=0x6event=0x8,period=2000003,umask=0x60Load operations that miss the first DTLB level but hit the second and do not cause page walksitlb_misses.walk_completed_1gpage_walker_loads.dtlb_l1Number of DTLB page walker hits in the L3 + XSNP  Spec update: BDM69, BDM98Retired load uops which data sources were load uops missed L1 but hit FB due to preceding miss to the same cache line with data not ready  Supports address when precise (Precise event)offcore_response.all_pf_data_rd.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020120offcore_response.all_rfo.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0004offcore_response.demand_code_rd.supplier_none.snoop_hitmoffcore_response.demand_data_rd.supplier_none.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010002offcore_response.other.l3_hit.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0010offcore_response.pf_l3_code_rd.l3_hit.snoop_hit_no_fwdoffcore_response.pf_l3_rfo.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020100Randomly selected loads with latency value being above 8  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)offcore_response.all_data_rd.l3_miss.snoop_not_neededoffcore_response.all_pf_rfo.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000120offcore_response.corewb.l3_miss_local_dram.snoop_hitmoffcore_response.corewb.l3_miss_local_dram.snoop_hit_no_fwdoffcore_response.corewb.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000001offcore_response.other.l3_miss.snoop_missoffcore_response.other.l3_miss_local_dram.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104008000offcore_response.pf_l2_code_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000020offcore_response.pf_l3_data_rd.l3_hit.snoop_non_dramoffcore_response.all_reads.llc_miss.remote_hitmoffcore_response.all_requests.llc_miss.any_responsel2_dbus_busy_rd.selfl2_ifetch.self.s_statel2_ld.self.prefetch.s_stateevent=0x2d,period=200000,umask=0x41event=0x24,period=200000,umask=0x40l2_lock.self.mesievent=0x27,period=200000,umask=0x40Cycles no L2 cache requests are pendingl2_reject_busq.self.any.e_stateevent=0x2e,period=200000,umask=0x72Floating point assistsSIMD assists invokedsimd_comp_inst_retired.scalar_doubleevent=0xca,period=2000000,umask=0x8event=0xc7,period=2000000,umask=0x1event=0xb0,period=2000000,umask=0x80Floating point computational micro-ops executedevent=0x7,period=200000,umask=0x88event=0x7,period=200000,umask=0xfBus queue is emptyevent=0x6e,period=200000,umask=0x40bus_trans_def.selfevent=0x77,period=200000,umask=0x2bevent=0x77,period=200000,umask=0xbevent=0x88,period=2000000,umask=0x8cpu_clk_unhalted.coreCore cycles when core is not haltedDivide operations retiredevent=0x8,period=200000,umask=0x7event=0xc,period=2000000,umask=0x3page_walks.d_side_cyclesevent=0xd1,period=200003,umask=0x80Counts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that miss the L2 cacheoffcore_response.any_rfo.l2_miss.snoop_miss_or_no_snoop_neededoffcore_response.demand_data_rd.l2_miss.hit_other_core_no_fwdCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.partial_reads.l2_miss.anyevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600004000Counts partial cache line data writes to uncacheable write combining (USWC) memory region  that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.partial_writes.l2_miss.anyCounts data cacheline reads generated by hardware L2 cache prefetcher that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.sw_prefetch.l2_miss.anymisalign_mem_ref.load_page_splitevent=0xc4,period=200003Retired taken branch instructions (Precise event capable) (Must be precise)event=0xc4,period=200003,umask=0xebevent=0xc4,period=200003,umask=0xfeCore cycles when core is not halted  (Fixed event)Reference cycles when core is not halted  (Fixed event)uops_retired.msevent=0x81,period=200003,umask=0x4Counts uops retired that had a DTLB miss on load, store or either.  Note that when two distinct memory operations to the same page miss the DTLB, only one of them will be recorded as a DTLB miss  Supports address when precise (Must be precise)Counts data reads (demand & prefetch) true miss for the L2 cache with a snoop miss in the other processor moduleCounts bus lock and split lock requests miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts bus lock and split lock requests miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of writeback transactions caused by L1 or L2 cache evictions have any transaction responses from the uncore subsystemevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010004Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts demand reads for ownership (RFO) requests generated by a write to full data cache line miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts demand reads for ownership (RFO) requests generated by a write to full data cache line true miss for the L2 cache with a snoop miss in the other processor moduleCounts any data writes to uncacheable write combining (USWC) memory region  miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.sw_prefetch.any_responseCounts all L2 HW prefetcher requestsCounts the number of store RFO requests that hit the L2 cacheOffcore outstanding Demand Data Read transactions in uncore queue  Spec update: HSD78, HSD62, HSD61, HSM63, HSM80offcore_response.pf_l2_code_rd.l3_hit.any_responseCounts cycles DSB is delivered four uops. Set Cmask = 4This event counts the number cycles during which the Front-end allocated exactly zero uops to the Resource Allocation Table (RAT) while the Back-end of the processor is not stalled.  This event is counted on a per-core basis  Spec update: HSD135offcore_response.all_rfo.l3_miss.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00002Instructions retired from execution  Spec update: HSD140, HSD143The number of times that split load operations are temporarily blocked because all resources for handling the split accesses are in useCycles per core when uops are executed in port 6event=0x8,period=100003,umask=0x80Completed page walks due to store misses in one or more TLB levels of 4K page structureMisses at all ITLB levels that cause page walksNumber of ITLB page walker loads that hit in the L1+FBCount number of STLB flush attemptsRetired load uop whose Data Source was: forwarded from remote cache  Supports address when precise.  Spec update: HSM30 (Precise event)event=0x24,period=200003,umask=0x20mem_load_uops_llc_hit_retired.xsnp_missCounts demand code reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresNumber of DSB to MITE switchesCycles where a code-fetch stalled due to L1 instruction-cache miss or an iTLB missIncrement each cycle # of uops delivered to IDQ from MS by either DSB or MITE. Set Cmask = 1 to count cyclesCounts demand data reads that miss the LLC and the data returned from dramevent=0x14,cmask=1,edge=1,period=100003,umask=0x4Cycles that the divider is active, includes INT and FP. Set 'edge =1, cmask=1' to count the number of dividesevent=0x34,umask=0x01LLC lookup request that access cache and found line in S-state. Unit: uncore_cbox LLC lookup request that access cache and found line in E-state or S-state. Unit: uncore_cbox Counts all data/code/rfo reads (demand & prefetch) that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.pf_l2_data_rd.llc_hit.hit_other_core_no_fwdoffcore_response.all_code_rd.llc_miss.remote_hit_forwardCounts all demand & prefetch data reads that hits the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x6004003f7Counts all prefetch (that bring data to L2) code reads that miss the LLC  and the data returned from remote & local dramllc_misses.itom_writeStreaming stores (partial cache line). Derived from unc_c_tor_inserts.opcode.streaming_partial. Unit: uncore_cbox ItoM write hits (as part of fast string memcpy stores). Derived from unc_c_tor_inserts.opcode.itom_write_hit. Unit: uncore_cbox unc_q_txl_flits_g0.dataunc_p_freq_band0_transitions(UNC_P_FREQ_MAX_CURRENT_CYCLES / UNC_P_CLOCKTICKS) * 100.This event counts executed load operations with all the following traits: 1. addressing of the format [base + offset], 2. the offset is between 1 and 2047, 3. the address specified in the base register is in one page and the address [base+offset] is in anbr_misp_exec.taken_direct_near_callIncrements the number of flags-merge uops in flight each cycleresource_stalls2.bob_fullMemory page activates. Unit: uncore_imc Load misses at all DTLB levels that cause completed page walksCounts all the store micro-ops retiredThis event counts the number of store micro-ops retiredCounts the number of load micro-ops retired that miss in the L2  Supports address when precise (Precise event)event=0x4,period=200003,umask=0x10event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000083091offcore_response.any_pf_l2.l2_hit_far_tile_moffcore_response.any_read.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010008000Counts demand code reads and prefetch code reads that accounts for any responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080001Counts Demand cacheable data writes that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateCounts L1 data HW prefetches that accounts for any responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000402000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400040event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000020Counts Software Prefetches that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeCounts Software Prefetches that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0offcore_response.uc_code_reads.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800044event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181803091event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800070offcore_response.any_read.mcdram_nearCounts Demand cacheable data write requests  that accounts for responses from DDR (local and far)offcore_response.demand_code_rd.mcdramevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800002offcore_response.pf_l2_rfo.ddr_faroffcore_response.pf_software.mcdram_nearoffcore_response.uc_code_reads.mcdram_farCounts the number of unhalted reference clock cyclesThis event counts the number of micro-ops (uops) retired. The processor decodes complex macro instructions into a sequence of simpler uops. Most instructions are composed of one or two uops. Some instructions are decoded into longer sequences such as repeat instructions, floating point transcendental instructions, and assistsevent=0x5,edge=1,period=100003,umask=0x2l1d.replL1 data cache read in M stateevent=0x26,period=200000,umask=0xffL2 data prefetches in the I state (misses)event=0x24,period=200000,umask=0xc0l2_transactions.ifetchmem_inst_retired.latency_above_threshold_32event=0xcb,period=10000,umask=0x10mem_load_retired.other_core_l2_hit_hitmoffcore_response.any_data.llc_hit_other_core_hitoffcore_response.any_request.llc_hit_no_other_coreOffcore requests satisfied by a remote cacheOffcore writebacks to the LLC and not found in a sibling coreoffcore_response.corewb.local_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F77event=0xb7,period=100000,umask=0x1,offcore_rsp=0x877Offcore data reads, RFO's and prefetches satisfied by the IO, CSR, MMIO unitAll offcore demand data requestsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x804Offcore demand RFO requests satisfied by the LLC or local DRAMoffcore_response.other.remote_cache_hitmoffcore_response.pf_data_rd.io_csr_mmioOffcore prefetch data reads satisfied by the LLC and not found in a sibling coreoffcore_response.pf_data_rd.local_cache_dramoffcore_response.pf_ifetch.llc_hit_other_core_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x440event=0xb7,period=100000,umask=0x1,offcore_rsp=0x220Offcore prefetch requests satisfied by the IO, CSR, MMIO unitfp_comp_ops_exe.sse_double_precisionOffcore code reads that missed the LLCOffcore requests satisfied by the local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF808offcore_response.data_ifetch.any_dramOffcore code or data read requests satisfied by the local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF833event=0xb7,period=100000,umask=0x1,offcore_rsp=0x6003event=0xb7,period=100000,umask=0x1,offcore_rsp=0x6020br_inst_exec.indirect_non_callMispredicted return branches executedevent=0xc0,period=2000000,umask=0x2Retired floating-point operations (Precise Event)rat_stalls.anyOther Resource related stall cyclesssex_uops_retired.vector_integerevent=0xd1,period=2000000,umask=0x4event=0xb1,cmask=1,inv=1,period=2000000,umask=0x40event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080004offcore_response.demand_code_rd.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1001C0001offcore_response.demand_data_rd.l3_hit_s.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4001C0002offcore_response.demand_rfo.l3_hit_m.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100002offcore_response.demand_rfo.l4_hit_local_l4.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000048000event=0xc6,period=100007,umask=0x1,frontend=0x400806Number of times an HLE execution aborted due to hardware timer expirationDemand Data Read requests who miss L3 cacheoffcore_response.demand_code_rd.l3_hit_e.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20001C0002This event counts both direct and indirect near call instructions retired  Spec update: SKL091 (Precise event)Counts all the retired branch instructions that were mispredicted by the processor. A branch misprediction occurs when the processor incorrectly predicts the destination of the branch.  When the misprediction is discovered at execution, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct pathevent=0xa6,period=2000003,umask=0x40exe_activity.exe_bound_0_portsinst_decoded.decodersevent=0xd,period=2000003,umask=0x1Loads blocked due to overlapping with a preceding store that cannot be forwardedCounts false dependencies in MOB when the partial comparison upon loose net check and dependency was resolved by the Enhanced Loose net mechanism. This may not result in high performance penalties. Loose net checks can fail when loads and stores are 4k aliasedMemory_Data_TLBsIpDSB_Miss_Ret1000 * MEM_LOAD_RETIRED.FB_HIT / INST_RETIRED.ANY( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING + EPT.WALK_PENDING ) / ( 2 * CPU_CLK_UNHALTED.THREAD )Stores that miss the DTLB and hit the STLBCycles when at least one PMH is busy with a page walk for a store. EPT page walk duration are excluded in SkylakeCounts 1 per cycle for each PMH that is busy with a page walk for a store. EPT page walk duration are excluded in SkylakeCounts streaming store that miss L2Store address buffer fullALL_TAKEN_BRANCHES counts the number of all taken branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Must be precise)This event counts the number of micro-ops retired. The processor decodes complex macro instructions into a sequence of simpler micro-ops. Most instructions are composed of one or two micro-ops. Some instructions are decoded into longer sequences such as repeat instructions, floating point transcendental instructions, and assists. In some cases micro-op sequences are fused or whole instructions are fused into one micro-op. See other UOPS_RETIRED events for differentiating retired fused and non-fused micro-opsCounts demand & prefetch data reads that hit in the LLC and the snoops sent to sibling cores return clean responseCounts prefetch RFOs that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.pf_llc_code_rd.llc_hit.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) data reads that hit in the LLCThis event counts cycles during which the microcode sequencer assisted the front-end in delivering uops.  Microcode assists are used for complex instructions or scenarios that can't be handled by the standard decoder.  Using other instructions, if possible, will usually improve performance.  See the Intel 64 and IA-32 Architectures Optimization Reference Manual for more informationoffcore_requests.demand.read_codeREQUEST = DATA_IFETCH and RESPONSE = ANY_CACHE_DRAMREQUEST = DATA_IFETCH and RESPONSE = IO_CSR_MMIOREQUEST = DATA_IN and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = DEMAND_IFETCH and RESPONSE = ANY_CACHE_DRAMREQUEST = DEMAND_IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = DEMAND_RFO and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = PF_RFO and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5070Super Queue LRU hints sent to LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x30ffoffcore_response.corewb.any_dram_and_remote_fwdREQUEST = CORE_WB and RESPONSE = OTHER_LOCAL_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf801offcore_response.demand_ifetch.any_dram_and_remote_fwdsnoopq_requests_outstanding.codeevent=0xb0,period=100000,umask=0x20Offcore data reads, RFOs, and prefetches that HITM in a remote cacheRetired load instructions which data sources missed L3 but serviced from local DRAM  Supports address when precise (Precise event)event=0xd3,period=100007,umask=0x8Counts all demand & prefetch data reads that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0490offcore_response.all_pf_rfo.l3_hit.no_snoop_neededOFFCORE_RESPONSE.ALL_RFO.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.pf_l2_data_rd.l3_hit.no_snoop_neededoffcore_response.pf_l2_rfo.l3_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800122Counts all demand data writes (RFOs) that miss the L3 and the data is returned from local dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000020Counts number of cache lines that are dropped and not written back to L3 as they are deemed to be less likely to be reused shortlywrite requests to memory controller. Unit: uncore_imc event=0x83,ch_mask=0x01,fc_mask=0x07,umask=0x04,ch_mask=0x1func_cha_llc_victims.total_sevent=0x37,umask=0x04RspIFwd Snoop Responses Received. Unit: uncore_cha unc_cha_snoop_resp.rsp_fwd_wbPCIe Completion Buffer Inserts of completions with data: Part 0-3. Unit: uncore_iio Counts every peer to peer write request of 4 bytes of data made by IIO Part1 to the MMIO space of an IIO target. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busCounts every write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part0 by a unit on the main die (generally a core) or by another IIO unit. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busPeer to peer write request of up to a 64 byte transaction is made to IIO Part1 by a different IIO unit. Unit: uncore_iio unc_iio_txn_req_of_cpu.mem_read.part1Peer to peer write request of up to a 64 byte transaction is made by IIO Part2 to an IIO target. Unit: uncore_iio Counts when the M2M (Mesh to Memory) issues writes to the iMC (Memory Controller)Data Response packets that go direct to Intel UPI. Unit: uncore_upi Counts incoming FLITs (FLow control unITs) which bypassed the TxL(transmit) FLIT buffer and pass directly out the UPI Link. Generally, when data is transmitted across the Intel Ultra Path Interconnect (UPI), it will bypass the TxQ and pass directly to the link.  However, the TxQ will be used in L0p (Low Power) mode and (Link Layer Retry) LLR  mode, increasing latency to transfer out to the linkevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200491ocr.all_data_rd.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040491ocr.all_pf_data_rd.l3_hit_f.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100490ocr.all_pf_rfo.l3_hit_e.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C07F7ocr.all_reads.l3_hit_f.hit_other_core_no_fwdOCR.ALL_READS.L3_HIT_M.NO_SNOOP_NEEDED  OCR.ALL_READS.L3_HIT_M.NO_SNOOP_NEEDEDOCR.ALL_READS.L3_HIT_S.HIT_OTHER_CORE_NO_FWD  OCR.ALL_READS.L3_HIT_S.HIT_OTHER_CORE_NO_FWDOCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_FWDOCR.ALL_RFO.L3_HIT_E.HIT_OTHER_CORE_FWD  OCR.ALL_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080122ocr.all_rfo.l3_hit_f.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200122ocr.all_rfo.l3_hit_m.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040122ocr.demand_code_rd.l3_hit_f.any_snoopCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_M.ANY_SNOOPocr.demand_code_rd.l3_hit_m.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100001Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDocr.other.l3_hit_e.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400208000ocr.other.l3_hit_s.hit_other_core_fwdocr.pf_l1d_and_sw.l3_hit_e.snoop_missocr.pf_l1d_and_sw.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040400Counts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_S.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100010Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_E.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100020ocr.pf_l3_data_rd.l3_hit.snoop_missCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_E.HITM_OTHER_CORECounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDocr.pf_l3_rfo.l3_hit.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.all_data_rd.l3_hit_f.hitm_other_coreoffcore_response.all_data_rd.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_F.SNOOP_NONEoffcore_response.all_pf_rfo.l3_hit_m.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020120offcore_response.all_rfo.l3_hit_f.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_S.HITM_OTHER_COREoffcore_response.all_rfo.l3_hit_s.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400122This event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.SNOOP_HIT_WITH_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.SUPPLIER_NONE.HITM_OTHER_COREoffcore_response.demand_data_rd.l3_hit_f.snoop_missThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_S.SNOOP_MISSThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_F.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_M.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l1d_and_sw.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_S.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020400This event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.SUPPLIER_NONE.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l2_data_rd.l3_hit_e.snoop_missThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_E.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_F.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_hit_e.no_snoop_neededoffcore_response.pf_l2_rfo.l3_hit_f.hitm_other_coreoffcore_response.pf_l2_rfo.l3_hit_f.no_snoop_neededoffcore_response.pf_l2_rfo.l3_hit_s.no_snoop_neededoffcore_response.pf_l3_data_rd.l3_hit_m.any_snoopoffcore_response.pf_l3_data_rd.l3_hit_s.snoop_missThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_E.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_rfo.l3_hit_m.any_snoopoffcore_response.pf_l3_rfo.pmm_hit_local_pmm.snoop_not_needed100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) * ( ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , OFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DATA_RD ) / CPU_CLK_UNHALTED.THREAD - (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD)) / #( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) - ( ( ( 1 - ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) / ( ( 19 * (MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + 10 * ( (MEM_LOAD_L3_MISS_RETIRED.LOCAL_DRAM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_FWD * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + (MEM_LOAD_L3_MISS_RETIRED.REMOTE_HITM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) ) ) + ( 25 * ( MEM_LOAD_RETIRED.LOCAL_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) + 33 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) ) ) ) ) ) ) * (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) if ( 1000000 * ( MEM_LOAD_L3_MISS_RETIRED.REMOTE_PMM + MEM_LOAD_RETIRED.LOCAL_PMM ) > MEM_LOAD_RETIRED.L1_MISS ) else 0 ) ) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (( (20.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) - (3.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) ) * MEM_LOAD_RETIRED.L3_HIT * (1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) / 2) / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) + ( (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD)) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) )Average 3DXP Memory Bandwidth Use for reads [GB / sec]Number of SSE/AVX computational 512-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP14 RSQRT14 SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element.  The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsfp_arith_inst_retired2.128bit_packed_bf16fp_arith_inst_retired2.512bit_packed_bf16Intel AVX-512 computational 256-bit packed BFloat16 instructions retiredevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000491OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSocr.all_pf_data_rd.l3_miss_local_dram.any_snoopOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORE  OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREocr.all_pf_rfo.l3_miss_remote_hop1_dram.any_snoopOCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C0007F7OCR.ALL_READS.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORE  OCR.ALL_READS.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x6040007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000122ocr.all_rfo.l3_miss_remote_hop1_dram.any_snoopocr.demand_data_rd.l3_miss.no_snoop_neededCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.SNOOP_NONECounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDocr.demand_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.demand_rfo.l3_miss_remote_hop1_dram.no_snoop_neededocr.other.l3_miss.any_snoopCounts any other requests  OCR.OTHER.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDocr.pf_l1d_and_sw.l3_miss.hitm_other_coreCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.HITM_OTHER_CORE OCR.PF_L1D_AND_SW.L3_MISS.HITM_OTHER_CORECounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.SNOOP_MISSocr.pf_l1d_and_sw.l3_miss_local_dram.hit_other_core_no_fwdCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.pf_l2_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreocr.pf_l2_rfo.l3_miss.remote_hit_forwardCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.ANY_SNOOP OCR.PF_L3_DATA_RD.L3_MISS.ANY_SNOOPCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.PF_L3_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000080Counts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.NO_SNOOP_NEEDED OCR.PF_L3_RFO.L3_MISS.NO_SNOOP_NEEDEDoffcore_response.all_data_rd.l3_miss.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.SNOOP_MISSoffcore_response.all_reads.l3_miss_remote_hop1_dram.snoop_noneoffcore_response.all_rfo.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.HIT_OTHER_CORE_FWDoffcore_response.pf_l1d_and_sw.l3_miss_local_dram.hitm_other_coreoffcore_response.pf_l1d_and_sw.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDOCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDED OCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.all_reads.pmm_hit_local_pmm.any_snoopCounts all demand code reads  OCR.DEMAND_CODE_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDCounts demand data reads  OCR.DEMAND_DATA_RD.SUPPLIER_NONE.ANY_SNOOPocr.demand_rfo.pmm_hit_local_pmm.snoop_not_neededCounts any other requests OCR.OTHER.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.other.supplier_none.any_snoopocr.other.supplier_none.hit_other_core_no_fwdocr.other.supplier_none.snoop_missCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPunc_m_pmm_read_latencyunc_m_pmm_cmd1.rdNumber of phases a demand request has waited due to L1D Fill Buffer (FB) unavailablabilityCounts demand data reads that hit a cacheline in the L3 where a snoop was sentevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1E003C0020Counts miscellaneous requests, such as I/O and un-cacheable accesses that hit a cacheline in the L3 where a snoop was sentevent=0x32,period=100003,umask=0x2event=0xc7,period=100003,umask=0x1Slots_Utilizationevent=0xc8,period=100003,umask=0x4ocr.other.l3_missCounts miscellaneous requests, such as I/O and un-cacheable accesses that have any type of responseCounts return instructions retired (Precise event)event=0xd,period=500009,umask=0x1Increments when an entry is added to the Last Branch Record (LBR) array (or removed from the array in case of RETURNs in call stack mode). The event requires LBR to be enabled properlyevent=0xa4,period=10000003,umask=0x1Number of uops executed on port 7 and 8Counts cycles when at least 2 micro-ops are executed from any thread on physical coreBranches;PGOMemoryTLBocr.reads_to_core.l3_hit.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00400ocr.demand_data_rd.remote_pmmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x703000477Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by PMM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeevent=0xd3,umask=0x02unc_m_hclockticksevent=0xea,umask=0x04unc_cha_tor_inserts.iaTOR Occupancy : All requests from IO Devices. Unit: uncore_cha unc_cha_tor_occupancy.io_hitunc_cha_tor_inserts.io_miss_itomevent=0x35,umask=0xC897FE01event=0x35,umask=0xC887FF01TOR Inserts : DRds issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0xC8F3FF04event=0x83,ch_mask=0x02,fc_mask=0x07,umask=0x80event=0xc0,ch_mask=0x20,fc_mask=0x07,umask=0x01Multi-socket cacheline Directory Lookups : Found in any state. Unit: uncore_m2m event=0x2d,umask=0x08unc_u_clockticksTOR Inserts : ItoMs issued by IO Devices to remotely HOMed memory. Unit: uncore_cha Counts the number of load uops retired that hit in the L2 cache  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3001F803C0000Counts demand instruction fetches and L1 instruction cache prefetches that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by the L3 cache where no snoop was needed to satisfy the requestocr.demand_data_and_l1pf_rd.l3_hit.snoop_missCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedocr.hwpf_l2_data_rd.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1001F803C0000Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by the L3 cache where no snoop was needed to satisfy the requestCounts the number of BACLEARS due to a return branchocr.hwpf_l2_code_rd.l3_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000470This event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.DRAMocr.hwpf_l2_code_rd.dramevent=0xe8,period=200003Counts the number of core cycles while the core is not in a halt state.  The core enters the halt state when it is running the HLT instruction. The core frequency may change from time to time. For this reason this event may have a changing ratio with regards to time. This event uses a programmable general purpose performance countertopdown_be_bound.allevent=0x74,period=1000003,umask=0x4Counts the number of issue slots every cycle that were not delivered by the frontend due to BACLEARSCounts the number of issue slots every cycle that were not delivered by the frontend due to BACLEARS, which occurs when the Branch Target Buffer (BTB) prediction or lack thereof, was corrected by a later branch predictor in the frontend. Includes BACLEARS due to all branch types including conditional and unconditional jumps, returns, and indirect branchesCounts the number of issue slots every cycle that were not delivered by the frontend due to BTCLEARSevent=0x71,period=1000003,umask=0x8topdown_fe_bound.frontend_latencyCounts the number of integer divide uops retired (Precise event)TOR Inserts : CRd_Prefs issued by iA Cores that Missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsData requested of the CPU : Card reading from DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Counts the number of page walks outstanding in the page miss handler (PMH) for loads every cycle.  A page walk is outstanding from start till PMH becomes idle again (ready to serve next walk). Includes EPT-walk intervalsCounts the number of page walks completed due to stores whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 4K pages.  Includes page walks that page faultevent=0x85,period=200003,umask=0x8Counts the number of memory retired ops that missed in the second level TLB  Supports address when precise (Precise event)Counts the number of store ops retired that miss in the second level TLB  Supports address when precise (Precise event)Counts the number of load ops retired that hit in DRAM  Supports address when preciseAverage number of Uops issued by front-end when it issued something. Unit: cpu_core Average data fill bandwidth to the L2 cache [GB / sec]. Unit: cpu_core Counts the number of issue slots  that were not consumed by the backend due to frontend stalls. Unit: cpu_atom Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwarded. Unit: cpu_atom mem_load_completed.l1_miss_anyRetired load instructions whose data sources were hits in L3 without snoops required  Supports address when precise (Precise event). Unit: cpu_core Number of completed demand load requests that missed the L1, but hit the FB(fill buffer), because a preceding miss to the same cacheline initiated the line to be brought into L1, but data is not yet ready in L1  Supports address when precise (Precise event). Unit: cpu_core event=0xe5,period=1000003,umask=0x3Retired memory uops for any access. Unit: cpu_core fp_arith_dispatched.port_1Retired Instructions who experienced Instruction L2 Cache true miss (Precise event). Unit: cpu_core event=0xc6,period=100007,umask=0x1,frontend=0x610006Uops not delivered by IDQ when backend of the machine is not stalled. Unit: cpu_core Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were not supplied by the L3 cache. Unit: cpu_atom mem_trans_retired.store_sampleRetired instructions with at least 1 store uop. This PEBS event is the trigger for stores sampled by the PEBS Store Facility  Supports address when precise (Must be precise). Unit: cpu_core event=0xc1,period=100003,umask=0x4event=0x2a,period=100003,umask=0x1,offcore_rsp=0x10800Counts the number of unhalted core clock cycles. Unit: cpu_atom machine_clears.mrn_nukeThis event is deprecated. Refer to new event ARITH.DIV_ACTIVE. Unit: cpu_core arith.div_activeTaken conditional branch instructions retired (Precise event). Unit: cpu_core event=0xad,period=500009,umask=0x80False dependencies in MOB due to partial compare on address. Unit: cpu_core event=0x3,period=100003,umask=0x88unc_m_act_count_wrLoads that miss the DTLB and hit the STLB. Unit: cpu_core event=0x12,period=100003,umask=0x4event=0x2a,period=100003,umask=0x1,offcore_rsp=0x1830004477Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by a cache on a remote socket where a snoop hit a modified line in another core's caches which forwarded the dataThis event is deprecated. Refer to new event OFFCORE_REQUESTS_OUTSTANDING.DATA_RDevent=0x3,umask=0x0000000008DRAM RD_CAS and WR_CAS Commands. : Pseudo Channel 0. Unit: uncore_imc event=0x83,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000002event=0x83,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000001Read requests from a remote socket. Unit: uncore_cha event=0x35,umask=0x00c8177e01TOR Inserts for DRd Pref misses from local IA targeting remote memory. Unit: uncore_cha unc_upi_txl_basic_hdr_match.ncsevent=0x31,umask=0x0000000001event=0x31,umask=0x0000000002RxQ Flit Buffer Bypassed : Slot 2. Unit: uncore_upi event=0x84,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000004event=0xc0,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000001event=0xc1,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000004event=0x5d,umask=0x000000000aM3UPI CMS Clockticks. Unit: uncore_m3upi event=0x35,umask=0x00c001ff01event=0x35,umask=0x00ccc7fd01event=0x35,umask=0x00c001ff04unc_cha_tor_inserts.loc_allunc_cha_tor_inserts.rem_snpsTOR Inserts; ItoM hits from local IO. Unit: uncore_cha TOR Inserts; CRd from local IA. Unit: uncore_cha unc_cha_tor_occupancy.allunc_cha_tor_occupancy.irq_iaunc_cha_tor_occupancy.ia_miss_drd_opt_prefevent=0x36,umask=0x00c817ff01All LLC lines in E state that are victimized on a fill. Unit: uncore_cha event=0x35,umask=0x00c837fd01event=0x35,umask=0x00c8f3fd04unc_cha_tor_inserts.ia_llcprefcodeTOR Inserts; LLCPrefCode from local IA. Unit: uncore_cha TOR Inserts : CRd issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha unc_cha_tor_occupancy.ia_specitomevent=0x36,umask=0x00c877de01unc_cha_tor_occupancy.ia_miss_wcilf_pmmevent=0x36,umask=0x00c8668601event=0x64,umask=0x80event=0x6d,umask=0x01The number of far control transfers retired including far call/jump/return, IRET, SYSCALL and SYSRET, plus exceptions and interrupts. Far control transfers are not subject to branch predictionex_ret_near_retRemote Link Controller Outbound Packet Types: Data (32B): Remote Link Controller 2event=0x187,umask=0x38Total number uOps assigned to all fpu pipesfpu_pipe_assignment.total2The number of operations (uOps) dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number uOps assigned to pipe 2Add/subtract Opsevent=0x3,umask=0x04ls_misal_accessesde_dis_dispatch_token_stalls0.agsq_token_stallremote_outbound_data_controller_0 + remote_outbound_data_controller_1 + remote_outbound_data_controller_2 + remote_outbound_data_controller_3The number of instruction fetches that miss in both the L1 and L2 TLBs. Instruction fetches to a 4KB pageStore-to-load conflicts: A load was unable to complete due to a non-forwardable conflict with an older store. Most commonly, a load's address range partially but not completely overlaps with an uncompleted older store. Software can avoid this problem by using same-size and same-alignment loads and stores when accessing the same data. Vector/SIMD code is particularly susceptible to this problem; software should construct wide vector stores by manipulating vector elements in registers using shuffle/blend/swap instructions prior to storing to memory, instead of using narrow element-by-element storesls_locks.non_spec_lockls_pref_instr_disp.prefetchSoftware Prefetch Data Cache Fills by Data Source. DRAM or IO from this thread's die.  From DRAM (home node local)event=0x5a,umask=0x01event=0xaa,umask=0x02bp_l1_tlb_miss_l2_tlb_miss.coalesced_4kop_cache_hit_miss.all_op_cache_accessesOp Cache Miss. Counts Op Cache micro-tag hit/miss eventsevent=0x44,umask=0x10event=0x5fevent=0x44,umask=0x48event=0x44,umask=0x50l1_data_cache_fills_allls_any_fills_from_sys.far_cachels_any_fills_from_sys.all_dram_ioevent=0x52,umask=0x03l2_request_g1.all_dcl2_pf_hit_l2.l2_up_downevent=0x71,umask=0x10event=0x71,umask=0x40Instruction cache lines (64 bytes) fulfilled from system memory or another cacheAverage sampled latency when data is sourced from DRAM in the same NUMA node. Unit: amd_l3 l3_xi_sampled_latency_requests.ext_nearCycles with no retire for any reasonevent=0x25f,umask=0x7feWrite data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 1local_socket_inf1_inbound_data_beats_ccm3event=0x5de,umask=0x7ffData beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 6event=0x41e,umask=0xbferemote_socket_inf1_outbound_data_beats_ccm4local_socket_outbound_data_beats_link2Data beats (64 bytes) for local socket outbound data from inter-socket xGMI link 3Retired SSE and AVX bottom-executing ops. Bottom-executing ops wait for all older ops to retire before executingRetired SSE and AVX serializing ops of all typesfp_ops_retired_by_width.x87_uops_retiredfp_ops_retired_by_width.pack_128_uops_retiredRetired scalar floating-point square root opsevent=0xa,umask=0x40sse_avx_ops_retired.sse_avx_subRetired SSE, AVX and MMX integer ops of all typesevent=0xd,umask=0x04event=0xd,umask=0x09event=0xd,umask=0x0apacked_int_op_type.int128_logicalevent=0xd,umask=0x0dL1 DTLB misses with L2 DTLB hits for 1G pagesde_src_op_disp.loop_bufferNumber of ops dispatched to the integer execution unitNumber of cycles dispatch is stalled for store queue tokensIn each cycle counts ops unable to dispatch because of back-end stallsretiring_fastpathAll L2 cache hitsAverage L3 read miss latency (in core clocks)FR_RETIRED_RESYNCSL1_DCACHE_ACCESSEVENT_28HEVENT_73HEVENT_90HEVENT_94HEVENT_C0HEVENT_C4HEVENT_EBHINSTR_MICRO_TLB_MISS_STALLEVENT_111HEVENT_14AHEVENT_15DHEVENT_1A4HEVENT_1DFHEVENT_233HEVENT_2ABHEVENT_2B0HEVENT_2D9HEVENT_2F0HEVENT_328HEVENT_354HEVENT_394HEVENT_3ADHEVENT_3B6HEVENT_3DBHL2D_CACHE_STL3D_CACHE_ALLOCATEsbsx_wr_reqsbsx_wdb_occ_cnt_ovflrnd_s1_rdata_beatsrni_txreq_flits_retriedcxla_rx_tlp_link2cxla_rx_cxs_link0clkdiv2_training_requestclkdiv2_bk_open_trackerCYCLEINSTR_COMPLETEDL1_DATA_SNOOPSCACHE_INHIBITED_STORESSS_SM_INSTR_COMPLETEDCYCLES_TWO_INSTR_COMPLETEDSTORE_MERGE_GATHERDSSALL_INSTR_COMPLETEDL2_STORE_HITSL3_LOAD_HITSBORDQ_FULLDATA_L1_CACHE_CASTOUTSEXT_INPU_INTR_LATENCY_CYCLESPMC1_OVERFLOWL2_CACHE_ALLOCATIONSDVT2_DETECTEDstoredimm-turnaroundAMD_K8truels_not_halted_cycl3_comb_clstr_state.request_missBRANCH_INSTRUCTION_RETIRED{"type": "pmcdetach"GenuineIntel-6-47GenuineIntel-6-2AGenuineIntel-6-8FFrontend_BoundThis category represents fraction of slots where no uops are being delivered due to a lack of required resources for accepting new uops in the Backend. Backend is the portion of the processor core where the out-of-order scheduler dispatches ready uops into their respective execution units; and once completed these uops get retired according to program order. For example; stalls due to data-cache misses or stalls due to the divider unit being overloaded are both categorized under Backend Bound. Backend Bound is further divided into two main categories: Memory Bound and Core BoundInstructions per FP Arithmetic instruction (lower number means higher occurrence rate)INST_RETIRED.ANYSummary;TmaL1L1D_PEND_MISS.PENDING / L1D_PEND_MISS.PENDING_CYCLESL3 cache true misses per kilo instruction for retired demand loadsHPC;SummaryFraction of cycles where both hardware Logical Processors were activeC2_Pkg_ResidencyL1D data line replacementsl2_lines_in.eDemand requests to L2 cacheevent=0x24,period=200003,umask=0x30This event counts the number of requests from the L2 hardware prefetchers that miss L2 cacheRetired load uops which data sources were HitM responses from shared L3. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM100 (Precise event)Retired load uop whose Data Source was: Remote cache HITM (Precise Event)  Supports address when precise.  Spec update: BDE70This is a precise version (that is, uses PEBS) of the event that counts line-splitted load uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K)  Supports address when precise (Precise event)event=0xf4,period=100003,umask=0x10fp_arith_inst_retired.scalar_doublebaclears.anyhle_retired.commitmem_trans_retired.load_latency_gt_128Number of times the TSX watchdog signaled an RTM abortevent=0x5d,period=2000003,umask=0x1tx_exec.misc3Speculative and retired macro-unconditional branches excluding calls and indirectsNot taken macro-conditional branchesevent=0x3c,period=2000003,umask=0x2event=0xa3,cmask=4,period=2000003,umask=0x4Counts number of cycles nothing is executed on any execution portExecution stalls while memory subsystem has an outstanding loadevent=0,period=2000003,umask=0x1This is a precise version (that is, uses PEBS) of the event that counts instructions retired  Spec update: BDM11, BDM55 (Must be precise)This event counts all not software-prefetch load dispatches that hit the fill buffer (FB) allocated for the hardware prefetchevent=0x4c,period=100003,umask=0x1uops_dispatched_port.port_4uops_dispatched_port.port_7event=0xb1,inv=1,period=2000003,umask=0x2uops_executed.cycles_ge_3_uops_execuops_executed_port.port_0uops_issued.single_mulevent=0xc2,period=2000003,umask=0x1event=0unc_c_llc_lookup.anyllc_misses.mmio_writellc_references.code_llc_prefetchevent=0x36,umask=0x3,filter_opc=0x182read requests to home agent. Unit: uncore_ha llc_misses.mem_readllc_misses.mem_writeCycles all ranks are in critical thermal throttle. Unit: uncore_imc This is an occupancy event that tracks the number of cores that are in C0.  It can be used by itself to get the average number of cores in C0, with threshholding to generate histograms, or with other PCU events and occupancy triggering to capture other details. Unit: uncore_pcu event=0x80,occ_sel=2unc_p_freq_max_limit_thermal_cyclesevent=0x49,period=100003,umask=0xeStore misses in all DTLB levels that cause completed page walks (2M/4M)  Spec update: BDM69event=0xbc,period=2000003,umask=0x21event=0xbd,period=100007,umask=0x20Retired store uops that split across a cacheline boundary  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0240Counts all prefetch data readsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0120offcore_response.all_pf_rfo.l3_hit.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020122offcore_response.corewb.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0008offcore_response.demand_data_rd.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000028000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020040offcore_response.pf_l2_code_rd.supplier_none.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020040Counts all prefetch (that bring data to L2) RFOs have any response typeoffcore_response.all_pf_code_rd.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0240offcore_response.all_pf_code_rd.l3_miss_local_dram.snoop_missoffcore_response.all_pf_code_rd.l3_miss_local_dram.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000090offcore_response.all_pf_rfo.l3_miss_local_dram.snoop_hit_no_fwdoffcore_response.demand_code_rd.l3_miss.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000002offcore_response.other.l3_hit.snoop_non_dramoffcore_response.other.l3_miss_local_dram.snoop_missoffcore_response.pf_l2_data_rd.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000020L3 Lookup read request that access cache and found line in E or S-stateRetired load uop whose Data Source was: forwarded from remote cache  Supports address when precise.  Spec update: BDE70 (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0604000122L1 Data Cacheable reads and writesL1 Data reads and writesevent=0x2c,period=200000,umask=0x42event=0x29,period=200000,umask=0x72l2_lines_in.self.demandevent=0x30,period=200000,umask=0x44l2_reject_busq.self.demand.s_statesimd_inst_retired.vectorSaturated arithmetic instructions retiredSIMD saturated arithmetic micro-ops executedevent=0xb3,period=2000000,umask=0x1cycles_icache_mem_stalled.icache_mem_stalledevent=0x80,period=200000,umask=0x1event=0x5,period=200000,umask=0x89prefetch.software_prefetchOutstanding cacheable data read bus requests durationbus_trans_p.selfevent=0x66,period=200000,umask=0x40Memory cluster signals to block micro-op dispatch for any reasonext_snoop.all_agents.cleanAll macro unconditional branch instructions, excluding calls and indirectsMicro-op reissues on a store-load collision (At Retirement)itlb.missesCounts load uops retired where the cache line containing the data was in the modified state of another core or modules cache (HITM).  More specifically, this means that when the load address was checked by other caching agents (typically another processor) in the system, one of those caching agents indicated that they had a dirty copy of the data.  Loads that obtain a HITM response incur greater latency than most is typical for a load.  In addition, since HITM indicates that some other processor had this data in its cache, it implies that the data was shared between processors, or potentially was a lock or semaphore value.  This event is useful for locating sharing, false sharing, and contended locks  Supports address when precise (Must be precise)Counts load uops retired that hit in the L2 cache  Supports address when precise (Must be precise)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x36000032b7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000022Counts demand reads for ownership (RFO) requests generated by a write to full data cache line that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredoffcore_response.pf_l1_data_rd.l2_miss.hit_other_core_no_fwdCounts reads for ownership (RFO) requests generated by L2 prefetcher that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.pf_l2_rfo.l2_miss.anyCounts reads for ownership (RFO) requests generated by L2 prefetcher that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000044800Counts data cache lines requests by software prefetch instructions that hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200001000event=0xcb,period=200003,umask=0x2br_inst_retired.all_taken_branchesbr_misp_retired.taken_jccevent=0xe,period=200003Integer divide uops retired. (Precise Event Capable) (Must be precise)itlb.missLoad uops retired that missed the DTLB (Precise event capable)  Supports address when precise (Must be precise)event=0xd0,period=200003,umask=0x12Duration of page-walks in cyclesevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000013091Counts data reads (demand & prefetch) outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts requests to the uncore subsystem have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts requests to the uncore subsystem true miss for the L2 cache with a snoop miss in the other processor moduleCounts reads for ownership (RFO) requests (demand & prefetch) outstanding, per cycle, from the time of the L2 miss to when any response is receivedoffcore_response.bus_locks.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000008Counts demand cacheable data reads of full cache lines have any transaction responses from the uncore subsystemCounts demand cacheable data reads of full cache lines true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand reads for ownership (RFO) requests generated by a write to full data cache line outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts data cache lines requests by software prefetch instructions true miss for the L2 cache with a snoop miss in the other processor moduleThis event counts when new data lines are brought into the L1 Data cache, which cause other lines to be evicted from the cacheThis event counts each cache miss condition for references to the last level cacheoffcore_response.demand_data_rd.l3_hit.hitm_other_core( UOPS_EXECUTED.CORE / 2 / (( cpu@UOPS_EXECUTED.CORE\,cmask\=1@ / 2 ) if #SMT_on else cpu@UOPS_EXECUTED.CORE\,cmask\=1@) ) if #SMT_on else UOPS_EXECUTED.CORE / (( cpu@UOPS_EXECUTED.CORE\,cmask\=1@ / 2 ) if #SMT_on else cpu@UOPS_EXECUTED.CORE\,cmask\=1@)Number of times an HLE execution aborted due to any reasons (multiple categories may count as one) (Precise event)offcore_response.all_reads.l3_miss.local_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100400002This event counts cycles during which no instructions were executed in the execution stage of the pipelineAn external snoop hits a modified line in some processor core. Unit: uncore_cbox Each cycle count number of valid entries in Coherency Tracker queue from allocation till deallocation. Aperture requests (snoops) appear as NC decoded internally and become coherent (snoop L3, access memory)ITLB misses that hit STLB (4K)Counts the number of Extended Page Table walks from the DTLB that hit in the L2Counts all data/code/rfo reads (demand & prefetch) miss the L3 and the data is returned from local dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x083FC007F7Counts any demand and L1 HW prefetch data load requests to L2mem_load_uops_retired.llc_hitmem_load_uops_retired.llc_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x000107F7event=0x11,period=2000003,umask=0x1DSB Fill encountered > 3 DSB linesLoads with latency value being above 256 (Must be precise)Loads with latency value being above 8 (Must be precise)uops_dispatched_port.port_0_coreCycles which a Uop is dispatched on port 4LLC lookup request that access cache and found line in M-stateoffcore_response.all_pf_data_rd.llc_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0090Counts all data/code/rfo reads (demand & prefetch) that hit in the LLCoffcore_response.demand_data_rd.llc_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc00244Counts demand data reads that miss in the LLCLLC misses for PCIe non-snoop writes (full line). Derived from unc_c_tor_inserts.miss_opcode.pcie_write. Unit: uncore_cbox PCIe non-snoop writes (partial). Derived from unc_c_tor_inserts.opcode.pcie_partial_write. Unit: uncore_cbox Read requests to memory controller. Derived from unc_m_cas_count.rd. Unit: uncore_imc event=0xc,edge=1unc_p_freq_ge_2000mhz_transitionsevent=0xc5,period=400009,umask=0x10This event counts the number of instructions retired from execution. For instructions that consist of multiple micro-ops, this event counts the retirement of the last micro-op of the instruction. Counting continues during hardware interrupts, traps, and inside interrupt handlersCounts the cycles of stall due to lack of load buffers(UNC_C_TOR_OCCUPANCY.MISS_ALL / UNC_C_CLOCKTICKS) * 100.offcore_response.any_data_rd.l2_hit_far_tileoffcore_response.any_data_rd.l2_hit_far_tile_mCounts Demand cacheable data and L1 prefetch data read requests  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000022offcore_response.bus_locks.l2_hit_this_tile_eoffcore_response.demand_rfo.l2_hit_near_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000014000Counts L1 data HW prefetches that accounts for responses which hit its own tile's L2 with data in F stateCounts L1 data HW prefetches that accounts for responses which hit its own tile's L2 with data in S stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010040event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400020offcore_response.pf_software.outstandingoffcore_response.uc_code_reads.l2_hit_this_tile_moffcore_response.any_code_rd.ddr_farCounts Demand cacheable data and L1 prefetch data read requests  that accounts for data responses from MCDRAM LocalCounts any Read request  that accounts for responses from DDR (local and far)offcore_response.any_read.ddr_nearoffcore_response.any_read.mcdram_faroffcore_response.partial_reads.non_dramoffcore_response.partial_writes.ddr_nearCounts all the retired locked loads. It does not include stores because we would double count if we count storesCounts the total number of core cycles for all the page walks. The cycles for page walks started in speculative path will also be includedL1 data cache read in E statel1d_cache_ld.i_stateevent=0x42,period=2000000,umask=0x8event=0x28,period=100000,umask=0x2l2_rqsts.prefetch_missevent=0xf0,period=200000,umask=0x2Memory instructions retired above 512 clocks (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x811All offcore code readsOffcore code reads satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x80FFoffcore_response.any_request.local_cache_dramOffcore request = all data, response = any cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1833Offcore data reads, RFO's and prefetches that HIT in a remote cacheOffcore demand data requests satisfied by the LLC and HIT in a sibling coreoffcore_response.demand_data_rd.any_locationOffcore demand data reads that HIT in a remote cacheOffcore demand code reads that HITM in a remote cacheoffcore_response.demand_rfo.any_cache_dramOffcore demand RFO requests satisfied by a remote cacheoffcore_response.other.io_csr_mmiooffcore_response.pf_data.llc_hit_other_core_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1030offcore_response.pf_data.remote_cache_hitmoffcore_response.pf_data_rd.any_cache_dramoffcore_response.pf_data_rd.remote_cache_hitAll offcore prefetch code readsoffcore_response.prefetch.remote_cache_dramevent=0xf4,period=2000000,umask=0x10store_blocks.at_retfp_mmx_trans.to_mmxevent=0x12,period=200000,umask=0x2event=0xfd,period=200000,umask=0x8Offcore data reads satisfied by any DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6044event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4008event=0xb7,period=100000,umask=0x1,offcore_rsp=0xF877offcore_response.data_in.any_llc_missOffcore request = all data, response = any LLC missOffcore demand data requests that missed the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4003offcore_response.demand_rfo.any_llc_missOffcore demand RFO requests that missed the LLCoffcore_response.pf_data_rd.any_llc_missoffcore_response.pf_ifetch.any_dramoffcore_response.pf_ifetch.remote_dramOffcore prefetch RFO requests satisfied by any DRAMoffcore_response.pf_rfo.local_dramoffcore_response.prefetch.any_dramThread responded HIT to snoopbr_inst_exec.anyIndirect return branches executedbr_misp_exec.indirect_non_callbr_misp_exec.non_callsbr_misp_exec.takenlsd.activeFlag stall cyclesUops executed on port 2 (core count)uops_executed.port3_coreevent=0xe,any=1,cmask=1,inv=1,period=2000000,umask=0x1event=0xe,any=1,cmask=1,period=2000000,umask=0x1uops_retired.macro_fusedl2_lines_out.useless_prefCounts the number of offcore outstanding RFO (store) transactions in the super queue (SQ) every cycle. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation). See corresponding Umask under OFFCORE_REQUESTSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100001offcore_response.demand_rfo.l3_hit_e.any_snoopoffcore_response.demand_rfo.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1001C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0108000offcore_response.other.l4_hit_local_l4.any_snoopoffcore_response.other.l4_hit_local_l4.spl_hitCounts retired Instructions that experienced DSB (Decode stream buffer i.e. the decoded instruction-cache) miss (Precise event)Retired instructions that are fetched after an interval where the front-end had at least 2 bubble-slots for a period of 2 cycles which was not interrupted by a back-end stall (Precise event)event=0x83,period=200003,umask=0x2Counts the number of uops delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Counting includes uops that may 'bypass' the IDQhle_retired.aborted_eventsoffcore_response.demand_data_rd.l3_miss.snoop_hitmoffcore_response.demand_data_rd.l4_hit_local_l4.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000002Core crystal clock cycles when the thread is unhaltedCounts cycles during which no uops were executed on all ports and Reservation Station (RS) was not emptyDemand load dispatches that hit L1D fill buffer (FB) allocated for software prefetch100 * ( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * INT_MISC.CLEAR_RESTEER_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) )Mem;MemoryBW;OffcoreTotal pipeline cost of instruction fetch related bottlenecks by large code footprint programs (i-side cache; TLB and BTB misses)Big_CodeBig_Code_SMT ( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) * ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * INT_MISC.CLEAR_RESTEER_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) ) * (4 * CPU_CLK_UNHALTED.THREAD) / BR_MISP_RETIRED.ALL_BRANCHESevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680003091Counts any request that miss L2 with a snoop miss responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000004Counts data cacheline reads generated by L2 prefetchers that miss L2rehabq.ld_block_st_forwardLoads blocked due to store forward restriction (Precise event)Stalls due to Memory orderingJCC counts the number of conditional branch (JCC) instructions retired. Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)Counts the number of mispredicted taken JCC branch instructions retired (Precise event)This event counts the number of load ops retired that had DTLB miss (Precise event)event=0xd4,period=100007,umask=0x2offcore_response.all_pf_code_rd.llc_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10040Counts prefetch (that bring data to LLC only) code reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts prefetch (that bring data to LLC only) RFOs that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts all prefetch (that bring data to LLC only) data reads that miss the LLC  and the data returned from dramActually retired uops. (Precise Event - PEBS) (Precise event)offcore_requests.demand.rfoevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff11REQUEST = ANY RFO and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HIToffcore_response.data_in.all_local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x750offcore_response.pf_rfo.local_dram_and_remote_cache_hitREQUEST = PREFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = ANY_REQUEST and RESPONSE = ANY_DRAM AND REMOTE_FWDevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf822REQUEST = PREFETCH and RESPONSE = OTHER_LOCAL_DRAMevent=0xb3,cmask=1,period=2000000,umask=0x2itlb_misses.walk_cyclesOffcore data reads, RFOs, and prefetches satisfied by the remote DRAMCounts all demand & prefetch data reads that hit in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0122Counts all prefetch (that bring data to L2) RFOs that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.all_pf_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts all demand & prefetch RFOs that miss the L3 and the data is returned from remote dramCounts demand data reads that miss the L3 and the data is returned from local dramCounts prefetch (that bring data to L2) data reads that miss the L3 and the data is returned from remote dramoffcore_response.pf_l3_data_rd.l3_miss.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00100Core cycles where the core was running in a manner where Turbo may be clipped to the Non-AVX turbo scheduleCore cycles where the core was running with power-delivery for baseline license level 0.  This includes non-AVX codes, SSE, AVX 128-bit, and low-current AVX 256-bit codesL2Evicts;Mem;Serverunc_m_act_count.wrevent=0x83,ch_mask=0x01,fc_mask=0x07,umask=0x04event=0x83,ch_mask=0x02,fc_mask=0x07,umask=0x04unc_iio_data_req_of_cpu.mem_write.part3event=0x50,umask=0x20unc_cha_rxc_irq1_reject.pa_matchevent=0x19,umask=0x80event=0x3d,umask=0x02PCIe Completion Buffer Inserts of completions with data: Part 3event=0xc0,ch_mask=0x08,fc_mask=0x07,umask=0x01Peer to peer read request for 4 bytes made by a different IIO unit to IIO Part2. Unit: uncore_iio event=0xc0,ch_mask=0x02,fc_mask=0x07,umask=0x02event=0xc1,ch_mask=0x01,fc_mask=0x07,umask=0x04Write request of up to a 64 byte transaction is made to IIO Part2 by the CPU. Unit: uncore_iio event=0x84,ch_mask=0x02,fc_mask=0x07,umask=0x08unc_i_coherent_ops.rfoevent=0x10,umask=0x8BL Ingress (from CMS) Allocations. Unit: uncore_m2m unc_upi_rxl_flits.all_dataevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0491OCR.ALL_DATA_RD.L3_HIT_M.ANY_SNOOP  OCR.ALL_DATA_RD.L3_HIT_M.ANY_SNOOPOCR.ALL_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWD  OCR.ALL_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0490OCR.ALL_PF_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000407F7ocr.all_rfo.l3_hit.hitm_other_coreOCR.ALL_RFO.L3_HIT.SNOOP_MISS OCR.ALL_RFO.L3_HIT.SNOOP_MISSocr.all_rfo.l3_hit_s.snoop_missCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_E.ANY_SNOOPCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_M.HITM_OTHER_COREocr.demand_code_rd.l3_hit_s.hit_other_core_no_fwdCounts demand data reads OCR.DEMAND_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.DEMAND_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDCounts demand data reads OCR.DEMAND_DATA_RD.L3_HIT.NO_SNOOP_NEEDED OCR.DEMAND_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDCounts demand data reads OCR.DEMAND_DATA_RD.L3_HIT.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040002ocr.demand_rfo.l3_hit_m.no_snoop_neededocr.demand_rfo.l3_hit_s.no_snoop_neededCounts any other requests OCR.OTHER.L3_HIT.HIT_OTHER_CORE_FWD OCR.OTHER.L3_HIT.HIT_OTHER_CORE_FWDocr.other.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100208000ocr.other.l3_hit_m.any_snoopCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_HIT.SNOOP_MISSCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_E.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100400Counts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_S.NO_SNOOP_NEEDEDCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDocr.pf_l2_data_rd.l3_hit_f.hit_other_core_no_fwdCounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_HIT.NO_SNOOP_NEEDED OCR.PF_L2_RFO.L3_HIT.NO_SNOOP_NEEDEDCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_S.ANY_SNOOPocr.pf_l3_data_rd.l3_hit.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080100Counts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400120offcore_response.all_reads.l3_hit_e.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_F.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.all_rfo.l3_hit_f.hitm_other_coreoffcore_response.all_rfo.l3_hit_s.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_rfo.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_F.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_E.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_F.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_E.ANY_SNOOPoffcore_response.pf_l1d_and_sw.l3_hit_e.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_M.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400010This event is deprecated. Refer to new event OCR.PF_L2_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDoffcore_response.pf_l3_data_rd.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_rfo.l3_hit_f.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000491ocr.all_pf_rfo.l3_miss.hit_other_core_fwdOCR.ALL_READS.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_READS.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDOCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDED  OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000004Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.DEMAND_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDocr.other.l3_miss_local_dram.hitm_other_coreCounts any other requests OCR.OTHER.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDocr.pf_l1d_and_sw.l3_miss.no_snoop_neededCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.REMOTE_HITMCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000020ocr.pf_l2_rfo.l3_miss_remote_hop1_dram.snoop_noneocr.pf_l3_data_rd.l3_miss.snoop_missCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDocr.pf_l3_rfo.l3_miss_local_dram.snoop_noneocr.pf_l3_rfo.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.all_data_rd.l3_miss.hitm_other_coreoffcore_response.all_data_rd.l3_miss.hit_other_core_fwdoffcore_response.all_data_rd.l3_miss_remote_hop1_dram.snoop_missoffcore_response.all_pf_data_rd.l3_miss_local_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEoffcore_response.all_pf_rfo.l3_miss.hit_other_core_no_fwdoffcore_response.all_pf_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdoffcore_response.all_reads.l3_miss_local_dram.snoop_missoffcore_response.all_reads.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDoffcore_response.demand_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDOCR.ALL_DATA_RD.SUPPLIER_NONE.HITM_OTHER_CORE  OCR.ALL_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREocr.all_pf_rfo.any_responseocr.all_reads.supplier_none.hit_other_core_no_fwdocr.all_rfo.supplier_none.snoop_missocr.demand_code_rd.pmm_hit_local_pmm.snoop_not_neededocr.demand_code_rd.supplier_none.hit_other_core_fwdocr.demand_data_rd.supplier_none.hit_other_core_no_fwdocr.demand_rfo.any_responseIntel Optane DC persistent memory bandwidth total (MB/sec). Derived from unc_m_pmm_rpq_inserts. Unit: uncore_imc Regular reads(RPQ) commands for Intel Optane DC persistent memory. Unit: uncore_imc All dirty line misses to Near Memory(DRAM cache) in Memory Mode. Unit: uncore_imc event=0x48,cmask=1,period=1000003,umask=0x1event=0x24,period=200003,umask=0xc8event=0xd0,period=1000003,umask=0x83Counts retired load instructions whose data sources were HitM responses from shared L3  Supports address when precise (Precise event)event=0xd1,period=1000003,umask=0x1ocr.demand_rfo.l3_hit.snoop_sentocr.hwpf_l2_rfo.l3_hit.snoop_not_neededevent=0x60,period=1000003,umask=0x4event=0xf4,period=100003,umask=0x4Number of SSE/AVX computational 128-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP14 RSQRT14 SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementocr.hwpf_l2_rfo.local_dramCounts streaming stores that DRAM supplied the requestevent=0xc5,period=50021,umask=0x1Cycle counts are evenly distributed between active threads in the Coreevent=0xa3,cmask=12,period=1000003,umask=0xcinst_retired.stall_cyclesCounts the number of Top-down Microarchitecture Analysis (TMA) method's  slots where no micro-operations were being issued from front-end to back-end of the machine due to lack of back-end resourcesevent=0,period=10000003,umask=0x4Number of uops executed on port 4 and 9Branches;InsTypeocr.demand_code_rd.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x808000004Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by a cache on a remote socket where a snoop hit a modified line in another core's caches which forwarded the dataCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that hit a modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeUOPS_RETIRED.SLOTS / INST_RETIRED.ANYCounts full cacheline writes (ItoM) that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallycore_snoop_response.i_hit_fseHit snoop reply without sending the data, line invalidatedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x703000002ocr.hwpf_l3.remoteCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeunc_m_rpq_inserts.pch1unc_m_rpq_occupancy_pch1event=0x50,umask=0x0cevent=0x35,umask=0xC80FFE01TOR Inserts : All requests from IO Devices that missed the LLC. Unit: uncore_cha TOR Inserts : ItoMs issued by IO Devices that Hit the LLC. Unit: uncore_cha TOR Occupancy : DRds issued by iA Cores targeting PMM Mem that Missed the LLC. Unit: uncore_cha event=0xc0,ch_mask=0x40,fc_mask=0x07,umask=0x01event=0xc1,ch_mask=0x80,fc_mask=0x07,umask=0x04unc_iio_txn_req_of_cpu.mem_write.part4event=0x84,ch_mask=0x80,fc_mask=0x07,umask=0x04unc_iio_txn_req_of_cpu.cmpd.part7event=0xc2,ch_mask=0x40,fc_mask=0x04,umask=0x03Counts the number of core requests (demand and L1 prefetchers) rejected by the L2 queue (L2Q) due to a full conditionCounts the number of first level data cacheline (dirty) evictions caused by misses, stores, and prefetchesevent=0x34,period=200003,umask=0x38Counts all code reads that were supplied by the L3 cache where a snoop was sent but the snoop missedCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by the L3 cache where a snoop was sent but the snoop missedocr.demand_data_and_l1pf_rd.l3_hit.snoop_hitmCounts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedCounts uncached memory reads that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedCounts uncached memory reads that were supplied by the L3 cache where a snoop was sent but the snoop missedCounts the number of bus locks a core issued its self (e.g. lock to UC or Split Lock) and does not include cache locks. Counts on a per core basisCounts the number of hardware interrupts received by the processorocr.all_code_rd.any_responseCounts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)ocr.prefetches.any_responseevent=0x74,period=1000003,umask=0x40Counts the number of issue slots every cycle that were not delivered by the frontend due to BTCLEARS, which occurs when the Branch Target Buffer (BTB) predicts a taken branchCounts the number of issue slots every cycle that were not delivered by the frontend due to Instruction Table Lookaside Buffer (ITLB) missesCounts snoop filter capacity evictions for entries tracking exclusive lines in the cores? cache.? Snoop filter capacity evictions occur when the snoop filter is full and evicts an existing entry to track a new entry.? Does not count clean evictions such as when a core?s cache replaces a tracked cacheline with a new cachelineTOR Occupancy : DRd_Opts issued by iA Cores that hit the LLC. Unit: uncore_cha PCIe Completion Buffer Inserts of completions with data : Part 0-7Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 3Counts the number of page walks completed due to loads (including SW prefetches) whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 4K pages. Includes page walks that page faultevent=0x49,period=2000003,umask=0x20Counts the number of page walks outstanding in the page miss handler (PMH) for stores every cycleInstructions Per Cycle (per Logical Processor). Unit: cpu_core 100 * LD_BLOCKS.DATA_UNKNOWN / MEM_UOPS_RETIRED.ALL_LOADSCycles with L1D load Misses outstanding. Unit: cpu_core event=0x44,period=200003,umask=0x1mem_uop_retired.anyDEMAND_DATA_RD & L3_HIT & SNOOP_HIT_WITH_FWD. Unit: cpu_core event=0x40,period=100003,umask=0x1Counts number of SSE/AVX computational scalar single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT RCP FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. Unit: cpu_core event=0x79,cmask=6,period=2000003,umask=0x4Cycles when no uops are not delivered by the IDQ when backend of the machine is not stalled. Unit: cpu_core ld_head.pgwalk_at_retevent=0x5,period=1000003,umask=0xa0Counts randomly selected loads when the latency from first dispatch to completion is greater than 128 cycles  Supports address when precise (Must be precise). Unit: cpu_core event=0x2d,cmask=1,period=1000003,umask=0x1Counts the number of issue slots every cycle that were not delivered by the frontend due to BTCLEARS. Unit: cpu_atom Instruction decoders utilized in a cycle. Unit: cpu_core uops_decoded.dec0_uopsUops executed on ports 2, 3 and 10. Unit: cpu_core Cycles with retired uop(s). Unit: cpu_core event=0xc2,period=2000003,umask=0x4,frontend=0x8Retirement slots used. Unit: cpu_core unc_m_dram_page_hit_rdunc_m_dram_page_empty_rdNumber of page walks outstanding for a demand load in the PMH each cycle. Unit: cpu_core event=0x13,period=100003,umask=0x10event=0x26,period=200003,umask=0x1event=0x2a,period=100003,umask=0x1,offcore_rsp=0x4003C4477event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3FBFC00004This event is deprecated. Refer to new event ARITH.FPDIV_ACTIVENumber of all retired NOP instructionsDRAM RD_CAS and WR_CAS Commands. : DRAM WR_CAS commands w/o auto-pre. Unit: uncore_imc unc_m_cas_count.pch1event=0x35,umask=0x00cc43ff04TOR Inserts for DRd misses from local IA targeting remote memory. Unit: uncore_cha event=0x2,umask=0x0000000047unc_upi_rxl_flits.slot0Matches on Transmit path of a UPI Port : Non-Coherent Bypass, Match Opcode. Unit: uncore_upi unc_upi_rxl_basic_hdr_match.ncb_opcevent=0x83,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000080event=0x32,umask=0x0000000104unc_m2m_tracker_occupancy.ch0Tracker Occupancy : Channel 1. Unit: uncore_m2m Data Prefetches Dropped. Unit: uncore_m2m Prefetch CAM Inserts : XPT - All Channels. Unit: uncore_m2m : XPT - All Channels. Unit: uncore_m2m FlowQ Generated Prefetch. Unit: uncore_m3upi event=0x50,umask=0x0000000003unc_cha_tor_inserts.evictTOR Inserts : IRQ - Non iA. Unit: uncore_cha TOR Inserts : All from Local IO. Unit: uncore_cha event=0x35event=0x35,umask=0x00c88ffe01unc_cha_tor_occupancy.snps_from_remevent=0x36,umask=0x00c817fd01event=0x36,umask=0x00c807fd01event=0x36,umask=0x00c001ff04event=0x36,umask=0x0000000008unc_cha_tor_occupancy.wbqTOR Occupancy : Match the Opcode in b[29:19] of the extended umask field. Unit: uncore_cha TOR Occupancy : Just NonCoherent. Unit: uncore_cha event=0x36,umask=0x00c887fe01TOR Occupancy; DRd Opt Pref from local IA. Unit: uncore_cha event=0x36,umask=0x00c80fff01event=0x35,umask=0x00ccd7fe01event=0x35,umask=0x00C8670A01event=0x36,umask=0x00c8978601TOR Occupancy : DRd_Prefs issued by iA Cores targeting DDR Mem that Missed the LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_drd_pref_remote_ddrunc_cha_tor_occupancy.ia_miss_crd_pref_remoteunc_cha_tor_occupancy.ia_hit_itomevent=0x36,umask=0x00cc23ff04unc_cha_tor_occupancy.io_miss_itomcachenearbp_l1_btb_correctl2_request_g1.ls_rd_blk_c_sevent=0x60,umask=0xf9event=0x64,umask=0x40Core to L2 cacheable request access status (not including L2 Prefetch). Data cache read hit on shared line in L2l3_comb_clstr_state.other_l3_miss_typsevent=0x90,umask=0x00The number of branch instructions retired. This includes all types of architectural control flow changes, including exceptions and interruptsevent=0xc4event=0x1cf,umask=0x01event=0,umask=0x02All FLOPSL1 DTLB Miss or Reload off all sizesevent=0x76100%L2 Cache Hits from L1 Data Cache Misses(xi_sys_fill_latency * 16) / xi_ccx_sdp_req1.all_l3_miss_req_typsMixed SSE/AVX StallsThe number of instruction fetches that hit in the L1 ITLB. Instruction fetches to a 4KB pagebp_l1_tlb_miss_l2_tlb_missTotal number of fp uOps. The number of operations (uOps) dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPSTotal number uOps assigned to pipe 3fp_ret_sse_avx_ops.mult_flopsfp_disp_faults.ymm_spill_faultls_refills_from_sys.ls_mabresp_rmt_dramDemand Data Cache Fills by Data Source. Hit in cache; Remote CCX and the address's Home Node is on a different dieevent=0x43,umask=0x01Hardware Prefetch Data Cache Fills by Data Source. From DRAM (home node remote)event=0xaa,umask=0xffCycles where a dispatch group is valid but does not get dispatched due to a token stall. Taken branch buffer resource stallDispatch of a single op that performs a memory store. Counts the number of operations dispatched to the LS unit. Unit Masks ADDedls_misal_loads.ma4kevent=0x59,umask=0x04event=0x44,umask=0x03l1_data_cache_fills_from_external_ccx_cacheDemand data cache fills from extension memoryAny data cache fills from cache of another CCX when the address was in the same NUMA nodeHardware prefetch data cache fills from cache of another CCX when the address was in a different NUMA nodeL2 cache requests: instruction cache readsL2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L2Stream (fetch additional sequential lines into L2 cache)l2_pf_miss_l2_hit_l3.l1_strideL2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L1Stream (fetch additional sequential lines into L1 cache)l3_xi_sampled_latency.near_cacheevent=0x25f,umask=0x7ffevent=0x8df,umask=0x7feevent=0x81f,umask=0xbfeData beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 2Data beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 4remote_socket_inf0_inbound_data_beats_ccm1remote_socket_inf0_outbound_data_beats_ccm4event=0xc9f,umask=0xf3efp_ret_x87_fp_ops.allfp_ops_retired_by_width.scalar_uops_retiredRetired scalar floating-point divide opsevent=0xa,umask=0x20fp_ops_retired_by_type.vector_cmpevent=0xa,umask=0x70Retired vector floating-point shuffle ops (may include instructions not necessarily thought of as including shuffles e.g. horizontal add, dot product, and certain MOV instructions)sse_avx_ops_retired.sse_avx_shaRetired 128-bit packed floating-point ops of other typesRetired 256-bit packed floating-point convert opspacked_int_op_type.int128_macRetired 128-bit packed integer multiply-accumulate opsevent=0xd,umask=0x06packed_int_op_type.int128_cmpevent=0xd,umask=0xffL1 DTLB misses with L2 DTLB misses (page-table walks are requested) for 1G pagesde_no_dispatch_per_slot.no_ops_from_frontendFraction of dispatched ops that did not retireFraction of dispatch slots used by ops that retiredbackend_bound * d_ratio(ex_no_retire.load_not_complete, ex_no_retire.not_complete)backend_bound_cpuExecution-time branch misprediction ratio (non-speculative)L1 demand data cache fills from DRAM or MMIO in the same NUMA noderemote_socket_upstream_write_beats_iom0 + remote_socket_upstream_write_beats_iom1 + remote_socket_upstream_write_beats_iom2 + remote_socket_upstream_write_beats_iom3local_socket_inf0_inbound_data_beats_ccm0 + local_socket_inf1_inbound_data_beats_ccm0 + local_socket_inf0_inbound_data_beats_ccm1 + local_socket_inf1_inbound_data_beats_ccm1 + local_socket_inf0_inbound_data_beats_ccm2 + local_socket_inf1_inbound_data_beats_ccm2 + local_socket_inf0_inbound_data_beats_ccm3 + local_socket_inf1_inbound_data_beats_ccm3 + local_socket_inf0_inbound_data_beats_ccm4 + local_socket_inf1_inbound_data_beats_ccm4 + local_socket_inf0_inbound_data_beats_ccm5 + local_socket_inf1_inbound_data_beats_ccm5 + local_socket_inf0_inbound_data_beats_ccm6 + local_socket_inf1_inbound_data_beats_ccm6 + local_socket_inf0_inbound_data_beats_ccm7 + local_socket_inf1_inbound_data_beats_ccm7remote_socket_outbound_data_from_cpulocal_socket_outbound_data_beats_link0 + local_socket_outbound_data_beats_link1 + local_socket_outbound_data_beats_link2 + local_socket_outbound_data_beats_link3 + local_socket_outbound_data_beats_link4 + local_socket_outbound_data_beats_link5 + local_socket_outbound_data_beats_link6 + local_socket_outbound_data_beats_link7Memory controller data bus utilizationDC_L1_DTLB_MISS_AND_L2_DTLB_HITIC_REFILL_FROM_SYSTEMMEM_READL2_CACHE_ACCESSEVENT_49HEVENT_5AHEVENT_5BHEVENT_5DHEVENT_6AHEVENT_ABHEVENT_BBHEVENT_BFHEVENT_C8HEVENT_CEHEVENT_EEHL1D_CACHE_WBBR_IMMED_RETIREDEVENT_141HEVENT_14DHEVENT_17AHEVENT_19EHEVENT_235HEVENT_27EHEVENT_283HEVENT_2A2HEVENT_2A9HEVENT_2B3HEVENT_2CFHEVENT_2D8HEVENT_2E4HEVENT_2F2HEVENT_30CHEVENT_33CHEVENT_33DHEVENT_35FHEVENT_380HEVENT_385HEVENT_3DFHEVENT_3EAHEVENT_3F3HEXC_DABORTDTLB_WALKSTREX_SPEChnf_pocq_addrhazhnf_txrsp_stallhni_rdt_wr_occ_cnt_ovflrni_txreq_flits_replayedcxra_ext_rsp_stallcxla_rx_cxs_link2cxla_avg_sz_rx_cxs_dw_beatclkdiv2_prank_turnaround_activateVIU1_INSTR_WAIT_CYCLESL1_DATA_SNOOP_HIT_CASTOUTFP_THREE_QUARTERS_FPSCR_RENAMES_BUSYDST_STREAM_1_CACHE_LINE_FETCHESFXU0_IDLE_FXU1_IDLEFXU_MARKED_INSTR_FINISHEDADDERBRANCHES_FINISHEDBTB_BRANCH_MISPRED_FROM_DIRECTIONCYCLES_DECODE_STALLEDCYCLES_BRANCH_ISSUE_STALLEDDECORATED_STORESK8-exclusivemodifiedtag-snoopsse-reclass-microfaultsXSCALEPPC7450INTEL_SANDYBRIDGE_XEON%s:
%s
{"type": "sysexit"%s, "pid": "%d", "start": "0x%016jx", "end": "0x%016jx"}
GenuineIntel-6-45bdwde metricsIpFLOPl1d_pend_miss.fb_fullThis event counts L2 fill requests that access L2 cacheevent=0x63,period=2000003,umask=0x2Retired load uops with L2 cache misses as data sources. Uses PEBS  Supports address when precise (Precise event)This event counts the demand RFO (read for ownership) requests including regular RFOs, locks, ItoMCycles with any input/output SSE or FP assistevent=0xca,period=100003,umask=0x4icache.ifdata_stallThis event counts the number of cycles 4  uops were  delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Counting includes uops that may bypass the IDQidq.all_dsb_cycles_any_uopsidq.mite_uopsevent=0x79,period=2000003,umask=0x4event=0x9c,cmask=4,period=2000003,umask=0x1Number of times HLE aborted and was not due to the abort conditions in subevents 3-6event=0xcd,period=1009,umask=0x1,ldlat=0x80This event counts speculative cache line split store-address (STA) uops dispatched to the L1 cachetx_exec.misc2br_inst_exec.taken_indirect_jump_non_call_retevent=0x88,period=200003,umask=0xa0event=0x89,period=200003,umask=0xc1Mispredicted conditional branch instructions retired. (Precise Event - PEBS) (Precise event)event=0xd,any=1,cmask=1,period=2000003,umask=0x3rs_events.empty_cyclesuops_dispatched_port.port_3Cycles per thread when uops are executed in port 7Cycles at least 1 micro-op is executed from any thread on physical coreCycles when Resource Allocation Table (RAT) does not issue Uops to Reservation Station (RS) for the threadMemory controller clock ticks. Unit: uncore_imc event=0x2,umask=0x4Store misses that miss the  DTLB and hit the STLB (4K)Store misses in all DTLB levels that cause completed page walks (1G)  Spec update: BDM69event=0xbc,period=2000003,umask=0x11This event counts retired load uops which data sources were HitM responses from a core on same socket (shared L3)  Supports address when precise.  Spec update: BDM100 (Precise event)offcore_response.all_pf_code_rd.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020001offcore_response.pf_l2_data_rd.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0200offcore_response.pf_l3_code_rd.l3_hit.snoop_noneoffcore_response.pf_l3_code_rd.supplier_none.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0100offcore_response.all_data_rd.l3_miss_local_dram.snoop_hitmoffcore_response.all_pf_data_rd.l3_hit.snoop_non_dramoffcore_response.corewb.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0020offcore_response.pf_l2_rfo.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000100Taken branch instructions retired (Precise event)event=0x22,umask=0x44L3 Lookup read request that access cache and found line in I-stateunc_cbo_cache_lookup.read_mesioffcore_response.all_code_rd.llc_miss.any_responseoffcore_response.all_code_rd.llc_miss.local_dramCounts all prefetch (that bring data to LLC only) RFOs miss in the L3l2_data_rqsts.self.e_stateevent=0x29,period=200000,umask=0x7fl2_ld_ifetch.self.m_statel2_ld_ifetch.self.s_statel2_lock.self.m_stateModified lines evicted from the L2 cacheevent=0x30,period=200000,umask=0x54l2_rqsts.self.any.mesievent=0x2e,period=200000,umask=0x41L2 cache demand requests from this core that missed the L2l2_rqsts.self.demand.s_stateCISC macro instructions decodedmisalign_mem_ref.splitStreaming SIMD Extensions (SSE) PrefetchT0 instructions executedHITM signal assertedbus_trans_brd.selfext_snoop.this_agent.cleanreissue.overlap_store.arCycles issue is stalled due to div busyDTLB misses due to store operationsevent=0xd1,period=200003,umask=0x10Load uops retired that split a cache-line (Precise event capable)  Supports address when precise (Must be precise)offcore_response.any_pf_data_rd.l2_miss.snoop_miss_or_no_snoop_neededoffcore_response.any_read.l2_hitCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that true miss for the L2 cache with a snoop miss in the other processor moduleoffcore_response.any_rfo.l2_miss.hit_other_core_no_fwdCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000010event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000010offcore_response.sw_prefetch.l2_miss.snoop_miss_or_no_snoop_neededevent=0xe9,period=200003,umask=0x1Counts the number of times the Microcode Sequencer (MS) starts a flow of uops from the MSROM. It does not count every time a uop is read from the MSROM.  The most common case that this counts is when a micro-coded instruction is encountered by the front end of the machine.  Other cases include when an instruction encounters a fault, trap, or microcode assist of any sort that initiates a flow of uops.  The event will count MS startups for uops that are speculative, and subsequently cleared by branch mispredict or a machine clearCounts near indirect call or near indirect jmp branch instructions retired (Must be precise)Retired mispredicted branch instructions (Precise event capable) (Must be precise)br_misp_retired.jccCounts mispredicted branch instructions retired that were near indirect call or near indirect jmp, where the target address taken was not what the processor predicted (Must be precise)Counts the number of issue slots per core cycle that were not consumed because of a full resource in the backend.  Including but not limited to resources such as the Re-order Buffer (ROB), reservation stations (RS), load/store buffers, physical registers, or any other needed machine resource that is currently unavailable.   Note that uops must be available for consumption in order for this event to fire.  If a uop is not available (Instruction Queue is empty), this event will not countevent=0xd0,period=200003,umask=0x11Counts the number of writeback transactions caused by L1 or L2 cache evictions miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts data cacheline reads generated by hardware L2 cache prefetcher true miss for the L2 cache with a snoop miss in the other processor moduleCounts data cacheline reads generated by hardware L2 cache prefetcher true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts page walks completed due to demand data stores whose address translations missed in the TLB and were mapped to 2M or 4M pages.  The page walks can end with or without a page faultCounts all L2 code requestsDemand Data Read requests that hit L2 cache  Spec update: HSD78, HSM80event=0xd3,period=100003,umask=0x1offcore_response.demand_data_rd.l3_hit.hit_other_core_no_fwdNumber of uops delivered to IDQ from any pathIncrement each cycle # of uops delivered to IDQ when MS_busy by DSB. Set Cmask = 1 to count cycles. Add Edge=1 to count # of deliveryevent=0xcd,period=100003,umask=0x1,ldlat=0x20Randomly selected loads with latency value being above 512  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)Randomly selected loads with latency value being above 8  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100400122Counts all prefetch (that bring data to LLC only) data reads miss in the L3Cycles where at least 2 uops were executed per-thread  Spec update: HSD144, HSD30, HSM31Cycles where at least 3 uops were executed per-thread  Spec update: HSD144, HSD30, HSM31Cycles per core when uops are executed in port 4Cycles which a uop is dispatched on port 6 in this threadL3 Lookup external snoop request that access cache and found line in MESI-state. Unit: uncore_cbox Demand load Miss in all translation lookaside buffer (TLB) levels causes a page walk that completes (2M/4M)Misses in all ITLB levels that cause completed page walksevent=0xbc,period=2000003,umask=0x42Counts the number of Extended Page Table walks from the ITLB that hit in memoryevent=0x28,period=200003,umask=0xfNot rejected writebacks from L1D to L2 cache lines in E statel2_rqsts.pf_hitfp_comp_ops_exe.sse_packed_singleevent=0x10,period=2000003,umask=0x40Counts 256-bit packed double-precision floating-point instructionsCycles with pending L1 cache miss loads. Set AnyThread to count per coreCycles which a Uop is dispatched on port 1Uops dispatched to port 2, loads and stores per core (speculative and retired)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0010offcore_response.pf_l2_data_rd.llc_hit.snoop_missevent=0x35,umask=0x3,filter_opc=0x1e6Counts the number of cycles that the uncore transitioned to a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band2=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Derived from unc_p_freq_band2_cycles. Unit: uncore_pcu unc_p_freq_ge_3000mhz_transitionsAll retired store uops (Precise event)This event counts the cycles attributed to a switch from the Decoded Stream Buffer (DSB), which holds decoded instructions, to the legacy decode pipeline.  It excludes cycles when the back-end cannot  accept new micro-ops.  The penalty for these switches is potentially several cycles of instruction starvation, where no micro-ops are delivered to the back-endCases of cancelling valid DSB fill not because of exceeding way limitevent=0xb6,period=100003,umask=0x1Each cycle there was a miss-pending demand load this thread and no uops dispatched, increment by 1. Note this is in DCU and connected to Umask 1 and 2. Miss Pending demand load should be deduced by OR-ing increment bits of DCACHE_MISS_PEND.PENDINGEach cycle there was a MLC-miss pending demand load and no uops dispatched on this thread (i.e. Non-completed valid SQ entry allocated for demand load and waiting for Uncore), increment by 1. Note this is in MLC and connected to Umask 0 and 2Retired instructions experiencing ITLB missesevent=0x5b,period=2000003,umask=0xcevent=0x36,umask=0x3mem_uops_retired.l2_miss_loadsCounts Demand cacheable data and L1 prefetch data read requests  that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002003091Counts any Prefetch requests that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeoffcore_response.any_read.l2_hit_far_tile_e_fCounts Demand cacheable data write requests  that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeoffcore_response.bus_locks.l2_hit_far_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400400Counts Bus locks and split lock requests that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000001Counts Demand cacheable data writes that accounts for responses which hit its own tile's L2 with data in E stateCounts Demand cacheable data writes that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400080Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000080offcore_response.pf_l1_data_rd.l2_hit_near_tileevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800401000Counts UC code reads (valid only for Outstanding response type)  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000200Counts Demand code reads and prefetch code read requests  that accounts for data responses from DRAM Localoffcore_response.any_code_rd.mcdram_farCounts any request that accounts for data responses from DRAM Localoffcore_response.any_rfo.ddr_faroffcore_response.any_rfo.ddr_nearevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000001Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for data responses from MCDRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200040Counts Software Prefetches that accounts for data responses from DRAM Farevent=0xc5,period=200003,umask=0xfdThis event counts the number of core cycles when no uops are allocated and the alloc pipe is stalled waiting for a mispredicted branch to retireevent=0x51,period=2000000,umask=0x8l1d_cache_st.m_stateevent=0x41,period=2000000,umask=0x8l1d_prefetch.missl2_data_rqsts.prefetch.m_statel2_lines_out.prefetch_cleanL2 instruction fetch hitsl2_transactions.prefetchAll L2 demand store RFOsevent=0xb,period=2000000,umask=0x10,ldlat=0x0mem_inst_retired.latency_above_threshold_64Memory instructions retired above 8 clocks (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x211event=0xb7,period=100000,umask=0x1,offcore_rsp=0x1FFOffcore RFO requests satisfied by any cache or DRAMOffcore writebacks to any cache or DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8008Offcore writebacks to the IO, CSR, MMIO unitOffcore writebacks that HIT in a remote cacheOffcore code or data read requests satisfied by any cache or DRAMoffcore_response.data_ifetch.remote_cache_hitoffcore_response.data_in.remote_cacheOffcore demand data requests satisfied by any cache or DRAMoffcore_response.demand_data.any_locationOffcore demand data requests satisfied by the LLC and not found in a sibling coreOffcore demand data requests that HITM in a remote cacheoffcore_response.demand_data_rd.llc_hit_other_core_hitoffcore_response.demand_rfo.remote_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x880event=0xb7,period=100000,umask=0x1,offcore_rsp=0x8030offcore_response.pf_ifetch.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x240All offcore prefetch RFO requestsOffcore code reads satisfied by the local DRAMOffcore requests satisfied by any DRAMOffcore writebacks to a remote DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2033offcore_response.demand_data.any_dramOffcore demand data requests satisfied by the local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6004Offcore prefetch data reads that missed the LLCoffcore_response.pf_ifetch.local_dramoffcore_response.prefetch.local_draml1i.hitsL1I instruction fetch hitsevent=0xc4,period=200000,umask=0x1event=0,period=2000000,umask=0x0ild_stall.anyLoops that can't stream from the instruction queueevent=0xa2,period=2000000,umask=0x40event=0xa2,period=2000000,umask=0x10event=0xc7,period=200000,umask=0x8event=0xd1,period=2000000,umask=0x8uops_executed.port015_stall_cyclesCycles Uops are not retiring (Precise Event)DTLB load miss page walks completeDTLB missesCounts duration of L1D miss outstanding in cyclesCounts the total number of requests from the L2 hardware prefetchersRetired load instructions missed L2 cache as data sources  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020004Counts demand data readshave any response typeoffcore_response.demand_data_rd.l3_hit_m.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC01C0002offcore_response.demand_rfo.l3_hit_m.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0020002offcore_response.demand_rfo.supplier_none.spl_hitoffcore_response.other.l3_hit_e.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80408000Number of PREFETCHT1 or PREFETCHT2 instructions executedCounts Decode Stream Buffer (DSB)-to-MITE switch true penalty cycles. These cycles do not include uops routed through because of the switch itself, for example, when Instruction Decode Queue (IDQ) pre-allocation is unavailable, or Instruction Decode Queue (IDQ) is full. SBD-to-MITE switch true penalty cycles happen after the merge mux (MM) receives Decode Stream Buffer (DSB) Sync-indication until receiving the first MITE uop. MM is placed before Instruction Decode Queue (IDQ) to merge uops being fed from the MITE and Decode Stream Buffer (DSB) paths. Decode Stream Buffer (DSB) inserts the Sync-indication whenever a Decode Stream Buffer (DSB)-to-MITE switch occurs.Penalty: A Decode Stream Buffer (DSB) hit followed by a Decode Stream Buffer (DSB) miss can cost up to six cycles in which no uops are delivered to the IDQ. Most often, such switches from the Decode Stream Buffer (DSB) to the legacy pipeline cost 02 cyclesevent=0xc6,period=100007,umask=0x1,frontend=0x14event=0xc6,period=100007,umask=0x1,frontend=0x420006Counts retired instructions that are delivered to the back-end after a front-end stall of at least 8 cycles. During this period the front-end delivered no uops (Precise event)Counts retired Instructions that experienced STLB (2nd level TLB) true miss (Precise event)event=0x83,period=200003,umask=0x1event=0x60,cmask=1,period=2000003,umask=0x10event=0x60,cmask=6,period=2000003,umask=0x10event=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000080001offcore_response.demand_rfo.l3_hit_m.snoop_non_dramCounts when there is a transition from ring 1, 2 or 3 to ring 0Counts cycles during which a total of 2 uops were executed on all ports and Reservation Station (RS) was not emptyRetirement slots usedMem;MemoryTLB;_SMTInstruction_Fetch_BW100 * ( (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * INT_MISC.CLEAR_RESTEER_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) ) - (100 * (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * ( (ICACHE_64B.IFTAG_STALL / CPU_CLK_UNHALTED.THREAD) + (( ICACHE_16B.IFDATA_STALL + 2 * cpu@ICACHE_16B.IFDATA_STALL\,cmask\=1\,edge@ ) / CPU_CLK_UNHALTED.THREAD) + (9 * BACLEARS.ANY / CPU_CLK_UNHALTED.THREAD) ) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))))UOPS_EXECUTED.THREAD / (( UOPS_EXECUTED.CORE_CYCLES_GE_1 / 2 ) if #SMT_on else UOPS_EXECUTED.CORE_CYCLES_GE_1)Fed;FetchBWJumpCounts completed page walks  (4K sizes) caused by demand data loads. This implies address translations missed in the DTLB and further levels of TLB. The page walk can end with or without a faultitlb_misses.walk_activeCycles when at least one PMH is busy with a page walk for code (instruction fetch) request. EPT page walk duration are excluded in Skylake microarchitectureCounts completed page walks (all page sizes) caused by a code fetch. This implies it missed in the ITLB (Instruction TLB) and further levels of TLB. The page walk can end with or without a faultCounts any rfo reads (demand & prefetch) that miss L2Counts demand and DCU prefetch instruction cacheline that have any response typeCounts DCU hardware prefetcher data read that miss L2Counts DCU hardware prefetcher data read that hit in the other module where modified copies were found in other core's L1 cacheLoad uops that split cache line boundary (Precise event)The BACLEARS event counts the number of times the front end is resteered, mainly when the Branch Prediction Unit cannot provide a correct prediction and this is corrected by the Branch Address Calculator at the front end.  The BACLEARS.COND event counts the number of JCC (Jump on Condtional Code) baclearsNON_RETURN_IND counts the number of near indirect JMP and near indirect CALL branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)This event counts the number of instructions that retire execution. For instructions that consist of multiple micro-ops, this event counts the retirement of the last micro-op of the instruction. The counter continues counting during hardware interrupts, traps, and inside interrupt handlersDuration of D-side page-walks in core cyclesoffcore_response.pf_l2_code_rd.llc_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0040offcore_response.all_pf_code_rd.llc_miss.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400200Offcore demand data read requestsREQUEST = DATA_IFETCH and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DEMAND_DATA_RD and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = DEMAND_DATA_RD and RESPONSE = REMOTE_CACHE_HITMREQUEST = OTHER and RESPONSE = ANY_CACHE_DRAMoffcore_response.any_ifetch.any_dram_and_remote_fwdevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf844event=0xb7,period=100000,umask=0x1,offcore_rsp=0x3001REQUEST = DEMAND_IFETCH and RESPONSE = OTHER_LOCAL_DRAMoffcore_response.prefetch.other_local_dramsnoopq_requests.dataOutstanding snoop data requestsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x27FFevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2740event=0xd3,period=100007,umask=0x2Counts all demand code reads that hit in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0080Counts all demand & prefetch data reads that miss the L3 and the modified data is transferred from remote cacheoffcore_response.all_pf_data_rd.l3_miss.remote_hit_forwardCounts all prefetch data reads that miss the L3 and clean or shared data is transferred from remote cacheCounts all prefetch data reads that miss the L3 and the data is returned from remote dramCounts all demand code reads that miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800400offcore_response.pf_l2_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdoffcore_response.pf_l2_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00080Counts all prefetch (that bring data to LLC only) RFOs that miss the L3 and the data is returned from local or remote dramPower_License0_UtilizationPower_License0_Utilization_SMTunc_m_wpq_insertsPCI Express bandwidth reading at IIO. Derived from unc_iio_data_req_of_cpu.mem_read.part0. Unit: uncore_iio Counts the number of cycles either the local or incoming distress signals are asserted.  Incoming distress includes up, dn and acrossRead request from a remote socket which hit in the HitMe Cache to a line In the E state. Unit: uncore_cha RspI Snoop Responses Received. Unit: uncore_cha unc_iio_comp_buf_occupancy.cmpd.part3event=0xc0,ch_mask=0x01,fc_mask=0x07,umask=0x04unc_iio_data_req_by_cpu.peer_write.part2event=0xc0,ch_mask=0x08,fc_mask=0x07,umask=0x02Peer to peer read request for 4 bytes made by IIO Part3 to an IIO target. Unit: uncore_iio event=0x83,ch_mask=0x04,fc_mask=0x07,umask=0x02unc_iio_txn_req_of_cpu.peer_read.part0event=0x84,ch_mask=0x01,fc_mask=0x07,umask=0x08Counts every peer to peer read request of up to a 64 byte transaction made by IIO Part0 to the MMIO space of an IIO target. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busCounts every peer to peer read request of up to a 64 byte transaction made by IIO Part1 to the MMIO space of an IIO target. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busunc_iio_txn_req_of_cpu.peer_read.part3Occupancy of the IRP Fire and Forget (FAF) queue, a queue used for processing inbound reads in the IRPMulti-socket cacheline Directory update from S to I. Unit: uncore_m2m unc_m2m_imc_writes.partialCounts when the M2M (Mesh to Memory) issues partial writes to the iMC (Memory Controller).  It only counts normal priority non-isochronous writesCounts when the M2M (Mesh to Memory) promotes a outstanding request in the prefetch queue due to a subsequent demand read request that entered the M2M with the same address.  Explanatory Side Note: The Prefecth queue is made of CAM (Content Addressable Memory)OCR.ALL_DATA_RD.L3_HIT.HITM_OTHER_CORE OCR.ALL_DATA_RD.L3_HIT.HITM_OTHER_CORE OCR.ALL_DATA_RD.L3_HIT.HITM_OTHER_COREOCR.ALL_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWD  OCR.ALL_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDocr.all_data_rd.l3_hit_s.hit_other_core_fwdocr.all_pf_data_rd.l3_hit.hit_other_core_fwdOCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDocr.all_pf_data_rd.l3_hit.snoop_noneocr.all_pf_data_rd.l3_hit_e.hitm_other_coreOCR.ALL_PF_DATA_RD.L3_HIT_S.ANY_SNOOP  OCR.ALL_PF_DATA_RD.L3_HIT_S.ANY_SNOOPOCR.ALL_PF_DATA_RD.L3_HIT_S.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F802007F7OCR.ALL_READS.L3_HIT_F.HIT_OTHER_CORE_FWD  OCR.ALL_READS.L3_HIT_F.HIT_OTHER_CORE_FWDocr.all_reads.l3_hit_f.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F801007F7ocr.all_rfo.l3_hit_f.hit_other_core_no_fwdOCR.ALL_RFO.L3_HIT_S.HIT_OTHER_CORE_FWD  OCR.ALL_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDocr.demand_code_rd.l3_hit_e.hitm_other_coreocr.demand_code_rd.l3_hit_f.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200004Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_F.HITM_OTHER_COREocr.demand_rfo.l3_hit.hit_other_core_fwdocr.demand_rfo.l3_hit.snoop_noneCounts any other requests OCR.OTHER.L3_HIT.NO_SNOOP_NEEDED OCR.OTHER.L3_HIT.NO_SNOOP_NEEDEDocr.pf_l1d_and_sw.l3_hit_e.any_snoopocr.pf_l2_data_rd.l3_hit.hit_other_core_fwdocr.pf_l2_data_rd.l3_hit_f.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_HIT.HITM_OTHER_CORE OCR.PF_L2_RFO.L3_HIT.HITM_OTHER_COREocr.pf_l2_rfo.l3_hit_f.hit_other_core_fwdocr.pf_l2_rfo.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100020Counts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.PF_L3_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_HIT.SNOOP_MISSocr.pf_l3_data_rd.l3_hit_m.any_snoopCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_S.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0100Counts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_E.NO_SNOOP_NEEDEDCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_F.ANY_SNOOPoffcore_response.all_data_rd.l3_hit_f.hit_other_core_fwdoffcore_response.all_data_rd.l3_hit_s.any_snoopThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020491This event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.all_pf_data_rd.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_S.SNOOP_MISSoffcore_response.demand_code_rd.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_S.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDoffcore_response.demand_data_rd.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_S.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020002offcore_response.other.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_F.HIT_OTHER_CORE_NO_FWDoffcore_response.other.l3_hit_m.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800028000This event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.HIT_OTHER_CORE_FWDoffcore_response.pf_l1d_and_sw.l3_hit_f.hit_other_core_fwdoffcore_response.pf_l1d_and_sw.pmm_hit_local_pmm.any_snoopoffcore_response.pf_l1d_and_sw.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020010offcore_response.pf_l2_rfo.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020020This event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.pf_l3_data_rd.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400080This event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.pf_l3_rfo.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400100Counts once for each Intel AVX-512 computational 128-bit packed BFloat16 floating-point instruction retired. Applies to the XMM based VDPBF16PS instruction. Each count represents 16 computation operations. This event is only supported on products formerly named Cooper Lake and is not supported on products formerly named Cascade LakeOCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDocr.all_pf_data_rd.l3_miss.no_snoop_neededOCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDocr.all_reads.l3_miss.remote_hit_forwardevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC007F7ocr.all_reads.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000122ocr.demand_code_rd.l3_miss.remote_hitmocr.demand_code_rd.l3_miss_local_dram.snoop_miss_or_no_fwdocr.demand_code_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000001Counts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.SNOOP_NONEocr.demand_rfo.l3_miss_remote_hop1_dram.any_snoopocr.demand_rfo.l3_miss_remote_hop1_dram.snoop_missCounts any other requests OCR.OTHER.L3_MISS.NO_SNOOP_NEEDED OCR.OTHER.L3_MISS.NO_SNOOP_NEEDEDocr.other.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts any other requests  OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210008000ocr.pf_l1d_and_sw.l3_miss.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000400Counts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.SNOOP_MISSCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDocr.pf_l2_data_rd.l3_miss_remote_hop1_dram.snoop_noneocr.pf_l2_rfo.l3_miss.any_snoopCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDocr.pf_l3_data_rd.l3_miss.snoop_noneocr.pf_l3_data_rd.l3_miss_local_dram.hit_other_core_fwdCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDocr.pf_l3_rfo.l3_miss_local_dram.hitm_other_coreCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREocr.pf_l3_rfo.l3_miss_remote_hop1_dram.hitm_other_coreocr.pf_l3_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.all_pf_data_rd.l3_miss_remote_hop1_dram.snoop_missoffcore_response.all_reads.l3_miss_remote_hop1_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.all_rfo.l3_miss.hitm_other_coreoffcore_response.all_rfo.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.ANY_SNOOPocr.all_data_rd.supplier_none.snoop_noneOCR.ALL_PF_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWD  OCR.ALL_PF_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.all_reads.supplier_none.snoop_noneCounts all demand data writes (RFOs) OCR.DEMAND_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.pf_l2_data_rd.any_responseUnderfill readsunc_m_tagchk.hitAll Clean line misses to Near Memory(DRAM cache) in Memory Mode. Unit: uncore_imc unc_m2m_tag_hit.nm_ufill_hit_dirtyevent=0xc7,period=100003,umask=0x40Cycles when uops are being delivered to IDQ while MS is busy1000 * ( ( OFFCORE_REQUESTS.ALL_DATA_RD - OFFCORE_REQUESTS.DEMAND_DATA_RD ) + L2_RQSTS.ALL_DEMAND_MISS + L2_RQSTS.SWPF_MISS ) / INST_RETIRED.ANYC9_Pkg_Residencyevent=0x54,period=100003,umask=0x80Counts Core cycles where the core was running with power-delivery for license level 1.  This includes high current AVX 256-bit instructions as well as low current AVX 512-bit instructionsocr.hwpf_l1d_and_swpf.any_responseCounts hardware prefetch data reads (which bring data to L2)  that DRAM supplied the requestevent=0xc1,period=100003,umask=0x7br_misp_retired.cond_ntakencpu_clk_unhalted.distributedCycles after recovery from a branch misprediction or machine clear till the first uop is issued from the resteered pathEstimated number of Top-down Microarchitecture Analysis slots that got dropped due to non front-end reasonsload_hit_prefetch.swpfTMA slots available for an unhalted logical processor. Fixed counter - architectural eventevent=0xc2,cmask=10,inv=1,period=1000003,umask=0x2Counts all requests that miss L2 cacheFlopsCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that resulted in a snoop hit a modified line in another core's caches which forwarded the dataCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that resulted in a snoop that hit in another core, which did not forward the dataocr.reads_to_core.snc_cache.hitmCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that either hit a non-modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeFor every cycle, increments by the number of outstanding code read requests pending.  Code Read requests include both cacheable and non-cacheable Code Reads.   Requests are considered outstanding from the time they miss the core's L2 cache until the transaction completion message is sent to the requestorocr.reads_to_core.l3_miss_localCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by DRAM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those DRAM accesses that are controlled by the close SNC ClusterCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM attached to this socket, whether or not in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts DRAM accesses that are controlled by the close or distant SNC Clusterunc_m_dram_refresh.opportunisticunc_m_rpq_inserts.pch0PMM Commands : All. Unit: uncore_imc event=0xea,umask=0x02TOR Inserts : PCIRdCurs issued by IO Devices. Unit: uncore_cha Four byte data request of the CPU : Card reading from DRAM. Unit: uncore_iio event=0x84,ch_mask=0x20,fc_mask=0x07,umask=0x80unc_iio_txn_req_of_cpu.cmpd.part6event=0xc2,ch_mask=0x02,fc_mask=0x04,umask=0x03unc_i_clockticksevent=0x2d,umask=0x04Counts the number of cycles a core is stalled due to a store buffer being fullCounts the number of load ops retired that hit in DRAM  Supports address when precise (Precise event)Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by the L3 cache where a snoop was sent but the snoop missedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0040Counts L2 cache hardware prefetch data reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedtopdown_fe_bound.icacheevent=0xe6,period=200003,umask=0x4Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were not supplied by the L3 cacheocr.l2wb_m.l3_miss_localThis event is deprecated. Refer to new event MEM_BOUND_STALLS.LOAD_L2_HITCounts miscellaneous requests, such as I/O accesses, that have any type of responseocr.uc_rd.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100184000000cpu_clk_unhalted.ref_tsc_pCounts the number of cycles the integer divider is busy.  Does not imply a stall waiting for the dividerCounts the total number of instructions retired (Precise event)event=0x74,period=1000003topdown_be_bound.non_mem_schedulerTOR Inserts : DRd_Opts issued by iA Cores that hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : WCiLs issued by iA Cores that Missed the LLC. Unit: uncore_cha TOR Inserts : WCiLs issued by iA Cores that Missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : All requests from IO Devices that missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Occupancy : RFOs issued by iA Cores that Missed the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsPCIe Completion Buffer Inserts of completions with data : Part 0 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 4Data requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Data requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 7Data requested of the CPU : Card writing to DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Responses to snoops of any type (code, data, invalidate) that hit M line in the IIO cacheClockticks in the UBOX using a dedicated 48-bit Fixed CounterCounts the number of times the machine was unable to find a translation in the Instruction Translation Lookaside Buffer (ITLB) and a new translation was filled into the ITLB. The event is speculative in nature, but will not count translations (page walks) that are begun and not finished, or translations that are finished but not filled into the ITLB(cycles / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 Fraction of Physical Core issue-slots utilized by this Logical Processor. Unit: cpu_core Actual per-core usage of the Floating Point execution units (regardless of the vector width). Unit: cpu_core UOPS_ISSUED.ANY / cpu_core@UOPS_ISSUED.ANY\,cmask\=1@CLKS_PFraction of cycles spent in Kernel mode. Unit: cpu_atom mem_scheduler_block.ld_bufevent=0xd0,period=1000003,umask=0x5,ldlat=0x4Counts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 512 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom Counts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 64 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom This event is deprecated. Refer to new event OFFCORE_REQUESTS_OUTSTANDING.DATA_RD. Unit: cpu_core Counts all microcode FP assists. Unit: cpu_core Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 64 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core Counts the number of cycles that the head (oldest load) of the load buffer is stalled due to other block cases when load subsequently retires when load subsequently retires. Unit: cpu_atom Counts demand data reads that have any type of response. Unit: cpu_core Counts the number of issue slots every cycle that were not delivered by the frontend due to frontend bandwidth restrictions due to decode, predecode, cisc, and other limitations. Unit: cpu_atom Cycle counts are evenly distributed between active threads in the Core. Unit: cpu_core Core crystal clock cycles. Cycle counts are evenly distributed between active threads in the Core. Unit: cpu_core Number of instructions retired. General Counter - architectural event (Precise event). Unit: cpu_core unc_m_dram_page_hit_wrevent=0x1fincoming write request page status is Page Miss. Unit: uncore_imc unc_m_act_count_totalevent=0x13,period=100003,umask=0x8event=0x11,period=100003,umask=0x10Counts retired load instructions with local Intel Optane DC persistent memory as the data source and the data request missed L3  Supports address when preciseCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that missed the L3 Cache and were supplied by the local socket (DRAM or PMM), whether or not in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts PMM or DRAM accesses that are controlled by the close or distant SNC Cluster.  It does not count misses to the L3 which go to Local CXL Type 2 Memory or Local Non DRAMevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x12380event=0xec,period=2000003,umask=0x20event=0xad,period=1000003,umask=0x20Uops executed on ports 2, 3 and 10This event is deprecated. Refer to new event UOPS_EXECUTED.STALLSevent=0x5,umask=0x00000000f0event=0x2,umask=0x00000000ffevent=0x3,umask=0x0000000022unc_m_pre_count.ufillevent=0x5,umask=0x0000000040DRAM RD_CAS and WR_CAS Commands. : Pseudo Channel 1. Unit: uncore_imc Write request of 4 bytes made by IIO Part1 to Memory. Unit: uncore_iio event=0x83,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000001event=0x83,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000002event=0x84,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000002event=0x84,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000002event=0x35,umask=0x00c80ffe01event=0x35,umask=0x00C896FE01Valid Flits Received : LLCTRL. Unit: uncore_upi unc_upi_txl_basic_hdr_match.ncbevent=0x84,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000004event=0xd5,fc_mask=0x04,umask=0x00000000ffevent=0x19,umask=0x07event=0x58,umask=0x0000000004Read requests made into the CHA. Unit: uncore_cha event=0x53,umask=0x0000000001unc_cha_tor_inserts.loc_iaTOR Inserts : All from Local iA. Unit: uncore_cha unc_cha_tor_inserts.hitunc_cha_tor_inserts.premorph_opcunc_cha_tor_inserts.noncohunc_cha_tor_inserts.io_hit_rfoTOR Inserts;ItoM from Local IA. Unit: uncore_cha unc_cha_tor_occupancy.loc_allunc_cha_tor_occupancy.rem_allunc_cha_tor_occupancy.io_hit_pcirdcurTOR Inserts : CRd issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_remote_wcilf_pmmevent=0x36,umask=0x00cd47ff01TOR Occupancy : DDR Access. Unit: uncore_cha l2_request_g1.rd_blk_levent=0x60,umask=0x04event=0x60,umask=0x02event=0x62,umask=0x01l2_wcb_req.wcb_writeRetired Branch Instructions Mispredictedfpu_pipe_assignment.dual3fp_ret_sse_avx_ops.allevent=0x3,umask=0x80event=0x3,umask=0x40event=0x5,umask=0x08L1 DTLB Reload of a page of 4K sizeTotal Page Table Walks on D-sideic_oc_mode_switch.oc_ic_mode_switchl2_cache_hits_from_ic_missesl2_cache_hits_from_l2_hwpfevent=0xaa,umask=0x03dram_channel_data_controller_0 + dram_channel_data_controller_1 + dram_channel_data_controller_2 + dram_channel_data_controller_3 + dram_channel_data_controller_4 + dram_channel_data_controller_5 + dram_channel_data_controller_6 + dram_channel_data_controller_7ls_hw_pf_dc_fill.ls_mabresp_rmt_cachels_hw_pf_dc_fill.ls_mabresp_lcl_dramHardware Prefetch Data Cache Fills by Data Source. From DRAM (home node local)event=0x28f,umask=0x07L1 DTLB Miss. DTLB reload to a 4K page that missed the L2 TLBls_sw_pf_dc_fills.mem_io_remoteHardware Prefetch Data Cache Fills by Data Source. From DRAM or IO connected in same nodede_dis_dispatch_token_stalls1.store_queue_rsrc_stallde_dis_cops_from_decoder.disp_op_type.any_integer_dispatch + de_dis_cops_from_decoder.disp_op_type.any_fp_dispatchL2 branch prediction overrides existing prediction (speculative)event=0x43,umask=0xffls_any_fills_from_sys.local_ccxls_any_fills_from_sys.remote_cachels_hw_pf_dc_fills.allL2 cache requests of all typesCore to L2 cache requests (not including L2 prefetch) with status: instruction cache hit modifiable line in L2event=0x71,umask=0x04l2_pf_miss_l2_hit_l3.l1_regionL2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L2NextLine (fetch the next line into L2 cache)L3 cache fill requests sourced from DRAM in the same NUMA node. Unit: amd_l3 event=0xd6,umask=0x01Reference cycles (P0 frequency) not in halt event=0x1c2remote_processor_read_data_beats_cs8Read data beats (64 bytes) for remote processor at Coherent Station (CS) 8event=0x29f,umask=0xbfeevent=0x41f,umask=0x7feData beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 5local_socket_inf1_outbound_data_beats_ccm0event=0x51f,umask=0xbffevent=0x55f,umask=0xbffevent=0xc5f,umask=0xf3elocal_socket_outbound_data_beats_link6Retired SSE and AVX floating-point bfloat multiply-accumulate ops (each operation is counted as 2 ops)Retired vector floating-point multiply opsRetired MMX integer pack opsevent=0xb,umask=0x0ffp_pack_ops_retired.fp256_cvtevent=0xc,umask=0xf0Retired 128-bit packed integer multiply opsRetired 128-bit packed integer ops of other typesfp_disp_faults.sse_avx_allumc_data_slot_clks.wrInstruction fetches that hit in the L1 ITLB for 4k or coalesced pages. A coalesced page is a 16k page created from four adjacent 4k pagesbad_speculation_pipeline_restartsAll L2 cache accessesL1 demand data cache fills from local L2 cacheevent=0xc0,umask=0x00event=0x12uncore_sys_ddr_pmuFR_RETIRED_NEAR_RETURNSFR_DECODER_EMPTYFR_FPU_EXCEPTIONSEVENT_1EHEVENT_3EHEVENT_48HEVENT_64HEVENT_74HEVENT_8AHEVENT_B3HEVENT_ECHEVENT_F8HCOHERENT_LINEFILL_MISSCBUS_ACCESS_LDEVENT_108HEVENT_112HEVENT_184HEVENT_18FHEVENT_1AAHEVENT_1D6HEVENT_1D8HEVENT_1DCHEVENT_1F6HEVENT_1FEHEVENT_20EHEVENT_237HEVENT_238HEVENT_246HEVENT_265HEVENT_268HEVENT_26AHEVENT_278HEVENT_290HEVENT_295HEVENT_298HEVENT_2ECHEVENT_2EEHEVENT_331HEVENT_349HEVENT_39EHEVENT_3A6HEVENT_3B2HEVENT_3ECHEXC_SVCL2D_TLB_REFILLhnf_sf_evictionssbsx_rd_axi_trkr_occ_cnt_ovflsbsx_cmo_axi_trkr_occ_cnt_ovflsbsx_arvalid_no_arreadyrnd_rxdat_flitsrnd_txreq_flits_retriedrni_rdb_hybridcxra_req_trk_occcxra_rd_dat_buf_occcxra_dat_pcrd_stalls_lnk1cxla_avg_sz_tx_cxs_dw_beatL1_DATA_TOTAL_HITSSUCCESSFUL_DST_TABLE_SEARCHESLSU_MISALIGN_STALLFP_ONE_QUARTER_FPSCR_RENAMES_BUSYFP_ONE_HALF_FPSCR_RENAMES_BUSYLOAD_INSTR_COMPLETEDBUS_READS_WRITES_NOT_RETRIEDONE_PLUS_INSTR_COMPLETEDSNOOP_HITSDLFB_LOAD_MISS_CYCLESSTASH_BUSY_1L2_COHERENT_LINE_INVALIDATIONSFPU_DENORM_INPUT_CYCLESSOFT-ntapostwrszbyteK8CMN600_PMUINTEL_IVYBRIDGETCfrontend{"type": "thr_create"GenuineIntel-6-4FGenuineIntel-6-4CGenuineIntel-6-2FGenuineIntel-6-55-[01234]This category represents fraction of slots where no uops are being delivered due to a lack of required resources for accepting new uops in the Backend. Backend is the portion of the processor core where the out-of-order scheduler dispatches ready uops into their respective execution units; and once completed these uops get retired according to program order. For example; stalls due to data-cache misses or stalls due to the divider unit being overloaded are both categorized under Backend Bound. Backend Bound is further divided into two main categories: Memory Bound and Core Bound. SMT version; use when SMT is enabled and measuring per logical CPUTotal issue-pipeline slots (per-Physical Core till ICL; per-Logical Processor ICL onward)Branch_Misprediction_CostBranch_Misprediction_Cost_SMTMemory-Level-Parallelism (average number of L1 miss demand load when there is at least one such miss. Per-Logical Processor)Average data fill bandwidth to the L1 data cache [GB / sec]L3_Cache_Fill_BWL2 cache hits per kilo instruction for all demand loads  (including speculative)Page_Walks_Utilization(cstate_pkg@c2\-residency@ / msr@tsc@) * 100(cstate_pkg@c3\-residency@ / msr@tsc@) * 100This event counts the total number of RFO (read for ownership) requests to L2 cache. L2 RFO requests include both L1D demand RFO misses as well as L1D RFO prefetchesAll L2 requestsevent=0xf0,period=200003,umask=0x2longest_lat_cache.referenceRetired load uops with L1 cache hits as data sources. (Precise Event - PEBS)  Supports address when precise (Precise event)event=0xd0,period=100003,umask=0x41mem_uops_retired.split_storesDemand RFO requests including regular RFOs, locks, ItoMThis event counts the number of cases when the offcore requests buffer cannot take more entries for the core. This can happen when the superqueue does not contain eligible entries, or when L1D writeback pending FIFO requests is full.
Note: Writeback pending FIFO has six entriesOffcore outstanding code reads transactions in SuperQueue (SQ), queue to uncore, every cycle  Spec update: BDM76This event counts the number of offcore outstanding Demand Data Read transactions in the super queue (SQ) every cycle. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor. See the corresponding Umask under OFFCORE_REQUESTS.
Note: A prefetch promoted to Demand is counted from the promotion point  Spec update: BDM76event=0xc7,period=2000003,umask=0x10This event counts the number of transitions from AVX-256 to legacy SSE when penalty is applicable  Spec update: BDM30This event counts the number of transitions from legacy SSE to AVX-256 when penalty is applicable  Spec update: BDM30icache.missesCycles when uops are being delivered to Instruction Decode Queue (IDQ) from MITE pathThis event counts the number of deliveries to Instruction Decode Queue (IDQ) initiated by Decode Stream Buffer (DSB) while the Microcode Sequencer (MS) is busy. Counting includes uops that may bypass the IDQevent=0x79,cmask=1,edge=1,period=2000003,umask=0x30idq_uops_not_delivered.coreevent=0xcd,period=101,umask=0x1,ldlat=0x200Number of times an RTM execution aborted due to none of the previous 4 categories (e.g. interrupt)Taken speculative and retired indirect branches excluding calls and returnsThis event counts taken speculative and retired indirect calls including both register and memory indirectcpu_clk_thread_unhalted.ref_xclkevent=0xa3,cmask=12,period=2000003,umask=0xccycle_activity.stalls_l2_pendingThis event counts the number of instructions retired from execution. For instructions that consist of multiple micro-ops, this event counts the retirement of the last micro-op of the instruction. Counting continues during hardware interrupts, traps, and inside interrupt handlers. 
Notes: INST_RETIRED.ANY is counted by a designated fixed counter, leaving the four (eight when Hyperthreading is disabled) programmable counters available for other events. INST_RETIRED.ANY_P is counted by a programmable counter and it is an architectural performance event. 
Counting: Faulting executions of GETSEC/VM entry/VM Exit/MWait will not count as retired instructionsPrecise instruction retired event with HW to reduce effect of PEBS shadow in IP distribution  Spec update: BDM11, BDM55 (Must be precise)event=0xa8,cmask=1,period=2000003,umask=0x1Cycles per thread when uops are executed in port 6uops_executed_port.port_3_coreevent=0xa1,any=1,period=2000003,umask=0x8event=0xa1,any=1,period=2000003,umask=0x40llc_misses.data_readLLC prefetch misses for RFO. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox event=0x35,umask=0x3,filter_opc=0x19e(UNC_M_POWER_CRITICAL_THROTTLE_CYCLES / UNC_M_DCLOCKTICKS) * 100.event=0x2,umask=0x1dtlb_load_misses.walk_completedevent=0x49,period=100003,umask=0x20event=0x4f,period=2000003,umask=0x10event=0xbc,period=2000003,umask=0x12This event counts retired load uops which data sources were load uops missed L1 but hit a fill buffer due to a preceding miss to the same cache line with the data not ready.
Note: Only two data-sources of L1/FB are applicable for AVX-256bit  even though the corresponding AVX load could be serviced by a deeper level in the memory hierarchy. Data source is reported for the Low-half load  Supports address when precise (Precise event)This event counts retired load uops which data sources were misses in the mid-level (L2) cache. Counting excludes unknown and UC data source  Supports address when precise (Precise event)Retired load uops that split across a cacheline boundary  Supports address when precise (Precise event)event=0xb0,period=100003,umask=0x80event=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0091offcore_response.all_pf_data_rd.supplier_none.snoop_not_neededoffcore_response.demand_code_rd.any_responseoffcore_response.demand_rfo.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80028000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020010offcore_response.pf_l3_code_rd.supplier_none.any_snoopoffcore_response.pf_l3_data_rd.l3_hit.snoop_hit_no_fwdoffcore_response.all_data_rd.l3_miss_local_dram.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0002offcore_response.pf_l2_code_rd.l3_miss_local_dram.any_snoopoffcore_response.pf_l3_data_rd.l3_miss.snoop_hit_no_fwdDirect and indirect macro near call instructions retired (captured in ring 3) (Precise event)This event counts return instructions retired (Precise event)unc_cbo_cache_lookup.read_munc_cbo_cache_lookup.write_mL3 Lookup write request that access cache and found line in MESI-stateNumber of Core coherent Data Read entries allocated in DirectData modeoffcore_response.pf_llc_rfo.llc_hit.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x087FC007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00002L1 Data line replacementsCycles the L2 cache data bus is busyL2 cache missesl2_rqsts.self.demand.mesisimd_comp_inst_retired.scalar_singleevent=0xca,period=2000000,umask=0x2simd_sat_uop_exec.arsimd_uop_type_exec.arithmetic.arevent=0xb3,period=2000000,umask=0xa0simd_uop_type_exec.unpack.sevent=0xaa,period=2000000,umask=0x3misalign_mem_ref.rmw_bubbleBus cycles when a LOCK signal is assertedbus_trans_any.all_agentsevent=0x65,period=200000,umask=0x40bus_trans_ifetch.all_agentsevent=0x6c,period=200000,umask=0xe0event=0x67,period=200000,umask=0xe0thermal_tripbr_inst_decodedbr_inst_retired.anybr_inst_type_retired.condMispredicted cond branch instructions retiredMispredicted indirect calls, including both register and memory indirectevent=0x12,period=2000000,umask=0x1reissue.any.aritlb.hitevent=0xc,period=2000000,umask=0x2event=0x2e,period=200003,umask=0x4fCounts memory load uops retired where the data is retrieved from DRAM.  Event is counted at retirement, so the speculative loads are ignored.  A memory load can hit (or miss) the L1 cache, hit (or miss) the L2 cache, hit DRAM, hit in the WCB or receive a HITM response  Supports address when precise (Must be precise)Store uops retired (Precise event capable)  Supports address when precise (Must be precise)Counts requests to the uncore subsystem that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredoffcore_response.corewb.l2_miss.hit_other_core_no_fwdCounts the number of writeback transactions caused by L1 or L2 cache evictions that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040001offcore_response.demand_rfo.l2_miss.anyevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000800offcore_response.pf_l1_data_rd.l2_miss.hitm_other_coreCounts data cache line reads generated by hardware L1 data cache prefetcher that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000010References per ICache line. This event counts differently than Intel processors based on Silvermont microarchitectureReferences per ICache line that are available in the ICache (hit). This event counts differently than Intel processors based on Silvermont microarchitecturemisalign_mem_ref.store_page_splitevent=0x3,period=200003,umask=0x2MS uops retired (Precise event capable) (Must be precise)mem_uops_retired.dtlb_missCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) have any transaction responses from the uncore subsystemevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000004800Instructions retired - using Reduced Skid PEBS feature (Must be precise)Page walk completed due to a demand data store to a 4K pageCounts page walks completed due to instruction fetches whose address translations missed in the TLB and were mapped to 2M or 4M pages.  The page walks can end with or without a page faultCounts the number of demand Data Read requests, initiated by load instructions, that hit L2 cache  Spec update: HSD78, HSM80Cycles when offcore outstanding Demand Data Read transactions are present in SuperQueue (SQ), queue to uncore  Spec update: HSD78, HSD62, HSD61, HSM63, HSM80offcore_response.demand_code_rd.l3_hit.hitm_other_coreApproximate counts of AVX & AVX2 256-bit instructions, including non-arithmetic instructions, loads, and stores.  May count non-AVX instructions that employ 256-bit operations, including (but not necessarily limited to) rep string instructions that use 256-bit loads and stores for optimized performance, XSAVE* and XRSTOR*, and operations that transition the x87 FPU data registers between x87 and MMXSpeculative cache-line split load uops dispatched to L1DNumber of times an RTM execution aborted due to incompatible memory type  Spec update: HSD65Number of times a HLE transactional region aborted due to a non XRELEASE prefixed instruction writing to an elided lock in the elision bufferUnhalted core cycles when the thread is not in ring 0This event counts all mispredicted branch instructions retired. This is a precise event (Must be precise)Number of loads missed L2  Spec update: HSM63, HSM80Non-SW-prefetch load dispatches that hit fill buffer allocated for S/W prefetchCount cases of saving new LBR records by hardwareevent=0x34,umask=0x48ITLB misses that hit STLB. No page walkevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0600400004Counts the number of lines brought into the L1 data cacheDirty L2 cache lines evicted by the MLC prefetcherRequests from the L2 hardware prefetchers that hit L2 cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10008Loads with latency value being above 4 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x6004001b3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400001Execution stalls while L2 cache miss load* is outstandingCycles per thread when uops are dispatched to port 1Number of slow LEA or similar uops allocated. Such uop has 3 sources (e.g. 2 sources + immediate) regardless if as a result of LEA instruction or notunc_arb_trk_occupancy.cycles_over_half_fullevent=0x80,cmask=10,umask=0x01event=0x81,umask=0x80A snoop invalidates a non-modified line in some processor coreLLC lookup request that access cache and found line in I-state. Unit: uncore_cbox unc_cbo_cache_lookup.extsnp_filterunc_cbo_cache_lookup.any_request_filterevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c03f7Counts demand data reads that hit in the LLC and sibling core snoop returned a clean responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0080Counts all demand code reads that miss the LLC  and the data forwarded from remote cacheoffcore_response.demand_data_rd.llc_miss.any_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x87f820001event=0x35,umask=0x1,filter_opc=0x19cunc_p_freq_max_current_cyclesCounts the number of cycles that the uncore was running at a frequency greater than or equal to 2Ghz. Derived from unc_p_freq_band1_cycles. Unit: uncore_pcu dtlb_load_misses.demand_ld_walk_durationpartial_rat_stalls.slow_lea_windowCounts the number of MEC requests that were not accepted into the L2Q because of any L2  queue reject condition. There is no concept of at-ret here. It might include requests due to instructions in the speculative pathoffcore_response.any_code_rd.l2_hit_this_tile_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000044offcore_response.any_data_rd.l2_hit_this_tile_moffcore_response.any_pf_l2.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x08000832f7Counts any Read request  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x40000032f7offcore_response.any_request.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000408000offcore_response.any_request.l2_hit_near_tile_e_foffcore_response.any_request.l2_hit_this_tile_eoffcore_response.any_request.l2_hit_this_tile_soffcore_response.demand_data_rd.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400002event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010080Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses which hit its own tile's L2 with data in S stateoffcore_response.partial_writes.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180100offcore_response.partial_writes.l2_hit_this_tile_eevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000100Counts L1 data HW prefetches that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts L2 code HW prefetches that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180040event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000401000offcore_response.pf_software.l2_hit_this_tile_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080200event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100408000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800022event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800022offcore_response.bus_locks.ddr_farCounts demand code reads and prefetch code reads that accounts for data responses from DRAM Faroffcore_response.demand_rfo.mcdram_nearevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000080Counts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for data responses from DRAM Farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181801000Fixed Counter: Counts the number of unhalted core clock cyclesCounts all nukesevent=0x63,period=2000000,umask=0x1event=0x41,period=2000000,umask=0x4event=0x4e,period=200000,umask=0x1L1 writebacks to L2 in I state (misses)All L2 data requestsevent=0x26,period=200000,umask=0x10event=0x24,period=200000,umask=0x10Memory instructions retired above 1024 clocks (Precise Event)event=0xcb,period=200000,umask=0x40mem_load_retired.l1d_hitevent=0xf,period=40000,umask=0x2All offcore data readsoffcore_response.any_data.io_csr_mmioOffcore code reads satisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3844Offcore requests satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x38FFevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1008offcore_response.data_ifetch.remote_cacheoffcore_response.demand_data.io_csr_mmioOffcore demand data requests satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F01All offcore demand code readsoffcore_response.demand_ifetch.llc_hit_no_other_coreOffcore demand RFO requests satisfied by a remote cache or remote DRAMOffcore other requests satisfied by the LLC and HIT in a sibling coreOffcore prefetch data requests that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1040event=0x10,period=2000000,umask=0x4Offcore prefetch data reads satisfied by any DRAMio_transactionsload_dispatch.rscpu_clk_unhalted.total_cyclesInstructions written to instruction queueCycles instructions are written to the instruction queueuops_decoded.stall_cyclesevent=0xb1,any=1,cmask=1,edge=1,inv=1,period=2000000,umask=0x1fevent=0xb1,any=1,period=2000000,umask=0x4event=0xc8,period=200000,umask=0x20Retired stores that miss the DTLB (Precise Event)event=0xf2,period=200003,umask=0x4Counts L2 writebacks that access L2 cacheNumber of retired load instructions that (start a) miss in the 2nd-level TLB (STLB)  Supports address when precise (Precise event)Retired load instructions missed L1 cache as data sources  Supports address when precise (Precise event)Counts retired load instructions with at least one uop that missed in the L1 cache  Supports address when precise (Precise event)Counts the number of offcore outstanding cacheable Core Data Read transactions in the super queue every cycle. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation). See corresponding Umask under OFFCORE_REQUESTSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x401C0004offcore_response.demand_code_rd.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2001C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0100001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2001C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40040002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400108000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200108000Number of PREFETCHW instructions executedfrontend_retired.latency_ge_2_bubbles_ge_2event=0xc6,period=100007,umask=0x1,frontend=0x402006icache_64b.iftag_hithle_retired.aborted_memCounts randomly selected loads when the latency from first dispatch to completion is greater than 128 cycles  Supports address when precise (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000080004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000002offcore_response.other.l3_miss.any_snoopNumber of times an RTM execution aborted due to uncommon conditionsCounts cycles during which a total of 1 uop was executed on all ports and Reservation Station (RS) was not emptyCycles total of 4 uops are executed on all ports and Reservation Station (RS) was not emptyPrecise instruction retired event with HW to reduce effect of PEBS shadow in IP distribution  Spec update: SKL091, SKL044 (Must be precise)Core cycles the Resource allocator was stalled due to recovery from an earlier branch misprediction or machine clear eventCounts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 7INST_RETIRED.ANY / MEM_INST_RETIRED.ALL_STORESFraction of branches that are taken conditionalsCallRetFB_HPKICounts completed page walks  (4K sizes) caused by demand data stores. This implies address translations missed in the DTLB and further levels of TLB. The page walk can end with or without a faultCounts cycles that fetch is stalled due to an outstanding ICache miss. That is, the decoder queue is able to accept bytes, but the fetch unit is unable to provide bytes due to an ICache miss.  Note: this event is not the same as the total number of cycles spent retrieving instruction cache lines from the memory hierarchy.
Counts cycles that fetch is stalled due to any reason. That is, the decoder queue is able to accept bytes, but the fetch unit is unable to provide bytes.  This will include cycles due to an ITLB miss, ICache miss and other eventsrehabq.any_ldThe NO_ALLOC_CYCLES.NOT_DELIVERED event is used to measure front-end inefficiencies, i.e. when front-end of the machine is not delivering micro-ops to the back-end and the back-end is not stalled. This event can be used to identify if the machine is truly front-end bound.  When this event occurs, it is an indication that the front-end of the machine is operating at less than its theoretical peak performance.  Background: We can think of the processor pipeline as being divided into 2 broader parts: Front-end and Back-end. Front-end is responsible for fetching the instruction, decoding into micro-ops (uops) in machine understandable format and putting them into a micro-op queue to be consumed by back end. The back-end then takes these micro-ops, allocates the required resources.  When all resources are ready, micro-ops are executed. If the back-end is not ready to accept micro-ops from the front-end, then we do not want to count these as front-end bottlenecks.  However, whenever we have bottlenecks in the back-end, we will have allocation unit stalls and eventually forcing the front-end to wait until the back-end is ready to receive more UOPS. This event counts the cycles only when back-end is requesting more uops and front-end is not able to provide them. Some examples of conditions that cause front-end efficiencies are: Icache misses, ITLB misses, and decoder restrictions that limit the front-end bandwidthThis event counts every cycle when a D-side (walks due to a load) page walk is in progress. Page walk duration divided by number of page walks is the average duration of page-walksRetired load uops with locked access. (Precise Event - PEBS) (Precise event)Counts data/code/rfo reads (demand & prefetch) that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts demand code reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts prefetch (that bring data to L2) RFOs that hit in the LLC and the snoops sent to sibling cores return clean responseoffcore_response.pf_llc_rfo.llc_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10200offcore_response.pf_l2_data_rd.llc_miss.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400010This event counts loads that followed a store to the same address, where the data could not be forwarded inside the pipeline from the store to the load.  The most common reason why store forwarding would be blocked is when a load's address range overlaps with a preceeding smaller uncompleted store.  See the table of not supported store forwards in the Intel 64 and IA-32 Architectures Optimization Reference Manual.  The penalty for blocked store forwarding is that the load must wait for the store to complete before it can be issuedCycles offcore demand code read busyoffcore_requests_outstanding.demand.rfo_not_emptyREQUEST = DATA_IN and RESPONSE = REMOTE_CACHE_HITMREQUEST = DEMAND_IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f02offcore_response.demand_rfo.local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8050REQUEST = PF_DATA and RESPONSE = LLC_HIT_OTHER_CORE_HITMoffcore_response.pf_data_rd.local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5020offcore_response.prefetch.all_local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf810REQUEST = PF_RFO and RESPONSE = ANY_LLC_MISSevent=0xb4,period=100000,umask=0x4DTLB miss large page walksevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2733mem_uncore_retired.remote_hitmevent=0xb1,cmask=1,edge=1,inv=1,period=2000000,umask=0x3fRetired load instructions whose data sources was remote HITM  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0002OFFCORE_RESPONSE.PF_L1D_AND_SW.L3_HIT.SNOOP_HIT_WITH_FWDevent=0xc7,period=2000003,umask=0x40Counts number of SSE/AVX computational 512-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT14 RCP14 FM(N)ADD/SUB. FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementoffcore_response.all_pf_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdoffcore_response.all_rfo.l3_miss.snoop_miss_or_no_fwdCounts all demand code reads that miss the L3 and the modified data is transferred from remote cacheCounts prefetch (that bring data to L2) data reads that miss the L3 and clean or shared data is transferred from remote cachePower_SMTPower_License2_Utilization_SMTevent=0x4,umask=0x1event=0x37,umask=0x08event=0x50,umask=0x10Counts every write request of 4 bytes of data made to the MMIO space of a card on IIO Part1 by a unit on the main die (generally a core) or by another IIO unit. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busCounts ever peer to peer read request for 4 bytes of data made by a different IIO unit to the MMIO space of a card on IIO Part2. Does not include requests made by the same IIO unit. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busevent=0xc0,ch_mask=0x08,fc_mask=0x07,umask=0x08Peer to peer write request of 4 bytes made to IIO Part2 by a different IIO unit. Unit: uncore_iio unc_iio_txn_req_by_cpu.peer_read.part1unc_i_cache_total_occupancy.memTotal IRP occupancy of inbound read and write requests. Unit: uncore_irp event=0x24Counts cycles when direct to core mode (which bypasses the CHA) was disabledunc_m2m_directory_update.s2aevent=0x38,umask=0x10BL Ingress (from CMS) Occupancy. Unit: uncore_m2m event=0x3,umask=0x0Fevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080490ocr.all_pf_data_rd.l3_hit_f.snoop_noneOCR.ALL_PF_RFO.L3_HIT.HITM_OTHER_CORE OCR.ALL_PF_RFO.L3_HIT.HITM_OTHER_CORE OCR.ALL_PF_RFO.L3_HIT.HITM_OTHER_COREOCR.ALL_PF_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.all_pf_rfo.l3_hit_m.snoop_noneOCR.ALL_PF_RFO.L3_HIT_M.SNOOP_NONEOCR.ALL_READS.L3_HIT.ANY_SNOOP OCR.ALL_READS.L3_HIT.ANY_SNOOP OCR.ALL_READS.L3_HIT.ANY_SNOOPOCR.ALL_READS.L3_HIT_M.HIT_OTHER_CORE_FWD  OCR.ALL_READS.L3_HIT_M.HIT_OTHER_CORE_FWDOCR.ALL_READS.L3_HIT_S.SNOOP_NONEocr.all_rfo.l3_hit.no_snoop_neededOCR.ALL_RFO.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_RFO.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_RFO.L3_HIT.NO_SNOOP_NEEDEDOCR.ALL_RFO.L3_HIT_E.SNOOP_MISSocr.all_rfo.l3_hit_s.hit_other_core_fwdOCR.ALL_RFO.L3_HIT_S.NO_SNOOP_NEEDED  OCR.ALL_RFO.L3_HIT_S.NO_SNOOP_NEEDEDCounts all demand code reads OCR.DEMAND_CODE_RD.L3_HIT.ANY_SNOOP OCR.DEMAND_CODE_RD.L3_HIT.ANY_SNOOPCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_E.HIT_OTHER_CORE_FWDocr.demand_code_rd.l3_hit_e.snoop_missocr.demand_code_rd.l3_hit_f.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200004Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_F.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100002Counts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_M.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100010ocr.pf_l2_rfo.l3_hit_e.any_snoopocr.pf_l2_rfo.l3_hit_s.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0080ocr.pf_l3_data_rd.l3_hit_e.no_snoop_neededocr.pf_l3_data_rd.l3_hit_f.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100080Counts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_HIT.ANY_SNOOP OCR.PF_L3_RFO.L3_HIT.ANY_SNOOPocr.pf_l3_rfo.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.SNOOP_NONEoffcore_response.all_data_rd.l3_hit_e.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400491This event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDoffcore_response.all_pf_data_rd.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDoffcore_response.all_pf_data_rd.supplier_none.hit_other_core_fwdoffcore_response.all_pf_rfo.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.all_reads.supplier_none.hit_other_core_no_fwdoffcore_response.all_rfo.l3_hit_f.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_F.NO_SNOOP_NEEDEDoffcore_response.all_rfo.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400004This event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_S.HITM_OTHER_COREoffcore_response.demand_rfo.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_E.SNOOP_MISSoffcore_response.other.l3_hit_f.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.SNOOP_NONEoffcore_response.pf_l1d_and_sw.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020400This event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_E.SNOOP_MISSoffcore_response.pf_l3_rfo.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_F.NO_SNOOP_NEEDEDAverage latency of data read request to external 3D X-Point memory [in nanoseconds]. Accounts for demand loads and L1/L2 data-read prefetchesocr.all_data_rd.l3_miss.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000491ocr.all_pf_data_rd.l3_miss.any_snoopocr.all_pf_data_rd.l3_miss.hit_other_core_fwdOCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000490ocr.all_pf_data_rd.l3_miss_local_dram.hit_other_core_fwdOCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORE  OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.all_pf_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F840007F7ocr.all_reads.l3_miss_remote_dram.snoop_miss_or_no_fwdocr.all_rfo.l3_miss.remote_hitmocr.all_rfo.l3_miss_local_dram.hit_other_core_fwdocr.demand_code_rd.l3_miss.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000004ocr.demand_code_rd.l3_miss_remote_hop1_dram.snoop_missocr.demand_code_rd.l3_miss_remote_hop1_dram.snoop_noneCounts any other requests OCR.OTHER.L3_MISS.ANY_SNOOP OCR.OTHER.L3_MISS.ANY_SNOOPocr.other.l3_miss.snoop_noneocr.other.l3_miss_remote_hop1_dram.hit_other_core_fwdocr.pf_l1d_and_sw.l3_miss.any_snoopocr.pf_l1d_and_sw.l3_miss.remote_hit_forwardCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.ANY_SNOOPCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.pf_l2_rfo.l3_miss_remote_hop1_dram.hitm_other_coreCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.NO_SNOOP_NEEDED OCR.PF_L3_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDocr.pf_l3_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.HITM_OTHER_COREoffcore_response.all_reads.l3_miss.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDoffcore_response.all_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPoffcore_response.demand_data_rd.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSoffcore_response.pf_l2_data_rd.l3_miss.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.pf_l2_rfo.l3_miss.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_data_rd.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPoffcore_response.pf_l3_rfo.l3_miss_local_dram.hitm_other_coreOCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONE OCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEOCR.ALL_PF_RFO.ANY_RESPONSE have any response typeOCR.ALL_PF_RFO.SUPPLIER_NONE.ANY_SNOOP  OCR.ALL_PF_RFO.SUPPLIER_NONE.ANY_SNOOPCounts all demand code reads OCR.DEMAND_CODE_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.demand_data_rd.supplier_none.snoop_noneocr.demand_rfo.supplier_none.hitm_other_coreCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDocr.pf_l1d_and_sw.supplier_none.hit_other_core_no_fwdocr.pf_l3_rfo.pmm_hit_local_pmm.snoop_noneIntel Optane DC persistent memory bandwidth write (MB/sec). Derived from unc_m_pmm_wpq_inserts. Unit: uncore_imc UNC_M_PMM_BANDWIDTH.TOTALunc_m_tagchk.miss_cleanevent=0x2c,umask=0x08Counts Software prefetch requests that miss the L2 cache. Accounts for PREFETCHNTA and PREFETCHT0/1/2 instructions when FB is not fullCounts core-originated cacheable requests that miss the L3 cache (Longest Latency cache). Requests include data and code reads, Reads-for-Ownership (RFOs), speculative accesses and hardware prefetches to the L1 and L2.  It does not include hardware prefetches to the L3, and may not count other types of requests to the L3event=0xd1,period=100021,umask=0x4ocr.demand_data_rd.l3_hit.anyocr.demand_data_rd.l3_hit.snoop_hitmCounts demand data reads that hit a cacheline in the L3 where a snoop was not needed to satisfy the requestCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit a cacheline in the L3 where a snoop hit in another core, data forwarding is not requiredCounts memory transactions sent to the uncoreCounts memory transactions sent to the uncore including requests initiated by the core, all L3 prefetches, reads resulting from page walks, and snoop responsesCounts instruction fetch tag lookups that miss in the instruction cache (L1I). Counts at 64-byte cache-line granularity. Accounts for both cacheable and uncacheable accessesCounts the number of cycles where optimal number of uops was delivered to the Instruction Decode Queue (IDQ) from the MITE (legacy decode pipeline) path. During these cycles uops are not being delivered from the Decode Stream Buffer (DSB)event=0xa3,cmask=2,period=1000003,umask=0x2Counts Unfriendly TSX abort triggered by a vzeroupper instructionCounts demand data reads that DRAM supplied the requestCounts cycles when divide unit is busy executing divide or square root operations. Accounts for integer and floating-point operationsCounts taken conditional branch instructions retired (Precise event)Indirect near branch instructions retired (excluding returns) (Precise event)All miss-predicted indirect branch instructions retired (excluding RETs. TSX aborts is considered indirect branch) (Precise event)event=0xec,period=2000003,umask=0x2lsd.cycles_oktopdown.backend_bound_slotsNumber of uops executed on port 0uops_dispatched.port_4_9Number of page walks outstanding for a demand load in the PMH each cyclemem_load_l3_hit_retired.xsnp_fwdCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that resulted in a snoop hit a modified line in another core's caches which forwarded the dataCounts demand data reads that were not supplied by the local socket's L1, L2, or L3 cachesocr.hwpf_l3.l3_miss_localCounts demand data reads that were supplied by DRAM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those DRAM accesses that are controlled by the close SNC ClusterCounts hardware prefetch (which bring data to L2) that have any type of responseocr.reads_to_core.remote_dramocr.reads_to_core.snc_pmmevent=0x10,umask=0x01event=0x35,umask=0xC80FFD01unc_cha_tor_inserts.ia_miss_crdTOR Inserts : All requests from IO Devices that hit the LLC. Unit: uncore_cha TOR Occupancy : All requests from IO Devices that hit the LLC. Unit: uncore_cha TOR Occupancy : All requests from IO Devices that missed the LLC. Unit: uncore_cha event=0x35,umask=0xC897FD01event=0x35,umask=0xCCC7FF01TOR Inserts : SpecItoMs issued by iA Cores. Unit: uncore_cha event=0x35,umask=0xC8168601unc_cha_llc_lookup.data_readClockticks of the integrated IO (IIO) traffic controller. Unit: uncore_iio event=0x83,ch_mask=0x40,fc_mask=0x07,umask=0x80event=0xc1,ch_mask=0x10,fc_mask=0x07,umask=0x01event=0xc1,ch_mask=0x40,fc_mask=0x07,umask=0x04event=0x84,ch_mask=0x80,fc_mask=0x07,umask=0x01event=0x84,ch_mask=0x20,fc_mask=0x07,umask=0x04unc_iio_comp_buf_inserts.cmpd.part4PCIe Completion Buffer Occupancy of completions with data : Part 1. Unit: uncore_iio unc_cha_tor_inserts.io_itom_remoteCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by the L3 cacheCounts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by the L3 cacheocr.l2wb_m.l3_hitocr.reads_to_core.l3_hit.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x101F803C0000Counts the number of misaligned load uops that are 4K page splits (Precise event)ocr.corewb_m.l3_missCounts miscellaneous requests, such as I/O accesses, that were not supplied by the L3 cacheCounts streaming stores which modify only part of a 64 byte cacheline that were not supplied by the L3 cacheCounts the number of unhalted cycles a core is blocked due to an accepted lock issued by other cores. Counts on a per core basisCounts all code reads that have any type of responseocr.hwpf_l2_code_rd.outstandingCounts the number of taken JCC (Jump on Conditional Code) branch instructions retired (Precise event)topdown_fe_bound.branch_resteerCounts the number of issue slots every cycle that were not delivered by the frontend due to a latency related stalls including BACLEARs, BTCLEARs, ITLB misses, and ICache missesTOR Inserts : CLFlushes issued by iA Cores : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : CRds issued by iA Cores that Hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x35,umask=0xC8A7FD01TOR Inserts : CRds issued by iA Cores that Missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : RFOs issued by iA Cores that Missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsunc_cha_tor_inserts.ia_miss_wilPCIe Completion Buffer Occupancy : Part 6 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 6Data requested of the CPU : Card reading from DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Counts the number of page walks completed due to load DTLB misses to a 2M or 4M pageept.epde_missCounts the number of cycles a core is stalled due to an instruction cache or Translation Lookaside Buffer (TLB) access which hit in the L2 cacheInstructions per (near) call (lower number means higher occurrence rate). Unit: cpu_core Instructions per FP Arithmetic Scalar Single-Precision instruction (lower number means higher occurrence rate). Unit: cpu_core L2 cache hits per kilo instruction for all demand loads  (including speculative). Unit: cpu_core event=0xd0,period=1000003,umask=0x5,ldlat=0x8Number of cycles a demand request has waited due to L1D due to lack of L2 resources. Unit: cpu_core L2 cache hits when fetching instructions, code reads. Unit: cpu_core Retired load instructions with L2 cache hits as data sources  Supports address when precise (Precise event). Unit: cpu_core event=0x21,period=100003,umask=0x8event=0x40,period=100003,umask=0x4event=0xc6,period=100007,umask=0x1,frontend=0x601006Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 32 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 4 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core machine_clears.slowNumber of near branch instructions retired that were mispredicted and taken (Precise event). Unit: cpu_core Reference cycles when the core is not in halt state. Unit: cpu_core inst_retired.macro_fusedNumber of Uops delivered by the LSD. Unit: cpu_core event=0xe0,period=400009,umask=0x20event=0xb2,period=2000003,umask=0x4incoming read request page status is Page Miss. Unit: uncore_imc Code miss in all TLB levels causes a page walk that completes. (2M/4M). Unit: cpu_core Code miss in all TLB levels causes a page walk that completes. (4K). Unit: cpu_core Completed demand load uops that miss the L1 d-cacheCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that hit in the L3 or were snooped from another core's caches on the same socketCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that resulted in a snoop hit a modified line in another core's caches which forwarded the dataFP_ARITH_INST_RETIRED2.SCALAR_HALFevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x730000001event=0x2a,period=100003,umask=0x1,offcore_rsp=0x73C004477Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM attached to this socket, whether or not in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts DRAM accesses that are controlled by the close or distant SNC ClusterCore clocks when the thread is in the C0.1 light-weight slower wakeup time but more power saving optimized stateCounts core clocks when the thread is in the C0.1 light-weight slower wakeup time but more power saving optimized state.  This state can be entered via the TPAUSE or UMWAIT instructionsNumber of slots of TMA method that were wasted due to incorrect speculation. It covers all types of control-flow or data-related mis-speculationsAll DRAM read CAS commands issued (does not include underfills). Unit: uncore_imc event=0x5,umask=0x00000000c2event=0x83,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000001event=0x83,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000002event=0x83,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000002unc_iio_txn_req_of_cpu.peer_write.part5TOR Inserts for ItoMCacheNears from IO devices. Unit: uncore_cha Valid Flits Sent : Protocol Header. Unit: uncore_upi unc_upi_txl_basic_hdr_match.ncb_opcevent=0x83,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000004event=0x84,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000001event=0x84,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000001event=0x21,umask=0x0310unc_m2m_prefcam_demand_drops.ch0_xptevent=0x55,umask=0x0000000001OSB Snoop Broadcast : Local InvItoE. Unit: uncore_cha TOR Inserts : WBQ. Unit: uncore_cha unc_cha_tor_inserts.local_tgtTOR Inserts; RFO from local IA. Unit: uncore_cha event=0x35,umask=0x00c897ff01unc_cha_tor_inserts.ia_clflushoptTOR Occupancy : SF/LLC Evictions. Unit: uncore_cha TOR Occupancy; ITOM hits from local IO. Unit: uncore_cha event=0x36,umask=0x00c827ff01unc_cha_tor_inserts.ia_miss_drd_pref_pmmunc_cha_tor_inserts.ia_miss_drd_pref_local_ddrunc_cha_tor_inserts.ia_miss_drd_pref_remote_ddrevent=0x35,umask=0x00cc27ff01event=0x35,umask=0x00C87FDE01unc_cha_tor_inserts.ia_miss_wcilf_pmmTOR Inserts : WCiLFs issued by iA Cores targeting PMM that missed the LLC. Unit: uncore_cha event=0x36,umask=0x00c80f7e01unc_cha_tor_occupancy.ia_itomcachenearTOR Occupancy : WCiLs issued by iA Cores. Unit: uncore_cha unc_cha_tor_occupancy.pmmunc_cha_tor_inserts.ddrPCU PCLK Clockticks. Unit: uncore_pcu The number of 32B fetch windows tried to read the L1 IC and missed in the full tagAll L2 Cache Requests (Breakdown 1 - Common). Data cache shared readsl2_request_g2.smc_invalLS to L2 WCB write requests. LS (Load/Store unit) to L2 WCB (Write Combining Buffer) write requestsevent=0x63,umask=0x04Core to L2 cacheable request access status (not including L2 Prefetch). Data cache shared read hit in L2Core to L2 cacheable request access status (not including L2 Prefetch). Data cache read hit in L2event=0xc9remote_outbound_data_controller_2event=0x87,umask=0x38Total number multi-pipe uOps assigned to pipe 2fpu_pipe_assignment.dual1The number of operations (uOps) and dual-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number uOps assigned to all pipesfpu_pipe_assignment.total3event=0x2,umask=0x07fp_ret_sse_avx_ops.dp_mult_flopsls_l1_d_tlb_miss.tlb_reload_4k_l2_hitbranch_predictionL3 Accesses. Unit: amd_l3 All TLBs Flushedls_bad_status2.stli_otherevent=0x37event=0x43,umask=0x10event=0xae,umask=0x04ex_ret_msprd_brnch_instr_dir_msmtchRetired Mispredicted Branch Instructions due to Direction Mismatchls_dmnd_fills_from_sys.ext_cache_remotels_sw_pf_dc_fills.ext_cache_localde_dis_dispatch_token_stalls1.load_queue_rsrc_stallls_any_fills_from_sys.local_allls_any_fills_from_sys.far_allAny data cache fills from extension memoryls_sw_pf_dc_fills.local_l2Hardware prefetch data cache fills from either DRAM or MMIO in a different NUMA node (same or different socket)Core to L2 cache requests (not including L2 prefetch) with status: instruction cache request miss in L2Core to L2 cache requests (not including L2 prefetch) for instruction cache hitsCore to L2 cache requests (not including L2 prefetch) for data and instruction cache missesevent=0x64,umask=0xffL2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L2Stream (fetch additional sequential lines into L2 cache)l2_pf_miss_l2_hit_l3.l2_next_lineevent=0x71,umask=0x20l2_pf_miss_l2_l3.l2_burstlocal_processor_read_data_beats_cs10event=0x9f,umask=0x7ffRead data beats (64 bytes) for remote processor at Coherent Station (CS) 2Read data beats (64 bytes) for remote processor at Coherent Station (CS) 4Write data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 1local_socket_inf1_inbound_data_beats_ccm1Data beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 5remote_socket_inf0_inbound_data_beats_ccm4remote_socket_inf1_inbound_data_beats_ccm4remote_socket_inf0_outbound_data_beats_ccm7local_socket_outbound_data_beats_link7Retired x87 floating-point add and subtract opsevent=0xa,umask=0x06event=0xa,umask=0x07event=0xa,umask=0x09fp_ops_retired_by_type.scalar_otherRetired scalar floating-point ops of other typesfp_ops_retired_by_type.scalar_allRetired vector floating-point square root opsRetired MMX integer MOV opssse_avx_ops_retired.sse_avx_shufflesse_avx_ops_retired.sse_avx_logicalevent=0xc,umask=0xffevent=0xd,umask=0x02Retired 128-bit packed integer subtract opsevent=0xd,umask=0x90packed_int_op_type.int256_movevent=0xd,umask=0xf0Floating-point dispatch faults for XMM fillsmemory controllerL1 DTLB misses with L2 DTLB misses (page-table walks are requested) for 2M pagesde_dis_ops_from_decoder.disp_op_type.any_integer_dispatchd_ratio((6 * cpu@de_no_dispatch_per_slot.no_ops_from_frontend\,cmask\=0x6@), total_dispatch_slots)L2 cache accesses from L1 instruction cache misses (including prefetch)dram_write_data_for_local_processorLocal socket upstream DMA read datacyclesFR_RETIRED_BRANCHES_MISPREDICTEDEVENT_0CHEVENT_2DHEVENT_3DHEVENT_56HEVENT_77HEVENT_8CHEVENT_A9HEVENT_B1HEVENT_C3HEVENT_C5HEVENT_D3HEVENT_F1HWRITE_STALLL2D_CACHEBR_RETURN_RETIREDUNALIGNED_LDST_RETIREDEVENT_114HEVENT_12BHEVENT_135HEVENT_163HEVENT_164HEVENT_172HEVENT_1AFHEVENT_1BEHEVENT_224HEVENT_24EHEVENT_253HEVENT_2D2HEVENT_2E5HEVENT_2EBHEVENT_2FEHEVENT_31DHEVENT_323HEVENT_330HEVENT_342HEVENT_3C6HEVENT_3F8HL1D_CACHE_STST_SPECEXC_UNDEFhni_wvalid_no_wreadycxla_tx_cxs_link0TLBIE_SNOOPSL1_INSTR_CACHE_RELOADSGPR_ISSUE_QUEUE_STALLEDMISPREDICTED_BRANCHESL3_INSTR_CACHE_MISSESRUN_CYCLESCYCLESBUS_HIGHBRANCH_INSTRS_COMPLETEDSTORE_UOPS_COMPLETEDCYCLES_LRU_SCHED_STALLEDCRIT_INPUT_INTR_TAKENL2_CACHE_DATA_DIRTY_HITSFPU_FINISHedgecycles-to-completeEDGETAGGINGCASCADEINTEL_SKYLAKEl3_slice_mask{"type": "proc_create"GenuineIntel-6-6AThe ratio of Executed- by Issued-UopsCoreIPC_SMTRet;SMT;TmaL1_SMTDSB;Fed;FetchBWL1 cache true misses per kilo instruction for retired demand loads1000 * L2_RQSTS.DEMAND_DATA_RD_HIT / INST_RETIRED.ANYC3_Pkg_Residencyl2_demand_rqsts.wb_hitl2_lines_in.sThis event counts L2 or L3 HW prefetches that access L2 cache including rejectsmem_load_uops_l3_hit_retired.xsnp_hitmThis is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were L3 Hit and a cross-core snoop missed in the on-pkg core cache  Supports address when precise.  Spec update: BDM100 (Precise event)event=0xd1,period=100003,umask=0x40Retired load uops misses in L1 cache as data sources. Uses PEBS  Supports address when precise (Precise event)Hit in last-level (L3) cache. Excludes Unknown data-source. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM100 (Precise event)Retired store uops that split across a cacheline boundary. (Precise Event - PEBS)  Supports address when precise (Precise event)offcore_requests.demand_code_rdevent=0xb0,period=100003,umask=0x4offcore_requests_outstanding.demand_code_rdoffcore_requests_outstanding.demand_rfoevent=0x60,period=2000003,umask=0x4fp_arith_inst_retired.256b_packed_singleMicro-op dispatches cancelled due to insufficient SIMD physical register file read portsevent=0x79,cmask=1,period=2000003,umask=0x18This event counts cycles during which uops are being delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Counting includes uops that may bypass the IDQThis event counts loads with latency value being above 64  Spec update: BDM100, BDM35 (Must be precise)mem_trans_retired.load_latency_gt_8Speculative and retired macro-conditional branchesThis event counts both taken and not taken speculative and retired macro-conditional branch instructionsevent=0x88,period=200003,umask=0xc8This event counts taken speculative and retired macro-conditional branch instructionsevent=0xc4,period=400009,umask=0x20event=0x89,period=200003,umask=0x81This is a precise version (that is, uses PEBS) of the event that counts mispredicted conditional branch instructions retired (Precise event)This event counts the number of instructions (EOMs) retired. Counting covers macro-fused instructions individually (that is, increments by two)  Spec update: BDM61event=0xa8,cmask=4,period=2000003,umask=0x1lsd.cycles_activeThis event counts cycles during which no uops were dispatched from the Reservation Station (RS) per threaduops_executed.threadCycles per core when uops are exectuted in port 1Actually retired uops. (Precise Event - PEBS)  Supports address when precise (Precise event)uops_retired.stall_cycles1llc_misses.pcie_non_snoop_writeevent=0x1,umask=0x3event=0x4Counts the number of cycles when temperature is the upper limit on frequency. Unit: uncore_pcu Counts the number of cycles when the OS is the upper limit on frequency. Unit: uncore_pcu dtlb_store_misses.miss_causes_a_walkdtlb_store_misses.walk_completed_4kitlb_misses.stlb_hitCode misses that miss the  DTLB and hit the STLB (2M)Misses in all ITLB levels that cause completed page walks  Spec update: BDM69event=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020240Counts prefetch RFOsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020122offcore_response.corewb.supplier_none.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0004offcore_response.demand_code_rd.supplier_none.snoop_missoffcore_response.demand_data_rd.supplier_none.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080028000offcore_response.pf_l2_code_rd.l3_hit.any_snoopoffcore_response.pf_l2_code_rd.l3_hit.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0020offcore_response.pf_l2_rfo.supplier_none.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020080offcore_response.pf_l3_data_rd.supplier_none.snoop_missNumber of SSE/AVX computational 128-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 2 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000122offcore_response.all_rfo.l3_miss.snoop_missoffcore_response.corewb.l3_miss.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000008offcore_response.demand_code_rd.l3_miss.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000010offcore_response.pf_l2_data_rd.l3_miss_local_dram.snoop_not_neededoffcore_response.pf_l3_code_rd.l3_miss_local_dram.any_snoopoffcore_response.pf_l3_code_rd.l3_miss_local_dram.snoop_hitmnumber of near branch instructions retired that were mispredicted and taken (Precise event)L3 Lookup any request that access cache and found line in MESI-stateCounts all data/code/rfo reads (demand & prefetch) hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all demand & prefetch RFOs hit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts all demand data writes (RFOs) hit in the L3L1 Cacheable Data Readsevent=0x40,period=2000000,umask=0xa2Cycles L2 address bus is in useevent=0x30,period=200000,umask=0x78l2_reject_busq.self.prefetch.mesil2_rqsts.self.demand.e_stateevent=0x2a,period=200000,umask=0x44event=0x2a,period=200000,umask=0x42event=0xcb,period=200000,umask=0x1Retired Streaming SIMD Extensions 2 (SSE2) scalar-double instructionsmisalign_mem_ref.ld_splitNonzero segbase ld-op-st 1 bubbleStreaming SIMD Extensions (SSE) Prefetch NTA instructions executedAny Software prefetchevent=0x7f,period=200000,umask=0x40event=0x60,period=200000,umask=0xe0event=0x69,period=200000,umask=0xe0event=0x6b,period=200000,umask=0x40bus_trans_pwr.all_agentsbr_inst_retired.any1All indirect branches that have a return mnemonicbr_missp_type_retired.ind_callcpu_clk_unhalted.core_pevent=0x3,period=200000,umask=0x7fevent=0xc,period=200000,umask=0x3event=0xd0,period=200003,umask=0x42Counts data reads (demand & prefetch) that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x10000032b7Counts requests to the uncore subsystem that hit the L2 cacheCounts reads for ownership (RFO) requests (demand & prefetch) that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000002event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000002offcore_response.partial_streaming_stores.l2_miss.snoop_miss_or_no_snoop_neededevent=0x86,period=200003br_inst_retired.ind_callRetired mispredicted conditional branch instructions (Precise event capable) (Must be precise)event=0xc5,period=200003,umask=0xfeCycles a divider is busycycles_div_busy.idivInstructions retired (Fixed event)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000400offcore_response.bus_locks.outstandingCounts demand cacheable data reads of full cache lines hit the L2 cacheCounts data cache line reads generated by hardware L1 data cache prefetcher true miss for the L2 cache with a snoop miss in the other processor moduleevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000010event=0xc3,period=20003Integer divide uops retired (Precise Event Capable) (Must be precise)Counts once per cycle for each page walk only while traversing the Extended Page Table (EPT), and does not count during the rest of the translation.  The EPT is used for translating Guest-Physical Addresses to Physical Addresses for Virtual Machine Monitors (VMMs).  Average cycles per walk can be calculated by dividing the count by number of walksoffcore_response.demand_rfo.l3_hit.hit_other_core_no_fwdRandomly selected loads with latency value being above 64  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x01004007F7offcore_response.demand_code_rd.l3_miss.local_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00010offcore_response.pf_l2_rfo.l3_miss.any_responseNumber of times an HLE transactional execution aborted due to an unsupported read alignment from the elision bufferNumber of integer move elimination candidate uops that were not eliminatedCycles allocation is stalled due to resource related reason  Spec update: HSD135Counts number of cycles no uops were dispatched to be executed on this thread  Spec update: HSD144, HSD30, HSM31Counts the number of micro-ops retired. Use Cmask=1 and invert to count active cycles or stalled cycles (Precise event)uncore otherLoad misses in all DTLB levels that cause page walksCounts the number of Extended Page Table walks from the DTLB that hit in the L3offcore_response.demand_code_rd.llc_hit.hit_other_core_no_fwdhsx metricsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00010l2_l1d_wb_rqsts.allAny MLC or LLC HW prefetch accessing L2, including rejectsoffcore_response.demand_data_rd.llc_hit.no_snoop_neededNumber of Instruction Cache, Streaming Buffer and Victim Cache Misses. Includes UC accessesivb metricsuops_dispatched_port.port_2_coreCounts the number of coherent and in-coherent requests initiated by IA cores, processor graphic units, or LLCunc_cbo_cache_lookup.mRemote cache HITMoffcore_response.other.portio_mmio_ucevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0080ivt metricsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x600400004offcore_response.demand_code_rd.llc_miss.remote_hit_forwardCounts all prefetch (that bring data to LLC only) code reads that miss in the LLCCycles where receiving QPI link is in half-width mode. Unit: uncore_qpi Memory page conflicts. Unit: uncore_imc freq_ge_1200mhz_cycles %l1d.evictionoffcore_response.all_demand_mlc_pref_reads.llc_miss.any_responseinsts_written_to_iq.instsResource stalls2 control structures full for physical registersOccupancy counter for memory read queue. Unit: uncore_imc Counts the total number of L2 cache referencesmem_uops_retired.l1_miss_loadsoffcore_response.any_code_rd.l2_hit_far_tile_e_foffcore_response.any_code_rd.l2_hit_near_tileoffcore_response.any_data_rd.l2_hit_this_tile_eCounts any Prefetch requests that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateCounts any request that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000022event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000022Counts Demand cacheable data writes that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeoffcore_response.partial_writes.any_responseCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for any responseoffcore_response.pf_l1_data_rd.l2_hit_this_tile_fCounts any Prefetch requests that accounts for data responses from DRAM Faroffcore_response.any_request.ddr_faroffcore_response.any_request.ddr_nearoffcore_response.any_rfo.mcdramevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800001Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses from DDR (local and far)Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for data responses from MCDRAM Far or Other tile L2 hit faroffcore_response.pf_l1_data_rd.ddr_farThis event counts the number of retired store that experienced a cache line boundary split(Precise Event). Note that each spilt should be counted only onceevent=0x1,umask=0x01mcdram bandwidth write (CPU traffic only) (MB/sec). Unit: uncore_edc_eclk l1d_all_ref.anyL1 data cache stores in S statel2_data_rqsts.prefetch.i_statel2_rqsts.loadsL2 RFO transactionsl2_write.rfo.i_stateMemory instructions retired above 0 clocks (Precise Event)event=0xb,period=50,umask=0x10,ldlat=0x800Offcore code or data read requests satisfied by a remote cacheOffcore request = all data, response = remote cache or dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x833offcore_response.demand_data.remote_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3803offcore_response.demand_data_rd.llc_hit_no_other_coreoffcore_response.demand_ifetch.any_locationevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x802Offcore other requests satisfied by the IO, CSR, MMIO unitoffcore_response.pf_data_rd.any_locationoffcore_response.pf_data_rd.remote_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1010event=0xb7,period=100000,umask=0x1,offcore_rsp=0x8020offcore_response.prefetch.any_locationevent=0x6,period=200000,umask=0x8offcore_response.any_ifetch.any_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF8FFoffcore_response.any_request.remote_dramoffcore_response.corewb.any_llc_missOffcore other requests satisfied by any DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4030event=0xe8,period=2000000,umask=0x2Branch instructions executedbr_inst_exec.indirect_near_callevent=0x89,period=2000,umask=0x4event=0x17,period=2000000,umask=0x1inst_retired.mmxROB read port stalls cyclesROB full stall cyclesReservation Station full stall cyclesuops_decoded.esp_foldingL1D miss outstandings duration in cyclesl2_lines_out.useless_hwpfCounts requests from the L1/L2/L3 hardware prefetchers or Load software prefetches that miss L2 cacheCounts the RFO (Read-for-Ownership) requests that miss L2 cacheCounts core-originated cacheable requests that miss the L3 cache (Longest Latency cache). Requests include data and code reads, Reads-for-Ownership (RFOs), speculative accesses and hardware prefetches from L1 and L2. It does not include all misses to the L3  Spec update: SKL057Core-originated cacheable demand requests that refer to L3  Spec update: SKL057Retired store instructions that split across a cacheline boundary  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040004offcore_response.demand_code_rd.l3_hit_s.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40080001offcore_response.demand_rfo.l3_hit_s.any_snoopoffcore_response.other.l3_hit_s.snoop_not_neededoffcore_response.other.l4_hit_local_l4.snoop_missCounts once for most SIMD 256-bit packed single computational precision floating-point instructions retired. Counts twice for DPP and FM(N)ADD/SUB instructions retiredevent=0xc6,period=100007,umask=0x1,frontend=0x13Number of times we entered an HLE region. Does not count nested transactionsCounts randomly selected loads when the latency from first dispatch to completion is greater than 8 cycles  Supports address when precise (Must be precise)event=0xb0,period=100003,umask=0x10offcore_response.demand_code_rd.l3_hit_s.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x203C400004offcore_response.demand_rfo.l3_hit_s.snoop_non_dramoffcore_response.other.l3_miss_local_dram.spl_hitThis event counts taken branch instructions retired  Spec update: SKL091 (Precise event)exe_activity.1_ports_utilCounts cycles that the Instruction Length decoder (ILD) stalls occurred due to dynamically changing prefix length of the decoded instruction (by operand size prefix instruction 0x66, address size prefix instruction 0x67 or REX.W for Intel64). Count is proportional to the number of prefixes in a 16B-line. This may result in a three-cycle penalty for each LCP (Length changing prefix) in a 16-byte chunkCounts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 11 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD)Total pipeline cost of (external) Memory Bandwidth related bottlenecksMemory_LatencyBranching_OverheadL3_Cache_Access_BWEach cycle count number of all Core outgoing valid entries. Such entry is defined as valid from its allocation till first of IDI0 or DRS0 messages is sent out. Accounts for Coherent and non-coherent trafficCounts completed page walks (1G page sizes) caused by a code fetch. This implies it missed in the ITLB (Instruction TLB) and further levels of TLB. The page walk can end with or without a faultLoads missed L2 (Precise event)offcore_response.any_code_rd.l2_miss.anyevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000044Any reissued store uopsCycles the divider is busy.  Does not imply a stall waiting for the dividerevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0100Outstanding offcore demand code readsCycles offcore demand data read busyevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5044REQUEST = ANY IFETCH and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = ANY RFO and RESPONSE = ANY_CACHE_DRAMREQUEST = ANY RFO and RESPONSE = REMOTE_CACHE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5033event=0xb7,period=100000,umask=0x1,offcore_rsp=0xff03REQUEST = DEMAND_DATA and RESPONSE = LLC_HIT_NO_OTHER_COREoffcore_response.demand_data.local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f01REQUEST = DEMAND_DATA_RD and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = DEMAND_IFETCH and RESPONSE = LOCAL_CACHEREQUEST = DEMAND_IFETCH and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DEMAND_RFO and RESPONSE = IO_CSR_MMIOevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x850REQUEST = PF_RFO and RESPONSE = IO_CSR_MMIOREQUEST = PF_IFETCH and RESPONSE = ANY_CACHE_DRAMREQUEST = ANY_REQUEST and RESPONSE = ANY_LLC_MISSREQUEST = DEMAND_DATA and RESPONSE = ANY_DRAM AND REMOTE_FWDoffcore_response.pf_rfo.any_dram_and_remote_fwdevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3020Outstanding snoop code requestsevent=0x8,period=200000,umask=0x4event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2744event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2703event=0xb7,period=100000,umask=0x1,offcore_rsp=0xF850Load instructions retired local dram and remote cache HIT data sources (Precise Event)Counts prefetch RFOs that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.all_pf_rfo.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0122OFFCORE_RESPONSE.DEMAND_CODE_RD.L3_HIT.SNOOP_HIT_WITH_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10100offcore_response.pf_l3_rfo.l3_hit.no_snoop_neededCounts all prefetch data reads that miss the L3 and the data is returned from local dramCounts prefetch RFOs that miss the L3 and the data is returned from local or remote dramCounts prefetch RFOs that miss the L3 and the data is returned from remote dramCounts demand data reads that miss the L3 and the data is returned from remote dramCounts all demand data writes (RFOs) that miss in the L3offcore_response.pf_l1d_and_sw.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts L1 data cache hardware prefetch requests and software prefetch requests that miss the L3 and the data is returned from remote dramoffcore_response.pf_l2_data_rd.l3_miss.remote_hitmoffcore_response.pf_l2_data_rd.l3_miss.snoop_miss_or_no_fwdCounts all prefetch (that bring data to L2) RFOs that miss the L3 and the data is returned from local dramCounts all prefetch (that bring data to L2) RFOs that miss the L3 and the data is returned from remote dramCore cycles where the core was running with power-delivery for license level 2 (introduced in Skylake Server michroarchtecture).  This includes high current AVX 512-bit instructions100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) * ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , OFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DATA_RD ) / CPU_CLK_UNHALTED.THREAD - (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD)) / #(CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (( (20.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) - (3.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) ) * MEM_LOAD_RETIRED.L3_HIT * (1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) / 2) / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) + ( (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD)) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) )Fraction of Core cycles where the core was running with power-delivery for license level 1.  This includes high current AVX 256-bit instructions as well as low current AVX 512-bit instructionsIoBW;Mem;SoC;Serverevent=0x4,umask=0xFMMIO writes. Unit: uncore_cha read requests from home agent. Unit: uncore_cha UPI interconnect send bandwidth for payload. Unit: uncore_upi PCI Express bandwidth reading at IIO. Unit: uncore_iio Normal priority reads issued to the memory controller from the CHA. Unit: uncore_cha unc_cha_llc_victims.total_fevent=0x39,umask=0x08event=0xd5,fc_mask=0x04,umask=0x02Read request for 4 bytes made by the CPU to IIO Part0. Unit: uncore_iio Counts every read request for 4 bytes of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part0. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busCounts every read request for 4 bytes of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part1. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busevent=0xc0,ch_mask=0x01,fc_mask=0x07,umask=0x02Read request for up to a 64 byte transaction is made by the CPU to IIO Part2. Unit: uncore_iio event=0xc1,ch_mask=0x01,fc_mask=0x07,umask=0x01unc_iio_txn_req_of_cpu.peer_read.part2Counts every peer to peer read request of up to a 64 byte transaction made by IIO Part3 to the MMIO space of an IIO target. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busCounts traffic in which the M2M (Mesh to Memory) to iMC (Memory Controller) bypass was not takenMulti-socket cacheline Directory update from I to A. Unit: uncore_m2m unc_m2m_directory_update.s2iAD Ingress (from CMS) Queue Inserts. Unit: uncore_m2m AD Ingress (from CMS) OccupancyClocks of the Intel Ultra Path Interconnect (UPI). Unit: uncore_upi Counts Data Response (DRS) packets that attempted to go direct to Intel Ultra Path Interconnect (UPI) bypassing the CHA Counts null FLITs (80 bit FLow control unITs) received from any of the 3 Intel Ultra Path Interconnect (UPI) Receive Queue slots on this UPI unitOCR.ALL_DATA_RD.L3_HIT_E.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200491ocr.all_pf_data_rd.l3_hit.snoop_hit_with_fwdocr.all_pf_rfo.l3_hit.snoop_hit_with_fwdocr.all_pf_rfo.l3_hit_e.hit_other_core_fwdOCR.ALL_PF_RFO.L3_HIT_M.NO_SNOOP_NEEDED  OCR.ALL_PF_RFO.L3_HIT_M.NO_SNOOP_NEEDEDocr.all_reads.l3_hit_e.hitm_other_coreOCR.ALL_READS.L3_HIT_F.HITM_OTHER_CORE  OCR.ALL_READS.L3_HIT_F.HITM_OTHER_COREOCR.ALL_RFO.L3_HIT.ANY_SNOOP OCR.ALL_RFO.L3_HIT.ANY_SNOOP OCR.ALL_RFO.L3_HIT.ANY_SNOOPOCR.ALL_RFO.L3_HIT_E.HITM_OTHER_CORE  OCR.ALL_RFO.L3_HIT_E.HITM_OTHER_CORECounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_F.HIT_OTHER_CORE_FWDocr.demand_data_rd.l3_hit.snoop_hit_with_fwdocr.demand_rfo.l3_hit_e.snoop_noneocr.demand_rfo.l3_hit_f.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200002Counts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_S.NO_SNOOP_NEEDEDCounts any other requests OCR.OTHER.L3_HIT.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800048000ocr.other.l3_hit_m.no_snoop_neededCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_S.HITM_OTHER_COREocr.pf_l2_data_rd.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080010ocr.pf_l2_rfo.l3_hit.snoop_missCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_F.HITM_OTHER_CORECounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_F.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200080ocr.pf_l3_data_rd.l3_hit_m.no_snoop_neededocr.pf_l3_data_rd.l3_hit_s.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_E.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.all_pf_data_rd.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDoffcore_response.all_pf_rfo.l3_hit_s.snoop_noneoffcore_response.all_pf_rfo.supplier_none.hitm_other_coreoffcore_response.all_reads.l3_hit_e.any_snoopThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_E.ANY_SNOOPoffcore_response.all_reads.supplier_none.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000207F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800207F7offcore_response.all_rfo.l3_hit_f.hit_other_core_no_fwdoffcore_response.all_rfo.l3_hit_f.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_F.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_S.SNOOP_NONEoffcore_response.demand_code_rd.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.SNOOP_NONEoffcore_response.demand_data_rd.l3_hit_f.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_E.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_M.SNOOP_MISSoffcore_response.other.l3_hit.hitm_other_coreoffcore_response.other.l3_hit_e.no_snoop_neededoffcore_response.other.l3_hit_m.no_snoop_neededoffcore_response.pf_l1d_and_sw.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.pf_l2_data_rd.l3_hit_m.snoop_noneoffcore_response.pf_l2_data_rd.supplier_none.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_M.NO_SNOOP_NEEDEDoffcore_response.pf_l2_rfo.l3_hit_s.hit_other_core_no_fwdoffcore_response.pf_l2_rfo.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.pf_l3_rfo.l3_hit_f.no_snoop_neededoffcore_response.pf_l3_rfo.l3_hit_s.any_snoopocr.all_data_rd.l3_miss.remote_hit_forwardevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000120ocr.all_reads.l3_miss.hitm_other_coreOCR.ALL_READS.L3_MISS.SNOOP_NONE OCR.ALL_READS.L3_MISS.SNOOP_NONEocr.all_reads.l3_miss_local_dram.hit_other_core_no_fwdOCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSocr.all_rfo.l3_miss.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000122Counts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDocr.demand_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000001ocr.demand_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.demand_rfo.l3_miss.remote_hit_forwardocr.demand_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts any other requests OCR.OTHER.L3_MISS.REMOTE_HITMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000400ocr.pf_l2_data_rd.l3_miss.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000010Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDocr.pf_l2_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.pf_l3_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000080Counts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.ANY_SNOOP OCR.PF_L3_RFO.L3_MISS.ANY_SNOOPCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.HITM_OTHER_COREoffcore_response.all_pf_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.all_rfo.l3_miss_remote_hop1_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEoffcore_response.demand_code_rd.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.HITM_OTHER_COREocr.all_data_rd.supplier_none.any_snoopOCR.ALL_PF_DATA_RD.ANY_RESPONSE have any response typeocr.all_reads.pmm_hit_local_pmm.snoop_noneocr.demand_code_rd.pmm_hit_local_pmm.any_snoopocr.demand_rfo.pmm_hit_local_pmm.any_snoopCounts all demand data writes (RFOs) OCR.DEMAND_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPocr.other.pmm_hit_local_pmm.any_snoopocr.other.pmm_hit_local_pmm.snoop_not_neededCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.SUPPLIER_NONE.ANY_SNOOPCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.pf_l3_data_rd.supplier_none.any_snoopl2_rqsts.swpf_hitCounts demand instruction fetches and L1 instruction cache prefetches that hit a cacheline in the L3 where a snoop was sentevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1E003C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C0400ocr.hwpf_l2_data_rd.l3_hit.snoop_missCounts hardware prefetch RFOs (which bring data to L2) that hit a cacheline in the L3 where a snoop was sent but no other cores had the dataFor every cycle, increments by the number of outstanding data read requests pending.  Data read requests include cacheable demand reads and L2 prefetches, but do not include RFOs, code reads or prefetches to the L3.  Reads due to page walks resulting from any request type will also be counted.  Requests are considered outstanding from the time they miss the core's L2 cache until the transaction completion message is sent to the requestorCounts the number of PREFETCHNTA instructions executedassists.fpevent=0xc7,period=100003,umask=0x20Number of SSE/AVX computational scalar double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0xc6,period=100007,umask=0x1,frontend=0x500806BR_INST_RETIRED.COND_NTAKEN / BR_INST_RETIRED.ALL_BRANCHES( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING ) / ( 2 * CPU_CLK_UNHALTED.DISTRIBUTED )event=0xc9,period=100003,umask=0x80Counts demand instruction fetches and L1 instruction cache prefetches that DRAM supplied the requestCounts demand data reads that have any type of responseCounts conditional branch instructions retired (Precise event)Far branch instructions retired (Precise event)Counts near indirect branch instructions retired excluding returns. TSX abort is an indirect branch (Precise event)Counts number of near branch instructions retired that were mispredicted and taken (Precise event)event=0xa3,cmask=16,period=1000003,umask=0x10event=0xa3,cmask=5,period=1000003,umask=0x5int_misc.all_recovery_cyclestopdown.slotsNumber of uops executed on port 2 and 3Page walks completed due to a demand data store to a 4K pageCycles the superQ cannot take any more entriesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1008000004Counts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that either hit a non-modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeicx metricsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F04400477event=0xb7,period=100003,umask=0x1,offcore_rsp=0x703C00001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x73C0000022LM Tag Check : Miss, existing data may be evicted to Far Memory. Unit: uncore_imc event=0xea,umask=0x01event=0x35,umask=0xC817FD01TOR Inserts : All requests from iA Cores that Missed the LLC. Unit: uncore_cha TOR Inserts : RFO_Prefs issued by iA Cores. Unit: uncore_cha event=0x35,umask=0xC8170A01unc_cha_tor_occupancy.ia_miss_drd_ddrevent=0x36,umask=0xC8178601event=0x35,umask=0xC8F3FF04TOR Occupancy : PCIRdCurs issued by IO Devices. Unit: uncore_cha event=0x34,umask=0x1BC1FFevent=0x83,ch_mask=0x80,fc_mask=0x07,umask=0x80unc_iio_txn_req_by_cpu.mem_read.part6PCIe Completion Buffer Occupancy of completions with data : Part 2. Unit: uncore_iio event=0x17TOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x34,period=200003,umask=0x10event=0x34,period=200003,umask=0x1Counts the number of retired loads that hit in the L3 cache, in which a snoop was required and modified data was forwarded from another core  Supports address when precise (Precise event)Counts the number of load uops retired that miss in the L2 cache  Supports address when precise (Precise event)ocr.all_code_rd.l3_hit.snoop_hitmCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_HIT.SNOOP_MISSCounts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedbaclears.indirectCounts the number of BACLEARS due to a direct, unconditional jumpThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_MISSocr.l2wb_m.l3_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003000000000000Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that have any type of responseCounts L1 data cache hardware prefetches and software prefetches (except PREFETCHW and PFRFO) that have any type of responseocr.l2wb_m.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000100000000000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200000010000Counts the number of machine clears due to memory ordering in which an internal load passes an older store within the same CPUCounts the number of machine clears due typically to program modifying data (self modifying code) within 1K of a recently fetched code pageevent=0x73,period=1000003,umask=0x4topdown_retiring.allDRAM Precharge commands. : Counts the number of DRAM Precharge commands sent on this channelCounts when a normal (Non-Isochronous) full line write is issued from the CHA to any of the memory controller channelsTOR Inserts : DRd_Opt_Prefs issued by iA Cores. Unit: uncore_cha TOR Inserts : PCIRdCurs issued by IO Devices that hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Occupancy : DRd_Opt_Prefs issued by iA Cores that hit the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsTOR Occupancy : PCIRdCurs issued by IO Devices that missed the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsPCIe Completion Buffer Inserts of completions with data : Part 0 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Number Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Counts the number of page walks completed due to store DTLB misses to any page sizeCounts the number of page walks completed due to store DTLB misses to a 4K pageCounts the number of Extended Page Directory Pointer Entry hitsCounts the number of page walks outstanding for an Extended Page table walk including GTLB hits per cycle.  The Extended Page Directory cache is used by Virtual Machine operating systems while the guest operating systems use the standard TLB cachesCounts the number of page walks completed due to instruction fetch misses to a 1G pageCounts the number of BACLEARS due to a non-indirect, non-conditional jumpInstructions per Branch (lower number means higher occurrence rate). Unit: cpu_atom Ratio of all branches which mispredict. Unit: cpu_atom 100 * UOPS_RETIRED.FPDIV / UOPS_RETIRED.ALLevent=0x4,period=20003,umask=0x1event=0x2a,period=100003,umask=0x1,offcore_rsp=0x10003C0001Counts demand data reads that resulted in a snoop hit in another cores caches, data forwarding is required as the data is modified. Unit: cpu_core Number of switches from DSB or MITE to the MS. Unit: cpu_core event=0x47,cmask=9,period=1000003,umask=0x9event=0x2a,period=100003,umask=0x1,offcore_rsp=0x10001Counts the total number of issue slots every cycle that were not consumed by the backend due to backend stalls. Unit: cpu_atom Direct and indirect near call instructions retired (Precise event). Unit: cpu_core inst_retired.rep_iterationUops that RAT issues to RS. Unit: cpu_core Cycles without actually retired uops. Unit: cpu_core unc_m_dram_page_miss_wrunc_mc0_rdcas_count_freerununc_mc0_wrcas_count_freerunevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x104000004event=0x2a,period=100003,umask=0x1,offcore_rsp=0x708000001event=0x2a,period=100003,umask=0x1,offcore_rsp=0x700C04477CPU_CLK_UNHALTED.PAUSE_INSTINT_MISC.UNKNOWN_BRANCH_CYCLESevent=0x5,umask=0x00000000c4event=0xe4,umask=0x0000000002Precharge due to write on page miss. Unit: uncore_imc unc_m_pmm_rpq_occupancy.no_gnt_sch0event=0x83,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000002event=0x36,umask=0x00c8178a01Valid Flits Sent : Slot NULL or LLCRD Empty. Unit: uncore_upi event=0x2,umask=0x0000000027unc_upi_rxl_basic_hdr_match.ncbevent=0xc2,ch_mask=0x04,fc_mask=0x07,umask=0x0000000004event=0x21,umask=0x0308unc_m2m_prefcam_demand_drops.xpt_allchTOR Inserts; DRd hits from local IA. Unit: uncore_cha TOR Inserts : PRQ - Non IOSF. Unit: uncore_cha TOR Inserts : All from Remote. Unit: uncore_cha unc_cha_tor_inserts.missTOR Inserts : MMIO Access. Unit: uncore_cha TOR Inserts : Match the PreMorphed Opcode in b[29:19] of the extended umask field. Unit: uncore_cha TOR Inserts; CRd Pref hits from local IA. Unit: uncore_cha event=0x35,umask=0x00cc43fd04TOR Inserts; DRd Opt from local IA. Unit: uncore_cha event=0x36,umask=0x00c88ffe01unc_cha_tor_occupancy.ia_rfo_prefevent=0x36,umask=0x00c8a7ff01TOR Inserts : DRd PTEs issued by iA Cores. Unit: uncore_cha event=0x35,umask=0x00c8168a01event=0x35,umask=0x00C8970A01unc_cha_tor_inserts.ia_hit_itomevent=0x36,umask=0x00c80efe01TOR Occupancy : ItoMs issued by iA Cores. Unit: uncore_cha The number of pipeline restarts caused by invalidating probes that hit on the instruction stream currently being executed. This would happen if the active instruction stream was being modified by another processor in an MP system - typically a highly unlikely eventevent=0x61,umask=0x20event=0xcaex_ret_mmx_fp_instr.sse_instrThe number of MMX, SSE or x87 instructions retired. The UnitMask allows the selection of the individual classes of instructions as given in the table. Each increment represents one complete instruction. Since this event includes non-numeric instructions it is not suitable for measuring MFLOPS. MMX instructionsevent=0x847,umask=0x02event=0x7,umask=0x38dram_channel_data_controller_3The number of operations (uOps) and dual-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number multi-pipe uOps assigned to pipe 2This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Double precision add/subtract FLOPSNumber of Scalar Ops optimizedfp_num_mov_elim_scal_op.sse_mov_opsfp_retired_ser_ops.sse_ctrl_retLS MAB allocates by type - DC prefetcherls_pref_instr_disp.load_prefetch_wevent=0x52,umask=0x01de_dis_dispatch_token_stalls0.retire_token_stalll3_accessesThe number of micro-ops retired. This count includes all processor activity (instructions, exceptions, interrupts, microcode assists, etc.). The number of events logged per cycle can vary from 0 to 8event=0xe,umask=0x08Floating Point Dispatch Faults. YMM fill faultRetired lock instructions. High speculative cacheable lock speculation succeededls_ret_cpuidL1 DTLB Miss. DTLB reload to a 4K page that hit in the L2 TLBHardware Prefetch Data Cache Fills by Data Source. Local L2 hitCore to L2 cacheable request access status (not including L2 Prefetch). Data cache read hit non-modifiable line in L2ic_tag_hit_miss.instruction_cache_hitevent=0x9a,umask=0xffevent=0x1c7All Allocations. Counts when a LS pipe allocates a MAB entryevent=0x47,umask=0x02Software Prefetch Data Cache Fills by Data Source. From DRAM or IO connected in same nodeSoftware Prefetch Data Cache Fills by Data Source. From cache of different CCX in same nodeevent=0xab,umask=0x08Any FP dispatch. Types of Oops Dispatched from DecoderL2 Cache Hits from L2 Cache HWPFL3 Cache Accesses. Unit: amd_l3 L1 Data Cache Fills: AllMacro-ops DispatchedDemand data cache fills from local L2 cachels_dmnd_fills_from_sys.alternate_memoriesAny data cache fills from local L2 cacheAny data cache fills from either DRAM or MMIO in a different NUMA node (same or different socket)ls_sw_pf_dc_fills.alternate_memoriesL2 cache requests of common types from L1 data cache (including prefetches)l2_cache_req_stat.ic_hit_in_l2l2_pf_hit_l2.l2_strideevent=0x70,umask=0x80l3_lookup_state.l3_missls_not_halted_p0_cyc.p0_freq_cycWrite data beats (64 bytes) for local processor at Coherent Station (CS) 5remote_processor_read_data_beats_cs2Write data beats (64 bytes) for remote processor at Coherent Station (CS) 3event=0x19f,umask=0xbffWrite data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 2event=0x55e,umask=0x7fflocal_socket_inf1_outbound_data_beats_ccm3remote_socket_inf0_outbound_data_beats_ccm2Retired SSE and AVX floating-point add and subtract opsfp_ret_sse_avx_ops.bfloat_mac_flopsfp_ops_retired_by_type.scalar_addevent=0xb,umask=0x03Retired MMX integer multiply-accumulate opsRetired SSE and AVX integer SHA opsfp_pack_ops_retired.fp128_cvtfp_pack_ops_retired.fp256_mulevent=0xd,umask=0x01event=0xd,umask=0x05Retired 256-bit packed integer add opsevent=0xd,umask=0xa0L1 DTLB misses with L2 DTLB misses (page-table walks are requested) for all page sizesInstruction fetches that miss in both the L1 and L2 ITLBs (page-table walks are requested) for coalesced pages. A coalesced page is a 16k page created from four adjacent 4k pagesevent=0x85,umask=0x0ftotal_dispatch_slotsLocal socket inbound data to the CPU (e.g. read data)remote_socket_inbound_data_to_cpuEstimated memory write bandwidthFP_DISPATCHED_FPU_FAST_FLAG_OPSLS_RETIRED_CFLUSH_INSTRUCTIONSBU_FILL_REQUEST_L2_MISSIC_L1_ITLB_MISS_AND_L2_ITLB_HITIC_L1_ITLB_MISS_AND_L2_ITLB_MISSFR_RETIRED_TAKEN_BRANCHES_MISPREDICTEDFR_DISPATCH_STALL_FOR_SERIALIZATIONNB_SIZED_COMMANDSINSTR_SPECL1_DCACHE_HASH_MISSEVENT_06HEVENT_0EHEVENT_1DHEVENT_42HEVENT_55HEVENT_5EHEVENT_78HEVENT_7EHEVENT_A7HEVENT_A8HEVENT_C1HEVENT_D0HEVENT_D9HEVENT_F2HINSTR_CACHE_DEPENDENT_STALLFLOATING_POINT_INSTR_RENAMEDISBL1D_CACHEL1D_TLB_REFILLEVENT_19AHEVENT_1BFHEVENT_1DEHEVENT_1E2HEVENT_202HEVENT_24FHEVENT_264HEVENT_26DHEVENT_27CHEVENT_2B1HEVENT_2E8HEVENT_2F3HEVENT_329HEVENT_3A0HEVENT_3B1HEVENT_3BAHEVENT_3BBHEVENT_3D4HEXC_HVCxp_txflit_stallsbsx_txrsp_retryackcxra_chi_dat_upload_stallscxla_rx_cxs_link1VPU_INSTR_COMPLETEDDTLB_HW_SEARCH_CYCLES_OVER_THRESHOLDLSU_ALIAS_VS_CSQFP_LOAD_SINGLE_INSTR_COMPLETED_IN_LSULSU_MISALIGN_STORE_COMPLETEDCACHEABLE_STORE_MERGE_TO_32_BYTESSECOND_SPECULATIVE_BRANCH_BUFFER_RESOLVED_CORRECTLYL3SQ_FULL_CYCLESRAQ_FULL_CYCLESFXU0_BUSY_FXU1_BUSYUOPS_DECODEDSTASH_REQUESTS_L2packed-sse-sse2THRESHOLDSYSWIDEINTEL_SKYLAKE_XEONINTEL_ICELAKEHygonGenuineLLC_REFERENCE%s, "pmcid": "0x%08x", "event": "%d", "flags": "0x%08x", "evname": "%s"}
GenuineIntel-6-8[CD]HygonGenuine-24-00This category represents fraction of slots wasted due to incorrect speculations. This include slots used to issue uops that do not eventually get retired and slots for which the issue-pipeline was blocked due to recovery from earlier incorrect speculation. For example; wasted work due to miss-predicted branches are categorized under Bad Speculation category. Incorrect data speculation followed by Memory Ordering Nukes is another example. SMT version; use when SMT is enabled and measuring per logical CPURetiring( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )Branches;Fed;FetchBW;Frontend;PGOIpArith_AVX128Actual Average Latency for L1 data-cache miss demand load instructions (in core cycles). Latency may be overestimated for multi-load instructions - e.g. repeat stringsBranches;OSevent=0x51,period=2000003,umask=0x1event=0x24,period=200003,umask=0xe1Core-originated cacheable demand requests missed L3This event counts core-originated cacheable demand requests that miss the last level cache (LLC). Demand requests include loads, RFOs, and hardware prefetches from L1D, and instruction fetches from IFUmem_load_uops_l3_hit_retired.xsnp_noneevent=0xd1,period=2000003,umask=0x1event=0xd0,period=2000003,umask=0x82event=0xd0,period=100003,umask=0x11event=0x60,period=2000003,umask=0x8offcore_requests_outstanding.demand_data_rd_ge_6Number of SSE/AVX computational 256-bit packed single precision floating-point instructions retired.  Each count represents 8 computations. Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP RSQRT SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementfp_arith_inst_retired.scalarCycles where a code fetch is stalled due to L1 instruction-cache missidq.emptyThis event counts cycles during which uops initiated by Decode Stream Buffer (DSB) are being delivered to Instruction Decode Queue (IDQ) while the Microcode Sequencer (MS) is busy. Counting includes uops that may bypass the IDQThis event counts the number of uops initiated by Decode Stream Buffer (DSB) that are being delivered to Instruction Decode Queue (IDQ) while the Microcode Sequencer (MS) is busy. Counting includes uops that may bypass the IDQLoads with latency value being above 256  Spec update: BDM100, BDM35 (Must be precise)event=0xcd,period=2003,umask=0x1,ldlat=0x40Counts the number of times a class of instructions (e.g., vzeroupper) that may cause a transactional abort was executed inside a transactional regionNumber of times we could not allocate Lock Bufferevent=0xc4,period=400009This event counts far branch instructions retired  Spec update: BDW98br_inst_retired.near_callSpeculative and retired mispredicted macro conditional branchesThis event counts taken speculative and retired mispredicted indirect branches that have a return mnemonicThis is a fixed-frequency event programmed to general counters. It counts when the core is unhalted at 100 MhzCycles while L2 cache miss demand load is outstandingmove_elimination.int_not_eliminateduops_dispatched_port.port_1This event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 1Number of uops executed from any threaduops_executed.core_cycles_ge_4event=0xb1,cmask=4,period=2000003,umask=0x2uops_executed.stall_cyclesNumber of Multiply packed/scalar single precision uops allocatedevent=0xe,cmask=1,inv=1,period=2000003,umask=0x1event=0x35,umask=0x3,filter_opc=0x182(UNC_P_POWER_STATE_OCCUPANCY.CORES_C0 / UNC_P_CLOCKTICKS) * 100.power_state_occupancy.cores_c3 %dtlb_load_misses.stlb_hit_2mdtlb_store_misses.walk_completed_1gevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020090offcore_response.all_pf_rfo.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010008offcore_response.corewb.supplier_none.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C8000offcore_response.other.supplier_none.snoop_noneoffcore_response.pf_l2_code_rd.any_responseCounts all prefetch (that bring data to LLC only) code reads have any response typeoffcore_response.pf_l3_data_rd.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000091offcore_response.all_data_rd.l3_miss_local_dram.any_snoopoffcore_response.all_pf_code_rd.l3_miss.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000004offcore_response.demand_code_rd.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004008000offcore_response.pf_l2_code_rd.l3_miss_local_dram.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000040offcore_response.pf_l2_data_rd.l3_miss.snoop_missoffcore_response.pf_l2_data_rd.l3_miss.snoop_noneoffcore_response.pf_l2_data_rd.l3_miss_local_dram.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000010offcore_response.pf_l2_data_rd.supplier_none.snoop_non_dramoffcore_response.pf_l2_rfo.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020200offcore_response.pf_l3_rfo.l3_miss_local_dram.snoop_non_dramA cross-core snoop resulted from L3 Eviction which misses in some processor coreunc_cbo_xsnp_response.hitm_xcoreA cross-core snoop initiated by this Cbox due to processor core memory request which hits a modified line in some processor core. Unit: uncore_cbox unc_cbo_cache_lookup.any_iuncore_arb( ITLB_MISSES.WALK_DURATION + DTLB_LOAD_MISSES.WALK_DURATION + DTLB_STORE_MISSES.WALK_DURATION + 7 * ( DTLB_STORE_MISSES.WALK_COMPLETED + DTLB_LOAD_MISSES.WALK_COMPLETED + ITLB_MISSES.WALK_COMPLETED ) ) / ( 2 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ) )Counts all demand & prefetch data reads hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.all_rfo.llc_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00244offcore_response.all_data_rd.llc_miss.any_responseCounts all demand & prefetch data reads miss the L3 and the modified data is transferred from remote cacheoffcore_response.all_rfo.llc_miss.any_responseCounts all demand & prefetch RFOs miss the L3 and the data is returned from local dramCounts all demand data writes (RFOs) miss in the L3uncore_qpil1d_cache.all_cache_refl2_dbus_busy.selfl2_ld.self.prefetch.e_stateevent=0x29,period=200000,umask=0x54event=0x2e,period=200000,umask=0x78event=0x2e,period=200000,umask=0x54event=0x2e,period=200000,umask=0x51Retired Streaming SIMD Extensions (SSE) scalar-single instructionssimd_sat_uop_exec.sevent=0x10,period=2000000,umask=0x81Load splitsevent=0x5,period=200000,umask=0x94misalign_mem_ref.split.arStreaming SIMD Extensions (SSE) PrefetchT1 and PrefetchT2 instructions executedDeferred bus transactionsevent=0x77,period=200000,umask=0x21All non-indirect callsbr_inst_type_retired.unconduops_retired.stalled_cyclesevent=0x8,period=200000,umask=0x5event=0x30,period=200003event=0xd0,period=200003,umask=0x83event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000003091Counts data reads generated by L1 or L2 prefetchers that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_read.l2_miss.anyevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200008000Counts reads for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests (demand & prefetch) that true miss for the L2 cache with a snoop miss in the other processor moduleCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.full_streaming_stores.l2_miss.snoop_miss_or_no_snoop_neededCounts partial cache line data writes to uncacheable write combining (USWC) memory region  that miss the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000004000Counts the number of demand write requests (RFO) generated by a write to partial data cache line, including the writes to uncacheable (UC) and write through (WT), and write protected (WP) types of memory that miss the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000042000Counts data cache line reads generated by hardware L1 data cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000001000Counts data cache lines requests by software prefetch instructions that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts requests to the Instruction Cache (ICache)  for one or more bytes in an ICache Line and that cache line is not in the ICache (miss).  The event strives to count on a cache line basis, so that multiple accesses which miss in a single cache line count as one ICACHE.MISS.  Specifically, the event counts when straight line code crosses the cache line boundary, or when a branch target is to a new line, and that cache line is not in the ICache. This event counts differently than Intel processors based on Silvermont microarchitectureCounts cycles that fetch is stalled due to any reason. That is, the decoder queue is able to accept bytes, but the fetch unit is unable to provide bytes.  This will include cycles due to an ITLB miss, ICache miss and other eventsevent=0x86,period=200003,umask=0x1machine_clears.allevent=0xc2,period=2000003,umask=0x10event=0x5,period=200003,umask=0x3Counts data read, code read, and read for ownership (RFO) requests (demand & prefetch) miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) true miss for the L2 cache with a snoop miss in the other processor moduleoffcore_response.bus_locks.l2_hitCounts the number of writeback transactions caused by L1 or L2 cache evictions hit the L2 cacheCounts data cache line reads generated by hardware L1 data cache prefetcher outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher have any transaction responses from the uncore subsystemCounts reads for ownership (RFO) requests generated by L2 prefetcher hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache lines requests by software prefetch instructions miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0x49,period=2000003,umask=0x8Page walk completed due to an instruction fetch in a 2M or 4M pageDemand code read requests sent to uncoreoffcore_response.all_code_rd.l3_hit.hit_other_core_no_fwdhit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedhsw metrics( ITLB_MISSES.WALK_DURATION + DTLB_LOAD_MISSES.WALK_DURATION + DTLB_STORE_MISSES.WALK_DURATION ) / CPU_CLK_UNHALTED.THREADNumber of times an HLE execution aborted due to incompatible memory type  Spec update: HSD65This event counts cycles during which no instructions were allocated because no Store Buffers (SB) were availableCycles which a uop is dispatched on port 2 in this threadCycles which a uop is dispatched on port 4 in this threadunc_cbo_cache_lookup.extsnp_mesiA cross-core snoop resulted from L3 Eviction which hits a modified line in some processor coreevent=0x22,umask=0x24Load miss in all TLB levels causes a page walk that completes. (1G)This event counts cycles when the  page miss handler (PMH) is servicing page walks caused by DTLB load missespage_walker_loads.ept_itlb_l3event=0xd3,period=100003,umask=0x20offcore_response.pf_l2_data_rd.llc_miss.any_responseRetired load uops which data sources were data hits in LLC without snoops required (Precise event)offcore_response.all_reads.any_responseCounts demand data writes (RFOs) that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresNumber of transitions from AVX-256 to legacy SSE when penalty applicableoffcore_response.data_in_socket.llc_miss.local_dramCycles with pending L2 cache miss loadsNumber of loads missed L2Number of flags-merge uops being allocatedevent=0x22,umask=0x10Filter on cross-core snoops initiated by this Cbox due to processor core memory request. Unit: uncore_cbox unc_cbo_xsnp_response.eviction_filterdtlb_load_misses.large_page_walk_completedevent=0x5f,period=100003,umask=0x4Counts demand data reads that miss the LLC  and the data forwarded from remote cacheevent=0xd,filter_band2=30Not rejected writebacks from L1D to L2 cache lines in S stateRetired load uops which data sources were data hits in LLC without snoops requiredMispredicted not taken branch instructions retired (Precise event)event=0x3,period=100003,umask=0x1Loads delayed due to SB blocks, preceding store operations with known addresses but unknown dataCounts the number of cycles that the uncore transitioned to a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band2=XXX with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Derived from unc_p_freq_band2_cycles. Unit: uncore_pcu offcore_response.any_code_rd.l2_hit_this_tile_eevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000044Counts any Prefetch requests that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0offcore_response.any_read.l2_hit_far_tileCounts any Read request  that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeCounts Demand cacheable data write requests  that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080004offcore_response.demand_code_rd.l2_hit_this_tile_eCounts demand code reads and prefetch code reads that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400001event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000002offcore_response.partial_reads.any_responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400080Counts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000100offcore_response.pf_l2_code_rd.l2_hit_this_tile_eCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for any responseCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses which hit its own tile's L2 with data in S stateCounts Software Prefetches that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateoffcore_response.uc_code_reads.l2_hit_this_tile_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000200event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600070Counts any request that accounts for responses from DDR (local and far)offcore_response.any_rfo.mcdram_nearCounts demand code reads and prefetch code reads that accounts for data responses from DRAM Localoffcore_response.demand_data_rd.mcdram_nearCounts Software Prefetches that accounts for responses from MCDRAM (local and far)Counts the number of core cycles when no micro-ops are allocated, the IQ is empty, and no other condition is blocking allocationCounts the number of occurences a retired load gets blocked because its address overlaps with a store whose data is not readyrs_full_stall.allevent=0x4e,period=200000,umask=0x4l2_lines_out.prefetch_dirtyl2_transactions.loadevent=0x27,period=100000,umask=0xe0mem_inst_retired.latency_above_threshold_16384mem_inst_retired.storesevent=0xb2,period=100000,umask=0x1Offcore data reads satisfied by a remote cacheoffcore_response.any_request.any_cache_dramOffcore requests satisfied by any cache or DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x122Offcore RFO requests satisfied by the LLC or local DRAMoffcore_response.any_rfo.remote_cache_hitoffcore_response.data_ifetch.remote_cache_dramoffcore_response.data_in.remote_cache_dramoffcore_response.demand_data.llc_hit_other_core_hitmOffcore demand data requests satisfied by a remote cache or remote DRAMoffcore_response.demand_data_rd.local_cache_dramOffcore demand data reads that HITM in a remote cacheoffcore_response.demand_ifetch.llc_hit_other_core_hitOffcore demand RFO requests satisfied by any cache or DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF80offcore_response.pf_data.llc_hit_other_core_hitmOffcore prefetch data reads that HIT in a remote cacheoffcore_response.prefetch.io_csr_mmioOffcore prefetch requests satisfied by the LLC and not found in a sibling coreevent=0x10,period=2000000,umask=0x10event=0x12,period=200000,umask=0x1simd_int_64.packed_arithoffcore_response.any_data.local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2011Offcore writebacks to the local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6001offcore_response.other.remote_dramoffcore_response.pf_data_rd.remote_dramOffcore prefetch RFO requests satisfied by the local DRAML1I instruction fetch stall cyclesevent=0x80,period=2000000,umask=0x2baclear.clearevent=0x88,period=200000,umask=0x1event=0x88,period=200000,umask=0x40event=0x89,period=2000,umask=0x20ild_stall.regenRegen stall cyclesinst_queue_writesCycles when uops were delivered by the LSDCycles machine clear assertedevent=0xa2,period=2000000,umask=0x2resource_stalls.mxcsrresource_stalls.rs_fullCounts the number of lines that have been hardware prefetched but not used and now evicted by L2 cacheCounts the total number of RFO (read for ownership) requests to L2 cache. L2 RFO requests include both L1D demand RFO misses as well as L1D RFO prefetchesmem_inst_retired.all_loadsAll retired load instructions  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0040004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100001offcore_response.demand_rfo.l4_hit_local_l4.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000108000offcore_response.other.l3_hit_s.snoop_noneoffcore_response.other.l3_hit_s.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400408000Number of PREFETCHNTA instructions executedevent=0x32,period=2000003,umask=0x2Counts once for most SIMD 128-bit packed computational single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsRetired Instructions who experienced DSB miss (Precise event)Number of retired Instructions that experienced a critical DSB (Decode stream buffer i.e. the decoded instruction-cache) miss. Critical means stalls were exposed to the back-end as a result of the DSB miss (Precise event)event=0xc6,period=100007,umask=0x1,frontend=0x404006hle_retired.aborted_memtypeoffcore_response.demand_code_rd.l3_miss.any_snoopoffcore_response.demand_data_rd.l3_miss.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000040002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x404008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x44008000Conditional branch instructions retired  Spec update: SKL091 (Precise event)exe_activity.2_ports_utilevent=0xd,any=1,period=2000003,umask=0x1Counts the cycles when at least one uop is delivered by the LSD (Loop-stream detector)other_assists.anyevent=0xe,period=2000003,umask=0x2event=0xc2,period=2000003,umask=0x4Memory_Bandwidth_SMT100 * (( BR_INST_RETIRED.CONDITIONAL + 3 * BR_INST_RETIRED.NEAR_CALL + (BR_INST_RETIRED.NEAR_TAKEN - ( BR_INST_RETIRED.CONDITIONAL - BR_INST_RETIRED.NOT_TAKEN ) - 2 * BR_INST_RETIRED.NEAR_CALL) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))L1 cache true misses per kilo instruction for all demand loads (including speculative)1000 * L2_RQSTS.ALL_DEMAND_DATA_RD / INST_RETIRED.ANYCounts demand and DCU prefetch data read that have any response typeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000100Counts DCU hardware prefetcher data read that miss L2 with a snoop miss responseCounts data cacheline reads generated by L2 prefetchers that hit in the other module where modified copies were found in other core's L1 cacheThis event counts the number of store uops reissued from Rehabqrehabq.ld_splitsThis event counts the number of retired stores that are delayed because there is not a store address buffer availableStore uops that split cache line boundaryThis event counts the number of retire stores that experienced cache line boundary splitsThis event counts the number of reference cycles that the core is not in a halt state. The core enters the halt state when it is running the HLT instruction. In mobile systems the core frequency may change from time. This event is not affected by core frequency changes but counts as if the core is running at the maximum frequency all the timeRetired load uops which data sources were data hits in LLC without snoops required. (Precise Event - PEBS) (Precise event)All retired load uops. (Precise Event - PEBS) (Precise event)offcore_response.all_pf_code_rd.llc_hit.hit_other_core_no_fwdCounts all demand data reads event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0002offcore_response.demand_rfo.llc_hit.snoop_missCounts all prefetch (that bring data to L2) data reads that hit in the LLCCounts prefetch (that bring data to L2) RFOs that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.pf_l2_rfo.llc_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1f80400200Offcore RFO requestsoffcore_requests_outstanding.demand.read_code_not_emptyCycles offcore demand RFOs busyREQUEST = ANY_DATA read and RESPONSE = LLC_HIT_NO_OTHER_COREREQUEST = ANY IFETCH and RESPONSE = ANY_CACHE_DRAMoffcore_response.any_request.local_dram_and_remote_cache_hitREQUEST = PF_DATA and RESPONSE = IO_CSR_MMIOevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1050REQUEST = PF_RFO and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = PF_IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf811offcore_response.any_rfo.other_local_dramREQUEST = DATA_IFETCH and RESPONSE = REMOTE_DRAMREQUEST = DEMAND_DATA and RESPONSE = ANY_LLC_MISSoffcore_response.pf_data.other_local_dramREQUEST = PF_DATA and RESPONSE = REMOTE_DRAMCounts the number of lines that are evicted by L2 cache when triggered by an L2 cache fill. Those lines can be either in modified state or clean state. Modified lines may either be written back to L3 or directly written to memory and not allocated in L3.  Clean lines may either be allocated in L3 or droppedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10490offcore_response.demand_data_rd.l3_hit.no_snoop_neededoffcore_response.all_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800491Counts all demand & prefetch data reads that miss the L3 and the data is returned from remote dramCounts prefetch RFOs that miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000400Counts L1 data cache hardware prefetch requests and software prefetch requests that miss in the L3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00010Counts prefetch (that bring data to L2) data reads that miss the L3 and the data is returned from local dramCounts all prefetch (that bring data to LLC only) data reads that miss the L3 and the modified data is transferred from remote cacheCore cycles the core was throttled due to a pending power level requestevent=0xfe,period=100003,umask=0x2Fraction of Core cycles where the core was running with power-delivery for license level 2 (introduced in SKX).  This includes high current AVX 512-bit instructions1000000000 * ( UNC_M_RPQ_OCCUPANCY / UNC_M_RPQ_INSERTS ) / imc_0@event\=0x0@Mem;MemoryLat;SoC;ServerAverage IO (network or disk) Bandwidth Use for Reads [GB / sec]event=0x4,umask=0x2read requests from local home agent. Unit: uncore_cha 4BytesPCI Express bandwidth writing at IIO, part 3. Unit: uncore_iio event=0x5b,umask=0x01Snoop filter capacity evictions for E-state entries. Unit: uncore_cha PCIe Completion Buffer Inserts of completions with data: Part 2unc_iio_data_req_by_cpu.peer_read.part0Counts every peer to peer read request for 4 bytes of data made by IIO Part2 to the MMIO space of an IIO target. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busevent=0x83,ch_mask=0x08,fc_mask=0x07,umask=0x02Read request for up to a 64 byte transaction is made by the CPU to IIO Part3. Unit: uncore_iio event=0xc1,ch_mask=0x04,fc_mask=0x07,umask=0x02Peer to peer write request of up to a 64 byte transaction is made to IIO Part3 by a different IIO unit. Unit: uncore_iio event=0x84,ch_mask=0x04,fc_mask=0x07,umask=0x04unc_iio_txn_req_of_cpu.peer_read.part1Multi-socket cacheline Directory lookups (cacheline found in A state). Unit: uncore_m2m event=0x2e,umask=0x8event=0x56AD Egress (to CMS) Occupancyunc_m3upi_upi_prefetch_spawnunc_upi_direct_attempts.d2uNull FLITs received from any slot. Unit: uncore_upi Valid Flits Sent; Data. Unit: uncore_upi ocr.all_data_rd.l3_hit_f.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200491OCR.ALL_PF_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDOCR.ALL_PF_DATA_RD.L3_HIT_M.ANY_SNOOP  OCR.ALL_PF_DATA_RD.L3_HIT_M.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100490OCR.ALL_PF_RFO.L3_HIT.SNOOP_NONE OCR.ALL_PF_RFO.L3_HIT.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080120OCR.ALL_PF_RFO.L3_HIT_M.SNOOP_MISSOCR.ALL_READS.L3_HIT.SNOOP_NONE OCR.ALL_READS.L3_HIT.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000407F7ocr.all_reads.l3_hit_s.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080122Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_E.NO_SNOOP_NEEDEDocr.demand_code_rd.l3_hit_m.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0001Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_F.ANY_SNOOPocr.demand_data_rd.l3_hit_f.hit_other_core_fwdocr.demand_data_rd.l3_hit_f.snoop_missocr.demand_data_rd.l3_hit_s.snoop_noneocr.demand_rfo.l3_hit_m.hit_other_core_fwdCounts any other requests  OCR.OTHER.L3_HIT_E.HITM_OTHER_COREocr.other.l3_hit_f.hitm_other_coreocr.other.l3_hit_f.no_snoop_neededCounts any other requests  OCR.OTHER.L3_HIT_M.HITM_OTHER_CORECounts any other requests  OCR.OTHER.L3_HIT_S.NO_SNOOP_NEEDEDCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_M.HIT_OTHER_CORE_FWDocr.pf_l1d_and_sw.l3_hit_s.snoop_missocr.pf_l1d_and_sw.l3_hit_s.snoop_noneocr.pf_l2_data_rd.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_HIT.HIT_OTHER_CORE_FWD OCR.PF_L2_RFO.L3_HIT.HIT_OTHER_CORE_FWDocr.pf_l2_rfo.l3_hit_e.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200080ocr.pf_l3_data_rd.l3_hit_m.hitm_other_coreCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_E.HITM_OTHER_COREocr.pf_l3_rfo.l3_hit_f.hitm_other_coreCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_F.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200100This event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_M.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020490offcore_response.all_pf_rfo.l3_hit.hit_other_core_fwdoffcore_response.all_pf_rfo.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_S.NO_SNOOP_NEEDEDoffcore_response.all_reads.pmm_hit_local_pmm.snoop_noneoffcore_response.all_reads.supplier_none.snoop_missThis event is deprecated. Refer to new event OCR.ALL_READS.SUPPLIER_NONE.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_F.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDoffcore_response.all_rfo.supplier_none.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_S.SNOOP_MISSoffcore_response.demand_data_rd.pmm_hit_local_pmm.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020001This event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDoffcore_response.other.l3_hit_s.no_snoop_neededoffcore_response.other.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_E.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_F.HIT_OTHER_CORE_FWDoffcore_response.pf_l1d_and_sw.l3_hit_f.no_snoop_neededoffcore_response.pf_l1d_and_sw.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_RFO.SUPPLIER_NONE.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_RFO.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.SUPPLIER_NONE.SNOOP_MISSoffcore_response.pf_l3_rfo.l3_hit_s.snoop_noneoffcore_response.pf_l3_rfo.supplier_none.hit_other_core_fwdOCR.ALL_DATA_RD.L3_MISS.HITM_OTHER_CORE OCR.ALL_DATA_RD.L3_MISS.HITM_OTHER_CORE OCR.ALL_DATA_RD.L3_MISS.HITM_OTHER_COREocr.all_data_rd.l3_miss.remote_hitmocr.all_pf_data_rd.l3_miss_local_dram.hitm_other_coreOCR.ALL_PF_RFO.L3_MISS.HITM_OTHER_CORE OCR.ALL_PF_RFO.L3_MISS.HITM_OTHER_CORE OCR.ALL_PF_RFO.L3_MISS.HITM_OTHER_COREocr.all_pf_rfo.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000120ocr.all_reads.l3_miss_local_dram.hit_other_core_fwdocr.all_reads.l3_miss_local_dram.no_snoop_neededOCR.ALL_READS.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDED  OCR.ALL_READS.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000122ocr.all_rfo.l3_miss_local_dram.no_snoop_neededocr.all_rfo.l3_miss_remote_hop1_dram.snoop_missCounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.SNOOP_NONECounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORECounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.demand_rfo.l3_miss.hit_other_core_no_fwdCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.NO_SNOOP_NEEDED OCR.DEMAND_RFO.L3_MISS.NO_SNOOP_NEEDEDocr.other.l3_miss.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC008000Counts any other requests  OCR.OTHER.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREocr.pf_l1d_and_sw.l3_miss_local_dram.snoop_miss_or_no_fwdocr.pf_l2_data_rd.l3_miss.remote_hitmocr.pf_l2_data_rd.l3_miss_local_dram.any_snoopocr.pf_l2_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdocr.pf_l2_rfo.l3_miss.hit_other_core_fwdocr.pf_l2_rfo.l3_miss_local_dram.hitm_other_coreCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREocr.pf_l3_data_rd.l3_miss.remote_hit_forwardocr.pf_l3_rfo.l3_miss.hit_other_core_fwdocr.pf_l3_rfo.l3_miss.no_snoop_neededocr.pf_l3_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDoffcore_response.all_data_rd.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDoffcore_response.all_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSoffcore_response.all_reads.l3_miss_remote_dram.snoop_miss_or_no_fwdoffcore_response.all_reads.l3_miss_remote_hop1_dram.any_snoopThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDoffcore_response.demand_code_rd.l3_miss_remote_hop1_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREoffcore_response.pf_l2_data_rd.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDocr.all_data_rd.pmm_hit_local_pmm.snoop_not_neededOCR.ALL_READS.SUPPLIER_NONE.HITM_OTHER_CORE  OCR.ALL_READS.SUPPLIER_NONE.HITM_OTHER_COREocr.demand_data_rd.pmm_hit_local_pmm.any_snoopocr.pf_l1d_and_sw.supplier_none.hitm_other_coreCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.pf_l2_data_rd.pmm_hit_local_pmm.snoop_not_neededCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.SUPPLIER_NONE.ANY_SNOOPocr.pf_l3_rfo.supplier_none.any_snoopunc_m_pmm_wpq_insertsWrite Pending Queue Occupancy of all write requests for Intel Optane DC persistent memory. Unit: uncore_imc TOR Occupancy : DRds issued by iA Cores that Missed the LLC. Unit: uncore_cha Counts retired load instructions whose data sources were L3 and cross-core snoop hits in on-pkg core cache  Supports address when precise (Precise event)Counts hardware prefetches to the L3 only that hit a cacheline in the L3 where a snoop was sent or notidq.mite_cycles_okCycles MITE is delivering optimal number of UopsCounts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that was not supplied by the L3 cacheevent=0x54,period=100003,umask=0x20Counts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that DRAM supplied the requestocr.hwpf_l2_data_rd.any_responseNumber of occurrences where a microcode assist is invoked by hardwareevent=0xc5,period=50021,umask=0x11br_misp_retired.indirect_callTMA slots where uops got droppedNumber of retired PAUSE instructions. This event is not supported on first SKL and KBL productsevent=0x56,period=1000003,umask=0x1event=0x8,period=100003,umask=0x20Counts cycles when at least one PMH (Page Miss Handler) is busy with a page walk for a demand loadCounts page walks completed due to demand data stores whose address translations missed in the TLB and were mapped to 2M/4M pages.  The page walks can end with or without a page faultRetired load instructions with remote Intel Optane DC persistent memory as the data source where the data request missed all caches  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x830000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x808000477event=0xef,period=1000003,umask=0x2Counts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were supplied by PMM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeocr.hwpf_l3.any_responseocr.reads_to_core.local_dramevent=0x2,umask=0x04unc_cha_tor_inserts.ia_hit_llcprefrfoevent=0x35,umask=0xCCC7FE01unc_cha_tor_inserts.ia_miss_rfo_prefTOR Inserts; WCiL misses from local IA. Unit: uncore_cha TOR Inserts : DRds issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha event=0x35,umask=0xC8F3FD04unc_cha_tor_inserts.io_pcirdcurevent=0x84,ch_mask=0x04,fc_mask=0x07,umask=0x80event=0x84,ch_mask=0x08,fc_mask=0x07,umask=0x80event=0xc0,ch_mask=0x10,fc_mask=0x07,umask=0x04unc_iio_data_req_of_cpu.mem_write.part5event=0x83,ch_mask=0x10,fc_mask=0x07,umask=0x04event=0x84,ch_mask=0x10,fc_mask=0x07,umask=0x80unc_iio_comp_buf_inserts.cmpd.part7unc_iio_comp_buf_occupancy.cmpd.part5event=0xf,umask=0x04FAF RF full. Unit: uncore_irp event=0x37,umask=0x0720unc_m2p_clockticksCMS Clockticks. Unit: uncore_m2pcie core_reject_l2q.anyCounts the number of L2 Cache accesses that resulted in a miss from a front door request only (does not include rejects or recycles). Counts on a per core basisocr.demand_data_and_l1pf_rd.l3_hit.snoop_not_neededocr.full_streaming_wr.l3_hitocr.hwpf_l2_code_rd.l3_hit.snoop_missocr.uc_wr.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1002184000000ocr.full_streaming_wr.any_responseCounts modified writebacks from L1 cache that miss the L2 cache that have any type of responseevent=0x73,period=1000003,umask=0x6topdown_bad_speculation.mispredictevent=0x35,umask=0xC001FE01,config1=0x40e33event=0x35,umask=0xC827FD01TOR Inserts : DRd_Opt_Prefs issued by iA Cores that hit the LLC. Unit: uncore_cha TOR Inserts : WCiLF issued by iA Cores that Missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : RFO_Prefs issued by iA Cores : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices that hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x36,umask=0xC827FD01TOR Occupancy : All requests from IO Devices that missed the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsData requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Counts the number of page walks outstanding in the page miss handler (PMH) for loads every cycleCounts the number of page walks due to storse that miss the PDE (Page Directory Entry) cacheCounts the number of page walks outstanding in the page miss handler (PMH) for instruction fetches every cycleCounts the number of cycles a core is stalled due to a demand load which hit in the L2 cacheINST_RETIRED.ANY / cyclesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2104000001The ratio of Executed- by Issued-Uops. Unit: cpu_core Core actual clocks when any Logical Processor is active on the Physical Core. Unit: cpu_core Instructions per Store (lower number means higher occurrence rate). Unit: cpu_core Measured Average Frequency for unhalted processors [GHz]. Unit: cpu_core Giga Floating Point Operations Per Second. Unit: cpu_core 100 * SERIALIZATION.NON_C01_MS_SCB / ( 5 * CPU_CLK_UNHALTED.CORE )Read requests with true-miss in L2 cache.[This event is alias to L2_RQSTS.MISS]. Unit: cpu_core Retired instructions with at least 1 uncacheable load or lock  Supports address when precise (Precise event). Unit: cpu_core Counts randomly selected loads when the latency from first dispatch to completion is greater than 256 cycles  Supports address when precise (Must be precise). Unit: cpu_core core_power.license_2Counts the number of issue slots every cycle that were not delivered by the frontend due to BACLEARS. Unit: cpu_atom Cycles while L2 cache miss demand load is outstanding. Unit: cpu_core Cycles where the Store Buffer was full and no loads caused an execution stall. Unit: cpu_core Precise instruction retired with PEBS precise-distribution (Precise event). Unit: cpu_core uops_dispatched.port_2_3_10Read CAS command sent to DRAM. Unit: uncore_imc Load miss in all TLB levels causes a page walk that completes. (All page sizes). Unit: cpu_core event=0x13,period=100003,umask=0xeNumber of page walks outstanding for a store in the PMH each cycle. Unit: cpu_core Cycles when at least one PMH is busy with a page walk for code (instruction fetch) request. Unit: cpu_core event=0x11,period=100003,umask=0xeCounts read requests of any type with true-miss in the L2 cache. True-miss excludes L2 misses that were merged with ongoing L2 misses.[This event is alias to L2_REQUEST.MISS]event=0x2a,period=100003,umask=0x1,offcore_rsp=0x830004477fp_arith_inst_retired2.scalar_halfRetired instructions with at least 1 store uop. This PEBS event is the trigger for stores sampled by the PEBS Store Facility  Supports address when precise (Must be precise)event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F3FFC4477Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by PMM attached to this socket, whether or not in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts PMM accesses that are controlled by the close or distant SNC ClusterThis is a non-precise version (that is, does not use PEBS) of the event that counts mispredicted return instructions retired (Precise event)Number of slots in TMA method where no micro-operations were being issued from front-end to back-end of the machine due to lack of back-end resourcesNumber of uops dispatch to execution  port 0Number of uops dispatch to execution ports 5 and 11event=0x3,umask=0x00000000ffunc_m_pre_count.rd_pch1Write request of 4 bytes made by IIO Part3 to Memory. Unit: uncore_iio unc_iio_data_req_of_cpu.peer_write.part4event=0x84,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000002unc_iio_txn_req_of_cpu.peer_write.part4TOR Inserts for DRd misses from local IA targeting local memory. Unit: uncore_cha unc_upi_txl_flits.slot2unc_upi_txl_flits.llcrdValid Flits Sent : LLCTRL. Unit: uncore_upi RxQ Flit Buffer Bypassed : Slot 1. Unit: uncore_upi event=0x84,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000001event=0xc1,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000001event=0xc2,ch_mask=0x40,fc_mask=0x07,umask=0x0000000004D2U Sent. Unit: uncore_m3upi D2C Sent. Unit: uncore_m3upi TOR Inserts : All. Unit: uncore_cha event=0x35,umask=0x00c803fe04TOR Inserts; DRd Opt Pref misses from local IA. Unit: uncore_cha TOR Inserts;CLFlush from Local IA. Unit: uncore_cha unc_cha_tor_occupancy.local_tgtTOR Occupancy; DRd Opt misses from local IA. Unit: uncore_cha TOR Occupancy; RFO prefetch misses from local IA. Unit: uncore_cha TOR Occupancy; RFO hits from local IO. Unit: uncore_cha TOR Inserts : DRd PTEs issued by iA Cores that Missed the LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_hit_llcprefdataTOR Inserts : WbMtoIs issued by an iA Cores. Modified Write Backs. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_local_wcilf_pmmTOR Inserts : WCiLFs issued by iA Cores targeting PMM that missed the LLC - HOMed locally. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_drd_pref_pmmunc_cha_tor_occupancy.ia_miss_crd_remoteTOR Occupancy : WbMtoIs issued by iA Cores. Unit: uncore_cha unc_cha_tor_occupancy.ia_itomunc_cha_tor_occupancy.ia_miss_wcil_ddrunc_cha_tor_inserts.pmmThe number of 32B fetch windows transferred from IC pipe to DE instruction decoder (includes non-cacheable and cacheable fill responses)All L2 Cache Requests (Breakdown 1 - Common). Data cache state change requests. Request change to writable, check L2 for current statel2_wcb_req.wcb_closel2_cache_req_stat.ls_rd_blk_csCycles with fill pending from L2. Total cycles spent with one or more fill requests in flight from L2event=0x70,umask=0xffThe number of near returns retired that were not correctly predicted by the return address predictor. Each such mispredict incurs the same penalty as a mispredicted conditional branch instructionevent=0xcb,umask=0x04Single-precision divide/square root FLOPSfp_ret_sse_avx_ops.sp_add_sub_flopsCounts the number of stores dispatched to the LS unit. Unit Masks ADDedevent=0x45,umask=0x08event=0x4b,umask=0x01All L2 Cache Missesevent=0x64,umask=0x06fp_ret_sse_avx_ops.add_sub_flopsLS MAB Allocates by Type. StoresL1 DTLB Miss. DTLB reload to a 2M page that hit in the L2 TLBSoftware Prefetch Data Cache Fills by Data Source. From DRAM (home node remote)event=0xae,umask=0x08Core to L2 cacheable request access status (not including L2 Prefetch). Instruction cache request miss in L2. Use l2_cache_misses_from_ic_miss insteadThe number of valid fills into the ITLB originating from the LS Page-Table Walker. Tablewalk requests are issued for L1-ITLB and L2-ITLB misses. Walk for >4K Coalesced pageop_cache_hit_miss.op_cache_hitThe number of macro-ops retiredRetired Indirect Branch Instructions. The number of indirect branches retiredCounts the number of interrupts takenls_any_fills_from_sys.mem_io_localL1 DTLB Miss. DTLB reload to a 2M page that also missed in the L2 TLBHardware Prefetch Data Cache Fills by Data Source. From CCX Cache in different NodeL3 Misses (includes cacheline state change requests). Unit: amd_l3 ls_any_fills_from_sys.dram_io_farSoftware prefetch instructions dispatched (speculative) of type PrefetchW (move data to L1 cache and mark it modifiable)ls_sw_pf_dc_fills.local_ccxevent=0x5a,umask=0x80l2_cache_req_stat.dc_access_in_l2L2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L1Region (fetch additional lines into L1 cache when the data access for a given instruction tends to be followed by a consistent pattern of other accesses within a localized region)l2_pf_miss_l2_l3.l2_next_linel2_pf_miss_l2_l3.l2_up_downAverage sampled latency when data is sourced from extension memory (CXL) in a different NUMA node. Unit: amd_l3 l3_xi_sampled_latency.allCycles with no retire while the oldest op is waiting to be executedRetired fused instructionsRead data beats (64 bytes) for remote processor at Coherent Station (CS) 3Read data beats (64 bytes) for remote processor at Coherent Station (CS) 6remote_processor_write_data_beats_cs7local_socket_upstream_read_beats_iom0local_socket_upstream_read_beats_iom2local_socket_upstream_write_beats_iom0Read data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 1Data beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 1event=0x5df,umask=0xbfeData beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 2event=0x5df,umask=0xbffRetired MMX floating-point opsevent=0x8,umask=0x20event=0xa,umask=0x10sse_avx_ops_retired.mmx_shiftRetired SSE and AVX integer AES opspacked_int_op_type.int128_mulevent=0xd,umask=0x07packed_int_op_type.int256_addumc_cas_cmd.allNumber of memory store operations dispatched to the load-store unitevent=0x96de_src_op_disp.op_cacheL2 cache hits from L1 instruction cache missesl1_demand_data_cache_fills_from_far_memoryL2 data TLB misses and data page walkslocal_socket_outbound_data_from_all_linksIC_INSTRUCTION_FETCH_STALLFR_DISPATCH_STALLSFR_DISPATCH_STALL_WHEN_WAITING_FOR_ALL_TO_BE_QUIETNB_MEMORY_CONTROLLER_PAGE_ACCESS_EVENTTTBR_WRITEEVENT_0FHEVENT_21HEVENT_63HEVENT_76HEVENT_85HEVENT_8BHEVENT_C7HEVENT_CCHEVENT_EFHDSBPLE_REQUEST_COMPLETEDL1I_TLB_REFILLEVENT_13BHEVENT_149HEVENT_16BHEVENT_1A1HEVENT_1C6HEVENT_1E1HEVENT_210HEVENT_212HEVENT_21AHEVENT_228HEVENT_24CHEVENT_25EHEVENT_281HEVENT_297HEVENT_2BEHEVENT_2D3HEVENT_30AHEVENT_325HEVENT_340HEVENT_35AHEVENT_37CHEVENT_3FFHL2D_CACHE_WB_VICTIML3D_CACHEITLB_WALKdn_rxreq_dvmop_vmid_filteredhnf_pocq_reqs_recvdhnf_ld_st_swp_adq_fullhnf_stash_snp_sentrni_s1_rdata_beatsclkdiv2_queue_depthclkdiv2_activateINSTR_DISPATCHEDUNRESOLVED_BRANCHESMTSPR_INSTR_COMPLETEDFPU_DENORMALIZATIONLD_ST_TRUE_ALIAS_STALLL1_DATA_TOTAL_MISSESVTQ_LINE_FETCHVR_STALLSWAQ_FULL_CYCLESEXTERNAL_INTERVENTIONSMARKED_GROUP_COMPLETE_TIMEOUTTIMEBASE_EVENTGCT_EMPTY_BY_ICACHE_MISSSTASH_BUSY_2L2_CACHE_DIRTY_DATA_ALLOCATIONSLOAD_RETRIEScycles-in-requestsse-retype-microfaultspage-missprobe-missk8-dc-missk8-fr-taken-hardware-interruptsDOMWIDEINTEL_COREARMV7_CORTEX_A15/usr/src/lib/libpmc/libpmc_pmu_util.cunit: %s
%s, "pmcid": "0x%08x", "pid": "%d"}
GenuineIntel-6-1EAuthenticAMD-23-[[:xdigit:]]+This category represents fraction of slots where no uops are being delivered due to a lack of required resources for accepting new uops in the BackendActual per-core usage of the Floating Point execution units (regardless of the vector width). Values > 1 are possible due to Fused-Multiply Add (FMA) counting. SMT version; use when SMT is enabled and measuring per logical CPUCor;Flops;HPC_SMTFlops;FpVector;InsTypeC3_Core_ResidencyC7 residency percent per corel1d_pend_miss.pendingNot rejected writebacks that hit L2 cachel2_lines_in.allevent=0xf1,period=100003,umask=0x2event=0x24,period=200003,umask=0x27event=0x24,period=200003,umask=0x21This event counts transactions that access the L2 pipe including snoops, pagewalks, and so onevent=0xf0,period=200003,umask=0x4event=0xd3,period=100007,umask=0x10mem_load_uops_retired.l1_hitDemand and prefetch data readsDemand Data Read requests sent to uncoreoffcore_requests.demand_rfoevent=0x60,cmask=6,period=2000003,umask=0x1other_assists.sse_to_avxuop_dispatches_cancelled.simd_prfidq.ms_switchesidq_uops_not_delivered.cycles_le_2_uop_deliv.corehle_retired.aborted_misc2mem_trans_retired.load_latency_gt_512Unfriendly TSX abort triggered by a nest count that is too deeptx_mem.abort_hle_elision_buffer_not_emptyThis event counts both taken and not taken speculative and retired macro-unconditional branch instructions, excluding calls and indirectsThis event counts not taken macro-conditional branch instructionsevent=0x88,period=200003,umask=0x90event=0xc4,period=100007,umask=0x2event=0x89,period=200003,umask=0xa0event=0xa3,cmask=2,period=2000003,umask=0x2cycle_activity.cycles_mem_anyinst_retired.any_pevent=0x4c,period=100003,umask=0x2This event counts cases of saving new LBR records by hardware. This assumes proper enabling of LBRs and takes into account LBR filtering done by the LBR_SELECT registerThis event counts cycles during which the reservation station (RS) is empty for the thread.
Note: In ST-mode, not active thread should drive 0. This is usually caused by severely costly branch mispredictions, or allocator/FE issuesCounts end of periods where the Reservation Station (RS) was empty. Could be useful to precisely locate Frontend Latency Bound issuesuops_dispatched_port.port_0Cycles where at least 3 uops were executed per-threadllc_misses.data_llc_prefetchllc_references.pcie_readevent=0x1,umask=0x2power_channel_ppd %event=0xaCounts the number of cycles when power is the upper limit on frequency. Unit: uncore_pcu Retired load uop whose Data Source was: local DRAM either Snoop not needed or Snoop Miss (RspI)  Supports address when precise.  Spec update: BDE70, BDM100 (Precise event)offcore_response.all_data_rd.supplier_none.snoop_hit_no_fwdCounts all demand code reads have any response typeoffcore_response.demand_code_rd.supplier_none.any_snoopoffcore_response.demand_code_rd.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0002offcore_response.pf_l2_code_rd.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0200offcore_response.pf_l3_rfo.supplier_none.snoop_noneRandomly selected loads with latency value being above 256  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)Counts randomly selected loads with latency value being above 64  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000080offcore_response.pf_l3_rfo.l3_miss.snoop_noneunc_cbo_xsnp_response.miss_xcoreA cross-core snoop resulted from L3 Eviction which misses in some processor core. Unit: uncore_cbox event=0x34,umask=0x21L3 Lookup write request that access cache and found line in M-state. Unit: uncore_cbox unc_cbo_cache_lookup.read_esevent=0x80,umask=0x02Each cycle count number of 'valid' coherent Data Read entries that are in DirectData mode. Such entry is defined as valid when it is allocated till data sent to Core (first chunk, IDI0). Applicable for IA Cores' requests in normal case. Unit: uncore_arb event=0x81,umask=0x20event=0,umask=0x01event=0x29,period=200000,umask=0x78l2_lock.self.e_statel2_lock.self.s_statel2_m_lines_out.self.demandevent=0x30,period=200000,umask=0x72event=0x2e,period=200000,umask=0x58l2_rqsts.self.prefetch.s_stateevent=0xc7,period=2000000,umask=0x8simd_uop_type_exec.arithmetic.ssimd_uop_type_exec.pack.arevent=0xb3,period=2000000,umask=0x84simd_uop_type_exec.shift.arFXCH uops executedevent=0xe6,period=2000000,umask=0x1Cycles during which instruction fetches are  stalledicache.accessesNonzero segbase 1 bubbleevent=0x5,period=200000,umask=0x8aStreaming SIMD Extensions (SSE) PrefetchT1 instructions executedevent=0xc4,period=200000,umask=0x2All macro conditional branch instructionsMultiply operations executedPeriods no micro-ops retiredpage_walks.cyclesCycles code-fetch stalled due to an outstanding ICache missmem_load_uops_retired.wcb_hitCounts data reads (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data reads generated by L1 or L2 prefetchers that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredoffcore_response.demand_code_rd.l2_hitCounts demand cacheable data reads of full cache lines that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand reads for ownership (RFO) requests generated by a write to full data cache line that miss the L2 cacheoffcore_response.demand_rfo.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040800offcore_response.streaming_stores.l2_hitCounts any data writes to uncacheable write combining (USWC) memory region  that miss the L2 cacheoffcore_response.sw_prefetch.l2_hitCounts machine clears due to floating point (FP) operations needing assists.  For instance, if the result was a floating point denormal, the hardware clears the pipeline and reissues uops to produce the correct IEEE compliant denormal resultMS decode startsevent=0xc4,period=200003,umask=0xfbevent=0xc5,period=200003,umask=0xfbmem_uops_retired.dtlb_miss_loadsCounts data reads (demand & prefetch) have any transaction responses from the uncore subsystemCounts data reads (demand & prefetch) have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts requests to the uncore subsystem miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_request.outstandingevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000022Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000012000Counts data cache line reads generated by hardware L1 data cache prefetcher outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts data cacheline reads generated by hardware L2 cache prefetcher miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests generated by L2 prefetcher miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredInstructions retired (Fixed event) (Must be precise)itlb_misses.walk_pendingDemand Data Read requests  Spec update: HSD78, HSM80Demand data read requests sent to uncore  Spec update: HSD78, HSM80offcore_response.all_requests.l3_hit.any_responseoffcore_response.all_data_rd.l3_miss.any_responseStall cycles because IQ is fullAliasing occurs when a load is issued after a store and their memory addresses are offset by 4K.  This event counts the number of loads that aliased with a preceding store, resulting in an extended address check in the pipeline which can have a performance impactCycles which a uop is dispatched on port 3 in this threadNumber of multiply packed/scalar single precision uops allocatedL3 Lookup external snoop request that access cache and found line in E or S-stateStore misses in all DTLB levels that cause completed page walks. (1G)event=0xbc,period=2000003,umask=0x48event=0xbc,period=2000003,umask=0x81Number of ITLB page walker loads from memory  Spec update: HSD25offcore_response.pf_l2_rfo.llc_hit.any_responseRetired load uops with L1 cache hits as data sources (Precise event)Retired load uops which data sources following L1 data-cache miss (Precise event)Retired load uops with L2 cache hits as data sources (Precise event)offcore_response.all_code_rd.llc_hit.no_snoop_neededCounts all demand & prefetch RFOs that hit in the LLCoffcore_response.demand_code_rd.llc_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0002Cycles when Decode Stream Buffer (DSB) fill encounter more than 3 Decode Stream Buffer (DSB) linesNumber of instructions at retirementPrecise instruction retired event with HW to reduce effect of PEBS shadow in IP distribution (Must be precise)event=0xc1,period=100003,umask=0x80event=0xa1,period=2000003,umask=0xcCounts the number of allocated write entries, include full, partial, and LLC evictionsA snoop hits a modified line in some processor core. Unit: uncore_cbox A snoop hits a modified line in some processor coreevent=0x22,umask=0x40event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc20040rxl0p_power_cycles %unc_q_txl0p_power_cyclesunc_p_freq_ge_2000mhz_cyclesCounts the number of cycles that the uncore was running at a frequency greater than or equal to 4Ghz. Derived from unc_p_freq_band3_cycles. Unit: uncore_pcu event=0xac,period=2000003,umask=0xaidq_uops_not_delivered.cycles_ge_1_uop_deliv.coreevent=0x4e,period=2000003,umask=0x2Resource stalls due to Rob being full, FCSW, MXCSR and OTHERevent=0x5b,period=2000003,umask=0x4fevent=0x8,period=100003,umask=0x10This event counts load operations that miss the first DTLB level but hit the second and do not cause any page walks. The penalty in this case is approximately 7 cyclesCounts the loads retired that get the data from the other core in the same tile in M state  Supports address when precise (Precise event)mem_uops_retired.utlb_miss_loadsCounts any Prefetch requests that accounts for any responseoffcore_response.any_pf_l2.l2_hit_far_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400070Counts any request that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts Demand cacheable data write requests  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateoffcore_response.any_rfo.l2_hit_this_tile_mCounts Bus locks and split lock requests that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080002offcore_response.partial_reads.l2_hit_this_tile_sCounts Partial streaming stores (WC and should be programmed on PMC1) that accounts for any responseCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses which hit its own tile's L2 with data in M stateCounts L2 code HW prefetches that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000020020Counts Software Prefetches that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x01004032f7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x00802032f7offcore_response.any_request.mcdram_nearCounts Bus locks and split lock requests that accounts for data responses from MCDRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800004event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600001event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800080event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600100Counts L1 data HW prefetches that accounts for responses from DDR (local and far)Counts L1 data HW prefetches that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses from any NON_DRAM system address. This includes MMIO transactionsCounts UC code reads (valid only for Outstanding response type)  that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts the number of near indirect CALL branch instructions retired (Precise event)br_misp_retired.far_branchl1d_cache_ld.m_stateL1 data cache load locks in E stateevent=0x52,period=2000000,umask=0x1l2_data_rqsts.demand.mesil2_lines_in.anyL2 lines evictedL2 lines evicted by a demand requestevent=0x24,period=200000,umask=0x4event=0xf0,period=200000,umask=0x10l2_write.rfo.hitevent=0xb,period=10,umask=0x10,ldlat=0x2000mem_inst_retired.loadsmem_uncore_retired.local_drammem_uncore_retired.uncacheableevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4711Offcore data reads that HIT in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1044offcore_response.any_request.local_cacheoffcore_response.any_request.remote_cacheOffcore writebacks to the LLCOffcore writebacks to a remote cacheoffcore_response.data_ifetch.local_cache_dramoffcore_response.data_in.io_csr_mmioOffcore data reads, RFO's and prefetches satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1003offcore_response.demand_data_rd.local_cacheoffcore_response.demand_data_rd.remote_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF04Offcore demand code reads satisfied by a remote cache or remote DRAMOffcore demand RFO requests satisfied by the LLC  and HITM in a sibling coreoffcore_response.other.remote_cache_dramoffcore_response.pf_data.local_cacheoffcore_response.pf_ifetch.local_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1070event=0xb7,period=100000,umask=0x1,offcore_rsp=0x870event=0xfd,period=200000,umask=0x4offcore_response.any_request.local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x40FFevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6033Offcore demand code reads satisfied by the local DRAMOffcore demand RFO requests satisfied by any DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF880offcore_response.pf_data.any_dramevent=0xb8,period=100000,umask=0x4baclear.bad_targetbr_inst_exec.direct_near_callevent=0x89,period=20000,umask=0x40inst_decoded.dec0event=0xa8,cmask=1,inv=1,period=2000000,umask=0x1lsd_overflowevent=0xa2,period=2000000,umask=0x80ssex_uops_retired.scalar_singleuops_executed.port015Cycles no Uops were issuedCycles Uops are being retired (Precise event)Retired instructions that missed the ITLB (Precise Event)Requests from the L1/L2/L3 hardware prefetchers or Load software prefetches that miss L2 cachemem_inst_retired.anymem_inst_retired.lock_loadsRetired store instructions that miss the STLB  Supports address when precise (Precise event)Retired load instructions which data sources were L3 hit and cross-core snoop missed in on-pkg core cache  Supports address when precise (Precise event)mem_load_retired.l3_hitCycles with offcore outstanding Code Reads transactions in the SuperQueue (SQ), queue to uncoreCounts all demand code readshave any response typeoffcore_response.demand_code_rd.l3_hit_m.snoop_hit_no_fwdoffcore_response.demand_code_rd.l4_hit_local_l4.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40100001offcore_response.demand_rfo.supplier_none.snoop_hitmoffcore_response.other.l3_hit_e.any_snoopoffcore_response.other.l3_hit_s.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0028000event=0xc6,period=100007,umask=0x1,frontend=0x408006Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 16 cycles which was not interrupted by a back-end stall (Precise event)frontend_retired.latency_ge_256event=0xc6,period=100007,umask=0x1,frontend=0x15Counts randomly selected loads when the latency from first dispatch to completion is greater than 16 cycles  Supports address when precise (Must be precise)Counts randomly selected loads when the latency from first dispatch to completion is greater than 16 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)Counts randomly selected loads when the latency from first dispatch to completion is greater than 4 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000080002offcore_response.demand_rfo.l3_miss.snoop_hitmoffcore_response.demand_rfo.l3_miss_local_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000002offcore_response.demand_rfo.l4_hit_local_l4.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000088000This event counts return instructions retired  Spec update: SKL091 (Precise event)Core crystal clock cycles when at least one thread on the physical core is unhaltedevent=0xa6,period=2000003,umask=0x1Mem;MemoryLat;OffcoreBranching_Overhead_SMTCounts the number of flushes of the big or small ITLB pages. Counting include both TLB Flush (covering all sets) and TLB Set Clear (set-specific)Counts completed page walks (2M/4M page sizes) caused by a code fetch. This implies it missed in the ITLB (Instruction TLB) and further levels of TLB. The page walk can end with or without a faultCounts any request that have any response typeCounts demand and DCU prefetch instruction cacheline that are are outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts demand and DCU prefetch data read that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts demand and DCU prefetch RFOs that miss L2 with a snoop miss responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000040Counts code reads generated by L2 prefetchers that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts RFO requests generated by L2 prefetchers that miss L2 with a snoop miss responseThis event counts the number of retired memory operations with lock semantics. These are either implicit locked instructions such as the XCHG instruction or instructions with an explicit LOCK prefix (0xF0)CALL counts the number of near CALL branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)This event counts the number of core cycles while the core is not in a halt state. The core enters the halt state when it is running the HLT instruction. In mobile systems the core frequency may change from time to time. For this reason this event may have a changing ratio with regards to timeoffcore_response.all_code_rd.llc_hit.snoop_missoffcore_response.all_pf_rfo.llc_hit.any_responseCounts prefetch (that bring data to L2) code reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.pf_llc_rfo.llc_hit.hitm_other_coreREQUEST = DATA_IN_SOCKET and RESPONSE = LLC_MISS_LOCAL and SNOOP = ANY_LLC_HITThis event counts the number of retirement slots used each cycle.  There are potentially 4 slots that can be used each cycle - meaning, 4 micro-ops or 4 instructions could retire each cycle.  This event is used in determining the 'Retiring' category of the Top-Down pipeline slots characterization. (Precise Event - PEBS) (Precise event)Outstanding offcore demand RFOsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5022REQUEST = CORE_WB and RESPONSE = LLC_HIT_OTHER_CORE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f33REQUEST = DATA_IN and RESPONSE = ANY_LOCATIONREQUEST = DEMAND_DATA_RD and RESPONSE = ANY_LOCATIONREQUEST = DEMAND_IFETCH and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = PF_DATA and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HIToffcore_response.pf_ifetch.local_dram_and_remote_cache_hitREQUEST = PF_IFETCH and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f20event=0x5,period=200000,umask=0x2REQUEST = DEMAND_DATA_RD and RESPONSE = ANY_LLC_MISSREQUEST = DEMAND_IFETCH and RESPONSE = REMOTE_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf850REQUEST = PF_DATA_RD and RESPONSE = ANY_DRAM AND REMOTE_FWDREQUEST = PF_RFO and RESPONSE = ANY_DRAM AND REMOTE_FWDsnoopq_requests.invalidateevent=0x85,period=200000,umask=0x80event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5811event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2750event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5850event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0122offcore_response.demand_rfo.l3_hit.no_snoop_neededoffcore_response.pf_l3_data_rd.l3_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0080offcore_response.all_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000004Counts all prefetch (that bring data to LLC only) data reads that miss the L3 and the data is returned from local dramCounts all prefetch (that bring data to LLC only) data reads that miss the L3 and the data is returned from remote dramevent=0x28,period=200003,umask=0x40event=0xfe,period=100003,umask=0x4Instructions per FP Arithmetic AVX 512-bit instruction (lower number means higher occurrence rate)Rate of non silent evictions from the L2 cache per Kilo instructionFraction of Core cycles where the core was running with power-delivery for baseline license level 0Average IO (network or disk) Bandwidth Use for Writes [GB / sec]unc_m_cas_count.rd_regAll DRAM Read CAS Commands issued (does not include underfills). Unit: uncore_imc unc_m_wpq_occupancyuncore_chaunc_cha_requests.reads_localPCI Express bandwidth reading at IIO, part 3. Unit: uncore_iio event=0x83,ch_mask=0x02,fc_mask=0x07,umask=0x01Lines Victimized; Lines in E state. Unit: uncore_cha event=0x11,umask=0x01event=0x5c,umask=0x40PCIe Completion Buffer Inserts of completions with data: Part 0-3PCIe Completion Buffer Inserts of completions with data: Part 3. Unit: uncore_iio event=0xc0,ch_mask=0x02,fc_mask=0x07,umask=0x01Peer to peer write request of 4 bytes made by IIO Part0 to an IIO target. Unit: uncore_iio Counts every peer to peer read request for up to a 64 byte transaction of data made by a different IIO unit to the MMIO space of a card on IIO Part1. Does not include requests made by the same IIO unit. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busevent=0x84,ch_mask=0x02,fc_mask=0x07,umask=0x01unc_iio_txn_req_of_cpu.peer_write.part0unc_iio_txn_req_of_cpu.peer_write.part3event=0x2d,umask=0x8event=0x2e,umask=0x1event=0x31,umask=0x2OCR.ALL_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDED  OCR.ALL_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040491OCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDOCR.ALL_PF_DATA_RD.L3_HIT_E.ANY_SNOOP  OCR.ALL_PF_DATA_RD.L3_HIT_E.ANY_SNOOPocr.all_pf_data_rd.l3_hit_e.snoop_missocr.all_pf_data_rd.l3_hit_f.no_snoop_neededOCR.ALL_PF_DATA_RD.L3_HIT_M.HITM_OTHER_CORE  OCR.ALL_PF_DATA_RD.L3_HIT_M.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040490ocr.all_pf_data_rd.l3_hit_s.hit_other_core_fwdocr.all_pf_rfo.l3_hit.hit_other_core_fwdocr.all_pf_rfo.l3_hit_m.no_snoop_neededOCR.ALL_PF_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDOCR.ALL_READS.L3_HIT_F.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10000407F7ocr.all_reads.l3_hit_s.any_snoopOCR.ALL_RFO.L3_HIT_E.ANY_SNOOP  OCR.ALL_RFO.L3_HIT_E.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080122ocr.all_rfo.l3_hit_f.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040122Counts all demand code reads OCR.DEMAND_CODE_RD.L3_HIT.SNOOP_MISSocr.demand_code_rd.l3_hit_e.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200004ocr.demand_code_rd.l3_hit_s.snoop_missocr.demand_data_rd.l3_hit_e.any_snoopCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_E.HITM_OTHER_CORECounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDocr.demand_data_rd.l3_hit_m.hit_other_core_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_F.NO_SNOOP_NEEDEDocr.demand_rfo.l3_hit_f.snoop_noneocr.demand_rfo.l3_hit_s.hit_other_core_fwdCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_E.ANY_SNOOPCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDocr.pf_l2_data_rd.l3_hit_s.snoop_missCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200080Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_M.HITM_OTHER_CORECounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDocr.pf_l3_rfo.l3_hit_f.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.SUPPLIER_NONE.SNOOP_MISSoffcore_response.all_pf_data_rd.l3_hit_f.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020490offcore_response.all_pf_rfo.l3_hit_e.snoop_missoffcore_response.all_pf_rfo.l3_hit_f.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020120This event is deprecated. Refer to new event OCR.ALL_PF_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_E.SNOOP_NONEoffcore_response.all_reads.l3_hit_f.hit_other_core_no_fwdoffcore_response.all_reads.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.all_rfo.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_S.ANY_SNOOPoffcore_response.all_rfo.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_E.SNOOP_MISSoffcore_response.demand_data_rd.l3_hit_f.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_S.SNOOP_MISSoffcore_response.other.supplier_none.hitm_other_coreThis event is deprecated. Refer to new event OCR.OTHER.SUPPLIER_NONE.NO_SNOOP_NEEDEDoffcore_response.pf_l1d_and_sw.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_E.HITM_OTHER_COREoffcore_response.pf_l2_data_rd.l3_hit_e.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_E.SNOOP_NONEoffcore_response.pf_l3_data_rd.l3_hit_f.hitm_other_coreoffcore_response.pf_l3_data_rd.l3_hit_s.snoop_noneoffcore_response.pf_l3_rfo.l3_hit_s.hit_other_core_fwdoffcore_response.pf_l3_rfo.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDfp_arith_inst_retired2.256bit_packed_bf16ocr.all_data_rd.l3_miss.snoop_noneocr.all_data_rd.l3_miss_local_dram.hit_other_core_fwdocr.all_data_rd.l3_miss_local_dram.snoop_missOCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000120ocr.all_pf_rfo.l3_miss.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F900007F7OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000122OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDED  OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDCounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.REMOTE_HITMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000004ocr.demand_data_rd.l3_miss.snoop_noneCounts any other requests  OCR.OTHER.L3_MISS_LOCAL_DRAM.ANY_SNOOPocr.pf_l1d_and_sw.l3_miss_remote_hop1_dram.hit_other_core_fwdCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.PF_L2_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.NO_SNOOP_NEEDED OCR.PF_L2_RFO.L3_MISS.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000100offcore_response.all_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.ANY_SNOOPoffcore_response.all_reads.l3_miss_local_dram.any_snoopThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.HITM_OTHER_COREoffcore_response.demand_code_rd.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.demand_rfo.l3_miss_local_dram.hitm_other_coreoffcore_response.demand_rfo.l3_miss_local_dram.no_snoop_neededoffcore_response.demand_rfo.l3_miss_remote_hop1_dram.any_snoopoffcore_response.demand_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDoffcore_response.other.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.pf_l1d_and_sw.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDoffcore_response.pf_l3_data_rd.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDocr.all_pf_data_rd.pmm_hit_local_pmm.any_snoopOCR.ALL_PF_DATA_RD.SUPPLIER_NONE.SNOOP_MISSocr.all_pf_rfo.supplier_none.any_snoopocr.all_pf_rfo.supplier_none.hitm_other_coreOCR.ALL_READS.PMM_HIT_LOCAL_PMM.ANY_SNOOP OCR.ALL_READS.PMM_HIT_LOCAL_PMM.ANY_SNOOPOCR.ALL_READS.SUPPLIER_NONE.ANY_SNOOP  OCR.ALL_READS.SUPPLIER_NONE.ANY_SNOOPCounts all demand code reads  OCR.DEMAND_CODE_RD.SUPPLIER_NONE.HITM_OTHER_COREocr.demand_data_rd.pmm_hit_local_pmm.snoop_not_neededocr.demand_rfo.pmm_hit_local_pmm.snoop_noneocr.pf_l1d_and_sw.supplier_none.any_snoopCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPocr.pf_l2_rfo.supplier_none.hitm_other_coreCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.SUPPLIER_NONE.HITM_OTHER_COREocr.pf_l3_data_rd.supplier_none.hitm_other_coreCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.SUPPLIER_NONE.HITM_OTHER_COREunc_m_pmm_cmd1.allWritesUNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART0 + UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART1 + UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART2 + UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART3event=0x36,umask=0x21,config1=0x40433unc_m2m_tag_hit.nm_ufill_hit_cleanevent=0xd1,period=100021,umask=0x10ocr.demand_data_rd.l3_hit.snoop_sentCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit a cacheline in the L3 where a snoop was not needed to satisfy the requestevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1E003C0010event=0x60,cmask=1,period=1000003,umask=0x4event=0xc7,period=100003,umask=0x2event=0xab,period=100003,umask=0x2event=0xc6,period=100007,umask=0x1,frontend=0x500106event=0xc6,period=100007,umask=0x1,frontend=0x500206CPU_CLK_UNHALTED.DISTRIBUTEDevent=0xc8,period=100003,umask=0x8Counts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that have any type of responseCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that DRAM supplied the requestCycles the Backend cluster is recovering after a miss-speculation or a Store Buffer or Load Buffer drain stallmisc_retired.lbr_insertsresource_stalls.scoreboardCycles when at least one PMH is busy with a page walk for code (instruction fetch) requestOCR.DEMAND_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDocr.demand_data_rd.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1008000001Counts cacheable and non-cacheable code reads to the coreCounts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that were not supplied by the local socket's L1, L2, or L3 cachesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84400400Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were not supplied by the local socket's L1, L2, or L3 cachesCycles where the core is waiting on at least 6 outstanding demand data read requests known to have missed the L3 cache.  Note that this event does not capture all elapsed cycles while the requests are outstanding - only cycles from when the requests were known to have missed the L3 cacheHit snoop reply with data, line invalidatedocr.demand_rfo.pmmevent=0xd3,umask=0x01event=0xd3,umask=0x10Clockticks of the uncore caching and home agent (CHA). Unit: uncore_cha TOR Occupancy : CRds issued by iA Cores that Missed the LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_rfounc_cha_tor_inserts.io_hit_itomevent=0x35,umask=0xC816FE01unc_cha_tor_inserts.ia_miss_rfo_localunc_cha_tor_inserts.ia_specitomunc_cha_tor_inserts.io_hit_pcirdcurevent=0x83,ch_mask=0x08,fc_mask=0x07,umask=0x80event=0x83,ch_mask=0x20,fc_mask=0x07,umask=0x80unc_iio_txn_req_of_cpu.mem_read.part6PCIe Completion Buffer Inserts of completions with data: Part 4. Unit: uncore_iio event=0xd5,fc_mask=0x04,umask=0x80Total IRP occupancy of inbound read and write requests to coherent memory. Unit: uncore_irp Clockticks in the UBOX using a dedicated 48-bit Fixed Counter. Unit: uncore_ubox Number of kfclks. Unit: uncore_upi event=0x34,period=200003,umask=0x20Counts the number of cycles the core is stalled due to a demand load which hit in the LLC or other core with HITE/F/Mevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0004Counts L2 cache hardware prefetch code reads (written to the L2 cache only) that were supplied by the L3 cache where no snoop was needed to satisfy the requestCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedCounts the total number of requests to the instruction cache.  The event only counts new cache line accesses, so that multiple back to back fetches to the exact same cache line or byte chunk count as one.  Specifically, the event counts when accesses from sequential code crosses the cache line boundary, or when a branch target is moved to a new line or to a non-sequential byte chunk of the same lineocr.full_streaming_wr.l3_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000020ocr.l1wb_m.l3_miss_localevent=0x63,period=200003,umask=0x2event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3000000010000Counts modified writebacks from L1 cache and L2 cache that have any type of responseocr.hwpf_l2_rfo.outstandingCounts uncached memory reads that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)Counts the number of issue slots every cycle that were not consumed by the backend due to IEC or FPC RAT stalls, which can be due to FIQ or IEC reservation stalls in which the integer, floating point or SIMD scheduler is not able to accept uopsevent=0x71,period=1000003TOR Inserts : DRd_Opts issued by iA Cores : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsunc_cha_tor_inserts.ia_miss_drd_opt_prefFAF allocation -- sent to ADQCounts the number of page walks outstanding in the page miss handler (PMH) for instruction fetches every cycle.  A page walk is outstanding from start till PMH becomes idle again (ready to serve next walk)ld_blocks.dtlb_missCounts the number of cycles a core is stalled due to an instruction cache or Translation Lookaside Buffer (TLB) access which hit in the Last Level Cache (LLC) or other core with HITE/F/MCounts the number of memory ordering machine clears triggered by a snoop from an external agentFraction of branches that are non-taken conditionals. Unit: cpu_core UOPS_RETIRED.ALL / INST_RETIRED.ANYStore_Fwd_Blocks100 * LD_BLOCKS.4K_ALIAS / MEM_UOPS_RETIRED.ALL_LOADSINST_RETIRED.ANY / ( BR_INST_RETIRED.FAR_BRANCH / 2 )BR_MISP_RETIRED.ALL_BRANCHES / BR_INST_RETIRED.ALL_BRANCHESMicrocode_Uop_RatioPercentage of all uops which are IDiv uops. Unit: cpu_atom event=0xd0,period=1000003,umask=0x5,ldlat=0x200RFO requests that miss L2 cache. Unit: cpu_core Retired load instructions with locked access  Supports address when precise (Precise event). Unit: cpu_core Retired Instructions who experienced DSB miss (Precise event). Unit: cpu_core Retired instructions after front-end starvation of at least 2 cycles (Precise event). Unit: cpu_core event=0x79,cmask=1,edge=1,period=100003,umask=0x20Counts randomly selected loads when the latency from first dispatch to completion is greater than 64 cycles  Supports address when precise (Must be precise). Unit: cpu_core Counts the number of near CALL branch instructions retired (Precise event). Unit: cpu_atom Counts the number of machine clears due to a page fault.  Counts both I-Side and D-Side (Loads/Stores) page faults.  A page fault occurs when either the page is not present, or an access violation occurs (Precise event). Unit: cpu_atom Counts the number of issue slots every cycle that were not consumed by the backend due to certain allocation restrictions. Unit: cpu_atom event=0xad,period=500009,umask=0x1event=0xa8,cmask=6,period=2000003,umask=0x1event=0xb2,period=2000003,umask=0x40Any Rank at Warm state. Unit: uncore_imc unc_m_pre_count_page_missCounts all L2 requests.[This event is alias to L2_RQSTS.REFERENCES]Retired load instructions with remote Intel Optane DC persistent memory as the data source where the data request missed all cachesRetired memory uops for any accessevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x1008000004event=0x2a,period=100003,umask=0x1,offcore_rsp=0x1030000001fp_arith_inst_retired2.scalarFRONTEND_RETIRED.MS_FLOWS (Precise event)event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F3FC00002exe.amx_busyCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeAMX_OPS_RETIRED.BF16event=0xec,period=2000003,umask=0x70Uops executed on ports 4 and 9unc_m_pre_count.ufill_pch1unc_m_pmm_rpq_occupancy.gnt_wait_sch1unc_iio_txn_req_of_cpu.peer_write.part7event=0x36,umask=0x00c817fe01TOR Occupancy for DRd misses from local IA. Unit: uncore_cha event=0x36,umask=0x00c8178601unc_upi_rxl_flits.nullMatches on Receive path of a UPI Port : Non-Coherent Standard, Match Opcode. Unit: uncore_upi event=0xc0,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000001Number Transactions requested by the CPU : Core writing to Cards MMIO space. Unit: uncore_iio event=0x21,umask=0x0340Tracker Occupancy : Channel 0. Unit: uncore_m2m event=0x56,umask=0x0000000005unc_cha_tor_inserts.irq_iaevent=0x35,umask=0x0000000080unc_cha_tor_inserts.rem_allTOR Inserts; DRd Pref hits from local IA. Unit: uncore_cha TOR Inserts; LLCPrefRFO from local IA. Unit: uncore_cha event=0x35,umask=0x00c8077e01TOR Occupancy; ITOM misses from local IO. Unit: uncore_cha TOR Occupancy; DRd Opt hits from local IA. Unit: uncore_cha unc_cha_tor_occupancy.io_rfoTOR Occupancy; CRd Pref from local IA. Unit: uncore_cha TOR Inserts : WBEFtoEs issued by an IA Core.  Non Modified Write Backs. Unit: uncore_cha event=0x36,umask=0x00cccffe01TOR Inserts : DRd_Prefs issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed locally. Unit: uncore_cha TOR Inserts : DRd_Prefs issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha TOR Occupancy : WiLs issued by iA Cores that Missed LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_remote_wcil_pmmunc_cha_tor_occupancy.ddrevent=0x84l2_request_g2.ls_rd_sizedAll L2 Cache Requests (Breakdown 2 - Rare). Self-modifying code invalidatesevent=0x61,umask=0x01LS to L2 WCB cache line zeroing requests. LS (Load/Store unit) to L2 WCB (Write Combining Buffer) cache line zeroing requestsevent=0x64,umask=0x08ex_ret_brn_ind_mispevent=0xc7,umask=0x38event=0x147,umask=0x38event=0,umask=0x04The number of operations (uOps) dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number uOps assigned to pipe 0The number of x87 floating-point Ops that have retired. The number of events logged per cycle can vary from 0 to 8fp_ret_sse_avx_ops.sp_mult_flopsNumber of Ops that are candidates for optimization (have Z-bit either set or pass)ls_inef_sw_pref.data_pipe_sw_pf_dc_hitCycles where a dispatch group is valid but does not get dispatched due to a token stall. ALSQ 3 Tokens unavailablel2_cachel3_read_miss_latencyall_remote_links_outboundls_locks.spec_lock_hi_specde_dis_uops_from_decoderCycles where a dispatch group is valid but does not get dispatched due to a token stall. SC AGU dispatch stallCycles where a dispatch group is valid but does not get dispatched due to a token stall. ALSQ3_0_TokenStallic_tag_hit_miss.all_instruction_cache_accessesAdd/subtract FLOPs. This is a retire-based event. The number of retired SSE/AVX FLOPs. The number of events logged per cycle can vary from 0 to 64. This event requires the use of the MergeEvent since it can count above 15 events per cycle. See 2.1.17.3 [Large Increment per Cycle Events]. It does not provide a useful count without the use of the MergeEventSSE/AVX control word mispredict traps. The number of serializing Ops retiredL1 DTLB Miss. DTLB reload coalesced page that also missed in the L2 TLBevent=0x78,umask=0xffl1_data_cache_fills_from_memoryRetired conditional branch instructionsMiss Address Buffer (MAB) entries allocated by a Load-Store (LS) pipe for load-store allocationsls_any_fills_from_sys.near_cachels_inef_sw_pref.allevent=0x70,umask=0x40Average sampled latency from all data sources. Unit: amd_l3 Write data beats (64 bytes) for remote processor at Coherent Station (CS) 0remote_processor_write_data_beats_cs5event=0x1df,umask=0xbffevent=0x2df,umask=0xbfflocal_socket_inf0_inbound_data_beats_ccm6Data beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 6Data beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 7event=0x59e,umask=0x7ffevent=0x4df,umask=0x7ffevent=0x55f,umask=0x7ffevent=0x5df,umask=0x7ffremote_socket_inf1_inbound_data_beats_ccm3event=0x49f,umask=0xbffremote_socket_inf1_outbound_data_beats_ccm3Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 6local_socket_outbound_data_beats_link3Retired scalar floating-point subtract opsRetired scalar floating-point ops of all typesRetired floating-point ops of all typesevent=0xb,umask=0x04sse_avx_ops_retired.mmx_shuffleevent=0xc,umask=0x07packed_int_op_type.int128_aesevent=0xd,umask=0xb0Retired 256-bit packed integer shuffle ops (may include instructions not necessarily thought of as including shuffles e.g. horizontal add, dot product, and certain MOV instructions)event=0xe,umask=0x0fNumber of PRECHARGE commands sent for writesL1 DTLB misses with L2 DTLB misses (page-table walks are requested) for coalesced pages. A coalesced page is a 16k page created from four adjacent 4k pagesevent=0x1a0,umask=0x01event=0x1a0,umask=0x60d_ratio(de_no_dispatch_per_slot.no_ops_from_frontend - (6 * cpu@de_no_dispatch_per_slot.no_ops_from_frontend\,cmask\=0x6@), total_dispatch_slots)PipelineL2;bad_speculation_groupL2 cache hits from L1 data cache missesOp cache miss ratio for all fetchesL1 data cache fills from DRAM or MMIO in any NUMA nodeDDRC write commandsTotal cache hits. Unit: uncore_imc FP_CYCLES_WITH_NO_FPU_OPS_RETIREDLS_SEGMENT_REGISTER_LOADLS_MICROARCHITECTURAL_RESYNC_BY_SELF_MODIFYING_CODELS_LOCKED_OPERATIONDC_REFILL_FROM_SYSTEMFR_INTERRUPTS_MASKED_WHILE_PENDING_CYCLESFR_DISPATCH_STALL_WHEN_LS_IS_FULLPC_PROC_RETURNMEM_ERROREVENT_01HEVENT_1CHEVENT_37HEVENT_50HEVENT_7BHEVENT_91HEVENT_A2HEVENT_C2HEVENT_E6HEVENT_107HEVENT_109HEVENT_11DHEVENT_12CHEVENT_133HEVENT_13EHEVENT_1F3HEVENT_1F9HEVENT_1FAHEVENT_220HEVENT_22FHEVENT_241HEVENT_244HEVENT_269HEVENT_313HEVENT_346HEVENT_351HEVENT_372HEVENT_382HEVENT_39FHEVENT_3B7HEVENT_3D1HEVENT_3DEHEVENT_3E2HL1D_CACHE_INVALL2D_CACHE_WB_CLEANL1D_TLB_RDhnf_slc_evictionhnf_qos_hh_retryhni_rdt_rd_occ_cnt_ovflrnd_rdb_ordclkdiv2_waiting_for_miSTORE_INSTR_COMPLETEDLS_LM_INSTR_PIECESITLB_HW_SEARCH_CYCLES_OVER_THRESHOLDFAST_BTIC_HITL2_TOUCH_HITSL3_CACHE_CASTOUTSDTQ_FULL_CYCLESBUS_RETRY_DUE_TO_INTERVENTION_ORDERINGVPU_MARKED_INSTR_COMPLETEDFPU_LONG_INSTR_COMPLETION_STALLGCT_EMPTY_BY_BRANCH_MISS_PREDICTLOAD_MISS_LDQ_FULLADDRESS_COLLISION_CYCLESDATA_MMU_VSP_RELOADSL2_CACHE_DATA_ACCESSESmultiply-pipe-junk-opsinstructionsk8-fr-retired-x86-instructionsWRITEARMV7INTEL_WESTMEREINTEL_WESTMERE_EXstate errorBRANCH_MISSES_RETIRED=GenuineIntel-6-56v9v15pme_test_soc_sysBad_Speculation( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))UOPS_EXECUTED.THREAD / UOPS_ISSUED.ANYINST_RETIRED.ANY / MEM_UOPS_RETIRED.ALL_STORESLoad_Miss_Real_LatencyMem;MemoryBound;MemoryLatL2MPKI_AllL2 cache hits per kilo instruction for all request types (including speculative)Page_Walks_Utilization_SMTHPC;Mem;MemoryBW;SoCSoC(cstate_pkg@c6\-residency@ / msr@tsc@) * 100Cycles a demand request was blocked due to Fill Buffers inavailabilityl2_lines_out.demand_cleanClean L2 cache lines evicted by demandDemand Data Read miss L2, no rejectsl2_rqsts.referencesevent=0x24,period=200003,umask=0xffl2_trans.all_requestsevent=0x2e,period=100003,umask=0x4fevent=0xd0,period=100003,umask=0x42This is a precise version (that is, uses PEBS) of the event that counts load uops with true STLB miss retired to the architected path. True STLB miss is an uop triggering page walk that gets completed without blocks, and later gets retired. This page walk can end up with or without a fault  Supports address when precise (Precise event)This event counts the Demand Data Read requests sent to uncore. Use it in conjunction with OFFCORE_REQUESTS_OUTSTANDING to determine average latency in the uncoreOffcore outstanding Demand Data Read transactions in uncore queue  Spec update: BDM76Cycles with at least 6 offcore outstanding Demand Data Read transactions in uncore queue  Spec update: BDM76Number of SSE/AVX computational 128-bit packed double precision floating-point instructions retired.  Each count represents 2 computations. Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementfp_arith_inst_retired.packedNumber of SSE/AVX computational single precision floating-point instructions retired. Applies to SSE* and AVX*scalar, double and single precision floating-point: ADD SUB MUL DIV MIN MAX RCP RSQRT SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per element. ?Number of X87 assists due to input valueevent=0xca,period=100003,umask=0x2Cycles Decode Stream Buffer (DSB) is delivering 4 Uopsidq.all_mite_cycles_any_uopsidq.dsb_cyclesidq.ms_mite_uopsidq_uops_not_delivered.cycles_fe_was_okevent=0x9c,cmask=3,period=2000003,umask=0x1event=0xc9,period=2000003,umask=0x10Number of times a RTM caused a faultNumber of times RTM aborted and was not due to the abort conditions in subevents 3-6event=0x5d,period=2000003,umask=0x2event=0x54,period=2000003,umask=0x20Number of intervals between processor halts while thread is in ring 0This event counts when there is a transition from ring 1,2 or 3 to ring0This event counts taken speculative and retired direct near callsbr_misp_retired.conditionalcpu_clk_unhalted.ref_tscCore cycles when the thread is not in halt stateThis is an architectural event that counts the number of thread cycles while the thread is not in a halt state. The thread enters the halt state when it is running the HLT instruction. The core frequency may change from time to time due to power or thermal throttling. For this reason, this event may have a changing ratio with regards to wall clock timeevent=0xa3,cmask=8,period=2000003,umask=0x8cycle_activity.cycles_l2_pendingcycle_activity.stalls_totalNumber of instructions retired. General Counter   - architectural event  Spec update: BDM61event=0x3,period=100003,umask=0x8uops_executed.cycles_ge_1_uop_execCycles per core when uops are dispatched to port 3uops_executed_port.port_4_coreevent=0xa1,any=1,period=2000003,umask=0x80This is a precise version (that is, uses PEBS) of the event that counts the number of retirement slots used (Precise event)LLC misses - Uncacheable reads (from cpu) . Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox event=0x35,umask=0x3,filter_opc=0x191unc_p_clockticksevent=0x8,period=2000003,umask=0x10This event counts store misses in all DTLB levels that cause a completed page walk (2M and 4M page sizes). The page walk can end with or without a fault  Spec update: BDM69event=0x49,period=100003,umask=0x2event=0xae,period=100007,umask=0x1Flushing of the Instruction TLB (ITLB) pages, includes 4k/2M/4M pagespage_walker_loads.dtlb_l2This event counts the number of any STLB flush attempts (such as entire, VPID, PCID, InvPage, CR3 write, and so on)offcore_response.all_data_rd.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0091offcore_response.all_data_rd.supplier_none.any_snoopCounts all prefetch code reads have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0240offcore_response.all_pf_code_rd.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020090offcore_response.demand_rfo.l3_hit.snoop_not_neededoffcore_response.pf_l2_rfo.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0080offcore_response.pf_l3_rfo.l3_hit.snoop_missNumber of SSE/AVX computational scalar floating-point instructions retired; some instructions will count twice as noted below. Each count represents 1 computation operation.   Applies to SSE* and AVX* scalar double and single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT RCP FM(N)ADD/SUB. FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000090offcore_response.all_pf_data_rd.l3_miss_local_dram.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000090offcore_response.all_rfo.l3_miss.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000008offcore_response.other.l3_miss_local_dram.snoop_hit_no_fwdoffcore_response.pf_l2_rfo.l3_miss.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000100This event counts mispredicted conditional branch instructions retired (Precise event)L3 Lookup write request that access cache and found line in M-stateunc_arb_trk_occupancy.cycles_with_any_requestCounts all demand & prefetch RFOs hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.pf_llc_code_rd.llc_hit.any_responseoffcore_response.all_data_rd.llc_miss.local_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x087FC00091l1d_cache.stl2_ld_ifetch.self.e_statel2_ld_ifetch.self.mesievent=0x2b,period=200000,umask=0x41event=0x32,period=200000,umask=0x40L2 cache demand requests from this corel2_st.self.i_statefp_assist.arSIMD packed multiply micro-ops retiredsimd_uop_type_exec.pack.sdecode_stall.iq_fullNonzero segbase store 1 bubbleevent=0x5,period=200000,umask=0xaprefetch.prefetcht0event=0x7,period=200000,umask=0x84event=0x63,period=200000,umask=0x40event=0x68,period=200000,umask=0xe0cycles_int_masked.cycles_int_pending_and_maskedevent=0xc6,period=2000000,umask=0x2ext_snoop.all_agents.anyevent=0x3c,period=200000,umask=0x1event=0x13,period=2000000,umask=0x1Duration of page-walks in core cyclesfetch_stall.icache_fill_pending_cyclesCounts load uops retired that miss in the L2 cache  Supports address when precise (Must be precise)event=0xd0,period=200003,umask=0x81Memory uops retired that split a cache-line (Precise event capable)  Supports address when precise (Must be precise)Counts requests to the uncore subsystem that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts reads for ownership (RFO) requests (demand & prefetch) that hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000022Counts bus lock and split lock requests that have any transaction responses from the uncore subsystemevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000004event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000001offcore_response.full_streaming_stores.l2_hitevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600002000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000040010Counts data cacheline reads generated by hardware L2 cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredoffcore_response.pf_l2_data_rd.l2_miss.snoop_miss_or_no_snoop_neededCounts data reads (demand & prefetch) true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_pf_data_rd.any_responseCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.bus_locks.l2_miss.snoop_miss_or_no_snoop_neededoffcore_response.demand_code_rd.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010001event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010800Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts any data writes to uncacheable write combining (USWC) memory region  have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xc3,period=20003,umask=0x1Offcore outstanding code reads transactions in SuperQueue (SQ), queue to uncore, every cycle  Spec update: HSD62, HSD61, HSM63Counts all prefetch (that bring data to LLC only) code reads hit in the L3Instruction Decode Queue (IDQ) empty cycles  Spec update: HSD135event=0x14,period=2000003,umask=0x2Branch instructions at retirementThis event counts the number of retirement slots used each cycle.  There are potentially 4 slots that can be used each cycle - meaning, 4 uops or 4 instructions could retire each cycle (Precise event)event=0x34,umask=0x28An external snoop hits a non-modified line in some processor coreoffcore_response.demand_data_rd.llc_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00020event=0x28,period=200003,umask=0x1Not rejected writebacks that missed LLCevent=0xf2,period=100003,umask=0x2Clean L2 cache lines evicted by L2 prefetchevent=0x24,period=200003,umask=0x40RFOs that hit cache lines in M stateRetired load uops which data sources missed LLC but serviced from local dramoffcore_response.all_code_rd.llc_hit.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00010001Counts all demand data reads that hit in the LLCCycles per thread when load or STA uops are dispatched to port 3offcore_response.all_data_rd.llc_hit.snoop_missCounts prefetch (that bring data to L2) data reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts all demand code reads that miss the LLC  and the data returned from remote dramoffcore_response.demand_data_rd.llc_miss.remote_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc20080Streaming stores (full cache line). Derived from unc_c_tor_inserts.opcode.streaming_full. Unit: uncore_cbox Memory page activates for reads and writes. Unit: uncore_imc Memory controller clock ticks. Use to generate percentages for memory controller CYCLES events. Unit: uncore_imc unc_p_freq_band1_transitionsevent=0xe,edge=1,filter_band3=40Retired load uops that split across a cacheline boundary (Precise event)Cases of cancelling valid Decode Stream Buffer (DSB) fill not because of exceeding way limitdsb_fill.other_cancelAll (macro) branch instructions retired. (Precise Event - PEBS) (Must be precise)Direct and indirect mispredicted near call instructions retired (Precise event)This event counts loads that followed a store to the same address, where the data could not be forwarded inside the pipeline from the store to the load.  The most common reason why store forwarding would be blocked is when a load's address range overlaps with a preceeding smaller uncompleted store.  See the table of not supported store forwards in the Intel? 64 and IA-32 Architectures Optimization Reference Manual.  The penalty for blocked store forwarding is that the load must wait for the store to complete before it can be issuedevent=0xa2,period=2000003,umask=0xauops_dispatched.threadevent=0x36,umask=0xa,filter_opc=0x182event=0x86,period=200003,umask=0x4Counts the number of MEC requests from the L2Q that reference a cache line (cacheable requests) exlcuding SW prefetches filling only to L2 cache and L1 evictions (automatically exlcudes L2HWP, UC, WC) that were rejected - Multiple repeated rejects should be counted multiple timesCounts Demand code reads and prefetch code read requests  that accounts for responses which hit its own tile's L2 with data in S stateCounts Demand cacheable data and L1 prefetch data read requests  that accounts for any responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800083091event=0xb7,period=100007,umask=0x1,offcore_rsp=0x00000132f7offcore_response.any_read.l2_hit_this_tile_sevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400022offcore_response.any_rfo.l2_hit_far_tile_moffcore_response.any_rfo.l2_hit_this_tile_fCounts Bus locks and split lock requests that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400001offcore_response.demand_data_rd.l2_hit_this_tile_eevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400002Counts Demand cacheable data writes that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateoffcore_response.partial_reads.l2_hit_near_tile_e_foffcore_response.partial_streaming_stores.any_responseoffcore_response.pf_l2_code_rd.l2_hit_this_tile_foffcore_response.pf_l2_rfo.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800180020offcore_response.pf_l2_rfo.l2_hit_this_tile_foffcore_response.pf_software.l2_hit_far_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800044Counts Demand cacheable data and L1 prefetch data read requests  that accounts for responses from DDR (local and far)offcore_response.any_pf_l2.mcdram_faroffcore_response.any_read.mcdramevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000002Counts Demand cacheable data writes that accounts for data responses from DRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400002Counts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses from MCDRAM (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200100offcore_response.pf_l1_data_rd.mcdram_faroffcore_response.pf_l2_code_rd.ddroffcore_response.pf_software.mcdram_farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200200recycleq.ld_block_std_notreadyrs_full_stall.mecAll references to the L1 data cacheL2 requestsevent=0x24,period=200000,umask=0xaaL2 writeback to LLC transactionsl2_write.lock.mesievent=0xb,period=5000,umask=0x10,ldlat=0x20Retired loads that miss L1D and hit an previously allocated LFB (Precise Event)mem_uncore_retired.remote_cache_local_home_hitoffcore_response.any_ifetch.io_csr_mmiooffcore_response.any_ifetch.local_cacheoffcore_response.any_ifetch.remote_cacheoffcore_response.any_request.llc_hit_other_core_hitoffcore_response.any_request.remote_cache_dramOffcore requests satisfied by a remote cache or remote DRAMoffcore_response.any_rfo.any_locationOffcore RFO requests that HIT in a remote cacheoffcore_response.data_ifetch.llc_hit_other_core_hitoffcore_response.data_in.local_cacheOffcore demand data requests satisfied by the IO, CSR, MMIO unitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1803event=0xb7,period=100000,umask=0x1,offcore_rsp=0x701Offcore demand RFO requests satisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3880All offcore prefetch data requestsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8010Offcore prefetch data reads satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4720offcore_response.pf_rfo.remote_cache_hitmoffcore_response.prefetch.llc_hit_other_core_hitevent=0x6,period=200000,umask=0x4fp_assist.allevent=0x12,period=200000,umask=0x4event=0x12,period=200000,umask=0x20simd_int_128.unpackSIMD integer 64 bit arithmetic operationsmacro_insts.decodedevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6008offcore_response.demand_data.remote_dramoffcore_response.demand_rfo.local_dramOffcore prefetch code reads satisfied by any DRAMOffcore prefetch code reads satisfied by the local DRAMOffcore prefetch requests that missed the LLCSuper Queue full stall cyclesevent=0xe6,period=2000000,umask=0x2Call branches executedRetired near call instructions (Precise Event)br_misp_exec.anyevent=0xd2,period=2000000,umask=0xfAll RAT stall cyclesFPU control word write stall cyclesCycles Uops executed on any port (core count)Macro-fused Uops retired (Precise Event)Uop unfusions due to FP exceptionsDTLB load missesevent=0x85,period=200000,umask=0x1Requests from the L1/L2/L3 hardware prefetchers or Load software prefetchesNumber of retired store instructions that (start a) miss in the 2nd-level TLB (STLB)  Supports address when precise (Precise event)Counts retired load instructions with at least one uop that missed in the L3 cache  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0020004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020004offcore_response.demand_data_rd.l4_hit_local_l4.any_snoopoffcore_response.demand_data_rd.l4_hit_local_l4.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020002offcore_response.other.l3_hit_s.snoop_missRetired Instructions who experienced iTLB true miss (Precise event)Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 128 cycles which was not interrupted by a back-end stall (Precise event)Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 512 cycles which was not interrupted by a back-end stall (Precise event)hle_retired.aborted_timerevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20001C0001Number of hardware interrupts received by the processorCycles total of 1 uop is executed on all ports and Reservation Station was not emptyNumber of all retired NOP instructions  Spec update: SKL091, SKL044 (Precise event)Cycles the issue-stage is waiting for front-end to fetch from resteered path following branch misprediction or machine clear eventsUops inserted at issue-stage in order to preserve upper bits of vector registersRet_SMTIDQ.DSB_UOPS / (IDQ.DSB_UOPS + IDQ.MITE_UOPS + IDQ.MS_UOPS)( ITLB_MISSES.WALK_PENDING + DTLB_LOAD_MISSES.WALK_PENDING + DTLB_STORE_MISSES.WALK_PENDING + EPT.WALK_PENDING ) / ( 2 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ) )Counts completed page walks  (1G sizes) caused by demand data loads. This implies address translations missed in the DTLB and further levels of TLB. The page walk can end with or without a faultCounts demand data stores that caused a page walk of any page size (4K/2M/4M/1G). This implies it missed in all TLB levels, but the walk need not have completedoffcore_response.any_code_rd.l2_miss.snoop_missCounts writeback (modified to exclusive) that miss L2 with no details on snoop-related informationevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000002Counts demand and DCU prefetch RFOs that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000080Counts code reads generated by L2 prefetchers that miss L2 with a snoop miss responseD-side page-walksevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0122offcore_response.pf_ifetch.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1f80408fffoffcore_response.demand_rfo.llc_miss.dramCounts all prefetch (that bring data to LLC only) code reads that miss the LLC  and the data returned from dramOutstanding offcore demand data readsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xffffevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5008REQUEST = CORE_WB and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f08event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f77REQUEST = DATA_IFETCH and RESPONSE = ANY_LOCATIONREQUEST = DATA_IFETCH and RESPONSE = REMOTE_CACHE_HITMoffcore_response.demand_data.all_local_dram_and_remote_cache_hitREQUEST = DEMAND_DATA_RD and RESPONSE = IO_CSR_MMIOREQUEST = OTHER and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = PF_RFO and RESPONSE = LOCAL_CACHEREQUEST = ANY_DATA read and RESPONSE = OTHER_LOCAL_DRAMREQUEST = ANY RFO and RESPONSE = ANY_DRAM AND REMOTE_FWDREQUEST = CORE_WB and RESPONSE = ANY_DRAM AND REMOTE_FWDREQUEST = DATA_IN and RESPONSE = ANY_DRAM AND REMOTE_FWDREQUEST = DEMAND_DATA and RESPONSE = REMOTE_DRAMREQUEST = PF_RFO and RESPONSE = REMOTE_DRAMload_block.overlap_storeMispredicted retired branch instructions (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5801event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5840event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2720Retired load instructions whose data sources was forwarded from a remote cache  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0491offcore_response.all_pf_data_rd.l3_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10010offcore_response.pf_l2_rfo.l3_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00120offcore_response.pf_l1d_and_sw.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00100Fraction of Core cycles where the core was running with power-delivery for license level 2 (introduced in SKX).  This includes high current AVX 512-bit instructions. SMT version; use when SMT is enabled and measuring per logical CPUevent=0x35,umask=0x21,config1=0x40040e33Counts snoop filter capacity evictions for entries tracking exclusive lines in the cores cache. Snoop filter capacity evictions occur when the snoop filter is full and evicts an existing entry to track a new entry. Does not count clean evictions such as when a cores cache replaces a tracked cacheline with a new cachelineRsp*Fwd*WB Snoop Responses Received. Unit: uncore_cha Counts when a transaction with the opcode type Rsp*WB Snoop Response was received which indicates which indicates the data was written back to its home.  This is returned when a non-RFO request hits a cacheline in the Modified state. The Cache can either downgrade the cacheline to a S (Shared) or I (Invalid) state depending on how the system has been configured.  This response will also be sent when a cache requests E (Exclusive) ownership of a cache line without receiving data, because the cache must acquire ownershipunc_iio_comp_buf_inserts.cmpd.part2event=0xc2,ch_mask=0x08,fc_mask=0x4,umask=0x03unc_iio_data_req_by_cpu.mem_write.part0Write request of 4 bytes made to IIO Part2 by the CPU. Unit: uncore_iio event=0xc0,ch_mask=0x02,fc_mask=0x07,umask=0x08unc_iio_data_req_by_cpu.peer_read.part2unc_iio_txn_req_by_cpu.mem_read.part0event=0xc1,ch_mask=0x08,fc_mask=0x07,umask=0x01Counts every peer to peer write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part0 by a different IIO unit. Does not include requests made by the same IIO unit. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busCounts every peer to peer write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part1 by a different IIO unit. Does not include requests made by the same IIO unit. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busunc_iio_txn_req_of_cpu.mem_write.part3event=0x10,umask=0x10Inbound read requests received by the IRP and inserted into the FAF queue. Unit: uncore_irp unc_m2m_direct2core_takenMulti-socket cacheline Directory update from/to Any state. Unit: uncore_m2m Counts cycles when the receive side (Rx) of the Intel Ultra Path Interconnect(UPI) is in L0p power mode. L0p is a mode where we disable 60% of the UPI lanes, decreasing our bandwidth in order to save powerOCR.ALL_DATA_RD.L3_HIT.ANY_SNOOP OCR.ALL_DATA_RD.L3_HIT.ANY_SNOOP OCR.ALL_DATA_RD.L3_HIT.ANY_SNOOPocr.all_data_rd.l3_hit_e.hit_other_core_no_fwdocr.all_pf_data_rd.l3_hit_f.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040490ocr.all_pf_data_rd.l3_hit_m.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0120OCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F800407F7OCR.ALL_RFO.L3_HIT.SNOOP_HIT_WITH_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0004ocr.demand_code_rd.l3_hit_e.hit_other_core_fwdocr.demand_code_rd.l3_hit_s.any_snoopocr.demand_code_rd.l3_hit_s.snoop_noneCounts demand data reads OCR.DEMAND_DATA_RD.L3_HIT.ANY_SNOOP OCR.DEMAND_DATA_RD.L3_HIT.ANY_SNOOPocr.demand_data_rd.l3_hit_f.hitm_other_coreCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200001Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDocr.demand_data_rd.l3_hit_m.snoop_noneCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_HIT.HIT_OTHER_CORE_FWD OCR.DEMAND_RFO.L3_HIT.HIT_OTHER_CORE_FWDocr.demand_rfo.l3_hit_e.hitm_other_coreocr.demand_rfo.l3_hit_f.hit_other_core_fwdocr.demand_rfo.l3_hit_s.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C8000Counts any other requests  OCR.OTHER.L3_HIT_S.HIT_OTHER_CORE_NO_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_HIT.SNOOP_NONECounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_E.HIT_OTHER_CORE_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_F.ANY_SNOOPocr.pf_l1d_and_sw.l3_hit_m.no_snoop_neededCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_HIT.SNOOP_MISSocr.pf_l2_rfo.l3_hit_s.hit_other_core_fwdocr.pf_l3_data_rd.l3_hit_e.hit_other_core_no_fwdocr.pf_l3_data_rd.l3_hit_f.hitm_other_coreocr.pf_l3_data_rd.l3_hit_f.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080100ocr.pf_l3_rfo.l3_hit_e.snoop_missCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_M.HITM_OTHER_COREocr.pf_l3_rfo.l3_hit_m.no_snoop_neededCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_E.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_F.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.HITM_OTHER_COREoffcore_response.all_reads.l3_hit_m.any_snoopoffcore_response.all_reads.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_READS.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.SNOOP_NONEoffcore_response.all_rfo.l3_hit_s.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020122This event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.demand_code_rd.l3_hit_f.snoop_missThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_F.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_F.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDoffcore_response.other.l3_hit_e.hitm_other_coreThis event is deprecated. Refer to new event OCR.OTHER.SUPPLIER_NONE.SNOOP_MISSoffcore_response.pf_l1d_and_sw.l3_hit_e.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_F.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_data_rd.l3_hit_f.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_F.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEoffcore_response.pf_l2_rfo.l3_hit_f.snoop_missoffcore_response.pf_l2_rfo.supplier_none.hit_other_core_no_fwdoffcore_response.pf_l3_data_rd.l3_hit_s.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020080offcore_response.pf_l3_rfo.l3_hit.hit_other_core_fwdIDQ.DSB_UOPS / (IDQ.DSB_UOPS + LSD.UOPS + IDQ.MITE_UOPS + IDQ.MS_UOPS)Number of SSE/AVX computational 512-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 16 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP14 RSQRT14 SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsocr.all_pf_data_rd.l3_miss.hit_other_core_no_fwdocr.all_pf_data_rd.l3_miss.remote_hit_forwardOCR.ALL_PF_RFO.L3_MISS.ANY_SNOOP OCR.ALL_PF_RFO.L3_MISS.ANY_SNOOP OCR.ALL_PF_RFO.L3_MISS.ANY_SNOOPocr.all_pf_rfo.l3_miss_local_dram.hit_other_core_fwdOCR.ALL_PF_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_PF_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000120ocr.all_reads.l3_miss.hit_other_core_fwdocr.all_rfo.l3_miss.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000122Counts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.REMOTE_HIT_FORWARDCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDocr.demand_data_rd.l3_miss.hit_other_core_fwdCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.ANY_SNOOP OCR.DEMAND_RFO.L3_MISS.ANY_SNOOPocr.demand_rfo.l3_miss_local_dram.hitm_other_coreCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORECounts any other requests OCR.OTHER.L3_MISS.HIT_OTHER_CORE_FWD OCR.OTHER.L3_MISS.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90008000ocr.other.l3_miss_remote_hop1_dram.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90008000Counts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDCounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.PF_L2_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.REMOTE_HITMocr.pf_l3_rfo.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000100offcore_response.all_data_rd.l3_miss_local_dram.hit_other_core_fwdoffcore_response.all_pf_rfo.l3_miss_remote_hop1_dram.any_snoopoffcore_response.all_pf_rfo.l3_miss_remote_hop1_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.demand_code_rd.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSoffcore_response.demand_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.pf_l1d_and_sw.l3_miss_local_dram.any_snoopoffcore_response.pf_l1d_and_sw.l3_miss_local_dram.hit_other_core_no_fwdoffcore_response.pf_l1d_and_sw.l3_miss_remote_hop1_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.NO_SNOOP_NEEDEDocr.all_pf_data_rd.supplier_none.hit_other_core_no_fwdOCR.ALL_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWD  OCR.ALL_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.SUPPLIER_NONE.ANY_SNOOPocr.pf_l2_data_rd.supplier_none.any_snoopunc_m_pmm_rpq_occupancy.all6000000000nsevent=0xea,umask=0x1TOR Occupancy : DRds issued by iA Cores that Missed the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsl2_rqsts.swpf_missocr.hwpf_l2_rfo.l3_hit.snoop_hit_no_fwdCounts hardware prefetch RFOs (which bring data to L2) that hit a cacheline in the L3 where a snoop was not needed to satisfy the requestevent=0x32,period=100003,umask=0x1Counts the number of PREFETCHT0 instructions executedevent=0xc7,period=100003,umask=0x80Number of SSE/AVX computational scalar single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT RCP FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsRetired instructions after front-end starvation of at least 2 cycles (Precise event)Counts the number of cycles uops were delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) pathCycles when no uops are not delivered by the IDQ when backend of the machine is not stalledevent=0x9c,cmask=1,inv=1,period=1000003,umask=0x1icl metrics(cstate_pkg@c8\-residency@ / msr@tsc@) * 100Counts the number of times HLE commit succeededCounts streaming stores that was not supplied by the L3 cacheCounts hardware prefetch RFOs (which bring data to L2) that have any type of responseocr.streaming_wr.dramCounts the number of occurrences where a microcode assist is invoked by hardware Examples include AD (page Access Dirty), FP and AVX related assistscpu_clk_unhalted.ref_distributedCore crystal clock cycles. Cycle counts are evenly distributed between active threads in the CoreCycles optimal number of Uops delivered by the LSD, but did not come from the decodertopdown.br_mispredict_slotsuops_dispatched.port_2_3Counts the cycles for which the thread is active and the superQ cannot take any more entriesCounts hardware prefetches to the L3 only that hit in the L3 or were snooped from another core's caches on the same socketocr.reads_to_core.remote_cache.snoop_hitmCounts hardware and software prefetches to all cache levels that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallyCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were not supplied by the local socket's L1, L2, or L3 caches and were supplied by the local socketCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by PMM attached to this socket, whether or not in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts PMM accesses that are controlled by the close or distant SNC ClusterRemote INVITOE requests (exclusive ownership of a cache line without receiving data) sent to the CHA's home agent. Unit: uncore_cha event=0x50,umask=0x08TOR Inserts : All requests from iA Cores. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_llcprefrfounc_cha_tor_inserts.ia_miss_rfoevent=0x35,umask=0xCC43FE04event=0xc0unc_cha_tor_inserts.ia_rfo_prefTOR Occupancy : DRds issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_drd_remoteevent=0x36,umask=0xC8177E01event=0x35,umask=0xC8F3FE04unc_iio_data_req_of_cpu.cmpd.part2unc_iio_data_req_by_cpu.mem_write.part4unc_iio_txn_req_of_cpu.mem_read.part5event=0xc2,ch_mask=0x20,fc_mask=0x04,umask=0x03event=0xd5,fc_mask=0x04,umask=0x40event=0x2d,umask=0x01event=0x35,umask=0xCC42FF04event=0x35,umask=0xCC437F04Counts the number of (demand and L1 prefetchers) core requests rejected by the L2 queue (L2Q) due to a full or nearly full condition, which likely indicates back pressure from L2Q.  It also counts requests that would have gone directly to the External Queue (XQ), but are rejected due to a full or nearly full condition, indicating back pressure from the IDI link.  The L2Q may also reject transactions  from a core to ensure fairness between cores, or to delay a cores dirty eviction when the address conflicts incoming external snoops.  (Note that L2 prefetcher requests that are dropped are not counted by this event).  Counts on a per core basisevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0477ocr.uc_rd.l3_hit.snoop_hit_with_fwdc0_stalls.load_llc_hitCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that have any type of responseCounts the total number of branch instructions retired for all branch types (Precise event)ld_blocks.allCounts the number of issue slots every cycle that were not consumed by the backend due to branch mispredictstopdown_fe_bound.frontend_bandwidthevent=0x35,umask=0xC001FE01,config1=0x40041e33TOR Inserts : CRDs issued by iA Cores : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x35,umask=0xC8A7FE01event=0x35,umask=0xC87FDE01TOR Occupancy : All requests from iA Cores that Hit the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsPCIe Completion Buffer Occupancy : Part 7 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 7Number Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4event=0x8,period=200003,umask=0xecycles:k / cycles( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE ) / CPU_CLK_UNHALTED.DISTRIBUTEDInstructions per Branch (lower number means higher occurrence rate). Unit: cpu_core Counts the total number of issue slots that were not consumed by the backend because allocation is stalled due to a mispredicted jump or a machine clear. Unit: cpu_atom Cycle cost per DRAM hit. Unit: cpu_atom Percent of instruction miss cost that hit in DRAM. Unit: cpu_atom mem_scheduler_block.st_bufCounts the number of cycles that uops are blocked due to a store buffer full condition. Unit: cpu_atom Counts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 16 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom event=0xc1,period=1000003,umask=0x10memory_activity.stalls_l1d_missCount all other microcode assist beyond FP, AVX_TILE_MIX and A/D assists (counted by their own sub-events). This includes assists at uop writeback like AVX* load/store (non-FP) assists, Null Assist in SNC (due to lack of FP precision format convert with FMA3x3 uarch) or assists generated by ROB (like assists to due to Missprediction for FSW register - fixed in SNC). Unit: cpu_core Counts the number of issue slots every cycle that were not consumed by the backend due to fast nukes such as memory ordering and memory disambiguation machine clears. Unit: cpu_atom Counts the number of issue slots every cycle that were not delivered by the frontend due to latency related stalls including BACLEARs, BTCLEARs, ITLB misses, and ICache misses. Unit: cpu_atom Return instructions retired (Precise event). Unit: cpu_core event=0xe7,period=1000003,umask=0x13TMA slots wasted due to incorrect speculations. Unit: cpu_core Uops executed on ports 7 and 8. Unit: cpu_core unc_m_prefetch_rdunc_m_dram_thermal_warmWrite CAS command sent to DRAM. Unit: uncore_imc Counts every 64B read  request entering the Memory Controller 0 to DRAM (sum of all channels). Unit: uncore_imc event=0x5,period=1000003,umask=0x90Page walks completed due to a demand data load to a 2M/4M page. Unit: cpu_core event=0x13,cmask=1,period=100003,umask=0x10Page walks completed due to a demand data store to a 1G page. Unit: cpu_core event=0x11,period=100003,umask=0x20MEM_LOAD_L3_MISS_RETIRED.REMOTE_DRAM  Supports address when preciseevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x4003C0001Counts the total number of uops delivered by the Microcode Sequencer (MS)Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAMThis event is deprecated. Refer to new event ARITH.DIV_ACTIVEcpu_clk_unhalted.c0_waitNumber of cycles total of 0 uops executed on all ports, Reservation Station (RS) was not empty, the Store Buffer (SB) was not full and there was no outstanding loadA version of INST_RETIRED that allows for a precise distribution of samples across instructions retired. It utilizes the Precise Distribution of Instructions Retired (PDIR++) feature to fix bias in how retired instructions get sampled. Use on Fixed Counter 0 (Precise event)unc_m_cas_count.rd_pre_underfillunc_m_pre_count.pgt_pch1event=0x83,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000001event=0x83,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000002TOR Inserts for DRd Pref misses from local IA targeting local memory. Unit: uncore_cha Valid Flits Received : Slot 2. Unit: uncore_upi Valid Flits Received : Slot NULL or LLCRD Empty. Unit: uncore_upi event=0x83,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000004event=0xc2,ch_mask=0x01,fc_mask=0x07,umask=0x0000000004TOR Inserts; LLCPrefRFO hits from local IA. Unit: uncore_cha TOR Inserts; misses from Local IA. Unit: uncore_cha TOR Inserts : MMCFG Access. Unit: uncore_cha TOR Inserts; DRd Opt Pref hits from local IA. Unit: uncore_cha TOR Occupancy; Hits from local IA. Unit: uncore_cha TOR Occupancy : All from Local IO. Unit: uncore_cha unc_cha_tor_occupancy.missunc_cha_tor_occupancy.mmioevent=0x36,umask=0x00c887ff01event=0x36,umask=0x00C896FE01event=0x36,umask=0x00ccd7ff01TOR Inserts : DRd_Prefs issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x35,umask=0x00CD47FF01unc_cha_tor_inserts.io_wbmtoiunc_cha_tor_occupancy.ia_miss_drd_pref_local_pmmTOR Occupancy : ItoMCacheNears, indicating a partial write request, from IO Devices that missed the LLC. Unit: uncore_cha TOR Inserts : PMM Access. Unit: uncore_cha ic_fetch_stall.ic_stall_anyl2_cache_req_stat.ls_rd_blk_xCore to L2 cacheable request access status (not including L2 Prefetch). Data cache request miss in L2 (all types)event=0x64,umask=0xf6event=0x1,umask=0x80ex_ret_brn_tknamd_dfThe number of operations (uOps) dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number uOps assigned to pipe 1fp_ret_sse_avx_ops.sp_div_flopsevent=0x29,umask=0x04ls_l1_d_tlb_miss.allTotal Page Table Walks IC Type 1Cycles not in HaltOC Mode Switch. OC to IC mode switchevent=0x29,umask=0x07l2_pf_hit_l2 + l2_pf_miss_l2_hit_l3 + l2_pf_miss_l2_l3tlbuops_retiredL2 Branch Prediction Overrides Existing Prediction (speculative)ex_ret_cond_mispevent=0xe,umask=0x02ls_ret_cl_flushNumber of stores dispatched. Counts the number of operations dispatched to the LS unit. Unit Masks ADDedNumber of loads dispatched. Counts the number of operations dispatched to the LS unit. Unit Masks ADDedCount of dispatched Ops from DecoderThe number of times the instruction decoder overrides the predicted targetic_tag_hit_miss.instruction_cache_missevent=0x44,umask=0x02event=0x44,umask=0x01The number of 64B misaligned (i.e., cacheline crossing) loadsHardware Prefetch Data Cache Fills by Data Source. From DRAM or IO connected in different Nodels_hw_pf_dc_fills.mem_io_localde_dis_dispatch_token_stalls1.taken_brnch_buffer_rsrcde_dis_dispatch_token_stalls2.int_sch3_token_stallevent=0x44,umask=0xffdecoderRetired taken branch instructions mispredictedl2_request_g1.allevent=0x70,umask=0x01L2 prefetches accepted by the L2 pipeline which hit in the L2 cache of all typesl2_pf_miss_l2_hit_l3.l1_streamL3 cache fill requests sourced from extension memory (CXL) in the same NUMA node. Unit: amd_l3 local_processor_read_data_beats_cs3local_processor_read_data_beats_cs8local_processor_write_data_beats_cs0local_processor_write_data_beats_cs5event=0x1df,umask=0x7ffWrite data beats (64 bytes) for remote processor at Coherent Station (CS) 9local_socket_upstream_read_beats_iom3event=0x81f,umask=0x7ffRead data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 0local_socket_inf1_inbound_data_beats_ccm6remote_socket_inf0_inbound_data_beats_ccm3remote_socket_inf0_inbound_data_beats_ccm7event=0x5de,umask=0xbferemote_socket_inf0_outbound_data_beats_ccm1remote_socket_inf1_outbound_data_beats_ccm2local_socket_outbound_data_beats_link5event=0x8,umask=0x02Retired vector floating-point blend opsevent=0xb,umask=0x07event=0xc,umask=0x02event=0xc,umask=0x50fp_pack_ops_retired.fp256_logicalRetired 128-bit packed integer add opsRetired 256-bit packed integer logical opsNon-cacheable store commits cancelled due to the non-cacheable commit buffer being fullNumber of cycles dispatch is stalled for integer scheduler queue 2 tokensNumber of cycles dispatch is stalled for retire queue tokensde_no_dispatch_per_slot.smt_contentionL2 cache accesses from L2 cache hardware prefetcherl1_demand_data_cache_fills_from_same_ccxOutbound data from all links (local socket)umc_data_bus_utilizationumc_cas_cmd_read_ratioumc_cas_cmd_write_ratiod_ratio(umc_act_cmd.all * 1000, umc_mem_clk)uncore_imc_free_running.cache_missL1_DCACHE_REFILLPC_IMM_BRANCHEVENT_07HEVENT_26HEVENT_2EHEVENT_8DHEVENT_93HEVENT_ADHEVENT_B7HEVENT_E8HEVENT_FEHEVENT_120HEVENT_12DHEVENT_12FHEVENT_13FHEVENT_155HEVENT_165HEVENT_1A9HEVENT_28DHEVENT_2BCHEVENT_2E0HEVENT_31CHEVENT_326HEVENT_344HEVENT_358HEVENT_375HEVENT_38BHEVENT_3DCHEVENT_3EBHEVENT_3F9HBR_MIS_PRED_RETIREDL2D_TLB_WRhnf_cache_misshnf_slc_fill_invalid_wayhni_wdb_allocsbsx_awvalid_no_awreadycxla_tx_tlp_link2WRITE_THROUGH_STORESLSU_TOUCH_LINE_ALIAS_VS_FSQ_WB0_WB1LSU_INSTR_COMPLETEDBRANCH_LINK_STACK_MISPREDICTEDFOLDED_BRANCHESL2SQ_FULL_CYCLESGROUP_COMPLETEDDVT0_DETECTEDDVT4_DETECTEDUCPSOFTINTEL_ATOM_SILVERMONTINTEL_BROADWELLunknown eventdesc: %s
ex_ret_brnRESOURCE_STALLGenuineIntel-6-1Cv18GenuineIntel-6-97testcpuBad_Speculation_SMT1 - ( (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) + (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) )Instruction per taken branchBranch Misprediction Cost: Fraction of TMA slots wasted per non-speculative branch misprediction (retired JEClear)Bad;BrMispredicts_SMTInstructions per FP Arithmetic Scalar Single-Precision instruction (lower number means higher occurrence rate). May undercount due to FMA double countingINST_RETIRED.ANY / ( FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE )L2 cache misses per kilo instruction for all request types (including speculative)CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSCCPU_CLK_UNHALTED.THREAD_P:k / INST_RETIRED.ANY_P:kDRAM_BW_Usecbox@event\=0x36\,umask\=0x3\,filter_opc\=0x182@ / cbox@event\=0x36\,umask\=0x3\,filter_opc\=0x182\,thresh\=1@event=0x48,cmask=1,period=2000003,umask=0x2l1d_pend_miss.pending_cycles_anyevent=0xf2,period=100003,umask=0x5This event counts the number of demand Data Read requests (including requests from L1D hardware prefetchers). These loads may hit or miss L2 cache. Only non rejected loads are countedAll requests that miss L2 cacheThis event counts Demand Data Read requests that access L2 cache, including rejectsevent=0xf0,period=200003,umask=0x10This is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were hits in the mid-level (L2) cache  Supports address when precise.  Spec update: BDM35 (Precise event)This event counts cycles when offcore outstanding cacheable Core Data Read transactions are present in the super queue. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation). See corresponding Umask under OFFCORE_REQUESTS  Spec update: BDM76event=0x60,cmask=1,period=2000003,umask=0x4event=0xe6,period=100003,umask=0x1fThis event counts Decode Stream Buffer (DSB)-to-MITE switch true penalty cycles. These cycles do not include uops routed through because of the switch itself, for example, when Instruction Decode Queue (IDQ) pre-allocation is unavailable, or Instruction Decode Queue (IDQ) is full. SBD-to-MITE switch true penalty cycles happen after the merge mux (MM) receives Decode Stream Buffer (DSB) Sync-indication until receiving the first MITE uop. 
MM is placed before Instruction Decode Queue (IDQ) to merge uops being fed from the MITE and Decode Stream Buffer (DSB) paths. Decode Stream Buffer (DSB) inserts the Sync-indication whenever a Decode Stream Buffer (DSB)-to-MITE switch occurs.
Penalty: A Decode Stream Buffer (DSB) hit followed by a Decode Stream Buffer (DSB) miss can cost up to six cycles in which no uops are delivered to the IDQ. Most often, such switches from the Decode Stream Buffer (DSB) to the legacy pipeline cost 02 cyclesNumber of switches from DSB (Decode Stream Buffer) or MITE (legacy decode pipeline) to the Microcode SequencerLoads with latency value being above 16  Spec update: BDM100, BDM35 (Must be precise)mem_trans_retired.load_latency_gt_32Loads with latency value being above 4  Spec update: BDM100, BDM35 (Must be precise)This event counts loads with latency value being above 512  Spec update: BDM100, BDM35 (Must be precise)Loads with latency value being above 8  Spec update: BDM100, BDM35 (Must be precise)event=0x54,period=2000003,umask=0x2tx_mem.abort_hle_elision_buffer_unsupported_alignmentThis event counts the unhalted core cycles during which the thread is in the ring 0 privileged modeevent=0x5c,period=2000003,umask=0x2br_inst_retired.conditionalThis event counts both taken and not taken speculative and retired mispredicted branch instructionsbr_misp_exec.taken_conditionalReference cycles when the at least one thread on the physical core is unhalted (counts at 100 MHz rate)Cycles Uops delivered by the LSD, but didn't come from the decoderlsd.uopsThis event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 6Cycles where at least 4 uops were executed per-threaduops_retired.total_cyclesM line forwarded from remote cache along with writeback to memory. Unit: uncore_ha M line forwarded from remote cache with no writeback to memory. Unit: uncore_ha event=0x86event=0x80,occ_sel=3(UNC_P_POWER_STATE_OCCUPANCY.CORES_C6 / UNC_P_CLOCKTICKS) * 100.power_state_occupancy.cores_c6 %(UNC_P_FREQ_MAX_LIMIT_THERMAL_CYCLES / UNC_P_CLOCKTICKS) * 100.unc_p_freq_max_os_cyclesdtlb_load_misses.walk_completed_2m_4mStore misses in all DTLB levels that cause completed page walks  Spec update: BDM69event=0x85,period=100003,umask=0x20page_walker_loads.dtlb_memoryevent=0xbc,period=2000003,umask=0x22Miss in last-level (L3) cache. Excludes Unknown data-source  Supports address when precise.  Spec update: BDM100, BDE70 (Precise event)offcore_requests.all_requestsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020091offcore_response.all_pf_code_rd.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0090offcore_response.all_pf_data_rd.l3_hit.snoop_hit_no_fwdoffcore_response.corewb.supplier_none.snoop_hitmoffcore_response.demand_code_rd.supplier_none.snoop_not_neededCounts demand data readsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020001offcore_response.other.l3_hit.snoop_hitmCounts all prefetch (that bring data to LLC only) code readsCounts prefetch (that bring data to L2) data readsoffcore_response.pf_l2_rfo.l3_hit.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0080offcore_response.pf_l3_data_rd.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020100offcore_response.all_pf_code_rd.l3_miss_local_dram.any_snoopoffcore_response.all_pf_data_rd.l3_miss_local_dram.snoop_missoffcore_response.all_pf_rfo.l3_miss.snoop_hit_no_fwdoffcore_response.all_rfo.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0004offcore_response.demand_data_rd.l3_miss.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000040offcore_response.pf_l2_data_rd.l3_miss_local_dram.any_snoopoffcore_response.pf_l2_rfo.l3_miss.snoop_not_neededoffcore_response.pf_l3_code_rd.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000080offcore_response.pf_l3_rfo.l3_miss.snoop_not_neededoffcore_response.pf_l3_rfo.l3_miss_local_dram.any_snoopA cross-core snoop initiated by this Cbox due to processor core memory request which misses in some processor coreL3 Lookup read request that access cache and found line in I-state. Unit: uncore_cbox Cycles with at least one request outstanding is waiting for data return from memory controller. Account for coherent and non-coherent requests initiated by IA Cores, Processor Graphics Unit, or LLCbdx metricsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C07F7Counts all data/code/rfo reads (demand & prefetch)miss the L3 and the data is returned from local dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x063BC007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC08FFFqpi_ctl_bandwidth_tx(UNC_M_POWER_CHANNEL_PPD / UNC_M_CLOCKTICKS) * 100.event=0x29,period=200000,umask=0x74event=0x29,period=200000,umask=0x48event=0x29,period=200000,umask=0x51event=0x29,period=200000,umask=0x52All read requests from L1 instruction and data cachesevent=0x25,period=200000,umask=0x40Retired computational Streaming SIMD Extensions (SSE) packed-single instructionsevent=0xb0,period=2000000,umask=0x0SIMD packed arithmetic micro-ops executedFloating point computational micro-ops retired (Must be precise)event=0x10,period=2000000,umask=0x2event=0x80,period=200000,umask=0x3This event counts the cycles where 1 or more uops are issued by the micro-sequencer (MS), including microcode assists and inserted flows, and written to the IQevent=0x5,period=200000,umask=0x91Invalidate bus transactionsbus_trans_wb.selfevent=0x3a,period=200000,umask=0x0event=0x77,period=200000,umask=0x8br_inst_retired.takenAll indirect calls, including both register and memory indirectevent=0x89,period=200000,umask=0x4cpu_clk_unhalted.refReference cycles when core is not haltedevent=0x3,period=200000,umask=0x1Duration of D-side only page walksCounts when a modified (dirty) cache line is evicted from the data L1 cache and needs to be written back to memory.  No count will occur if the evicted line is clean, and hence does not require a writebackCounts memory requests originating from the core that reference a cache line in the L2 cachemem_load_uops_retired.hitmevent=0xb7,period=100007,umask=0x1event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200003091Counts data reads (demand & prefetch) that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of writeback transactions caused by L1 or L2 cache evictions that miss the L2 cacheCounts demand cacheable data reads of full cache lines that are outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that true miss for the L2 cache with a snoop miss in the other processor moduleCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts partial cache line data writes to uncacheable write combining (USWC) memory region  that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts partial cache line data writes to uncacheable write combining (USWC) memory region  that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000020Counts data cache lines requests by software prefetch instructions that miss the L2 cacheCounts data cache lines requests by software prefetch instructions that true miss for the L2 cache with a snoop miss in the other processor moduleevent=0xc2,period=2000003,umask=0x8Counts BACLEARS on Jcc (Jump on Conditional Code/Jump if Condition is Met) branchesbaclears.returnCounts the number of taken branch instructions retired (Must be precise)br_inst_retired.jccCounts near return branch instructions retired (Must be precise)Counts the number of demand and L1 prefetcher requests rejected by the L2Q due to a full or nearly full condition which likely indicates back pressure from L2Q. It also counts requests that would have gone directly to the XQ, but are rejected due to a full or nearly full condition, indicating back pressure from the IDI link. The L2Q may also reject transactions from a core to insure fairness between cores, or to delay a core's dirty eviction when the address conflicts with incoming external snoopsCounts demand cacheable data reads of full cache lines miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests generated by L2 prefetcher have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000014800Counts INST_RETIRED.ANY using the Reduced Skid PEBS feature that reduces the shadow in which events aren't counted allowing for a more unbiased distribution of samples across instructions retired (Must be precise)l1d_pend_miss.request_fb_fullDemand requests to L2 cache  Spec update: HSD78, HSM80Counts all L2 store RFO requestsRetired load uops missed L2. Unknown data source excluded  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)Retired load uops missed L3. Excludes unknown data source   Supports address when precise.  Spec update: HSD74, HSD29, HSD25, HSM26, HSM30 (Precise event)Counts demand data reads hit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedThis event counts uops delivered by the Front-end with the assistance of the microcode sequencer.  Microcode assists are used for complex instructions or scenarios that can't be handled by the standard decoder.  Using other instructions, if possible, will usually improve performanceThis event counts the number of memory ordering machine clears detected. Memory ordering machine clears can result from memory address aliasing or snoops from another hardware thread or core to data inflight in the pipeline.  Machine clears can have a significant performance impact if they are happening frequentlyoffcore_response.demand_data_rd.l3_miss.local_dramNumber of times a transactional abort was signaled due to a data capacity limitation for transactional writesNumber of times an HLE transactional execution aborted due to NoAllocatedElisionBuffer being non-zeroCounts the number of thread cycles while the thread is not in a halt state. The thread enters the halt state when it is running the HLT instruction. The core frequency may change from time to time due to power or thermal throttlingExecution stalls due to memory subsystemThis events counts the cycles where at least three uop were executed. It is counted per thread  Spec update: HSD144, HSD30, HSM31unc_cbo_xsnp_response.hitm_externalThis event counts store operations from a 4K page that miss the first DTLB level but hit the second and do not cause page walkspage_walker_loads.ept_dtlb_l2event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0600400122offcore_response.demand_rfo.llc_miss.local_dramRetired load uops which data sources were LLC hit and cross-core snoop missed in on-pkg core cache (Precise event)Retired load uops with L2 cache misses as data sources (Precise event)Counts demand & prefetch data reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.all_data_rd.llc_hit.no_snoop_neededevent=0x10,period=2000003,umask=0x10Number of SSE* or AVX-128 FP Computational scalar double-precision uops issued this cycleevent=0x10,period=2000003,umask=0x20event=0x10,period=2000003,umask=0x1other_assists.avx_storesimd_fp_256.packed_doubleevent=0xab,period=2000003,umask=0x1Counts LLC replacementsevent=0xa1,any=1,period=2000003,umask=0x30Filter on processor core initiated cacheable read requestsFilter on external snoop requests. Unit: uncore_cbox event=0x8,period=100003,umask=0x81Cycle count for an Extended Page table walk.  The Extended Page Directory cache is used by Virtual Machine operating systems while the guest operating systems use the standard TLB cachesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0040Counts prefetch (that bring data to L2) data reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0010Counts all prefetch (that bring data to LLC only) code reads that hit in the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x67f800010offcore_response.pf_l2_data_rd.llc_miss.remote_hit_forwardCounts prefetch (that bring data to LLC only) data reads that miss in the LLCllc_references.pcie_ns_readunc_c_tor_occupancy.miss_localevent=0xe,edge=1Retired store uops that split across a cacheline boundary (Precise event)Number of AVX-256 Computational FP double precision uops issued this cycleThis event counts the number of cycles spent executing performance-sensitive flags-merging uops. For example, shift CL (merge_arith_flags). For more details, See the Intel? 64 and IA-32 Architectures Optimization Reference Manualoffcore_response.any_pf_l2.any_responseCounts any Prefetch requests that accounts for responses which hit its own tile's L2 with data in F stateCounts any Read request  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x10000832f7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400022offcore_response.bus_locks.l2_hit_this_tile_sCounts Bus locks and split lock requests that accounts for responses which hit its own tile's L2 with data in S stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400004Counts demand code reads and prefetch code reads that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeoffcore_response.demand_code_rd.l2_hit_this_tile_foffcore_response.demand_rfo.l2_hit_near_tileoffcore_response.demand_rfo.l2_hit_this_tile_foffcore_response.partial_reads.outstandingoffcore_response.partial_writes.l2_hit_this_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800402000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004002000offcore_response.pf_software.l2_hit_far_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000200offcore_response.uc_code_reads.outstandingThis event counts the number of packed vector SSE, AVX, AVX2, and AVX-512 micro-ops retired (floating point, integer and store) except for loads (memory-to-register mov-type micro-ops), packed byte and word multipliesevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x00808032f7Counts Demand cacheable data write requests  that accounts for responses from MCDRAM (local and far)offcore_response.any_rfo.mcdram_farCounts Bus locks and split lock requests that accounts for data responses from DRAM Farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000004event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600004offcore_response.demand_data_rd.mcdramoffcore_response.partial_reads.ddrevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600080event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200080event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600200recycleq.any_ldCounts the total number of core cycles the Alloc pipeline is stalled when any one of the reservation stations is fullcache_lock_cycles.l1d_l2L1 data cache read in S stateevent=0x42,period=2000000,umask=0x4L2 data prefetches in the S stateL2 lines alloacatedL2 modified lines evicted by a demand requestmem_inst_retired.latency_above_threshold_0event=0xb,period=1000,umask=0x10,ldlat=0x80Retired loads that hit sibling core's L2 in modified or unmodified states (Precise Event)event=0xf,period=4000,umask=0x80Offcore data reads satisfied by the LLC and HIT in a sibling coreOffcore data reads satisfied by the LLC or local DRAMoffcore_response.any_ifetch.llc_hit_other_core_hitoffcore_response.any_ifetch.local_cache_dramOffcore code reads satisfied by a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4FFevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x222event=0xb7,period=100000,umask=0x1,offcore_rsp=0x422Offcore RFO requests satisfied by a remote cacheoffcore_response.data_in.local_cache_dramOffcore demand code reads that HIT in a remote cacheAll offcore demand RFO requestsoffcore_response.demand_rfo.io_csr_mmioevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1802event=0xb7,period=100000,umask=0x1,offcore_rsp=0x180offcore_response.other.llc_hit_other_core_hitOffcore other requests that HIT in a remote cacheoffcore_response.pf_data.remote_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x810event=0xb7,period=100000,umask=0x1,offcore_rsp=0x8070Offcore prefetch requests satisfied by the LLC and HIT in a sibling coreCacheable loads delayed with L1D block codeSSE2 integer Uops128 bit SIMD integer logical operationssimd_int_128.shuffle_movemacro_insts.fusions_decodedOffcore data reads that missed the LLCOffcore demand data reads satisfied by any DRAMevent=0x6c,period=2000000,umask=0x1load_dispatch.mobbr_misp_exec.condbr_misp_exec.indirect_near_callPartial register stall cyclesevent=0xc7,period=200000,umask=0x2uops_executed.core_active_cyclesCounts the number of lines that are evicted by L2 cache when triggered by an L2 cache fill. Those lines are in Modified state. Modified lines are written back to L3Counts the RFO (Read-for-Ownership) requests that hit L2 cacheRetired load instructions with L3 cache hits as data sources  Supports address when precise (Precise event)Counts the number of offcore outstanding demand rfo Reads transactions in the super queue every cycle. The 'Offcore outstanding' state of the transaction lasts from the L2 miss until the sending transaction completion to requestor (SQ deallocation). See the corresponding Umask under OFFCORE_REQUESTSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x401C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40040001offcore_response.demand_rfo.l3_hit_s.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4001C8000offcore_response.other.l3_hit_e.snoop_hit_no_fwdoffcore_response.other.l3_hit_m.snoop_hitmRetired instructions that are fetched after an interval where the front-end had at least 3 bubble-slots for a period of 2 cycles which was not interrupted by a back-end stall (Precise event)frontend_retired.latency_ge_64Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 8 cycles which was not interrupted by a back-end stall (Precise event)icache_64b.iftag_missCounts the number of cycles 4 uops were delivered to the Instruction Decode Queue (IDQ) from the MITE (legacy decode pipeline) path. Counting includes uops that may 'bypass' the IDQ. During these cycles uops are not being delivered from the Decode Stream Buffer (DSB)Counts cycles during which uops are being delivered to Instruction Decode Queue (IDQ) while the Microcode Sequencer (MS) is busy. Counting includes uops that may 'bypass' the IDQ. Uops maybe initiated by Decode Stream Buffer (DSB) or MITENumber of times an HLE execution aborted due to HLE-unfriendly instructions and certain unfriendly events (such as AD assists etc.)Counts the number of machine clears due to memory order conflicts  Spec update: SKL089event=0xb7,period=100003,umask=0x1,offcore_rsp=0x7C400004offcore_response.demand_data_rd.l3_miss.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x7C400002offcore_response.demand_rfo.l3_miss_local_dram.snoop_non_dramCounts the number of reference cycles when the core is not in a halt state. The core enters the halt state when it is running the HLT instruction or the MWAIT instruction. This event is not affected by core frequency changes (for example, P states, TM2 transitions) but has the same incrementing frequency as the time stamp counter. This event can approximate elapsed time while the core was not in a halt state. This event has a constant ratio with the CPU_CLK_UNHALTED.REF_XCLK event. It is counted on a dedicated fixed counter, leaving the four (eight when Hyperthreading is disabled) programmable counters available for other events. Note: On all current platforms this event stops counting during 'throttling (TM)' states duty off periods the processor is 'halted'.  The counter update is done at a lower clock rate then the core clock the overflow status bit for this counter may appear 'sticky'.  After the counter has overflowed and software clears the overflow status bit and resets the counter to less than MAX. The reset value to the counter is not clocked immediately so the overflow status bit will flip 'high (1)' and generate another PMI (if enabled) after which the reset value gets clocked into the counter. Therefore, software will get the interrupt, read the overflow status bit '1 for bit 34 while the counter value is less than MAX. Software should ignore this caseNumber of decoders utilized in a cycle when the MITE (legacy decode pipeline) fetches instructionspartial_rat_stalls.scoreboardevent=0xcc,period=2000003,umask=0x40Counts the number of x87 uops dispatched( BR_INST_RETIRED.CONDITIONAL - BR_INST_RETIRED.NOT_TAKEN )  / BR_INST_RETIRED.ALL_BRANCHES64 * OFFCORE_REQUESTS.ALL_REQUESTS / 1000000000 / duration_timeFill Buffer (FB) true hits per kilo instructions for retired demand loadsPage walk completed due to a demand data load to a 1G pagedtlb_store_misses.walk_activeAll LoadsThis event counts the number of load ops retiredoffcore_response.any_request.l2_miss.snoop_missCounts DCU hardware prefetcher data read that have any response typeUops with lock semanticsThis event counts the number of load uops retired (Precise Event) (Precise event)Retired load uops that miss the STLB. (Precise Event - PEBS) (Precise event)offcore_response.all_pf_code_rd.llc_hit.snoop_missoffcore_response.all_pf_rfo.llc_hit.hit_other_core_no_fwdCounts data/code/rfo reads (demand & prefetch) that hit in the LLC and the snoops sent to sibling cores return clean responseREQUEST = ANY_REQUEST and RESPONSE = LLC_MISS_LOCAL and SNOOP = DRAMevent=0xb0,period=100000,umask=0x8REQUEST = ANY IFETCH and RESPONSE = LOCAL_CACHEREQUEST = DEMAND_DATA and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DEMAND_IFETCH and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5002offcore_response.other.local_dram_and_remote_cache_hitREQUEST = PREFETCH and RESPONSE = LOCAL_CACHEsq_misc.lru_hintsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf808offcore_response.corewb.other_local_dramREQUEST = DATA_IFETCH and RESPONSE = OTHER_LOCAL_DRAMREQUEST = DATA_IN and RESPONSE = ANY_LLC_MISSREQUEST = OTHER and RESPONSE = OTHER_LOCAL_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2702offcore_response.all_pf_rfo.l3_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0001Counts L1 data cache hardware prefetch requests and software prefetch requests that have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10020offcore_response.all_pf_rfo.l3_miss.snoop_miss_or_no_fwdoffcore_response.demand_code_rd.l3_miss.remote_hitmoffcore_response.pf_l3_rfo.l3_miss.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000100Core cycles where the core was running with power-delivery for license level 1.  This includes high current AVX 256-bit instructions as well as low current AVX 512-bit instructionsFraction of Core cycles where the core was running with power-delivery for baseline license level 0.  This includes non-AVX codes, SSE, AVX 128-bit, and low-current AVX 256-bit codesFraction of Core cycles where the core was running with power-delivery for baseline license level 0.  This includes non-AVX codes, SSE, AVX 128-bit, and low-current AVX 256-bit codes. SMT version; use when SMT is enabled and measuring per logical CPUCounts CAS (Column Access Select) underfill read commands issued to DRAM due to a partial write, on a per channel basis.  CAS commands are issued to specify the address to read or write on DRAM, and this command counts underfill reads.  Partial writes must be completed by first reading in the underfill from DRAM and then merging in the partial write data before writing the full line back to DRAM. This event will generally count about the same as the number of partial writes, but may be slightly less because of partials hitting in the WPQ (due to a previous write request)event=0x20MMIO writes. Derived from unc_cha_tor_inserts.ia_miss. Unit: uncore_cha event=0x83,ch_mask=0x04,fc_mask=0x07,umask=0x01event=0x83,ch_mask=0x01,fc_mask=0x07,umask=0x01,ch_mask=0x1fUNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART0 +UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART1 +UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART2 +UNC_IIO_DATA_REQ_OF_CPU.MEM_WRITE.PART3unc_cha_llc_victims.total_eCounts the number of lines that were victimized on a fill.  This can be filtered by the state that the line was inLines Victimized; Lines in F State. Unit: uncore_cha Lines Victimized; Lines in S State. Unit: uncore_cha RspCnflct* Snoop Responses Received. Unit: uncore_cha event=0x5c,umask=0x01Counts clockticks of the 1GHz trafiic controller clock in the IIO unitPCIe Completion Buffer occupancy of completions with data: Part 3. Unit: uncore_iio unc_iio_data_req_by_cpu.mem_read.part2unc_iio_data_req_by_cpu.mem_read.part3Peer to peer write request of 4 bytes made to IIO Part3 by a different IIO unit. Unit: uncore_iio Write request of up to a 64 byte transaction is made by IIO Part3 to Memory. Unit: uncore_iio event=0x25Multi-socket cacheline Directory update from I to S. Unit: uncore_m2m unc_m2m_txc_bl_occupancy.allunc_upi_direct_attempts.d2cunc_upi_rxl_bypassed.slot0Counts incoming FLITs (FLow control unITs) whcih bypassed the slot2 RxQ buffer (Receive Queue)  and passed directly to the Egress.  This is a latency optimization, and should generally be the common case.  If this value is less than the number of FLITs transfered, it implies that there was queueing getting onto the ring, and thus the transactions saw higher latencyCounts retired load instructions with remote Intel Optane DC persistent memory as the data source and the data request missed L3 (AppDirect or Memory Mode) and DRAM cache(Memory Mode). Precise event  Supports address when precise (Precise event)ocr.all_data_rd.l3_hit.hit_other_core_no_fwdocr.all_data_rd.l3_hit_e.hitm_other_coreOCR.ALL_PF_DATA_RD.L3_HIT.HITM_OTHER_CORE OCR.ALL_PF_DATA_RD.L3_HIT.HITM_OTHER_CORE OCR.ALL_PF_DATA_RD.L3_HIT.HITM_OTHER_COREocr.all_pf_data_rd.l3_hit.no_snoop_neededocr.all_pf_data_rd.l3_hit_e.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200490ocr.all_pf_data_rd.l3_hit_s.hit_other_core_no_fwdocr.all_pf_rfo.l3_hit_f.hitm_other_coreOCR.ALL_READS.L3_HIT.SNOOP_MISS OCR.ALL_READS.L3_HIT.SNOOP_MISSOCR.ALL_READS.L3_HIT_E.ANY_SNOOP  OCR.ALL_READS.L3_HIT_E.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000807F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x802007F7OCR.ALL_READS.L3_HIT_M.SNOOP_NONEocr.all_rfo.l3_hit_e.any_snoopocr.all_rfo.l3_hit_m.any_snoopOCR.ALL_RFO.L3_HIT_M.HIT_OTHER_CORE_FWD  OCR.ALL_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100122Counts all demand code reads OCR.DEMAND_CODE_RD.L3_HIT.SNOOP_NONEocr.demand_code_rd.l3_hit_m.any_snoopocr.demand_code_rd.l3_hit_m.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200001ocr.demand_data_rd.l3_hit_m.hit_other_core_no_fwdocr.demand_rfo.l3_hit_m.any_snoopCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDocr.demand_rfo.l3_hit_m.snoop_noneCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C8000Counts any other requests  OCR.OTHER.L3_HIT_F.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80208000Counts any other requests  OCR.OTHER.L3_HIT_M.ANY_SNOOPocr.pf_l1d_and_sw.l3_hit.hit_other_core_fwdCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.PF_L2_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDocr.pf_l2_data_rd.l3_hit.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.pf_l3_data_rd.l3_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040080Counts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_HIT.NO_SNOOP_NEEDED OCR.PF_L3_RFO.L3_HIT.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0100ocr.pf_l3_rfo.l3_hit.snoop_missocr.pf_l3_rfo.l3_hit_e.hit_other_core_fwdCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_hit_e.any_snoopThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_hit_m.any_snoopoffcore_response.all_reads.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.SNOOP_HIT_WITH_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.all_reads.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_RFO.SUPPLIER_NONE.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_RFO.SUPPLIER_NONE.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020122This event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_E.ANY_SNOOPoffcore_response.demand_code_rd.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDoffcore_response.demand_data_rd.l3_hit_e.hit_other_core_fwdoffcore_response.demand_data_rd.l3_hit_f.no_snoop_neededoffcore_response.demand_data_rd.supplier_none.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDoffcore_response.other.l3_hit.snoop_hit_with_fwdThis event is deprecated. Refer to new event OCR.OTHER.PMM_HIT_LOCAL_PMM.SNOOP_NONEoffcore_response.pf_l1d_and_sw.l3_hit_e.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l2_data_rd.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400020offcore_response.pf_l3_data_rd.l3_hit_e.any_snoopoffcore_response.pf_l3_data_rd.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.SUPPLIER_NONE.HITM_OTHER_COREOCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEocr.all_data_rd.l3_miss_remote_hop1_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000490OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSocr.all_reads.l3_miss_remote_hop1_dram.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000122ocr.demand_code_rd.l3_miss_local_dram.no_snoop_neededocr.demand_code_rd.l3_miss_remote_hop1_dram.any_snoopocr.demand_code_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdocr.demand_code_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.SNOOP_MISSocr.demand_data_rd.l3_miss_remote_hop1_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000002Counts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.HIT_OTHER_CORE_FWD OCR.PF_L1D_AND_SW.L3_MISS.HIT_OTHER_CORE_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.PF_L1D_AND_SW.L3_MISS.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000400ocr.pf_l2_data_rd.l3_miss_local_dram.hitm_other_coreCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000080offcore_response.all_data_rd.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.SNOOP_NONEoffcore_response.demand_code_rd.l3_miss.hit_other_core_fwdoffcore_response.demand_code_rd.l3_miss.hit_other_core_no_fwdoffcore_response.demand_rfo.l3_miss.hit_other_core_fwdoffcore_response.other.l3_miss.hitm_other_coreoffcore_response.other.l3_miss.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l1d_and_sw.l3_miss.hit_other_core_fwdoffcore_response.pf_l1d_and_sw.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSOCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDED OCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDOCR.ALL_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDED  OCR.ALL_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.all_pf_data_rd.supplier_none.no_snoop_neededOCR.ALL_PF_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDocr.all_pf_rfo.supplier_none.snoop_missOCR.ALL_READS.SUPPLIER_NONE.HIT_OTHER_CORE_FWD  OCR.ALL_READS.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.all_rfo.pmm_hit_local_pmm.snoop_noneocr.demand_code_rd.supplier_none.hitm_other_coreCounts any other requests OCR.OTHER.PMM_HIT_LOCAL_PMM.ANY_SNOOPCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.SUPPLIER_NONE.NO_SNOOP_NEEDEDCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.pf_l3_data_rd.pmm_hit_local_pmm.any_snoopocr.pf_l3_rfo.supplier_none.hitm_other_coreCycles where DRAM ranks are in power down (CKE) mode+C37. Unit: uncore_imc Intel Optane DC persistent memory bandwidth read (MB/sec). Derived from unc_m_pmm_rpq_inserts. Unit: uncore_imc event=0x37,umask=0x8Dirty line read hits(Regular and RFO) to Near Memory(DRAM cache) in Memory Mode. Unit: uncore_m2m Cache lines that have been L2 hardware prefetched but not used by demand accessesCore-originated cacheable requests that missed L3  (Except hardware prefetches to the L3)Counts retired load instructions whose data sources were hits in L3 without snoops required  Supports address when precise (Precise event)ocr.demand_rfo.l3_hit.snoop_hitmocr.hwpf_l1d_and_swpf.l3_hit.snoop_missCounts hardware prefetch data reads (which bring data to L2)  that hit a cacheline in the L3 where a snoop was sent but no other cores had the dataevent=0x32,period=100003,umask=0x8event=0xc7,period=100003,umask=0x4Counts cycles during which uops are being delivered to Instruction Decode Queue (IDQ) while the Microcode Sequencer (MS) is busy. Uops maybe initiated by Decode Stream Buffer (DSB) or MITENumber of switches from DSB or MITE to the MS(BR_INST_RETIRED.NEAR_TAKEN - BR_INST_RETIRED.COND_TAKEN - 2 * BR_INST_RETIRED.NEAR_CALL) / BR_INST_RETIRED.ALL_BRANCHESCounts the number of times an HLE execution aborted due to unfriendly events (such as interrupts)Counts the number of Machine Clears detected dye to memory ordering. Memory Ordering Machine Clears may apply when a memory read may not conform to the memory ordering rules of the x86 architectureocr.hwpf_l2_data_rd.l3_missCounts demand data read requests that miss the L3 cacheCounts the number of times an RTM execution aborted due to HLE-unfriendly instructionsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x184008000Cycles when divide unit is busy executing divide or square root operationsCounts not taken branch instructions retired (Precise event)br_misp_retired.indirectevent=0xd,period=500009,umask=0x80Counts cycles where the pipeline is stalled due to serializing operationsCounts cycles when at least 3 micro-ops are executed from any thread on physical coreevent=0xa6,period=1000003,umask=0x80Core-originated cacheable requests that refer to L3 (Except hardware prefetches to the L3)ocr.demand_code_rd.snc_cache.hit_with_fwdCounts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that hit in the L3 or were snooped from another core's caches on the same socketCycles with outstanding code read requests pending.  Code Read requests include both cacheable and non-cacheable Code Reads.  Requests are considered outstanding from the time they miss the core's L2 cache until the transaction completion message is sent to the requestor( UNC_CHA_TOR_INSERTS.IO_HIT_ITOM + UNC_CHA_TOR_INSERTS.IO_MISS_ITOM + UNC_CHA_TOR_INSERTS.IO_HIT_ITOMCACHENEAR + UNC_CHA_TOR_INSERTS.IO_MISS_ITOMCACHENEAR ) * 64 / 1000000000 / duration_timeCounts demand instruction fetches and L1 instruction cache prefetches that were not supplied by the local socket's L1, L2, or L3 cachesCounts demand instruction fetches and L1 instruction cache prefetches that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline is homed locallyCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were not supplied by the local socket's L1, L2, or L3 cachesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F04400002Counts streaming stores that missed the local socket's L1, L2, and L3 cachesocr.demand_code_rd.snc_dramocr.demand_data_rd.remote_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x703000001ocr.reads_to_core.local_socket_pmmLocal read requests that miss the SF/LLC and remote read requests sent to the CHA's home agent. Unit: uncore_cha TOR Inserts : CRds issued by iA Cores that Missed the LLC. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_crdevent=0x35,umask=0xC88FFD01unc_cha_tor_inserts.ia_hit_rfo_prefevent=0x36,umask=0xC80FFF01TOR Inserts : DRds issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed locally. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_drd_remote_pmmunc_cha_tor_inserts.io_miss_pcirdcurTOR Occupancy : PCIRdCurs issued by IO Devices that missed the LLC. Unit: uncore_cha event=0x83,ch_mask=0x01,fc_mask=0x07,umask=0x80unc_iio_txn_req_of_cpu.cmpd.part3unc_iio_data_req_of_cpu.mem_write.part4event=0x83,ch_mask=0x40,fc_mask=0x07,umask=0x04unc_m2m_clockticksClockticks of the mesh to PCI (M2P). Unit: uncore_m2pcie event=0x35,umask=0xCD437F04Counts the number of cycles the core is stalled due to an instruction cache or TLB miss which hit in the LLC or other core with HITE/F/MCounts the number of cycles the core is stalled due to a demand load which hit in the Last Level Cache (LLC) or other core with HITE/F/Mevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0044ocr.all_code_rd.l3_hit.snoop_not_neededCounts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were supplied by the L3 cache where no snoop was needed to satisfy the requestCounts modified writebacks from L1 cache that miss the L2 cache that were supplied by the L3 cacheocr.reads_to_core.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0477ocr.partial_streaming_wr.l3_missCounts all hardware and software prefetches that were not supplied by the L3 cachebus_lock.allevent=0x63,period=200003,umask=0x1Counts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that have any type of responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000000000000477Counts the number of mispredicted JCC (Jump on Conditional Code) branch instructions retired (Precise event)Counts the number of retired loads that are blocked because its address partially overlapped with an older store (Precise event)topdown_be_bound.registerCounts the number of issue slots every cycle that were not consumed by the backend due to the reorder buffer being full (ROB stalls)TOR Inserts : All requests from iA Cores that Hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsPCIe Completion Buffer Inserts of completions with data : Part 2 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 2Data requested by the CPU : Core reporting completion of Card read from Core DRAM : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 5Number Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 5Clockticks of the mesh to PCI (M2P)Counts the number of page walks due to loads that miss the PDE (Page Directory Entry) cacheCounts the number of page walks completed due to load DTLB misses to a 1G pageCounts the number of page walks completed due to stores whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to any page size.  Includes page walks that page faultTotal issue-pipeline slots (per-Physical Core till ICL; per-Logical Processor ICL onward). Unit: cpu_core Backend_Bound_Aux100 * UOPS_RETIRED.IDIV / UOPS_RETIRED.ALLPercentage of all uops which are x87 uops. Unit: cpu_atom Counts the number of cycles the core is stalled due to an instruction cache or tlb miss which hit in the L2, LLC, DRAM or MMIO (Non-DRAM). Unit: cpu_atom Counts the number of cycles that uops are blocked for any of the following reasons:  load buffer, store buffer or RSV full. Unit: cpu_atom event=0xd0,period=1000003,umask=0x5,ldlat=0x20Counts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 4 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom This event is deprecated. Refer to new event L1D_PEND_MISS.L2_STALLS. Unit: cpu_core Retired load instructions that miss the STLB  Supports address when precise (Precise event). Unit: cpu_core Number of PREFETCHNTA instructions executed. Unit: cpu_core Counts randomly selected loads when the latency from first dispatch to completion is greater than 4 cycles  Supports address when precise (Must be precise). Unit: cpu_core event=0x73,period=1000003,umask=0x3Cycles while memory subsystem has an outstanding load. Unit: cpu_core Cycles total of 1 uop is executed on all ports and Reservation Station was not empty. Unit: cpu_core event=0x75,period=2000003,umask=0x1event=0xa4,period=10000003,umask=0x4Uops executed on port 6. Unit: cpu_core Incoming VC1 read request. Unit: uncore_imc Page walks completed due to a demand data load to a 4K page. Unit: cpu_core event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F003C4477event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F3FC04477Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that have any type of responseocr.reads_to_core.remote_memoryThis event counts the cycles the integer divider is busyCounts the number of reference cycles when the core is not in a halt state. The core enters the halt state when it is running the HLT instruction or the MWAIT instruction. This event is not affected by core frequency changes (for example, P states, TM2 transitions) but has the same incrementing frequency as the time stamp counter. This event can approximate elapsed time while the core was not in a halt state. It is counted on a dedicated fixed counter, leaving the eight programmable counters available for other events. Note: On all current platforms this event stops counting during 'throttling (TM)' states duty off periods the processor is 'halted'.  The counter update is done at a lower clock rate then the core clock the overflow status bit for this counter may appear 'sticky'.  After the counter has overflowed and software clears the overflow status bit and resets the counter to less than MAX. The reset value to the counter is not clocked immediately so the overflow status bit will flip 'high (1)' and generate another PMI (if enabled) after which the reset value gets clocked into the counter. Therefore, software will get the interrupt, read the overflow status bit '1 for bit 34 while the counter value is less than MAX. Software should ignore this caseevent=0x10,umask=0x0000000001event=0x3,umask=0x0000000004unc_iio_data_req_of_cpu.peer_write.part7TOR Inserts for DRd misses from local IA. Unit: uncore_cha Valid Flits Received : LLCRD Not Empty. Unit: uncore_upi unc_upi_txl_basic_hdr_match.ncs_opcevent=0x83,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000004event=0x84,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000001AD Ingress (from CMS) : AD Ingress (from CMS) Allocations. Unit: uncore_m2m event=0x35,umask=0x00c817fd01event=0x35,umask=0x00c807fe01TOR Inserts : SF/LLC Evictions. Unit: uncore_cha TOR Inserts : RRQ. Unit: uncore_cha TOR Inserts : Just Misses. Unit: uncore_cha event=0x36,umask=0x00c001ff01unc_cha_tor_occupancy.ia_hit_drdTOR Occupancy; RFO hits from local IA. Unit: uncore_cha TOR Occupancy; RFO misses from local IO. Unit: uncore_cha TOR Occupancy; ITOM from local IO. Unit: uncore_cha event=0x36,umask=0x00ccc7ff01unc_cha_tor_occupancy.ia_drd_prefunc_cha_tor_occupancy.ia_miss_rfo_pref_localevent=0x35,umask=0x00ccd7ff01TOR Inserts; LLCPrefCode misses from local IA. Unit: uncore_cha TOR Inserts : ItoMs issued by iA Cores that Missed LLC. Unit: uncore_cha TOR Inserts : WCiLF issued by iA Cores. Unit: uncore_cha event=0x35,umask=0x00C86FFF01event=0x36,umask=0x00c8968a01TOR Occupancy : WCiLFs issued by iA Cores targeting PMM that missed the LLC - HOMed remotely. Unit: uncore_cha event=0x36,umask=0x00c86e8601branchevent=0x94ic_fetch_stall.ic_stall_back_pressurel2_pf_miss_l2_hit_l3amd_l3Other L3 Miss Request Types. Unit: amd_l3 L3 Cache Miss Latency. Total cycles for all transactions divided by 16. Ignores SliceMask and ThreadMask. Unit: amd_l3 xi_ccx_sdp_req1.all_l3_miss_req_typsRetired Taken Branch Instructionsevent=0xc6The number of resync branches. These reflect pipeline restarts due to certain microcode assists and events such as writes to the active instruction stream, among other things. Each occurrence reflects a restart penalty similar to a branch mispredict. This is relatively rareThe number of operations (uOps) and dual-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number multi-pipe uOps assigned to pipe 1event=0,umask=0x0fevent=0x2,umask=0x02This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Double precision multiply FLOPSThe number of serializing Ops retired. SSE bottom-executing uOps retiredCounts the number of loads dispatched to the LS unit. Unit Masks ADDedL1 DTLB Miss of a page of 32K sizede_dis_dispatch_token_stalls0.alsq1_token_stallevent=0xaf,umask=0x01All L1 Data Cache AccessesAll L2 Cache AccessesL2 Cache Misses from L2 HWPFL2 Cache Hits from L1 Instruction Cache Missesic_fetch_miss_ratioL1 DTLB MissesThe number of instruction fetches that miss in both the L1 and L2 TLBs. Instruction fetches to a 2MB pageevent=0xd2event=0xe,umask=0x01event=0x59,umask=0x40Software Prefetch Data Cache Fills by Data Source. Local L2 hitevent=0xa9event=0x90ls_hw_pf_dc_fills.lcl_l2This event counts the in-flight L1 data cache misses (allocated Miss Address Buffers) divided by 4 and rounded down each cycle unless used with the MergeEvent functionality. If the MergeEvent is used, it counts the exact number of outstanding L1 data cache misses. See 2.1.17.3 [Large Increment per Cycle Events]l1_data_cache_fills_from_within_same_ccxRetired near returns (RET or RET Iw)ls_dmnd_fills_from_sys.far_cacheAny data cache fills from cache of another CCX when the address was in a different NUMA nodeSoftware prefetch data cache fills from extension memoryL2 cache requests of non-cacheable type (non-cached data and instructions reads, self-modifying code checks)l2_pf_hit_l2.l2_burstl2_pf_miss_l2_hit_l3.l2_streamL2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L2UpDown (fetch the next or previous line into L2 cache for all memory accesses)Op cache hitsl3_xi_sampled_latency_requests.dram_nearRetired microcoded instructionslocal_processor_read_data_beats_cs7event=0x2df,umask=0x7ffevent=0x5f,umask=0xbfeevent=0x11f,umask=0xbffRead data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 1Read data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 3Data beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 1Data beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 6Data beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 2local_socket_inf1_outbound_data_beats_ccm5Data beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 7event=0x41f,umask=0xbffevent=0xb9f,umask=0xf3eData beats (64 bytes) for local socket outbound data from inter-socket xGMI link 2Retired x87 control word mispredict traps due to mispredictions in RC or PC, or changes in exception mask bitsRetired scalar floating-point convert opsfp_ops_retired_by_type.vector_blendRetired MMX integer addRetired MMX integer ops of all typesevent=0xb,umask=0x30sse_avx_ops_retired.sse_avx_allevent=0xc,umask=0x08fp_pack_ops_retired.fp256_shufflepacked_int_op_type.int128_movUnit: uncore_umcpmc Number of memory clock cyclesL1 DTLB misses with L2 DTLB hits for 4k pagesL1 DTLB misses with L2 DTLB hits for 2M pagesde_no_dispatch_per_slot.backend_stallsd_ratio(de_no_dispatch_per_slot.no_ops_from_frontend, total_dispatch_slots)d_ratio(de_src_op_disp.all - ex_ret_ops, total_dispatch_slots)d_ratio(de_no_dispatch_per_slot.backend_stalls, total_dispatch_slots)Fraction of dispatch slots used by fastpath ops that retiredl2_cache_misses_from_l1_dc_missl2_cache_hits_from_l1_ic_missL2 cache hits from L2 cache hardware prefetcherL1 demand data cache fills from another CCX cache in the same NUMA nodeRemote socket inbound data to the CPU (e.g. read data)d_ratio(umc_pchg_cmd.all * 1000, umc_mem_clk)LS_BUFFER2_FULLLS_MICROARCHITECTURAL_LATE_CANCELFR_RETIRED_FAR_CONTROL_TRANSFERSNB_PROBE_RESULTL1_CACHE_ACCESS_NOCP15EVENT_70HEVENT_B8HEVENT_F3HEVENT_F9HJAVA_BYTECODEINSTR_MAIN_TLB_MISS_STALLPLE_CACHE_LINE_REQ_SKIPPEDBR_MIS_PREDL2D_CACHE_WBEXC_IRQEVENT_13AHEVENT_15AHEVENT_169HEVENT_1EDHEVENT_23FHEVENT_2C3HEVENT_2CEHEVENT_2D4HEVENT_2DEHEVENT_2E6HEVENT_315HEVENT_322HEVENT_333HEVENT_33AHEVENT_3CAHEVENT_3CEHEVENT_3E4HEVENT_3EDHSTREX_FAIL_SPECPC_WRITE_SPECdn_rxreq_snp_stalledhnf_snp_sent_untrksbsx_wr_req_trkr_occ_cnt_ovflsbsx_txrsp_stallrnd_rdb_hybridcxha_rdbbyp_occcxla_tx_cxs_link_credit_backpressureclkdiv2_lrank_turnaround_activateMFSPR_INSTR_COMPLETEDIU1_INSTR_COMPLETEDL3_CACHE_MISSESL1_EXTERNAL_INTERVENTIONSPREFETCH_ENGINE_COLLISION_VS_LOADBR_MARKED_INSTR_FINISHBRANCH_MISPREDSYSCALL_TRAP_INTRL2_LINEFILL_BUFFERCASTOUTS_RELEASEDDVT1_DETECTEDmmx-3dnowx87-reclass-microfaultsdram-controller-queue-bypassprobe-hit-dirty-with-memory-cancelic-missesTSCARMV7_CORTEX_A5ARMV7_CORTEX_A8BRANCH-MISSES-RETIREDoffcore_rspcmask%s, "pmcid": "0x%08x", "pid": "%d" "tid": "%d", "value": "0x%016jx"}
%s, "userdata": "0x%08x"}
v17GenuineIntel-6-A[56]( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD)Instructions Per Cycle (per Logical Processor)FP_Arith_UtilizationCor;Flops;HPCUOPS_EXECUTED.THREAD / (( cpu@UOPS_EXECUTED.CORE\,cmask\=1@ / 2 ) if #SMT_on else UOPS_EXECUTED.CYCLES_GE_1_UOP_EXEC)Instructions per FP Arithmetic AVX/SSE 128-bit instruction (lower number means higher occurrence rate). May undercount due to FMA double countingAverage per-core data fill bandwidth to the L3 cache [GB / sec]Giga Floating Point Operations Per SecondC6_Pkg_ResidencyL2 cache lines in E state filling L2Demand Data Read requestsL2 fill requests that access L2 cacheevent=0xd2,period=20011,umask=0x4mem_load_uops_l3_miss_retired.remote_hitmOffcore requests buffer cannot take more entries for this thread coreThis event counts cycles when offcore outstanding Demand Data Read transactions are present in the super queue (SQ). A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation)  Spec update: BDM76Offcore outstanding RFO store transactions in SuperQueue (SQ), queue to uncore  Spec update: BDM76event=0xc7,period=2000006,umask=0x15fp_assist.simd_outputNumber of times an HLE abort was attributed to a Memory condition (See TSX_Memory event for additional details)event=0xc8,period=2000003,umask=0x40event=0xcd,period=100007,umask=0x1,ldlat=0x20Number of times a disallowed operation caused an RTM abortNumber of times we entered an RTM region; does not count nested transactionscpl_cycles.ring0event=0x88,period=200003,umask=0xffbr_inst_exec.all_conditionalThis event counts both taken and not taken speculative and retired indirect branches excluding calls and return branchesevent=0x88,period=200003,umask=0x81This event counts all (macro) branch instructions retiredbr_inst_retired.far_branchevent=0xc4,period=100007,umask=0x40br_misp_exec.nontaken_conditionalTotal execution stallsresource_stalls.anyCycles stalled due to no eligible RS entry availableuops_dispatched_port.port_2Cycles per thread when uops are executed in port 5event=0xb1,cmask=3,period=2000003,umask=0x1uops_executed_port.port_7_coreNumber of flags-merge uops being allocated. Such uops considered perf sensitive
 added by GSR u-archevent=0x35,umask=0x1,filter_opc=0x19ellc_references.pcie_writeThis is an occupancy event that tracks the number of cores that are in C3.  It can be used by itself to get the average number of cores in C0, with threshholding to generate histograms, or with other PCU events and occupancy triggering to capture other details. Unit: uncore_pcu Load misses that miss the  DTLB and hit the STLB (2M)dtlb_load_misses.walk_completed_4kdtlb_load_misses.walk_durationThis event counts the number of DTLB flush attempts of the thread-specific entriesoffcore_response.all_pf_code_rd.supplier_none.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020004offcore_response.demand_rfo.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000018000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C8000offcore_response.pf_l2_code_rd.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020020Number of SSE/AVX computational packed floating-point instructions retired; some instructions will count twice as noted below. Applies to SSE* and AVX* packed double and single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000090offcore_response.all_pf_rfo.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000040offcore_response.pf_l3_code_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000200offcore_response.pf_l3_code_rd.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000100This event counts mispredicted return instructions retired (Precise event)This event counts the number of retirement slots used (Precise event)event=0x34,umask=0x11L3 Lookup read request that access cache and found line in M-state. Unit: uncore_cbox event=0x34,umask=0x2foffcore_response.all_reads.llc_hit.hit_other_core_no_fwdoffcore_response.all_rfo.llc_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC002008Bytesevent=0x22,period=200000,umask=0x40event=0x2d,period=200000,umask=0x42l2_reject_busq.self.any.s_statemem_load_retired.l2_hitsimd_comp_inst_retired.packed_singleevent=0xb3,period=2000000,umask=0x20simd_uop_type_exec.logical.arSIMD packed shift micro-ops retiredevent=0x10,period=2000000,umask=0x1event=0x87,period=2000000,umask=0x1ld-op-st splitsevent=0x5,period=200000,umask=0xfevent=0x5,period=200000,umask=0x92event=0x7b,period=200000,umask=0x20event=0x65,period=200000,umask=0xe0Burst read bus transactionsbus_trans_def.all_agentsbus_trans_inval.all_agentsevent=0xc4,period=2000000,umask=0x0br_inst_retired.pred_takenbr_inst_type_retired.cond_takendiv.sdata_tlb_misses.dtlb_missevent=0x31,period=200003Memory uop retired where cross core or cross module HITM occurred (Precise event capable)  Supports address when precise (Must be precise)Load uops retired that missed L1 data cache (Precise event capable)  Supports address when precise (Must be precise)offcore_response.any_data_rd.l2_miss.snoop_miss_or_no_snoop_neededCounts data reads generated by L1 or L2 prefetchers that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_read.l2_miss.snoop_miss_or_no_snoop_neededevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000008Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that are outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts partial cache line data writes to uncacheable write combining (USWC) memory region  that hit the L2 cacheCounts partial cache line data writes to uncacheable write combining (USWC) memory region  that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts partial cache line data writes to uncacheable write combining (USWC) memory region  that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache line reads generated by hardware L1 data cache prefetcher that miss the L2 cacheoffcore_response.pf_l2_rfo.l2_miss.hit_other_core_no_fwdCounts reads for ownership (RFO) requests generated by L2 prefetcher that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts data cache lines requests by software prefetch instructions that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredevent=0xc3,period=200003,umask=0x4BACLEARs asserted for return branchCounts requests to the Instruction Cache (ICache) for one or more bytes in an ICache Line.  The event strives to count on a cache line basis, so that multiple fetches to a single cache line count as one ICACHE.ACCESS.  Specifically, the event counts when accesses from straight line code crosses the cache line boundary, or when a branch target is to a new line.
This event counts differently than Intel processors based on Silvermont microarchitectureCounts Jcc (Jump on Conditional Code/Jump if Condition is Met) branch instructions retired that were taken and does not count when the Jcc branch instruction were not taken (Must be precise)Core cycles when core is not halted.  This event uses a (_P)rogrammable general purpose performance countercycles_div_busy.allmachine_clears.disambiguationevent=0xc3,period=200003,umask=0x8Duration of D-side page-walks in cyclesCounts every core cycle when a Data-side (walks due to a data operation) page walk is in progressCounts data reads generated by L1 or L2 prefetchers hit the L2 cacheCounts data reads generated by L1 or L2 prefetchers true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests (demand & prefetch) miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts demand reads for ownership (RFO) requests generated by a write to full data cache line have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand reads for ownership (RFO) requests generated by a write to full data cache line hit the L2 cacheCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000800event=0x8,period=200003,umask=0x8Counts page walks completed due to demand data loads (including SW prefetches) whose address translations missed in all TLB levels and were mapped to 1GB pages.  The page walks can end with or without a page faultevent=0x8,period=200003,umask=0x4Page walk completed due to an instruction fetch in a 1GB pagetlb_flushes.stlb_anyThis event counts the number of L2 cache lines brought into the L2 cache.  Lines are filled into the L2 cache when there was an L2 missOffcore outstanding demand data read transactions in SQ to uncore. Set Cmask=1 to count cycles  Spec update: HSD78, HSD62, HSD61, HSM63, HSM80offcore_response.all_rfo.l3_hit.hitm_other_coreevent=0xc6,period=2000003,umask=0x7offcore_response.all_requests.l3_miss.any_responseoffcore_response.pf_l3_rfo.l3_miss.any_responseNumber of times an RTM execution startedCounts the number of not taken branch instructions retiredevent=0x87,period=2000003,umask=0x4Cycles at least 3 micro-op is executed from any thread on physical core  Spec update: HSD30, HSM31Cycles when Resource Allocation Table (RAT) does not issue Uops to Reservation Station (RS) for all threadsNumber of DTLB page walker loads that hit in the L2event=0xf2,period=100003,umask=0x4l2_store_lock_rqsts.hit_mRetired load uops which data sources were HitM responses from shared LLC (Precise event)Retired load uops that miss the STLB. (Precise Event)Counts all demand & prefetch data reads that hit in the LLCoffcore_response.demand_code_rd.llc_hit.any_responseoffcore_response.demand_data_rd.llc_hit.any_responseCounts all demand data writes (RFOs) that hit in the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x18000fp_comp_ops_exe.x87Cycles DSB to MITE switches caused delayCounts cycles the IDQ is emptyCounts demand code reads that miss the LLC and the data returned from dramCycles weighted by number of requests pending in Coherency TrackerNumber of requests allocated in Coherency Tracker. Unit: uncore_arb Counts the number of coherent and in-coherent requests initiated by IA cores, processor graphic units, or LLC. Unit: uncore_arb Counts the number of LLC evictions allocatedA snoop misses in some processor coreevent=0x34,umask=0x20event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0090Counts prefetch data reads that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts prefetch data reads that hit in the LLC and sibling core snoop returned a clean responseoffcore_response.all_reads.llc_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0001Counts L2 hints sent to LLC to keep a line from being evicted out of the core cachesCounts demand data reads that miss the LLC  and the data returned from local dramoffcore_response.pf_l2_data_rd.llc_miss.remote_dramOccupancy for all LLC misses that are addressed to remote memory. Unit: uncore_cbox freq_band1_cycles %unc_p_freq_ge_4000mhz_transitionsAllocated L1D data cache lines in M statel2_l1d_wb_rqsts.hit_sThis event counts retired load uops that hit in the last-level (L3) cache without snoops requiredbr_misp_retired.near_callThis event counts the number of times that load operations are temporarily blocked because of older stores, with addresses that are not yet known. A load operation may incur more than one block of this typeevent=0xa2,period=2000003,umask=0xeCounts the number of load micro-ops retired that hit in the L2  Supports address when precise (Precise event)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800403091offcore_response.any_pf_l2.l2_hit_this_tile_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000070event=0xb7,period=100007,umask=0x1,offcore_rsp=0x00040032f7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x00100032f7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000088000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000022event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400004offcore_response.demand_code_rd.l2_hit_near_tileevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000004Counts demand cacheable data and L1 prefetch data reads that accounts for any responseoffcore_response.partial_reads.l2_hit_far_tile_e_fCounts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses which hit its own tile's L2 with data in M stateCounts Partial reads (UC or WC and is valid only for Outstanding response type).  that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts L1 data HW prefetches that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateCounts L1 data HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400040event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000040Counts L2 data RFO prefetches (includes PREFETCHW instruction) that provides no supplier detailsevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000081000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100403091event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000070Counts any Read request  that accounts for responses from MCDRAM (local and far)offcore_response.demand_data_rd.mcdram_farCounts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for data responses from DRAM LocalCounts L2 code HW prefetches that accounts for data responses from DRAM FarCounts the number of branch instructions retired (Precise event)event=0x3,period=200003,umask=0x20l1d.m_snoop_evictevent=0x43,period=2000000,umask=0x1l2_data_rqsts.demand.m_statel2_data_rqsts.demand.s_stateevent=0x24,period=200000,umask=0x2l2_transactions.l1d_wbevent=0x27,period=100000,umask=0x20mem_inst_retired.latency_above_threshold_2048event=0xcb,period=2000000,umask=0x1Load instructions retired IO (Precise Event)Offcore data reads satisfied by a remote cache or remote DRAMOffcore requests satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x477offcore_response.demand_data.any_cache_dramoffcore_response.demand_data.remote_cache_hitmAll offcore demand data readsOffcore demand code reads satisfied by the LLC and HIT in a sibling coreoffcore_response.demand_ifetch.llc_hit_other_core_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1804event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F02Offcore demand RFO requests satisfied by the LLC and HIT in a sibling coreOffcore prefetch data requests satisfied by any cache or DRAMOffcore prefetch data requests satisfied by the LLC and HIT in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x730Offcore prefetch data requests satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1830Offcore prefetch data reads satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3810event=0xb7,period=100000,umask=0x1,offcore_rsp=0x840Offcore prefetch RFO requests satisfied by the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x170offcore_response.prefetch.remote_cachefp_comp_ops_exe.mmxfp_comp_ops_exe.sse2_integerSIMD integer 64 bit packed multiply operationsOffcore demand data requests satisfied by any DRAMoffcore_response.demand_ifetch.remote_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2002offcore_response.pf_rfo.any_llc_missevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF820Loads dispatched from the MOBFalse dependencies due to partial address aliasingSegment rename stall cyclesevent=0xc4,period=200000,umask=0x4resource_stalls.rob_fullevent=0xb1,any=1,period=2000000,umask=0x8uops_issued.fusedRetired load instructions which data sources were HitM responses from shared L3  Supports address when precise (Precise event)Retired load instructions with L2 cache hits as data sources  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0080001offcore_response.demand_data_rd.l3_hit_s.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0080002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0088000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400048000offcore_response.other.supplier_none.spl_hitevent=0x32,period=2000003,umask=0x8event=0xc6,period=100007,umask=0x1,frontend=0x1cycle_activity.cycles_l3_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000040004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000002All (macro) branch instructions retired  Spec update: SKL091Number of instructions retired. General Counter - architectural event  Spec update: SKL091, SKL044int_misc.clear_resteer_cyclesevent=0xc1,period=100003,umask=0x3fBad;Branches;CodeGen;PGOPage walk completed due to a demand data load to a 2M/4M pageCounts demand and DCU prefetch instruction cacheline that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts code reads generated by L2 prefetchers that miss L2offcore_response.all_pf_rfo.llc_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0122offcore_response.pf_l2_code_rd.llc_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1f80400004Counts all prefetch (that bring data to L2) code reads that miss the LLC  and the data returned from dramCounts all prefetch (that bring data to LLC only) RFOs that miss the LLC  and the data returned from dramoffcore_requests.demand.read_dataevent=0xb0,period=100000,umask=0x4event=0x60,period=2000000,umask=0x8REQUEST = ANY_DATA read and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8ffoffcore_response.any_rfo.all_local_dram_and_remote_cache_hitREQUEST = CORE_WB and RESPONSE = IO_CSR_MMIOREQUEST = CORE_WB and RESPONSE = REMOTE_CACHE_HITMREQUEST = DEMAND_DATA and RESPONSE = LLC_HIT_OTHER_CORE_HITMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff80REQUEST = PF_DATA and RESPONSE = LLC_HIT_OTHER_CORE_HIToffcore_response.pf_ifetch.all_local_dram_and_remote_cache_hitREQUEST = PF_IFETCH and RESPONSE = LLC_HIT_NO_OTHER_COREREQUEST = PF_IFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = PF_IFETCH and RESPONSE = REMOTE_CACHE_HITMoffcore_response.any_data.other_local_dramREQUEST = ANY_DATA read and RESPONSE = REMOTE_DRAMREQUEST = ANY IFETCH and RESPONSE = ANY_LLC_MISSoffcore_response.demand_rfo.any_dram_and_remote_fwdevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf880REQUEST = PF_IFETCH and RESPONSE = REMOTE_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf870Retired load instructions which data sources missed L3 but serviced from remote dram  Supports address when precise (Precise event)offcore_response.demand_code_rd.l3_hit.snoop_hit_with_fwdoffcore_response.pf_l1d_and_sw.l3_hit.any_snoopCounts L1 data cache hardware prefetch requests and software prefetch requests that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts all prefetch (that bring data to LLC only) data reads that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.pf_l3_rfo.l3_hit.hitm_other_coreCounts prefetch RFOs that miss the L3 and the modified data is transferred from remote cacheCounts all demand & prefetch RFOs that miss the L3 and the modified data is transferred from remote cacheoffcore_response.all_rfo.l3_miss.remote_hit_forwardoffcore_response.demand_code_rd.l3_miss.remote_hit_forwardoffcore_response.demand_data_rd.l3_miss.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00002offcore_response.pf_l1d_and_sw.l3_miss.snoop_miss_or_no_fwdPower_License1_UtilizationWrite Pending Queue Occupancy. Unit: uncore_imc event=0x83,ch_mask=0x01,fc_mask=0x07,umask=0x01unc_cha_core_snp.evict_gtoneCounts the total number of requests coming from a unit on this socket for exclusive ownership of a cache line without receiving data (INVITOE) to the CHACounts when a a transaction with the opcode type RspCnflct* Snoop Response was received. This is returned when a snoop finds an existing outstanding transaction in a remote caching agent. This triggers conflict resolution hardware. This covers both the opcode RspCnflct and RspCnflctWbIunc_iio_comp_buf_inserts.cmpd.all_partsevent=0xc2,ch_mask=0x0f,fc_mask=0x4,umask=0x03PCIe Completion Buffer Inserts of completions with data: Part 0. Unit: uncore_iio Read request for 4 bytes made by the CPU to IIO Part1. Unit: uncore_iio event=0xc0,ch_mask=0x01,fc_mask=0x07,umask=0x08Peer to peer write request of 4 bytes made to IIO Part0 by a different IIO unit. Unit: uncore_iio unc_iio_data_req_of_cpu.peer_read.part1Counts every read request for up to a 64 byte transaction of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part2. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busPeer to peer read request for up to a 64 byte transaction is made by a different IIO unit to IIO Part1. Unit: uncore_iio event=0xc1,ch_mask=0x04,fc_mask=0x07,umask=0x08unc_iio_txn_req_by_cpu.peer_write.part2Counts every peer to peer write request of up to a 64 byte transaction of data made by IIO Part0 to the MMIO space of an IIO target. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busevent=0xf,umask=0x4uncore_m2mCounts when messages were sent direct to the Intel Ultra Path Interconnect (bypassing the CHA)event=0x2d,umask=0x1Counts when the a new entry is Received(RxC) and then added to the AD (Address Ring) Ingress Queue from the CMS (Common Mesh Stop).  This is generally used for reads, andevent=0x2Valid data FLITs received from any slot. Unit: uncore_upi unc_upi_txl_bypassedOCR.ALL_DATA_RD.L3_HIT_F.ANY_SNOOP  OCR.ALL_DATA_RD.L3_HIT_F.ANY_SNOOPOCR.ALL_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWD  OCR.ALL_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDocr.all_pf_data_rd.l3_hit_m.snoop_noneocr.all_pf_rfo.l3_hit.any_snoopOCR.ALL_PF_RFO.L3_HIT_F.SNOOP_MISSOCR.ALL_PF_RFO.L3_HIT_M.ANY_SNOOP  OCR.ALL_PF_RFO.L3_HIT_M.ANY_SNOOPOCR.ALL_READS.L3_HIT_E.HITM_OTHER_CORE  OCR.ALL_READS.L3_HIT_E.HITM_OTHER_COREOCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDOCR.ALL_RFO.L3_HIT_F.HIT_OTHER_CORE_FWD  OCR.ALL_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0004Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_E.ANY_SNOOPocr.demand_data_rd.l3_hit_m.snoop_missCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDocr.demand_rfo.l3_hit_f.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C8000ocr.other.l3_hit_f.any_snoopocr.other.l3_hit_f.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200400ocr.pf_l1d_and_sw.l3_hit_f.hitm_other_coreCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_M.HITM_OTHER_COREocr.pf_l2_data_rd.l3_hit_s.any_snoopCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_S.HITM_OTHER_COREocr.pf_l2_rfo.l3_hit_e.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080020Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040080ocr.pf_l3_data_rd.l3_hit_s.snoop_missCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_HIT.SNOOP_NONEocr.pf_l3_rfo.l3_hit_f.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.pf_l3_rfo.l3_hit_m.hit_other_core_fwdoffcore_response.all_data_rd.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPoffcore_response.all_pf_rfo.supplier_none.no_snoop_neededoffcore_response.all_reads.l3_hit_f.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_S.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4000207F7offcore_response.all_reads.supplier_none.snoop_noneThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_code_rd.l3_hit_f.hit_other_core_fwdoffcore_response.demand_code_rd.l3_hit_f.snoop_noneThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_F.SNOOP_NONEoffcore_response.demand_code_rd.l3_hit_s.hit_other_core_no_fwdoffcore_response.demand_code_rd.supplier_none.no_snoop_neededoffcore_response.demand_data_rd.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_E.SNOOP_MISSoffcore_response.demand_data_rd.l3_hit_f.snoop_noneThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_E.SNOOP_NONEThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.other.l3_hit_f.hitm_other_coreoffcore_response.other.l3_hit_s.hitm_other_coreoffcore_response.other.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_S.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.OTHER.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT.ANY_SNOOPoffcore_response.pf_l1d_and_sw.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.SUPPLIER_NONE.NO_SNOOP_NEEDEDoffcore_response.pf_l2_data_rd.l3_hit_f.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80400010This event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.SUPPLIER_NONE.SNOOP_MISSoffcore_response.pf_l2_rfo.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_F.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_S.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDoffcore_response.pf_l3_rfo.l3_hit_e.hit_other_core_no_fwdoffcore_response.pf_l3_rfo.pmm_hit_local_pmm.snoop_noneoffcore_response.pf_l3_rfo.supplier_none.no_snoop_neededocr.all_data_rd.l3_miss.hit_other_core_no_fwdocr.all_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000490OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDED  OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000120OCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_FWD OCR.ALL_PF_RFO.L3_MISS.HIT_OTHER_CORE_FWDocr.all_pf_rfo.l3_miss_local_dram.hitm_other_coreocr.all_pf_rfo.l3_miss_local_dram.hit_other_core_no_fwdocr.all_reads.l3_miss.any_snoopOCR.ALL_READS.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_READS.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDOCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOP  OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPOCR.ALL_RFO.L3_MISS.REMOTE_HITM OCR.ALL_RFO.L3_MISS.REMOTE_HITMOCR.ALL_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOP  OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPOCR.ALL_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDED  OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDOCR.ALL_RFO.L3_MISS_LOCAL_DRAM.SNOOP_NONECounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.SNOOP_MISSocr.demand_code_rd.l3_miss_local_dram.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000001ocr.demand_data_rd.l3_miss_remote_hop1_dram.any_snoopCounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.pf_l2_data_rd.l3_miss_remote_hop1_dram.any_snoopocr.pf_l2_rfo.l3_miss_local_dram.hit_other_core_no_fwdCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000020Counts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.HITM_OTHER_CORE OCR.PF_L3_DATA_RD.L3_MISS.HITM_OTHER_COREocr.pf_l3_data_rd.l3_miss_local_dram.no_snoop_neededocr.pf_l3_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDoffcore_response.all_pf_data_rd.l3_miss.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.ANY_SNOOPoffcore_response.demand_rfo.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.pf_l1d_and_sw.l3_miss_remote_hop1_dram.hitm_other_coreoffcore_response.pf_l2_data_rd.l3_miss_local_dram.hitm_other_coreoffcore_response.pf_l2_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISSoffcore_response.pf_l3_data_rd.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREoffcore_response.pf_l3_rfo.l3_miss_local_dram.no_snoop_neededocr.all_data_rd.supplier_none.hit_other_core_no_fwdOCR.ALL_PF_DATA_RD.SUPPLIER_NONE.ANY_SNOOP  OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.ANY_SNOOPocr.all_pf_rfo.pmm_hit_local_pmm.snoop_not_neededocr.all_rfo.supplier_none.any_snoopocr.demand_data_rd.any_responseUNC_M_PMM_RPQ_OCCUPANCY.ALL / UNC_M_PMM_RPQ_INSERTS / UNC_M_CLOCKTICKSUnderfill read commands for Intel Optane DC persistent memory. Unit: uncore_imc Write Pending Queue Occupancy of all write requests for Intel Optane DC persistent memoryTag Hit; Underfill Rd Hit from NearMem, Dirty  Lineevent=0x51,period=100003,umask=0x1Counts demand requests to L2 cacheCounts retired load instructions with L2 cache hits as data sources  Supports address when precise (Precise event)Counts hardware prefetch RFOs (which bring data to L2) that hit a cacheline in the L3 where a snoop hit in another core, data forwarding is not requiredNumber of SSE/AVX computational 128-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0x9c,period=1000003,umask=0x1Counts the number of times HLE abort was triggeredCounts hardware prefetch data reads (which bring data to L2)  that was not supplied by the L3 cacheevent=0x54,period=100003,umask=0x40ocr.demand_code_rd.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000010Counts taken conditional mispredicted branch instructions retired (Precise event)Core cycles the allocator was stalled due to recovery from earlier clear event for this threadevent=0xcc,period=100003,umask=0x20event=0xcc,period=100003,umask=0x40Number of uops decoded out of instructions exclusively fetched by decoder 0Demand Data Read transactions pending for off-core. Highly correlatedCounts cycles when the memory subsystem has an outstanding load. Increments by 4 for every such cycleDSB;FetchBWCounts demand instruction fetches and L1 instruction cache prefetches that resulted in a snoop hit a modified line in another core's caches which forwarded the dataCounts demand instruction fetches and L1 instruction cache prefetches that either hit a non-modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeocr.reads_to_core.l3_hitocr.reads_to_core.snc_cache.hit_with_fwdCounts both cacheable and non-cacheable code reads to the core(cstate_core@c1\-residency@ / msr@tsc@) * 100Counts responses to snoops indicating the line will now be (I)nvalidated: removed from this core's cache, after the data is forwarded back to the requestor and indicating the data was found unmodified in the (FE) Forward or Exclusive State in this cores caches cache.  A single snoop response from the core counts on all hyperthreads of the coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x700800002Counts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that were supplied by DRAMevent=0x45,umask=0x04unc_m_wpq_inserts.pch0unc_m_act_count.allevent=0x2,umask=0x1Cevent=0x35,umask=0xC001FF01unc_cha_tor_inserts.ia_hit_drdTOR Inserts : RFOs issued by iA Cores that Hit the LLC. Unit: uncore_cha TOR Occupancy : All requests from iA Cores. Unit: uncore_cha unc_cha_tor_occupancy.ia_missunc_cha_tor_inserts.ia_hit_crd_prefunc_cha_tor_inserts.ia_llcprefrfoevent=0x35,umask=0xCD43FF04unc_cha_tor_inserts.io_miss_itomcachenearunc_cha_tor_inserts.ia_miss_llcprefdataevent=0x35,umask=0xCCD7FE01unc_cha_tor_occupancy.io_miss_pcirdcurunc_iio_txn_req_of_cpu.cmpd.part0event=0xc0,ch_mask=0x80,fc_mask=0x07,umask=0x01event=0x84,ch_mask=0x40,fc_mask=0x07,umask=0x04event=0xd5,fc_mask=0x04,umask=0xffevent=0x8,period=100003,umask=0x8Counts the number of cacheable memory requests that access the Last Level Cache (LLC). Requests include demand loads, reads for ownership (RFO), instruction fetches and L1 HW prefetches. If the platform has an L3 cache, the LLC is the L3 cache, otherwise it is the L2 cache. Counts on a per core basisCounts the number of store uops retired  Supports address when precise (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0044This event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_HIT.SNOOP_HITMCounts L1 data cache hardware prefetches and software prefetches (except PREFETCHW and PFRFO) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0040Counts L2 cache hardware prefetch data reads (written to the L2 cache only) that were supplied by the L3 cache where no snoop was needed to satisfy the requestCounts streaming stores which modify only part of a 64 byte cacheline that were supplied by the L3 cacheCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by the L3 cacheocr.uc_rd.l3_hit.snoop_not_neededCounts uncached memory reads that were supplied by the L3 cache where no snoop was needed to satisfy the requestocr.hwpf_l2_data_rd.l3_miss_localCounts modified writebacks from L1 cache and L2 cache that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000000000000001Counts L2 cache hardware prefetch data reads (written to the L2 cache only) that were supplied by DRAMCounts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were supplied by DRAMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100000010000topdown_bad_speculation.alltopdown_bad_speculation.machine_clearstopdown_be_bound.mem_schedulerevent=0x74,period=1000003,umask=0x2Counts the number of uops that are from complex flows issued by the Microcode Sequencer (MS). This includes uops from flows due to complex instructions, faults, assists, and inserted flows (Precise event)uops_retired.x87Read Pending Queue Allocations : Counts the number of allocations into the Read Pending Queue.  This queue is used to schedule reads out to the memory controller and to track the requests.  Requests allocate into the RPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the HA to the iMC.  They deallocate after the CAS command has been issued to memory.  This includes both ISOCH and non-ISOCH requestsunc_cha_tor_inserts.ia_drd_opt_prefTOR Inserts : ItoMCacheNears, indicating a partial write request, from IO Devices that missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsunc_cha_tor_occupancy.ia_miss_drd_optData requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Number Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Coherent Ops : WbMtoI : Counts the number of coherency related operations servied by the IRPClockticks of the power control unit (PCU)itlb_misses.pde_cache_missCounts the number of page walks completed due to instruction fetch misses to a 2M or 4M pageCounts the number of load ops retired that miss in the second Level TLB  Supports address when precise (Precise event)Total number of retired Instructions, Sample with: INST_RETIRED.PREC_DIST. Unit: cpu_core L2 cache hits per kilo instruction for all request types (including speculative). Unit: cpu_core Utilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accesses. Unit: cpu_core Average number of parallel requests to external memory. Accounts for all requests. Unit: cpu_core CPU_CLK_UNHALTED.CORE_PINST_RETIRED.ANY / BR_INST_RETIRED.CALLInstructions per Load. Unit: cpu_atom Instructions per Store. Unit: cpu_atom Cycle cost per LLC hit. Unit: cpu_atom MEM_BOUND_STALLS.LOAD_LLC_HIT / MEM_LOAD_UOPS_RETIRED.L3_HITCounts the number of cycles the core is stalled due to a demand load miss which hit in DRAM or MMIO (Non-DRAM). Unit: cpu_atom Counts the number of store uops retired  Supports address when precise (Precise event). Unit: cpu_atom event=0xd0,period=1000003,umask=0x5,ldlat=0x80mem_uops_retired.load_latency_gt_16Retired load instructions that split across a cacheline boundary  Supports address when precise (Precise event). Unit: cpu_core Retired load instructions with L1 cache hits as data sources  Supports address when precise (Precise event). Unit: cpu_core Counts the number of floating point operations retired that required microcode assist. Unit: cpu_atom Counts the number of floating point divide uops retired (x87 and SSE, including x87 sqrt) (Precise event). Unit: cpu_atom event=0xb0,cmask=1,period=1000003,umask=0x1Uops delivered to IDQ while MS is busy. Unit: cpu_core Mispredicted conditional branch instructions retired (Precise event). Unit: cpu_core Cycles total of 4 uops are executed on all ports and Reservation Station was not empty. Unit: cpu_core misc2_retired.lfenceIncrements whenever there is an update to the LBR array. Unit: cpu_core unc_m_cas_count_wrStore misses in all TLB levels causes a page walk that completes. (All page sizes). Unit: cpu_core event=0x2a,period=100003,umask=0x1,offcore_rsp=0x73C000001arith.idiv_activeevent=0x5,umask=0x00000000ffWrite request of 4 bytes made by IIO Part0 to Memory. Unit: uncore_iio event=0x83,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000002Data requested of the CPU : Card writing to another Card (same or different stack). Unit: uncore_iio Read requests from a unit on this socket. Unit: uncore_cha event=0x36,umask=0x00c816fe01event=0x35,umask=0x00c8178a01unc_upi_rxl_flits.prothdrevent=0x84,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000001event=0xc0,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000001event=0xc0,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000001event=0xc1,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000001event=0x83,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000080event=0x54,umask=0x0000000002event=0x35,umask=0x00C001FFffevent=0x35,umask=0x00c001fe01TOR Inserts; Misses from local IO. Unit: uncore_cha TOR Inserts; DRd Opt hits from local IA. Unit: uncore_cha event=0x36,umask=0x00c001ff08event=0x36,umask=0x00C001FFffunc_cha_tor_occupancy.ipqunc_cha_tor_occupancy.ia_miss_rfo_prefunc_cha_tor_occupancy.ia_crd_prefTOR Inserts : DRd PTEs issued by iA Cores that Hit the LLC. Unit: uncore_cha event=0x36,umask=0x00c8f3ff04unc_cha_tor_inserts.ia_hit_llcprefdataevent=0x36,umask=0x00ccd7fd01TOR Occupancy; LLCPrefCode misses from local IA. Unit: uncore_cha event=0x35,umask=0x00C8978A01unc_cha_tor_inserts.ia_miss_crd_pref_remoteunc_cha_tor_inserts.ia_wbmtoievent=0x35,umask=0x00CC23FF04TOR Occupancy : DRds issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00c86f0601TOR Occupancy : ItoMCacheNears, indicating a partial write request, from IO Devices that hit the LLC. Unit: uncore_cha TOR Occupancy : PMM Access. Unit: uncore_cha The number of instruction fetches that miss in both the L1 and L2 TLBsl2_request_g1.cacheable_ic_readevent=0x64,umask=0x10All L3 Request Types. Unit: amd_l3 The number of taken branches that were retired. This includes all types of architectural control flow changes, including exceptions and interruptsex_div_countfp_ret_sse_avx_ops.dp_mult_add_flopsevent=0x4,umask=0x08ls_l1_d_tlb_miss.tlb_reload_2m_l2_missevent=0x46,umask=0x02event=0xaf,umask=0x02l1_dtlb_misses3e-5MiBevent=0x94,umask=0x04The number of instruction fetches that hit in the L1 ITLB. Instruction fetches to a 2MB pageevent=0x43,umask=0x08ls_sw_pf_dc_fill.ls_mabresp_lcl_cachels_sw_pf_dc_fill.ls_mabresp_lcl_l2Hardware Prefetch Data Cache Fills by Data Source. From another cache (home node remote)ls_hw_pf_dc_fill.ls_mabresp_lcl_l2Instruction Cache Refills from L2. The number of 64 byte instruction cache line was fulfilled from the L2 cachexi_ccx_sdp_req1Multiply-Accumulate FLOPs. Each MAC operation is counted as 2 FLOPS. This is a retire-based event. The number of retired SSE/AVX FLOPs. The number of events logged per cycle can vary from 0 to 64. This event requires the use of the MergeEvent since it can count above 15 events per cycle. See 2.1.17.3 [Large Increment per Cycle Events]. It does not provide a useful count without the use of the MergeEventHardware Prefetcher Allocations. Counts when a LS pipe allocates a MAB entryls_dmnd_fills_from_sys.lcl_l2Software Prefetch Instructions Dispatched (Speculative). PrefetchT0, T1 and T2 instructions. See docAPM3 PREFETCHlevelCycles where a dispatch group is valid but does not get dispatched due to a Token Stall. Also counts cycles when the thread is not selected to dispatch but would have been stalled due to a Token Stall. Integer Physical Register File resource stall. Integer Physical Register File, applies to all ops that have an integer destination registerevent=0x64,umask=0xf0d_ratio(op_cache_hit_miss.op_cache_miss, op_cache_hit_miss.all_op_cache_accesses)event=0x1c9Demand data cache fills from either DRAM or MMIO in the same NUMA nodeAny data cache fills from cache of another CCX when the address was in the same or a different NUMA nodeSoftware prefetch data cache fills from either DRAM or MMIO in the same NUMA nodels_sw_pf_dc_fills.far_cacheCore to L2 cache requests (not including L2 prefetch) with status: data cache store or state change hit in L2Core to L2 cache requests (not including L2 prefetch) for data and instruction cache hitsL2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L2Stride (fetch additional lines into L2 cache when each access is at a constant distance from the previous)event=0xad,umask=0x01l3_xi_sampled_latency_requests.dram_farevent=0xad,umask=0x08local_processor_read_data_beats_cs4local_processor_read_data_beats_cs6Read data beats (64 bytes) for local processor at Coherent Station (CS) 11event=0x9f,umask=0xbfeRead data beats (64 bytes) for remote processor at Coherent Station (CS) 5event=0x1f,umask=0xbffWrite data beats (64 bytes) for remote processor at Coherent Station (CS) 5Write data beats (64 bytes) for remote processor at Coherent Station (CS) 10event=0x8df,umask=0xbfeData beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 4Data beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 3event=0x4df,umask=0xbfffp_ret_x87_fp_ops.add_sub_opsRetired x87 floating-point divide and square root opssse_avx_ops_retired.mmx_mulevent=0xb,umask=0x09Retired MMX integer shift opsRetired MMX integer logical opsevent=0xc,umask=0x01fp_pack_ops_retired.fp128_allpacked_int_op_type.int128_shaRetired 128-bit packed integer compare opsevent=0xd,umask=0x0euncore_umcpmcNumber of ACTIVATE commands sent for writesumc_data_slot_clks.allL1 DTLB misses with L2 DTLB misses (page-table walks are requested) for 4k pagesInstruction fetches that miss in both the L1 and L2 ITLBs (page-table walks are requested) for all page sizesInstruction fetches that hit in the L1 ITLB for 2M pagesde_dis_ops_from_decoder.any_fp_dispatchNumber of cycles dispatch is stalled for integer physical register file tokensNumber of cycles dispatch is stalled for taken branch buffer tokensNumber of cycles dispatch is stalled for floating-point scheduler tokensNumber of cycles dispatch is stalled for integer scheduler queue 3 tokensAll L1 data cache fillsumc_mem_write_bandwidthumc_mem_bandwidthuncore_hisi_ddrc.flux_wcmdINSTR_RETIRED_ANYLS_RETIRED_CPUID_INSTRUCTIONSIC_RETURN_STACK_HITFR_RETIRED_FPU_INSTRUCTIONSFR_NUMBER_OF_BREAKPOINTS_FOR_DR0NB_MEMORY_CONTROLLER_BYPASS_SATURATIONCID_WRITECLOCK_CYCLESL2_STORE_MERGEDMEM_REPLAY_EVTEVENT_00HEVENT_09HEVENT_31HEVENT_7CHEVENT_95HEVENT_AAHEVENT_B2HEVENT_B4HEVENT_FCHDATA_ENGINE_CLOCK_ENABLEDEVENT_11EHEVENT_128HEVENT_14BHEVENT_15EHEVENT_180HEVENT_194HEVENT_1A3HEVENT_1A6HEVENT_1B0HEVENT_1CEHEVENT_1ECHEVENT_1F2HEVENT_213HEVENT_21FHEVENT_221HEVENT_247HEVENT_276HEVENT_287HEVENT_29EHEVENT_2A0HEVENT_2B8HEVENT_2CCHEVENT_2D1HEVENT_2DFHEVENT_2F1HEVENT_30FHEVENT_381HEVENT_396HEVENT_39BHEVENT_3A8HEVENT_3D3HEVENT_3EEHEVENT_3EFHL1D_TLB_REFILL_LDUNALIGNED_LD_SPECBUS_ACCESS_NOT_SHAREDhni_rdt_rd_allocsbsx_txdat_flitvcxha_chirsp_up_stallL1_DATA_SNOOP_HIT_CASTOUT_QUEUETAKEN_BRANCHES_PROCESSEDSS_SM_INSTR_PIECESSNOOP_MODIFIEDBUS_TAS_FOR_READSADDRESS_COLLISIONEXT_INPUT_INTR_TAKENLV2_VSL2_CACHE_DATA_ALLOCATIONSFPU_RESULT_STALL_CYCLESusrrdszbyteinterruptsINTEL_ATOMARMV7_CORTEX_A9THREADps->ps_state != PL_STATE_ERRORbr_misp_retired.all_branchesv10GenuineIntel-6-35GenuineIntel-6-1FGenuineIntel-6-96Branches;Fed;FetchBWIpCallBR_INST_RETIRED.ALL_BRANCHES / BR_INST_RETIRED.NEAR_TAKENIpArithInstructions per FP Arithmetic Scalar Single-Precision instruction (lower number means higher occurrence rate)Flops;FpScalar;InsTypeIpArith_AVX256Mem;MemoryBound;MemoryBWL2HPKI_LoadFraction of cycles spent in the Operating System (OS) Kernel modeMEM_Parallel_ReadsSocket_CLKSl1d.replacementevent=0x27,period=200003,umask=0x50L2 cache lines in I state filling L2event=0x24,period=200003,umask=0x24Core-originated cacheable demand requests that refer to L3This event counts retired load uops where the data came from local DRAM. This does not include hardware prefetches. This is a precise event  Supports address when precise.  Spec update: BDE70, BDM100event=0xd3,period=100007,umask=0x4Retired load uops that miss the STLB. (Precise Event - PEBS)  Supports address when precise (Precise event)fp_arith_inst_retired.128b_packed_singleevent=0xc7,period=2000003,umask=0x1fp_assist.simd_inputevent=0x79,cmask=1,period=2000003,umask=0x4event=0x9c,cmask=2,period=2000003,umask=0x1Number of times an HLE execution aborted due to various memory events (e.g., read/write capacity and conflicts)event=0xc8,period=2000003,umask=0x20event=0xcd,period=503,umask=0x1,ldlat=0x100Number of times an RTM execution aborted due to various memory events (e.g. read/write capacity and conflicts)Counts the number of times an instruction execution caused the transactional nest count supported to be exceededThis event counts not taken branch instructions retiredevent=0x3c,period=2000003Execution stalls while L1 cache miss demand load is outstandinginst_retired.x87event=0xc3,period=2000003,umask=0x1Cycles there was a Nuke. Account for both thread-specific and All Thread NukesNumber of integer Move Elimination candidate uops that were not eliminatedevent=0x5e,cmask=1,edge=1,inv=1,period=200003,umask=0x1event=0xb1,cmask=2,period=2000003,umask=0x2event=0xa1,any=1,period=2000003,umask=0x10Cycles per core when uops are exectuted in port 6uops_issued.flags_mergeevent=0xe,period=2000003,umask=0x20event=0x35,umask=0x1,filter_opc=0x180,filter_tid=0x3eunc_c_tor_occupancy.llc_data_readunc_h_requests.writes_localwrite requests to remote home agent. Unit: uncore_ha read requests to memory controller. Derived from unc_m_cas_count.rd. Unit: uncore_imc event=0x2,umask=0x8This is an occupancy event that tracks the number of cores that are in C6.  It can be used by itself to get the average number of cores in C0, with threshholding to generate histograms, or with other PCU events . Unit: uncore_pcu itlb_misses.walk_completed_4kRetired load uops misses in L1 cache as data sources  Supports address when precise (Precise event)This event counts retired load uops which data sources were misses in the nearest-level (L1) cache. Counting excludes unknown and UC data source  Supports address when precise (Precise event)offcore_response.all_pf_code_rd.supplier_none.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0090offcore_response.all_rfo.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0008offcore_response.demand_code_rd.l3_hit.snoop_not_neededoffcore_response.other.l3_hit.snoop_not_neededoffcore_response.pf_l2_data_rd.supplier_none.snoop_not_neededoffcore_response.pf_l2_rfo.l3_hit.snoop_hitmoffcore_response.pf_l2_rfo.supplier_none.snoop_hitmoffcore_response.pf_l2_rfo.supplier_none.snoop_missCounts prefetch (that bring data to LLC only) code reads have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0080offcore_response.pf_l3_data_rd.supplier_none.snoop_hitmoffcore_response.pf_l3_rfo.l3_hit.snoop_noneNumber of SSE/AVX computational double precision floating-point instructions retired; some instructions will count twice as noted below. Applies to SSE* and AVX* scalar and packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000240offcore_response.all_pf_data_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000001offcore_response.other.l3_miss_local_dram.snoop_noneoffcore_response.other.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020040offcore_response.pf_l3_data_rd.l3_miss_local_dram.snoop_hitmActually retired uops (Precise event)event=0x80,umask=0x01Each cycle count number of 'valid' coherent Data Read entries that are in DirectData mode. Such entry is defined as valid when it is allocated till data sent to Core (first chunk, IDI0). Applicable for IA Cores' requests in normal caseoffcore_response.all_reads.llc_miss.remote_dramModified cache lines evicted from the L1 data cacheModified cache lines allocated in the L1 data cacheevent=0x23,period=200000,umask=0x40event=0x29,period=200000,umask=0x4fl2_m_lines_out.self.anyRejected L2 cache requestsevent=0x30,period=200000,umask=0x71l2_rqsts.self.any.m_statel2_rqsts.self.demand.m_statemem_load_retired.l2_missevent=0xcd,period=100000,umask=0x0simd_inst_retired.scalar_singleSIMD saturated arithmetic micro-ops retiredsimd_uops_exec.arSIMD packed multiply micro-ops executedevent=0x87,period=2000000,umask=0x2event=0x5,period=200000,umask=0x8cprefetch.hw_prefetchL1 hardware prefetch requestevent=0x62,period=200000,umask=0x20bus_request_outstanding.selfevent=0x70,period=200000,umask=0xe0bus_trans_brd.all_agentsbus_trans_burst.selfbus_trans_io.selfPartial write bus transactionbus_trans_rfo.all_agentseist_transsegment_reg_loads.anybr_inst_retired.mispred_not_takenAll indirect branches that are not callsevent=0x89,period=200000,umask=0x11Mispredicted ind branches that are not callsevent=0x3,period=200000,umask=0xffreissue.overlap_storeevent=0x2,period=200000,umask=0x83Cycles no micro-ops retireddata_tlb_misses.dtlb_miss_ldCounts cycles that fetch is stalled due to an outstanding ICache miss. That is, the decoder queue is able to accept bytes, but the fetch unit is unable to provide bytes due to an ICache miss.  Note: this event is not the same as the total number of cycles spent retrieving instruction cache lines from the memory hierarchyCounts memory requests originating from the core that miss in the L2 cacheCounts data reads (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400003010event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000048000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000004event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400004000offcore_response.pf_l1_data_rd.l2_hitCounts reads for ownership (RFO) requests generated by L2 prefetcher that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0x80,period=200003,umask=0x1event=0xc4,period=200003,umask=0xf9Retired conditional branch instructions that were taken (Precise event capable) (Must be precise)event=0xc5,period=200003event=0xca,period=200003,umask=0x2Unfilled issue slots per cycle because of a full resource in the backendevent=0x3,period=200003,umask=0x4Loads blocked due to store forward restriction (Precise event capable) (Must be precise)ld_blocks.utlb_missLoads blocked because address in not in the UTLB (Precise event capable) (Must be precise)Counts data reads generated by L1 or L2 prefetchers miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests (demand & prefetch) hit the L2 cacheCounts the number of writeback transactions caused by L1 or L2 cache evictions miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts the number of writeback transactions caused by L1 or L2 cache evictions true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes true miss for the L2 cache with a snoop miss in the other processor moduleCounts data cache line reads generated by hardware L1 data cache prefetcher miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000011000event=0x49,period=200003,umask=0x10Page walks outstanding due to walking the EPT every cycleCounts all L2 HW prefetcher requests that missed L2offcore_response.all_data_rd.l3_hit.hitm_other_coreNumber of instructions retired. General Counter   - architectural event  Spec update: HSD11, HSD140L3 Lookup write request that access cache and found line in I-stateCompleted page walks in any TLB of any page size due to demand load missesDemand load Miss in all translation lookaside buffer (TLB) levels causes a page walk that completes (4K)Completed page walks due to misses in ITLB 4K page entriesNumber of DTLB page walker hits in the L2Number of DTLB page walker loads that hit in the L3  Spec update: HSD25Counts the number of Extended Page Table walks from the DTLB that hit in memoryCounts the number of Extended Page Table walks from the ITLB that hit in the L2Counts all demand & prefetch code reads that hit in the LLCoffcore_response.all_data_rd.llc_hit.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0001fp_comp_ops_exe.sse_scalar_doubleNumber of transitions from SSE to AVX-256 when penalty applicableCycles Allocation is stalled due to Resource Related reasonCounts cycles weighted by the number of requests waiting for data returning from the memory controller. Accounts for coherent and non-coherent requests initiated by IA cores, processor graphic units, or LLCA snoop invalidates a modified line in some processor core. Unit: uncore_cbox Filter on cross-core snoops initiated by this Cbox due to LLC evictionLLC lookup request that access cache and found line in E-stateevent=0x34,umask=0x04unc_cbo_cache_lookup.esDemand load Miss in all translation lookaside buffer (TLB) levels causes an page walk of any page sizeCounts all prefetch data reads that hit the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x107fc003f7Counts all demand code reads that miss the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x67fc00010LLC prefetch misses for code reads. Derived from unc_c_tor_inserts.miss_opcode.code. Unit: uncore_cbox event=0x35,umask=0x3,filter_opc=0x19cevent=0x35,umask=0x1,filter_opc=0x1c8Counts the number of cycles that the uncore was running at a frequency greater than or equal to 1.2Ghz. Derived from unc_p_freq_band0_cycles. Unit: uncore_pcu event=0xc,filter_band1=20unc_p_freq_ge_4000mhz_cyclesevent=0x27,period=200003,umask=0x4All retired load uops (Precise event)event=0x14,cmask=1,edge=1,period=100003,umask=0x1cycle_activity.cycles_no_dispatchEach cycle there was no dispatch for this thread, increment by 1. Note this is connect to Umask 2. No dispatch can be deduced from the UOPS_EXECUTED eventNumber of cycles waiting for the checkpoints in Resource Allocation Table (RAT) to be recovered after Nuke due to all other cases except JEClear (e.g. whenever a ucode assist is needed like SSE exception, memory disambiguation, etc...)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400044event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000403091Counts any Prefetch requests that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080070event=0xb7,period=100007,umask=0x1,offcore_rsp=0x10004032f7offcore_response.any_read.l2_hit_this_tile_eCounts Demand cacheable data write requests  that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000004Counts demand cacheable data and L1 prefetch data reads that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateCounts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080100event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000100Counts L1 data HW prefetches that accounts for responses which hit its own tile's L2 with data in E stateoffcore_response.uc_code_reads.l2_hit_far_tile_e_fCounts UC code reads (valid only for Outstanding response type)  that accounts for responses which hit its own tile's L2 with data in E stateCounts UC code reads (valid only for Outstanding response type)  that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts Bus locks and split lock requests that accounts for responses from MCDRAM (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800001Counts Demand cacheable data writes that accounts for data responses from DRAM FarCounts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for data responses from DRAM Faroffcore_response.partial_reads.mcdramoffcore_response.partial_reads.mcdram_farCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for data responses from DRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800020Counts Software Prefetches that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts Software Prefetches that accounts for data responses from MCDRAM Localoffcore_response.uc_code_reads.ddr_nearevent=0xcd,period=2000003,umask=0x1Counts the number of times that the machine clears due to program modifying data within 1K of a recently fetched code pageCounts the total number of core cycles when no micro-ops are allocated for any reasonmcdram bandwidth read (CPU traffic only) (MB/sec). Unit: uncore_edc_eclk event=0x2,umask=0x01L1D cache lines replaced in M stateevent=0x40,period=2000000,umask=0x2l2_data_rqsts.prefetch.mesil2_rqsts.prefetchesL2 RFO requestsevent=0xf0,period=200000,umask=0x20event=0x27,period=100000,umask=0x80L2 demand lock RFOs in M statel2_write.rfo.mesievent=0xb,period=200,umask=0x10,ldlat=0x200Offcore data reads satisfied by any cache or DRAMoffcore_response.any_data.local_cacheOffcore data reads that HITM in a remote cacheOffcore RFO requests satisfied by the LLCOffcore RFO requests that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F08event=0xb7,period=100000,umask=0x1,offcore_rsp=0x1077event=0xb7,period=100000,umask=0x1,offcore_rsp=0x733offcore_response.data_in.remote_cache_hitmOffcore demand data reads satisfied by the LLC or local DRAMoffcore_response.demand_ifetch.remote_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x402offcore_response.demand_rfo.remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1080event=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF30Offcore prefetch data requests satisfied by the IO, CSR, MMIO unitOffcore prefetch data reads that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1840event=0xb7,period=100000,umask=0x1,offcore_rsp=0x3840offcore_response.pf_rfo.any_locationoffcore_response.pf_rfo.llc_hit_other_core_hitmX87 Floating point assists (Precise Event)event=0xf7,period=20000,umask=0x4X87 Floating poiint assists for invalid input value (Precise Event)SSE and SSE2 FP Uopsevent=0x10,period=2000000,umask=0x20simd_int_128.packed_shiftsimd_int_64.packed_mpyOffcore data reads satisfied by the local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF844Offcore RFO requests that missed the LLCoffcore_response.pf_data_rd.any_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF840Offcore prefetch RFO requests that missed the LLCOffcore prefetch requests satisfied by a remote DRAMload_dispatch.rs_delayedsb_drain.anyConditional branch instructions executedevent=0x18,period=2000000,umask=0x1event=0x4c,period=200000,umask=0x1event=0xa8,cmask=1,period=2000000,umask=0x1rat_stalls.flagsevent=0xd2,period=2000000,umask=0x4Stack pointer instructions decodedUops decoded by Microcode Sequencerevent=0xb1,any=1,cmask=1,period=2000000,umask=0x3fUops executed on port 0Uops executed on port 1DTLB first level misses but second level hitevent=0x49,period=200000,umask=0x2Counts duration of L1D miss outstanding, that is each cycle number of Fill Buffers (FB) outstanding required by Demand Reads. FB either is held by demand loads, or it is held by non-demand loads and gets hit at least once by demand. The valid outstanding interval is defined until the FB deallocation by one of the following ways: from FB allocation, if FB is allocated by demand from the demand Hit FB, if it is allocated by hardware or software prefetch.Note: In the L1D, a Demand Read contains cacheable or noncacheable demand loads, including ones causing cache-line splits and reads due to page walks resulted from any request typeCounts the number of L2 cache lines filling the L2. Counting does not cover rejectsmem_inst_retired.stlb_miss_loadsCounts the Demand Data Read requests sent to uncore. Use it in conjunction with OFFCORE_REQUESTS_OUTSTANDING to determine average latency in the uncoreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10001C0004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x801C0004offcore_response.demand_code_rd.l3_hit_m.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400004offcore_response.demand_data_rd.l3_hit_e.snoop_hit_no_fwdoffcore_response.demand_data_rd.l3_hit_s.snoop_hitmoffcore_response.demand_rfo.l3_hit_m.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200408000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40028000Counts once for most SIMD scalar computational single precision floating-point instructions retired. Counts twice for DPP and FM(N)ADD/SUB instructions retiredevent=0xc6,period=100007,umask=0x1,frontend=0x11Counts retired Instructions that experienced iTLB (Instruction TLB) true miss (Precise event)Execution stalls while L3 cache miss demand load is outstandingevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x203C400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020002rtm_retired.aborted_timerNumber of times we entered an RTM region. Does not count nested transactionsThis event counts far branch instructions retired  Spec update: SKL091 (Precise event)event=0x3c,period=25003,umask=0x2Core crystal clock cycles when this thread is unhalted and the other thread is haltedCycles where no uops were executed, the Reservation Station was not empty, the Store Buffer was full and there was no outstanding loadThis event counts cycles during which the microcode scoreboard stalls happenNumber of macro-fused uops retired. (non precise)event=0x49,cmask=1,period=100003,umask=0x10Counts 1 per cycle for each PMH that is busy with a page walk for an instruction fetch request. EPT page walk duration are excluded in SkylakeThis event counts the number of load ops retired that got data from the other core or from the other module (Precise event)Counts demand and DCU prefetch instruction cacheline that miss L2 with a snoop miss responseCountsof demand RFO requests to write to partial cache lines that miss L2Counts data cacheline reads generated by L2 prefetchers that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedALL_BRANCHES counts the number of any mispredicted branch instructions retired. This umask is an architecturally defined event. This event counts the number of retired branch instructions that were mispredicted by the processor, categorized by type. A branch misprediction occurs when the processor predicts that the branch would be taken, but it is not, or vice-versa.  When the misprediction is discovered, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct path (Precise event)JCC counts the number of mispredicted conditional branches (JCC) instructions retired.  This event counts the number of retired branch instructions that were mispredicted by the processor, categorized by type. A branch misprediction occurs when the processor predicts that the branch would be taken, but it is not, or vice-versa.  When the misprediction is discovered, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct path (Precise event)Counts the number of mispredicted near indirect JMP and near indirect CALL branch instructions retired (Precise event)Retired load uops which data sources were load uops missed L1 but hit FB due to preceding miss to the same cache line with data not ready. (Precise Event - PEBS) (Precise event)This event counts line-splitted load uops retired to the architected path. A line split is across 64B cache-line which includes a page split (4K). (Precise Event - PEBS) (Precise event)Counts demand & prefetch code reads that hit in the LLC and the snoops sent to sibling cores return clean responseCounts data/code/rfo reads (demand & prefetch) that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.pf_llc_rfo.llc_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400090REQUEST = PF_DATA_RD and RESPONSE = LLC_MISS_LOCAL and SNOOP = DRAMREQUEST = PF_RFO and RESPONSE = LLC_MISS_LOCAL and SNOOP = DRAMREQUEST = PF_LLC_DATA_RD and RESPONSE = LLC_MISS_LOCAL and SNOOP = DRAMevent=0x60,cmask=1,period=2000000,umask=0x8REQUEST = ANY_DATA read and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = ANY_REQUEST and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HIToffcore_response.data_ifetch.all_local_dram_and_remote_cache_hitREQUEST = DATA_IN and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = PF_IFETCH and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = ANY_REQUEST and RESPONSE = REMOTE_DRAMREQUEST = CORE_WB and RESPONSE = REMOTE_DRAMoffcore_response.demand_rfo.other_local_dramREQUEST = PF_IFETCH and RESPONSE = ANY_DRAM AND REMOTE_FWDoffcore_response.prefetch.any_dram_and_remote_fwdevent=0x4f,period=2000000,umask=0x10event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2777offcore_response.pf_l1d_and_sw.l3_hit.hit_other_core_no_fwdoffcore_response.pf_l2_rfo.l3_hit.hit_other_core_no_fwdOFFCORE_RESPONSE.PF_L3_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDCounts all prefetch (that bring data to LLC only) RFOs that hit in the L3offcore_response.all_data_rd.l3_miss.any_snoopoffcore_response.all_data_rd.l3_miss.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00491offcore_response.all_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000080Counts all prefetch (that bring data to LLC only) RFOs that miss the L3 and the modified data is transferred from remote cacheevent=0x28,period=200003,umask=0x181000000000 * ( cha@event\=0x36\,umask\=0x21\,config\=0x40433@ / cha@event\=0x35\,umask\=0x21\,config\=0x40433@ ) / ( cha_0@event\=0x0@ / duration_time )Read Pending Queue Allocations. Unit: uncore_imc event=0x50,umask=0x04unc_upi_txl_flits.all_dataunc_iio_data_req_of_cpu.mem_read.part0PCI Express bandwidth writing at IIO, part 2. Unit: uncore_iio Core Cross Snoops Issued; Multiple Core Requests. Unit: uncore_cha unc_cha_dir_update.haunc_cha_fast_asserted.horzCounts when a transaction with the opcode type RspI Snoop Response was received which indicates the remote cache does not have the data, or when the remote cache silently evicts data (such as when an RFO: the Read for Ownership issued before a write hits non-modified data)RspSFwd Snoop Responses Received. Unit: uncore_cha Counts ever peer to peer read request for 4 bytes of data made by a different IIO unit to the MMIO space of a card on IIO Part1. Does not include requests made by the same IIO unit. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busevent=0x83,ch_mask=0x04,fc_mask=0x07,umask=0x08Counts every peer to peer read request for 4 bytes of data made by IIO Part3 to the MMIO space of an IIO target. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busCounts every read request for up to a 64 byte transaction of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part1. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busevent=0xc1,ch_mask=0x01,fc_mask=0x07,umask=0x02Counts every peer to peer write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part2 by a different IIO unit. Does not include requests made by the same IIO unit. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busevent=0x84,ch_mask=0x08,fc_mask=0x07,umask=0x08event=0x22,umask=0x2unc_m2m_direct2upi_not_taken_dirstateunc_m2m_directory_lookup.state_ievent=0x2e,umask=0x20unc_m2m_directory_update.anyunc_m2m_prefcam_demand_promotionsPrefecth requests that got turn into a demand request. Unit: uncore_m2m event=0x9unc_upi_rxl_flits.all_nullmem_load_retired.local_pmmocr.all_data_rd.l3_hit_f.no_snoop_neededocr.all_data_rd.l3_hit_f.snoop_missOCR.ALL_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDED  OCR.ALL_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100491OCR.ALL_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDED  OCR.ALL_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDOCR.ALL_DATA_RD.L3_HIT_S.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0490OCR.ALL_PF_DATA_RD.L3_HIT_E.SNOOP_MISSOCR.ALL_PF_RFO.L3_HIT_F.ANY_SNOOP  OCR.ALL_PF_RFO.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040120ocr.all_reads.l3_hit.hit_other_core_no_fwdocr.all_reads.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100122ocr.demand_code_rd.l3_hit_s.no_snoop_neededocr.demand_data_rd.l3_hit_m.any_snoopocr.demand_rfo.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100002ocr.pf_l1d_and_sw.l3_hit.hitm_other_coreCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_HIT.NO_SNOOP_NEEDED OCR.PF_L1D_AND_SW.L3_HIT.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100400ocr.pf_l1d_and_sw.l3_hit_s.hit_other_core_fwdCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_S.HIT_OTHER_CORE_FWDCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_HIT.ANY_SNOOP OCR.PF_L2_DATA_RD.L3_HIT.ANY_SNOOPocr.pf_l2_data_rd.l3_hit_e.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040010ocr.pf_l2_data_rd.l3_hit_m.snoop_noneCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_M.ANY_SNOOPocr.pf_l2_rfo.l3_hit_s.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100020ocr.pf_l2_rfo.l3_hit_s.snoop_noneocr.pf_l3_data_rd.l3_hit.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_M.ANY_SNOOPocr.pf_l3_data_rd.l3_hit_m.snoop_missocr.pf_l3_rfo.l3_hit.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080100Counts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_S.ANY_SNOOPoffcore_response.all_data_rd.l3_hit_s.snoop_missThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_E.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_F.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_M.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_S.SNOOP_MISSoffcore_response.all_pf_rfo.pmm_hit_local_pmm.any_snoopThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.SUPPLIER_NONE.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020120This event is deprecated. Refer to new event OCR.ALL_PF_RFO.SUPPLIER_NONE.SNOOP_NONEoffcore_response.all_reads.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.PMM_HIT_LOCAL_PMM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x804007F7offcore_response.all_rfo.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDoffcore_response.demand_code_rd.l3_hit_f.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDoffcore_response.other.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_M.ANY_SNOOPoffcore_response.pf_l2_data_rd.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020010offcore_response.pf_l2_rfo.l3_hit_e.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_RFO.SUPPLIER_NONE.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.HITM_OTHER_COREoffcore_response.pf_l3_rfo.l3_hit_e.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_F.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020100ocr.all_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdocr.all_pf_data_rd.l3_miss_local_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000490ocr.all_pf_rfo.l3_miss.snoop_missOCR.ALL_PF_RFO.L3_MISS.SNOOP_NONE OCR.ALL_PF_RFO.L3_MISS.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000120ocr.all_pf_rfo.l3_miss_local_dram.snoop_missOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.SNOOP_NONEOCR.ALL_READS.L3_MISS.REMOTE_HITM OCR.ALL_READS.L3_MISS.REMOTE_HITMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B8007F7OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000122Counts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.HITM_OTHER_CORE OCR.DEMAND_DATA_RD.L3_MISS.HITM_OTHER_CORECounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000001Counts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.REMOTE_HITMocr.pf_l1d_and_sw.l3_miss_local_dram.any_snoopCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000400ocr.pf_l2_rfo.l3_miss.snoop_missocr.pf_l2_rfo.l3_miss.snoop_noneocr.pf_l2_rfo.l3_miss_local_dram.no_snoop_neededCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.PF_L3_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPocr.pf_l3_data_rd.l3_miss_remote_hop1_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.REMOTE_HITMThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.all_pf_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdoffcore_response.all_reads.l3_miss.remote_hit_forwardThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.ANY_SNOOPoffcore_response.demand_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreoffcore_response.demand_data_rd.l3_miss_remote_hop1_dram.snoop_noneoffcore_response.other.l3_miss_local_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.pf_l1d_and_sw.l3_miss.snoop_missThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.pf_l2_rfo.l3_miss.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.REMOTE_HITMoffcore_response.pf_l3_rfo.l3_miss_remote_hop1_dram.hit_other_core_no_fwdocr.all_pf_data_rd.pmm_hit_local_pmm.snoop_noneocr.all_rfo.supplier_none.no_snoop_neededocr.pf_l2_data_rd.supplier_none.snoop_noneocr.pf_l2_rfo.supplier_none.no_snoop_neededocr.pf_l3_rfo.pmm_hit_local_pmm.snoop_not_neededocr.pf_l3_rfo.supplier_none.hit_other_core_fwdTag Check; CleanCounts all retired load instructions. This event accounts for SW prefetch instructions for loads  Supports address when precise (Precise event)ocr.demand_data_rd.l3_hit.snoop_hit_no_fwdCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit a cacheline in the L3 where a snoop hit in another cores caches, data forwarding is required as the data is modifiedCycles where at least 1 outstanding data read request is pendingevent=0xc7,period=100003,umask=0x8event=0x80,period=500009,umask=0x4C10_Pkg_ResidencyCounts hardware prefetch RFOs (which bring data to L2) that was not supplied by the L3 cacheSpeculatively counts the number of Transactional Synchronization Extensions (TSX) aborts due to a data capacity limitation for transactional readsCounts miscellaneous requests, such as I/O and un-cacheable accesses that DRAM supplied the requestevent=0xc4,period=400009,umask=0x11event=0xa3,cmask=1,period=1000003,umask=0x1TMA slots where no uops were being issued due to lack of back-end resourcesCounts the number of page walks outstanding for a store in the PMH (Page Miss Handler) each cycleCounts cycles when at least one PMH (Page Miss Handler) is busy with a page walk for a code (instruction fetch) requestmem_load_l3_hit_retired.xsnp_no_fwdRetired load instructions with local Intel Optane DC persistent memory as the data source where the data request missed all caches  Supports address when precise (Precise event)Counts demand data reads that were supplied by a cache on a remote socket where a snoop hit a modified line in another core's caches which forwarded the dataocr.demand_rfo.l3_hitocr.streaming_wr.l3_miss_localLine not found snoop replyHitM snoop reply with data, line kept in Shared statecore_snoop_response.s_hit_fseocr.demand_data_rd.snc_pmmCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were supplied by PMM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those PMM accesses that are controlled by the close SNC ClusterCounts hardware prefetches to the L3 only that were not supplied by the local socket's L1, L2, or L3 caches and the cacheline was homed in a remote socketevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x73C000477event=0xb7,period=100003,umask=0x1,offcore_rsp=0x730000477PMM Commands : Writes. Unit: uncore_imc event=0x35,umask=0xC817FE01TOR Inserts; DRd Pref misses from local IA. Unit: uncore_cha TOR Inserts : RFO_Prefs issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_drd_ddrevent=0x35,umask=0xC8170601Data requested by the CPU : Core reporting completion of Card read from Core DRAM. Unit: uncore_iio event=0xc0,ch_mask=0x20,fc_mask=0x07,umask=0x04unc_iio_txn_req_of_cpu.mem_write.part7event=0x84,ch_mask=0x10,fc_mask=0x07,umask=0x04event=0xc2,ch_mask=0x04,fc_mask=0x04,umask=0x03event=0xc2,ch_mask=0x08,fc_mask=0x04,umask=0x03unc_iio_comp_buf_inserts.cmpd.part6PCIe Completion Buffer Occupancy of completions with data : Part 0. Unit: uncore_iio : All Inserts Inbound (p2p + faf + cset). Unit: uncore_irp Multi-socket cacheline Directory Lookups : Found in I state. Unit: uncore_m2m M2M Reads Issued to iMC : PMM - All Channels. Unit: uncore_m2m Valid Flits Received : All Non Data. Unit: uncore_upi Valid Flits Sent : All Non Data. Unit: uncore_upi Counts the number of demand and prefetch transactions that the External Queue (XQ) rejects due to a full or near full condition which likely indicates back pressure from the IDI link.  The XQ may reject transactions from the L2Q (non-cacheable requests), BBL (L2 misses) and WOB (L2 write-back victims)event=0x34,period=200003,umask=0x8ocr.all_code_rd.l3_hit.snoop_hit_no_fwdocr.all_code_rd.l3_hit.snoop_hit_with_fwdCounts L2 cache hardware prefetch data reads (written to the L2 cache only) that were supplied by the L3 cacheCounts modified writeBacks from L2 cache that miss the L3 cache that were supplied by the L3 cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1002003C0000ocr.all_code_rd.l3_missocr.demand_data_and_l1pf_rd.l3_missCounts the number of unhalted cycles a core is blocked due to an accepted lock issued by other coresCounts the number of unhalted cycles a core is blocked due to an accepted lock it issued. Counts on a per core basisCounts L2 cache hardware prefetch data reads (written to the L2 cache only) that have any type of responseCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)Counts the total number of BTCLEARSCounts the number of unhalted reference clock cycles at TSC frequencyThis event is deprecated. Refer to new event TOPDOWN_BAD_SPECULATION.FASTNUKECounts the number of issue slots every cycle that were not consumed by the backend due to the physical register file unable to accept an entry (marble stalls)TOR Occupancy : RFOs issued by iA Cores : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsData requested of the CPU : CmpD - device sending completion to CPU request : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 1Data requested of the CPU : Card reading from DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 5event=0x49,period=200003,umask=0x8Counts the number of page walks completed due to instruction fetch misses to any page size( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE) ) / ( 2 * CPU_CLK_UNHALTED.DISTRIBUTED )Counts the total number of issue slots  that were not consumed by the backend due to backend stalls. Unit: cpu_atom Instructions Per Cycle. Unit: cpu_atom CPU_CLK_UNHALTED.CORE:k / CPU_CLK_UNHALTED.COREAverage CPU Utilization. Unit: cpu_atom mem_uops_retired.load_latency_gt_128event=0xd0,period=1000003,umask=0x5,ldlat=0x40Number of phases a demand request has waited due to L1D Fill Buffer (FB) unavailablability. Unit: cpu_core Number of PREFETCHT1 or PREFETCHT2 instructions executed. Unit: cpu_core Counts number of SSE/AVX computational 128-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 2 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. Unit: cpu_core event=0x79,cmask=1,period=2000003,umask=0x20event=0x47,cmask=2,period=1000003,umask=0x2event=0xcd,period=1000003,umask=0x2Counts demand data reads that have any type of response. Unit: cpu_atom Counts the number of machine clears due to memory ordering in which an internal load passes an older store within the same CPU. Unit: cpu_atom event=0xc3,period=1000003,umask=0x80Counts the number of issue slots every cycle that were not delivered by the frontend due to decode stalls. Unit: cpu_atom event=0x1cCounts every 64B write request entering the Memory Controller 1 to DRAM (sum of all channels). Each write request counts as a new request incrementing this counter. However, same cache line write requests (both full and partial) are combined to a single 64 byte data transfer to DRAM. Unit: uncore_imc Cycles when at least one PMH is busy with a page walk for a store. Unit: cpu_core Number of page walks outstanding for an outstanding code request in the PMH each cycle. Unit: cpu_core L2_LINES_OUT.NON_SILENTNumber of completed demand load requests that missed the L1 data cache including shadow misses (FB hits, merge to an ongoing L1D miss)FP_ARITH_DISPATCHED.PORT_5fp_arith_inst_retired2.512b_packed_halfCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were not supplied by the local socket's L1, L2, or L3 cachesevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x708000002ARITH.IDIV_ACTIVECycles with retired uop(s)UOPS_RETIRED.MSevent=0x5,umask=0x00000000D0Data requested of the CPU : Card writing to DRAM. Unit: uncore_iio event=0x83,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000002event=0x35,umask=0x00c8178601TOR Occupancy for DRds issued by iA Cores targeting PMM Mem that Missed the LLC. Unit: uncore_cha event=0x2,umask=0x0000000002event=0x4,umask=0x000000000fevent=0x5,umask=0x000000000fRxQ Flit Buffer Allocations : Slot 2. Unit: uncore_upi Tx Flit Buffer Bypassed. Unit: uncore_upi event=0x83,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000080event=0x58,umask=0x0000000002event=0x58,umask=0x0000000008event=0x55,umask=0x0000000002unc_cha_tor_inserts.mmcfgTOR Inserts : Just Local Targets. Unit: uncore_cha event=0x35,umask=0x00c887fd01TOR Inserts; RFO pref from local IA. Unit: uncore_cha TOR Inserts; DRd Pref from local IA. Unit: uncore_cha event=0x36,umask=0x0000000020unc_cha_tor_occupancy.ia_hit_crd_prefevent=0x36,umask=0x00c807ff01TOR Occupancy; LLCPrefRFO from local IA. Unit: uncore_cha TOR Occupancy; RdCur and FsRdCur hits from local IO. Unit: uncore_cha TOR Inserts : DRd_Prefs issued by iA Cores targeting PMM Mem that Missed the LLC. Unit: uncore_cha TOR Inserts : WCiLs issued by iA Cores targeting PMM that missed the LLC - HOMed remotely. Unit: uncore_cha TOR Occupancy : DRds issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha event=0x36,umask=0x00c8978a01TOR Occupancy : CRd issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha TOR Occupancy : CLFlushes issued by iA Cores. Unit: uncore_cha unc_cha_tor_occupancy.ia_wcilunc_cha_tor_occupancy.ia_miss_local_wcil_pmmevent=0x36,umask=0x00c86f0a01TOR Occupancy : WCiLs issued by iA Cores targeting DDR that missed the LLC. Unit: uncore_cha event=0x60,umask=0x10All L2 Cache Requests (Breakdown 1 - Common). PrefetchL2CmdAll L2 Cache Requests (Breakdown 2 - Rare). Data cache read sized non-cacheablel2_wcb_req.cl_zeroevent=0xc1Retired Near ReturnsThe number of near return instructions (RET or RET Iw) retiredremote_outbound_data_controller_0event=0x807,umask=0x02dram_channel_data_controller_0event=0x41,umask=0x08LS MAB allocates by type - loadsevent=0x45,umask=0x20ic_oc_mode_switch.ic_oc_mode_switchall_l2_cache_hitsL3 Misses (includes Chg2X). Unit: amd_l3 d_ratio(l2_cache_req_stat.ic_access_in_l2, bp_l1_tlb_fetch_hit + bp_l1_tlb_miss_l2_hit + bp_l1_tlb_miss_l2_miss)bp_l1_tlb_fetch_hit.if2mevent=0x94,umask=0x01Total number uOps assigned to pipe 1fp_disp_faults.x87_fill_faultNumber of interrupts takenDemand Data Cache Fills by Data Source. DRAM or IO from this thread's dieDemand Data Cache Fills by Data Source. Hit in cache; local CCX (not Local L2), or Remote CCX and the address's Home Node is on this thread's diels_l1_d_tlb_miss.tlb_reload_coalesced_page_hitCount of dispatched Ops from OpCacheThe number of valid fills into the ITLB originating from the LS Page-Table Walker. Tablewalk requests are issued for L1-ITLB and L2-ITLB misses. Walk to 4K pageInstruction Cache Miss. Counts various IC tag related hit and miss eventsevent=0x18e,umask=0x07Op Cache Hit. Counts Op Cache micro-tag hit/miss eventsLoad-op-Store Dispatch. Dispatch of a single op that performs a load from and store to the same memory address. Counts the number of operations dispatched to the LS unit. Unit Masks ADDedL1 DTLB Miss. DTLB reload to a 1G page that also missed in the L2 TLBevent=0x47,umask=0x01Cycles where a dispatch group is valid but does not get dispatched due to a token stall. No tokens for Integer Scheduler Queue 3 availablede_dis_dispatch_token_stalls2.int_sch1_token_stallevent=0x44,umask=0x14Demand data cache fills from all types of data sourcesAny data cache fills from all types of data sourcesL2 cache requests: prefetch directly into L2event=0x70,umask=0x10l2_pf_miss_l2_hit_l3.l2_burstL2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L1Stride (fetch additional lines into L1 cache when each access is a constant distance from the previous)L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of all typesl3_xi_sampled_latency_requests.ext_farRetired CLFLUSH instructionsCycles with no retire because thread arbitration did not select the threadRead data beats (64 bytes) for local processor at Coherent Station (CS) 2Read data beats (64 bytes) for local processor at Coherent Station (CS) 8local_processor_write_data_beats_cs1event=0x5f,umask=0x7ffremote_processor_read_data_beats_cs5event=0x8df,umask=0xbfflocal_socket_inf0_inbound_data_beats_ccm3event=0x59e,umask=0x7feData beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 3local_socket_inf1_inbound_data_beats_ccm4event=0x51e,umask=0x7ffData beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 0Data beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 0event=0x49e,umask=0xbffremote_socket_inf0_outbound_data_beats_ccm3Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 2Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 5Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 7Data beats (64 bytes) for local socket outbound data from inter-socket xGMI link 1event=0xc1f,umask=0xf3eRetired x87 floating-point multiply opsRetired SSE and AVX floating-point multiply-accumulate ops (each operation is counted as 2 ops)Retired floating-point ops of all widthsfp_ops_retired_by_type.scalar_sqrtRetired vector floating-point divide opsevent=0xa,umask=0xd0event=0xa,umask=0xffRetired MMX integer compare opsRetired SSE and AVX integer ops of other typesfp_pack_ops_retired.fp128_addevent=0xc,umask=0x03fp_pack_ops_retired.fp256_allevent=0xd,umask=0x08event=0xd,umask=0x20Retired 256-bit packed integer multiply opsNumber of ACTIVATE commands sentumc_pchg_cmd.rdNumber of clocks used by the data busCycles when the op queue is empty. Such cycles indicate that the front-end is not delivering instructions fast enoughOps fetched from op cache and dispatchedOps dispatched from any sourceTotal dispatch slots (upto 6 instructions can be dispatched in each cycle)l2_cache_misses_from_l1_ic_miss(umc_cas_cmd.wr * 64) / 1e6 / duration_timeumc_precharge_cmd_rateDC_MISALIGNED_DATA_REFERENCEDC_MICROARCHITECTURAL_LATE_CANCELBU_FILL_INTO_L2NB_HT_BUS1_BANDWIDTHL2_ACCESSPRED_BRANCH_EXEC_TAKENEVENT_0AHEVENT_18HEVENT_59HEVENT_87HEVENT_A4HEVENT_C6HPREDICTABLE_FUNCTION_RETURNLOAD_STORE_PIPEDMB_STALLEVENT_118HEVENT_132HEVENT_13CHEVENT_13DHEVENT_157HEVENT_158HEVENT_16AHEVENT_17DHEVENT_18BHEVENT_18CHEVENT_190HEVENT_197HEVENT_19DHEVENT_1A5HEVENT_1A7HEVENT_1BBHEVENT_1CFHEVENT_1F0HEVENT_226HEVENT_242HEVENT_249HEVENT_2DBHEVENT_2E3HEVENT_2F7HEVENT_303HEVENT_336HEVENT_33EHEVENT_388HEVENT_393HEVENT_3A2HEVENT_3BCHEVENT_3E1HEVENT_3E6HL1D_CACHE_WB_VICTIMDSB_SPECRC_LD_SPECL2D_TLBdn_rxreq_pici_dvmophni_arvalid_no_arreadysbsx_wvalid_no_wreadyrnd_rrt_occ_ovflrni_s0_rdata_beatscxla_tx_cxs_link1THRESHOLD_INSTR_QUEUE_ENTRIES_CYCLESL1_DATA_STORE_MISSL1_DATA_CACHE_OP_HITCYCLES_NO_INSTR_DISPATCHEDTOTAL_TRANSLATEDDATA_MMU_MISSSTASH_HITSSTASH_REQUESTS_L1SNOOP_THROTTLING_TURNED_ONCYCLES_COMPLETION_STALLED_NEXUS_FIFO_FULLdirty-l2-victimlow-op-pos-2postwrszdwordrdmodwrbranch-mispredictsUSERINTEL_SANDYBRIDGEINTEL_HASWELL_XEONUNKNOWNDISABLEDALLOCATED/usr/src/lib/libpmc/pmclog.cUNHALTED-CORE-CYCLESl3_request_g1.caching_l3_cache_accessesRESOURCE_STALLS_ANYinst_retired.anyevent{"type": "pmcattach"%s, "pmcid": "0x%08x", "event": "0x%08x", "flags": "0x%08x", "rate": "%jd"}
GenuineIntel-6-254 * CPU_CLK_UNHALTED.THREADFLOPcActual Average Latency for L1 data-cache miss demand load instructions (in core cycles)Mem;MemoryBWAverage Frequency Utilization relative nominal frequencyThis event counts the number of L2 cache lines filling the L2. Counting does not cover rejectsl2_lines_in.ievent=0x24,period=200003,umask=0xf8RFO requests to L2 cacheevent=0x24,period=200003,umask=0x44L2 or L3 HW prefetches that access L2 cachel2_trans.code_rdRFO requests that access L2 cacheevent=0xd1,period=100007,umask=0x20offcore_requests_outstanding.cycles_with_data_rdevent=0x60,cmask=1,period=2000003,umask=0x1Number of SSE/AVX computational scalar floating-point instructions retired. Applies to SSE* and AVX* scalar, double and single precision floating-point: ADD SUB MUL DIV MIN MAX RSQRT RCP SQRT FM(N)ADD/SUB. FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementfp_arith_inst_retired.singleidq.ms_cyclesThis event counts, on the per-thread basis, cycles when no uops are delivered to Resource Allocation Table (RAT). IDQ_Uops_Not_Delivered.core =4This event counts, on the per-thread basis, cycles when less than 1 uop is  delivered to Resource Allocation Table (RAT). IDQ_Uops_Not_Delivered.core >=3event=0x54,period=2000003,umask=0x1Number of times a TSX Abort was triggered due to release/commit but data and address mismatchNumber of times a TSX Abort was triggered due to attempting an unsupported alignment from Lock BufferCycles when L1 and L2 are locked due to UC or split lockSpeculative and retired indirect branches excluding calls and returnsTaken speculative and retired macro-conditional branchesReturn instructions retired. (Precise Event - PEBS) (Precise event)Mispredicted macro branch instructions retired. (Precise Event - PEBS) (Must be precise)This event counts the number of reference cycles when the core is not in a halt state. The core enters the halt state when it is running the HLT instruction or the MWAIT instruction. This event is not affected by core frequency changes (for example, P states, TM2 transitions) but has the same incrementing frequency as the time stamp counter. This event can approximate elapsed time while the core was not in a halt state. This event has a constant ratio with the CPU_CLK_UNHALTED.REF_XCLK event. It is counted on a dedicated fixed counter, leaving the four (eight when Hyperthreading is disabled) programmable counters available for other events. 
Note: On all current platforms this event stops counting during 'throttling (TM)' states duty off periods the processor is 'halted'.  This event is clocked by base clock (100 Mhz) on Sandy Bridge. The counter update is done at a lower clock rate then the core clock the overflow status bit for this counter may appear 'sticky'.  After the counter has overflowed and software clears the overflow status bit and resets the counter to less than MAX. The reset value to the counter is not clocked immediately so the overflow status bit will flip 'high (1)' and generate another PMI (if enabled) after which the reset value gets clocked into the counter. Therefore, software will get the interrupt, read the overflow status bit '1 for bit 34 while the counter value is less than MAX. Software should ignore this casecycle_activity.cycles_l1d_missExecution stalls while L2 cache miss demand load is outstandingInstructions retired from executionMaskmov false fault - counts number of time ucode passes through Maskmov flow due to instruction's mask being 0 while the flow was completed without raising a faultevent=0xcc,period=2000003,umask=0x20Cycles per thread when uops are executed in port 0event=0xa1,period=2000003,umask=0x8uops_dispatched_port.port_5event=0xb1,cmask=1,inv=1,period=2000003,umask=0x1uops_retired.allevent=0x37,umask=0x1M line evictions from LLC (writebacks to memory). Unit: uncore_cbox event=0x35,umask=0x3,filter_opc=0x187event=0x35,umask=0x1,filter_opc=0x1c8,filter_tid=0x3eunc_h_snoop_resp.rspcnflct(UNC_M_POWER_CHANNEL_PPD / UNC_M_DCLOCKTICKS) * 100.event=0x5dtlb_load_misses.stlb_hitdtlb_load_misses.stlb_hit_4kevent=0x8,period=2000003,umask=0x4Store operations that miss the first TLB level but hit the second and do not cause page walksitlb_misses.walk_completedCode miss in all TLB levels causes a page walk that completes. (4K)  Spec update: BDM69Mem;SoCAverage number of parallel requests to external memory. Accounts for all requestsRetired load uops with L1 cache hits as data sources  Supports address when precise (Precise event)Retired load uops which data sources were data hits in L3 without snoops required  Supports address when precise.  Spec update: BDM100 (Precise event)offcore_response.all_data_rd.l3_hit.snoop_hitmoffcore_response.all_pf_rfo.l3_hit.snoop_missoffcore_response.all_pf_rfo.supplier_none.snoop_not_neededCounts all demand & prefetch RFOs have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020008offcore_response.demand_code_rd.l3_hit.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020004offcore_response.demand_data_rd.supplier_none.snoop_missoffcore_response.pf_l2_code_rd.l3_hit.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0040offcore_response.pf_l2_code_rd.supplier_none.any_snoopoffcore_response.pf_l2_code_rd.supplier_none.snoop_noneoffcore_response.pf_l2_data_rd.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020100Randomly selected loads with latency value being above 64  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)offcore_response.all_pf_rfo.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000122offcore_response.corewb.l3_miss.snoop_noneoffcore_response.demand_data_rd.l3_miss.snoop_noneoffcore_response.demand_data_rd.l3_miss_local_dram.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020100event=0x34,umask=0x26offcore_response.demand_rfo.llc_hit.any_responseCounts all demand & prefetch code reads miss in the L3offcore_response.all_data_rd.llc_miss.remote_hit_forwardevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC007F7Counts all demand data writes (RFOs) miss the L3 and the modified data is transferred from remote cachel1d_cache.all_refl1d_cache.evictevent=0x2c,period=200000,umask=0x48L2 cacheable instruction fetch requestsl2_ld.self.any.i_statel2_ld.self.any.mesil2_ld.self.prefetch.i_stateevent=0x2d,period=200000,umask=0x48l2_reject_busq.self.demand.i_stateevent=0x2e,period=200000,umask=0x7fl2_rqsts.self.prefetch.mesievent=0x2a,period=200000,umask=0x41Floating point assists for retired operationsevent=0xaa,period=2000000,umask=0x2event=0xaa,period=2000000,umask=0x1event=0x7d,period=200000,umask=0x40bus_drdy_clocks.all_agentsbus_trans_mem.all_agentsPartial bus transactionsevent=0x66,period=200000,umask=0xe0event=0x88,period=2000000,umask=0x20Instructions retired (precise event) (Must be precise)mul.sMicro-op reissues on a store-load collisionuops_retired.anyitlb.flushmem_load_retired.dtlb_misspage_walks.i_side_walksl2_reject_xq.allevent=0xd0,period=200003,umask=0x21Counts requests to the uncore subsystem that have any transaction responses from the uncore subsystemoffcore_response.corewb.l2_miss.hitm_other_coreoffcore_response.demand_code_rd.l2_miss.snoop_miss_or_no_snoop_neededdecode_restriction.predecode_wronghw_interrupts.maskedbr_inst_retired.rel_callRetired mispredicted near indirect call instructions (Precise event capable) (Must be precise)event=0xc5,period=200003,umask=0x7eCounts the number of reference cycles that the core is not in a halt state. The core enters the halt state when it is running the HLT instruction.  In mobile systems the core frequency may change from time.  This event is not affected by core frequency changes but counts as if the core is running at the maximum frequency all the time.  This event uses fixed counter 2.  You cannot collect a PEBs record for this eventevent=0xca,period=200003Counts the number of issue slots per core cycle that were not consumed by the backend because allocation is stalled waiting for a mispredicted jump to retire or other branch-like conditions (e.g. the event is relevant during certain microcode flows).   Counts all issue slots blocked while within this window including slots where uops were not available in the Instruction QueueCounts machine clears due to memory disambiguation.  Memory disambiguation happens when a load which has been issued conflicts with a previous unretired store in the pipeline whose address was not known at issue time, but is later resolved to be the same as the load addressCounts data reads (demand & prefetch) hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts requests to the uncore subsystem outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts bus lock and split lock requests true miss for the L2 cache with a snoop miss in the other processor moduleCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache line reads generated by hardware L1 data cache prefetcher have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests generated by L2 prefetcher true miss for the L2 cache with a snoop miss in the other processor moduleevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200004800Page walk completed due to a demand load to a 4K pageRetired load uops misses in L1 cache as data sources  Supports address when precise.  Spec update: HSM30 (Precise event)Demand RFO read requests sent to uncore, including regular RFOs, locks, ItoMoffcore_response.pf_l2_data_rd.l3_hit.any_responseRandomly selected loads with latency value being above 128  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00040Counts prefetch (that bring data to L2) data reads miss in the L3Number of near branch instructions retired that were taken but mispredicted (Precise event)ild_stall.iq_fullNumber of instructions at retirement  Spec update: HSD11, HSD140Number of integer move elimination candidate uops that were eliminatedNumber of microcode assists invoked by HW upon uop writebackL3 Lookup external snoop request that access cache and found line in MESI-stateevent=0x22,umask=0x88event=0x22,umask=0x28Number of cache load STLB hits. No page walkStore misses in all DTLB levels that cause completed page walksStore misses in all DTLB levels that cause completed page walks (2M/4M)Store miss in all TLB levels causes a page walk that completes. (4K)page_walker_loads.ept_dtlb_l1event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0600400001l2_l1d_wb_rqsts.hit_eRetired load uops with locked access. (Precise Event)Cases when offcore requests buffer cannot take more entries for coreCounts demand & prefetch data reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresarith.fpu_divCycles per core when uops are dispatched to port 0This 48-bit fixed counter counts the UCLK cycles. Unit: uncore_arb A snoop invalidates a non-modified line in some processor core. Unit: uncore_cbox A snoop hits a non-modified line in some processor core. Unit: uncore_cbox event=0x22,umask=0x20Counts prefetch data reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.all_reads.llc_hit.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c03f7offcore_response.pf_l2_data_rd.llc_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x67f800244event=0xb7,period=100003,umask=0x1,offcore_rsp=0x87f8203f7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x67f800004Counts demand data reads that miss the LLC  and the data returned from remote & local dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc20001Counts prefetch (that bring data to L2) data reads  that miss the LLC  and the data returned from remote dramevent=0x35,umask=0x1,filter_opc=0x1e4unc_c_tor_occupancy.miss_remote(UNC_P_FREQ_BAND3_CYCLES / UNC_P_CLOCKTICKS) * 100.unc_p_freq_band3_transitionsCycles when 1 or more uops were delivered to the by the front endevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x600400077event=0x89,period=200003,umask=0xd0Each cycle there was a MLC-miss pending demand load this thread (i.e. Non-completed valid SQ entry allocated for demand load and waiting for Uncore), increment by 1. Note this is in MLC and connected to Umask 0event=0x3,period=100003,umask=0x10Aliasing occurs when a load is issued after a store and their memory addresses are offset by 4K.  This event counts the number of loads that aliased with a preceding store, resulting in an extended address check in the pipeline.  The enhanced address check typically has a performance penalty of 5 cyclesCounts all the load micro-ops retiredevent=0x4,period=200003,umask=0x20event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400044offcore_response.any_code_rd.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000044event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000044offcore_response.any_data_rd.l2_hit_far_tile_e_fCounts any Prefetch requests that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080070Counts any Read request  that accounts for any responseoffcore_response.any_request.l2_hit_far_tile_e_fCounts any request that accounts for responses which hit its own tile's L2 with data in S stateoffcore_response.any_rfo.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080001offcore_response.demand_rfo.l2_hit_far_tile_e_foffcore_response.partial_reads.l2_hit_this_tile_moffcore_response.partial_writes.l2_hit_far_tile_e_fCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses which hit its own tile's L2 with data in E stateoffcore_response.pf_software.l2_hit_near_tileCounts Software Prefetches that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateCounts Software Prefetches that accounts for responses which hit its own tile's L2 with data in E stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400200Counts UC code reads (valid only for Outstanding response type)  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateCounts all instruction fetches that miss the instruction cache or produce memory requests. An instruction fetch miss is counted only once and not once for every cycle it is outstandingoffcore_response.any_code_rd.mcdramCounts Demand code reads and prefetch code read requests  that accounts for data responses from MCDRAM Localoffcore_response.any_data_rd.mcdram_nearevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101008000offcore_response.demand_code_rd.ddr_farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400080event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400100offcore_response.pf_l2_code_rd.mcdram_farCounts Software Prefetches that accounts for responses from DDR (local and far)unc_m_cas_count.rdL1D cache lines allocated in the M stateL1D load lock accepted in fill bufferL1 writebacks to L2 in M statel2_write.lock.i_stateevent=0x27,period=100000,umask=0x10L2 demand lock RFOs in I state (misses)Offcore data reads satisfied by the IO, CSR, MMIO unitoffcore_response.any_data.remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x744Offcore code reads that HIT in a remote cacheoffcore_response.any_ifetch.remote_cache_hitmOffcore code or data read requests satisfied by a remote cache or remote DRAMoffcore_response.data_in.llc_hit_other_core_hitOffcore demand data requests that HIT in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x803event=0xb7,period=100000,umask=0x1,offcore_rsp=0x401event=0xb7,period=100000,umask=0x1,offcore_rsp=0x104offcore_response.demand_ifetch.remote_cacheOffcore prefetch data requests that HIT in a remote cacheAll offcore prefetch data readsoffcore_response.pf_data_rd.llc_hit_other_core_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4710Offcore prefetch RFO requests satisfied by any cache or DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x720offcore_response.pf_rfo.remote_cacheOffcore prefetch requests satisfied by the LLCstore_blocks.l1d_blockfp_assist.outputfp_comp_ops_exe.sse_fpfp_comp_ops_exe.sse_fp_scalarevent=0x19,period=2000000,umask=0x1Offcore RFO requests satisfied by any DRAMoffcore_response.demand_data_rd.any_llc_missevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF801offcore_response.demand_rfo.remote_draml1i.readsevent=0x7,period=200000,umask=0x1snoop_response.hitInstruction queue forced BACLEARbr_inst_exec.near_callsevent=0x88,period=20000,umask=0x30Retired branch instructions (Precise Event)Cycles when thread is not halted (programmable counter)event=0xc3,period=20000,umask=0x4rat_stalls.scoreboarduops_decoded.esp_syncuops_executed.core_stall_countUops executed on any port (core count)Cycles no Uops were issued on any threadevent=0xc2,cmask=1,period=2000000,umask=0x1event=0xf2,period=200003,umask=0x1Retired load instructions with locked access  Supports address when precise (Precise event)Counts the number of offcore outstanding Demand Data Read transactions in the super queue (SQ) every cycle. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor. See the corresponding Umask under OFFCORE_REQUESTS.Note: A prefetch promoted to Demand is counted from the promotion pointoffcore_response.demand_data_rd.l3_hit_m.spl_hitoffcore_response.demand_data_rd.l3_hit_s.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10001C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x401C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000400002offcore_response.demand_rfo.supplier_none.any_snoopoffcore_response.other.l3_hit_e.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80088000offcore_response.other.l3_hit_e.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0048000event=0xc6,period=100007,umask=0x1,frontend=0x400406frontend_retired.stlb_missCounts cycles during which uops are being delivered to Instruction Decode Queue (IDQ) from the MITE path. Counting includes uops that may 'bypass' the IDQevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFC400001offcore_response.other.l3_hit_m.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFC408000event=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC408000offcore_response.other.l3_miss.spl_hitevent=0x3c,period=25003,umask=0x1Number of cycles using always true condition applied to  PEBS instructions retired event  Spec update: SKL091, SKL044 (Must be precise)Counts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to port 5uops_issued.vector_width_mismatch100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) * ( ( (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) / ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (min( 9 * cpu@DTLB_LOAD_MISSES.STLB_HIT\,cmask\=1@ + DTLB_LOAD_MISSES.WALK_ACTIVE , max( CYCLE_ACTIVITY.CYCLES_MEM_ANY - CYCLE_ACTIVITY.CYCLES_L1D_MISS , 0 ) ) / CPU_CLK_UNHALTED.THREAD) / (max( ( CYCLE_ACTIVITY.STALLS_MEM_ANY - CYCLE_ACTIVITY.STALLS_L1D_MISS ) / CPU_CLK_UNHALTED.THREAD , 0 )) ) + ( (EXE_ACTIVITY.BOUND_ON_STORES / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (( 9 * cpu@DTLB_STORE_MISSES.STLB_HIT\,cmask\=1@ + DTLB_STORE_MISSES.WALK_ACTIVE ) / CPU_CLK_UNHALTED.THREAD) / #(EXE_ACTIVITY.BOUND_ON_STORES / CPU_CLK_UNHALTED.THREAD) ) ) Total penalty related to DSB (uop cache) misses - subset/see of/the Instruction_Fetch_BW BottleneckAll Storesevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000018008event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400008008event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000040The BACLEARS event counts the number of times the front end is resteered, mainly when the Branch Prediction Unit cannot provide a correct prediction and this is corrected by the Branch Address Calculator at the front end.  The BACLEARS.ANY event counts the number of baclears for any type of branchCounts all machine clearsThis event counts the number of times that a program writes to a code section. Self-modifying code causes a severe penalty in all Intel? architecture processorsThis event counts retired load uops that hit in the last-level cache (L3) and were found in a non-modified state in a neighboring core's private cache (same package).  Since the last level cache is inclusive, hits to the L3 may require snooping the private L2 caches of any cores on the same socket that have the line.  In this case, a snoop was required, and another L2 had the line in a non-modified state. (Precise Event - PEBS) (Precise event)Counts demand & prefetch code reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all prefetch code reads that hit in the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0120offcore_response.pf_llc_rfo.llc_hit.snoop_missCounts prefetch (that bring data to LLC only) RFOs that hit in the LLC and the snoops sent to sibling cores return clean responseoffcore_response.all_pf_data_rd.llc_miss.dramoffcore_response.pf_l_ifetch.llc_miss_local.dramOffcore demand code read requestsevent=0x60,cmask=1,period=2000000,umask=0x2offcore_response.any_rfo.local_dram_and_remote_cache_hitREQUEST = OTHER and RESPONSE = LOCAL_CACHEREQUEST = OTHER and RESPONSE = REMOTE_CACHE_HITMREQUEST = PF_DATA and RESPONSE = ANY_CACHE_DRAMREQUEST = PF_DATA and RESPONSE = LLC_HIT_NO_OTHER_COREREQUEST = PF_RFO and RESPONSE = LLC_HIT_NO_OTHER_COREoffcore_response.any_request.other_local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3050event=0xb7,period=100000,umask=0x1,offcore_rsp=0xf840REQUEST = PF_IFETCH and RESPONSE = ANY_LLC_MISSsnoopq_requests_outstanding.dataCycles snoop invalidate requests queueddtlb_misses.walk_cyclesevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x58FFevent=0xf,period=40000,umask=0x4Counts all demand & prefetch RFOs that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.demand_code_rd.l3_hit.no_snoop_neededoffcore_response.demand_data_rd.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00001Counts all demand data writes (RFOs) that miss the L3 and clean or shared data is transferred from remote cacheoffcore_response.pf_l1d_and_sw.l3_miss.remote_hitmCounts all prefetch (that bring data to LLC only) data reads that miss in the L3core_power.lvl1_turbo_licenseLLC misses - Uncacheable reads (from cpu) . Unit: uncore_cha unc_cha_requests.writes_localunc_iio_data_req_of_cpu.mem_read.part2event=0x37,umask=0x01Counts when a RFO (the Read for Ownership issued before a  write) request hit a cacheline in the S (Shared) stateunc_cha_requests.invitoe_remoteIngress (from CMS) Occupancy; IRQ. Unit: uncore_cha Counts number of entries in the specified Ingress queue in each cycleCounts snoop filter capacity evictions for entries tracking shared lines in the cores cache. Snoop filter capacity evictions occur when the snoop filter is full and evicts an existing entry to track a new entry. Does not count clean evictions such as when a cores cache replaces a tracked cacheline with a new cachelineevent=0xc2,ch_mask=0x01,fc_mask=0x4,umask=0x03Counts every peer to peer write request of 4 bytes of data made to the MMIO space of a card on IIO Part3 by a different IIO unit. Does not include requests made by the same IIO unit. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busRead request for up to a 64 byte transaction is made by the CPU to IIO Part1. Unit: uncore_iio Counts every write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part1 by a unit on the main die (generally a core) or by another IIO unit. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busCounts every write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part2 by a unit on the main die (generally a core) or by another IIO unit. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busCounts every write request of up to a 64 byte transaction of data made to the MMIO space of a card on IIO Part3 by a unit on the main die (generally a core) or by another IIO unit. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busRead request for up to a 64 byte transaction is made by IIO Part0 to Memory. Unit: uncore_iio event=0x84,ch_mask=0x01,fc_mask=0x07,umask=0x01Peer to peer write request of up to a 64 byte transaction is made by IIO Part0 to an IIO target. Unit: uncore_iio Counts every peer to peer write request of up to a 64 byte transaction of data made by IIO Part1 to the MMIO space of an IIO target.In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busunc_m2m_direct2core_not_taken_dirstateunc_m2m_directory_update.a2ievent=0x2e,umask=0x40event=0x38,umask=0x2event=0x57Count cases where flow control queue that sits between the Intel Ultra Path Interconnect (UPI) and the mesh spawns a prefetch to the iMC (Memory Controller)uncore_m3upievent=0x21unc_upi_rxl0p_power_cyclesocr.all_data_rd.l3_hit.hitm_other_coreocr.all_data_rd.l3_hit.snoop_missOCR.ALL_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWD  OCR.ALL_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080490OCR.ALL_PF_DATA_RD.L3_HIT_F.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100490ocr.all_pf_rfo.l3_hit_f.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100120OCR.ALL_PF_RFO.L3_HIT_S.NO_SNOOP_NEEDED  OCR.ALL_PF_RFO.L3_HIT_S.NO_SNOOP_NEEDEDOCR.ALL_READS.L3_HIT_E.HIT_OTHER_CORE_FWD  OCR.ALL_READS.L3_HIT_E.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000807F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800407F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10001007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C0122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0122OCR.ALL_RFO.L3_HIT_M.ANY_SNOOP  OCR.ALL_RFO.L3_HIT_M.ANY_SNOOPocr.all_rfo.l3_hit_m.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100122ocr.demand_code_rd.l3_hit.hit_other_core_fwdCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_F.HITM_OTHER_CORECounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_M.HIT_OTHER_CORE_FWDocr.demand_data_rd.l3_hit.hitm_other_coreocr.demand_data_rd.l3_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0001ocr.demand_data_rd.l3_hit_e.hitm_other_coreCounts any other requests OCR.OTHER.L3_HIT.HITM_OTHER_CORE OCR.OTHER.L3_HIT.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80088000Counts any other requests  OCR.OTHER.L3_HIT_F.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080400Counts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_E.HITM_OTHER_CORECounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_F.HITM_OTHER_COREocr.pf_l1d_and_sw.l3_hit_f.hit_other_core_fwdCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_M.ANY_SNOOPCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_S.ANY_SNOOPCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_E.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100080ocr.pf_l3_data_rd.l3_hit_s.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200100ocr.pf_l3_rfo.l3_hit_m.hit_other_core_no_fwdocr.pf_l3_rfo.l3_hit_m.snoop_missThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDoffcore_response.all_data_rd.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_F.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_S.SNOOP_NONEoffcore_response.all_pf_rfo.l3_hit_s.snoop_missoffcore_response.all_reads.l3_hit_e.hit_other_core_no_fwdoffcore_response.all_reads.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_S.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_E.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.demand_code_rd.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.ANY_RESPONSEoffcore_response.demand_rfo.l3_hit_e.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.SUPPLIER_NONE.HITM_OTHER_COREoffcore_response.demand_rfo.supplier_none.no_snoop_neededThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.OTHER.SUPPLIER_NONE.HIT_OTHER_CORE_FWDoffcore_response.other.supplier_none.hit_other_core_no_fwdoffcore_response.pf_l1d_and_sw.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.HITM_OTHER_COREoffcore_response.pf_l2_data_rd.l3_hit_e.hit_other_core_no_fwdoffcore_response.pf_l2_data_rd.l3_hit_m.snoop_missoffcore_response.pf_l2_data_rd.l3_hit_s.snoop_missThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_S.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l2_rfo.l3_hit_e.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_S.NO_SNOOP_NEEDEDoffcore_response.pf_l2_rfo.l3_hit_s.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400020offcore_response.pf_l2_rfo.pmm_hit_local_pmm.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020020This event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.pf_l3_data_rd.l3_hit_e.hitm_other_coreoffcore_response.pf_l3_data_rd.l3_hit_e.hit_other_core_fwdoffcore_response.pf_l3_data_rd.l3_hit_f.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800020080This event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDNumber of SSE/AVX computational 512-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB MUL DIV MIN MAX RCP14 RSQRT14 SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000490OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOP  OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPocr.all_pf_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdOCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.ALL_READS.L3_MISS.HIT_OTHER_CORE_NO_FWDocr.all_reads.l3_miss.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC0007F7ocr.all_reads.l3_miss_local_dram.any_snoopocr.all_reads.l3_miss_local_dram.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10040007F7OCR.ALL_RFO.L3_MISS.HITM_OTHER_CORE OCR.ALL_RFO.L3_MISS.HITM_OTHER_CORE OCR.ALL_RFO.L3_MISS.HITM_OTHER_COREOCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORE  OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.demand_data_rd.l3_miss.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000001ocr.pf_l1d_and_sw.l3_miss.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.PF_L2_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C000010ocr.pf_l2_data_rd.l3_miss.remote_hit_forwardevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000020ocr.pf_l3_data_rd.l3_miss_local_dram.snoop_noneocr.pf_l3_rfo.l3_miss.snoop_missThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.all_pf_rfo.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPoffcore_response.all_pf_rfo.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.all_pf_rfo.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.ANY_SNOOPoffcore_response.all_reads.l3_miss_local_dram.hit_other_core_fwdoffcore_response.all_reads.l3_miss_local_dram.snoop_miss_or_no_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.all_rfo.l3_miss_remote_hop1_dram.snoop_noneThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.REMOTE_HIT_FORWARDoffcore_response.demand_code_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.other.l3_miss_remote_dram.snoop_miss_or_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS.REMOTE_HITMOCR.ALL_READS.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWD  OCR.ALL_READS.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDOCR.ALL_RFO.SUPPLIER_NONE.ANY_SNOOP  OCR.ALL_RFO.SUPPLIER_NONE.ANY_SNOOPocr.all_rfo.supplier_none.snoop_noneOCR.ALL_RFO.SUPPLIER_NONE.SNOOP_NONECounts all demand code reads OCR.DEMAND_CODE_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPCounts demand data reads OCR.DEMAND_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDCounts any other requests  OCR.OTHER.SUPPLIER_NONE.ANY_SNOOPCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDCounts the number of cache lines that have been prefetched by the L2 hardware prefetcher but not used by demand access when evicted from the L2 cacheocr.demand_code_rd.l3_hit.snoop_hitmocr.demand_code_rd.l3_hit.snoop_sentocr.other.l3_hit.snoop_not_neededFor every cycle, increments by the number of outstanding data read requests pendingCounts number of SSE/AVX computational scalar single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SSE* and AVX* scalar single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT RCP FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementCounts the number of Decode Stream Buffer (DSB a.k.a. Uop Cache)-to-MITE speculative transitionsCounts retired Instructions who experienced Instruction L2 Cache true miss (Precise event)Counts retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 128 cycles which was not interrupted by a back-end stall (Precise event)Counts retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 4 cycles which was not interrupted by a back-end stall (Precise event)Counts cycles where a code fetch is stalled due to L1 instruction cache tag missNumber of times an HLE execution aborted due to any reasons (multiple categories may count as one)ocr.demand_data_rd.l3_missocr.demand_rfo.l3_missCounts the number of times we could not allocate Lock Bufferevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000001ocr.hwpf_l1d_and_swpf.local_dramocr.hwpf_l2_rfo.dramCounts hardware prefetch RFOs (which bring data to L2) that DRAM supplied the requestNot taken branch instructions retired (Precise event)br_inst_retired.indirectevent=0xc4,period=100003,umask=0x80uops_decoded.dec0uops_executed.cycles_ge_4Instructions Per Cycle (per physical core)ocr.demand_data_rd.remote_cache.snoop_hitmCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by a cache on a remote socket where a snoop was sent and data was returned (Modified or Not Modified)ocr.demand_rfo.l3_miss_localCounts miscellaneous requests, such as I/O and un-cacheable accesses that were not supplied by the local socket's L1, L2, or L3 cachescore_snoop_response.i_fwd_mHitM snoop reply with data, line invalidatedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F3FFC0002ocr.itom.remoteocr.reads_to_core.remoteunc_m_rpq_occupancy_pch0TOR Inserts : CRds issued by iA Cores that Hit the LLC. Unit: uncore_cha event=0x36,umask=0xC807FF01TOR Occupancy : DRds issued by iA Cores. Unit: uncore_cha TOR Inserts : RFOs issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha unc_cha_tor_occupancy.io_pcirdcurunc_iio_data_req_by_cpu.mem_read.part5event=0xc2,ch_mask=0x80,fc_mask=0x04,umask=0x03PCIe Completion Buffer Inserts of completions with data: Part 7. Unit: uncore_iio event=0xd5,fc_mask=0x04,umask=0x10unc_i_misc1.lost_fwdMulti-socket cacheline Directory Lookups : Found in A state. Unit: uncore_m2m event=0x2e,umask=0x01Counts the number of cacheable memory requests that miss in the LLC. Counts on a per core basismem_bound_stalls.ifetch_dram_hitmem_bound_stalls.load_dram_hitCounts the number of cycles the core is stalled due to a demand load miss which hit in DRAM or MMIO (Non-DRAM)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0477Counts the total number of BACLEARS, which occur when the Branch Target Buffer (BTB) prediction or lack thereof, was corrected by a later branch predictor in the frontend.  Includes BACLEARS due to all branch types including conditional and unconditional jumps, returns, and indirect branchesThis event is deprecated. Refer to new event BUS_LOCK.BLOCK_CYCLESCounts the number of retired JCC (Jump on Conditional Code) branch instructions retired, includes both taken and not taken branches (Precise event)This event is deprecatedevent=0x71,period=1000003,umask=0x1Write Pending Queue Allocations : Counts the number of allocations into the Write Pending Queue.  This can then be used to calculate the average queuing latency (in conjunction with the WPQ occupancy count).  The WPQ is used to schedule write out to the memory controller and to track the writes.  Requests allocate into the WPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the CHA to the iMC.  They deallocate after being issued to DRAM.  Write requests themselves are able to complete (from the perspective of the rest of the system) as soon they have posted to the iMCunc_cha_tor_inserts.ia_drd_optTOR Inserts; Data read from local IA that misses in the snoop filterTOR Inserts : RFO_Prefs issued by iA Cores that Missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0x36,umask=0xC8A7FF01TOR Occupancy : CRds issued by iA Cores that Missed the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsTOR Occupancy : DRd_Opt issued by iA Cores that missed the LLC. Unit: uncore_cha PCIe Completion Buffer Inserts of completions with data : Part 1 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 5event=0x85,period=2000003,umask=0x80Counts the number of page walks completed due to instruction fetches whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 2M or 4M pages.  Includes page walks that page faultCounts the number of cycles a core is stalled due to a demand load which hit in the Last Level Cache (LLC) or other core with HITE/F/MFraction of Uops delivered by the LSD (Loop Stream Detector; aka Loop Cache). Unit: cpu_core Fraction of branches of other types (not individually covered by other metrics in Info.Branches group). Unit: cpu_core L1 cache true misses per kilo instruction for all demand loads (including speculative). Unit: cpu_core Average CPU Utilization. Unit: cpu_core TOPDOWN_FE_BOUND.ALL / (5 * CPU_CLK_UNHALTED.CORE)Cycles_per_Demand_Load_L2_HitCycles_per_Demand_Load_L3_Hitoffcore_requests_outstanding.data_rdevent=0x5,period=1000003,umask=0xffCounts the number of machine clears due to program modifying data (self modifying code) within 1K of a recently fetched code page. Unit: cpu_atom Counts the number of issue slots every cycle that were not consumed by the backend due to branch mispredicts. Unit: cpu_atom event=0xe7,period=1000003,umask=0x3The number of times that split load operations are temporarily blocked because all resources for handling the split accesses are in use. Unit: cpu_core Counts the number of demand load dispatches that hit L1D fill buffer (FB) allocated for software prefetch. Unit: cpu_core Cycles where at least 3 uops were executed per-thread. Unit: cpu_core Counts the number of cycles that the head (oldest load) of the load buffer is stalled due to a DTLB miss when load subsequently retires. Unit: cpu_atom event=0x13,period=100003,umask=0x20event=0x13,period=100003,umask=0x2Retired load instructions which data sources missed L3 but serviced from local dram  Supports address when preciseevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x808000001Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that resulted in a snoop that hit in another core, which did not forward the dataevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x708004477Uops executed on ports 7 and 8event=0x3,umask=0x0000000080event=0x3,umask=0x0000000044event=0x50,umask=0x0000000030event=0x35,umask=0x00C8977E01Valid Flits Received : Slot 0. Unit: uncore_upi Tx Flit Buffer Occupancy. Unit: uncore_upi event=0x11,umask=0x0000000008event=0x84,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000004event=0xc2,ch_mask=0x20,fc_mask=0x07,umask=0x0000000004event=0xc2,ch_mask=0x80,fc_mask=0x07,umask=0x0000000004event=0x21,umask=0x0301TOR Inserts; CRd hits from local IA. Unit: uncore_cha event=0x35,umask=0x00c88ffd01event=0x35,umask=0x00c827fd01TOR Inserts; RFO pref misses from local IA. Unit: uncore_cha event=0x35,umask=0x00c803ff04unc_cha_tor_inserts.ia_drdevent=0x36,umask=0x00C001FF08event=0x36,umask=0x00c887fd01All LLC lines in S state that are victimized on a fill. Unit: uncore_cha event=0x35,umask=0x00cd43fe04event=0x35,umask=0x00c837fe01unc_cha_tor_inserts.ia_drdpteTOR Occupancy; LLCPrefData hits from local IA. Unit: uncore_cha TOR Inserts : DRd_Prefs issued by iA Cores targeting DDR Mem that Missed the LLC. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_crd_remoteevent=0x35,umask=0x00C88F7E01TOR Inserts : CRd_Prefs issued by iA Cores that Missed the LLC - HOMed remotely. Unit: uncore_cha event=0x35,umask=0x00C86F8A01TOR Inserts : CLFlushes issued by IO Devices. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_crd_pref_localunc_cha_tor_occupancy.ia_wbmtoievent=0x87,umask=0x02All L2 Cache Requests (Breakdown 1 - Common). L2 Prefetcher. All prefetches accepted by L2 pipeline, hit or miss. Types of PF and L2 hit/miss broken out in a separate perfmon eventl2_latency.l2_cycles_waiting_on_fillsevent=0x63,umask=0x20l2_cache_req_stat.ic_dc_miss_in_l2event=0,umask=0x40fpu_pipe_assignment.total1This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Single precision multiply-add FLOPS. Multiply-add counts as 2 FLOPSThis is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemes. Number of SSE Move Opsls_dispatch.store_dispatchls_tablewalker.dc_type1L2 DTLB Misses & Data page walks6.1e-5MiBTotal number of fp uOpsls_hw_pf_dc_fill.ls_mabresp_lcl_cacheCycles where a dispatch group is valid but does not get dispatched due to a token stall. Load queue resource stall. Applies to all ops with load semanticsCore to L2 cacheable request access status (not including L2 Prefetch). Data cache request miss in L2 (all types). Use l2_cache_misses_from_dc_misses insteadL1 ITLB Miss, L2 ITLB Hit. The number of instruction fetches that miss in the L1 ITLB but hit in the L2 ITLBls_mab_alloc.load_store_allocationsevent=0x44,umask=0x40ls_any_fills_from_sys.ext_cache_localAll L1 DTLB Misses or Reloads. Use l1_dtlb_misses insteadls_sw_pf_dc_fills.mem_io_localevent=0x5a,umask=0x04de_dis_cops_from_decoder.disp_op_type.any_integer_dispatchde_dis_dispatch_token_stalls1.fp_flush_recovery_stallL2 Cache Misses from L2 Cache HWPFRetired branch instructions mispredicted due to direction mismatchls_dmnd_fills_from_sys.local_l2Demand data cache fills from cache of another CCX when the address was in the same NUMA nodels_any_fills_from_sys.dram_io_nearAny data cache fills from either cache of another CCX, DRAM or MMIO when the address was in a different NUMA node (same or different socket)Software prefetch instructions dispatched (speculative) of type PrefetchNTA (move data with minimum cache pollution i.e. non-temporal access)ls_hw_pf_dc_fills.local_ccxl2_pf_hit_l2.allL2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L1Region (fetch additional lines into L1 cache when the data access for a given instruction tends to be followed by a consistent pattern of other accesses within a localized region)L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L2Stream (fetch additional sequential lines into L2 cache)event=0x72,umask=0x02Instruction cache lines (64 bytes) fulfilled from the L2 cachel3_xi_sampled_latency.dram_nearevent=0xac,umask=0x20local_processor_read_data_beats_cs0Read data beats (64 bytes) for local processor at Coherent Station (CS) 0event=0x5f,umask=0x7feevent=0x11f,umask=0x7feevent=0x21f,umask=0x7feRead data beats (64 bytes) for local processor at Coherent Station (CS) 10Write data beats (64 bytes) for local processor at Coherent Station (CS) 7local_processor_write_data_beats_cs9Write data beats (64 bytes) for remote processor at Coherent Station (CS) 4remote_socket_upstream_write_beats_iom1event=0x49e,umask=0x7feevent=0x49f,umask=0x7ffevent=0x51f,umask=0xbfeData beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 5Data beats (64 bytes) for local socket outbound data from inter-socket xGMI link 0local_socket_outbound_data_beats_link1fp_ret_x87_fp_ops.div_sqrt_opsRetired packed 256-bit floating-point opsfp_ops_retired_by_type.scalar_blendevent=0xa,umask=0x30Retired vector floating-point ops of other typesRetired MMX integer shuffle ops (may include instructions not necessarily thought of as including shuffles e.g. horizontal add, dot product, and certain MOV instructions)sse_avx_ops_retired.sse_avx_mulevent=0xb,umask=0x70fp_pack_ops_retired.fp128_otherRetired 256-bit packed floating-point multiply opsevent=0xd,umask=0xd0L1 DTLB misses for all page sizesNumber of cycles dispatch is stalled for integer scheduler queue 0 tokensFraction of dispatch slots that remained unused because the frontend did not supply enough instructions/opsl1_data_cache_fills_from_same_ccxlocal_socket_upstream_read_beats_iom0 + local_socket_upstream_read_beats_iom1 + local_socket_upstream_read_beats_iom2 + local_socket_upstream_read_beats_iom3local_socket_upstream_dma_read_dataumc_mem_read_bandwidthevent-hyphenUNC_CBO_TWO_HYPH. Unit: uncore_cbox BU_CPU_CLK_UNHALTEDEXC_TAKENL1_DCACHE_WBL1_DCACHE_NEON_ACCESSL2_CACHE_NEON_HITPMUEXTIN1_EVTEVENT_4FHEVENT_5CHEVENT_83HEVENT_BEHEVENT_E7HEVENT_10DHEVENT_113HEVENT_152HEVENT_188HEVENT_195HEVENT_1C2HEVENT_1D3HEVENT_20AHEVENT_21BHEVENT_24AHEVENT_254HEVENT_25CHEVENT_260HEVENT_262HEVENT_28BHEVENT_29FHEVENT_2A8HEVENT_2AEHEVENT_2B5HEVENT_308HEVENT_339HEVENT_347HEVENT_370HEVENT_3A1HEVENT_3CBHEVENT_3D8HEVENT_3E0HEXC_TRAP_PABORTdn_rxreq_snp_senthnf_slc_sf_cache_accesshni_rdt_wr_allocsbsx_txrsp_flitvrnd_wrt_occ_ovflcxha_sdb_occcxra_wr_dat_buf_occL3_EXTERNAL_INTERVENTIONSCACHEOPS_TRANSLATEDWRITE_THROUGH_STORES_TRANSLATEDSTORES_COMPLETED_ALLOCATED_TO_DLFBINSTR_L1_CACHE_LOCKSDLFB_RETRIES_TO_MBARL2_CACHE_INSTR_HITSDVT5_DETECTEDDVT6_DETECTEDDECORATED_LOADSownerpage-hitnonpostwrszdwordE500FREEmetric_expr: %s
config1basic_string%s, "pmcid": "0x%08x", "pid": "%d", "pathname": "%s"}
"0x%016jx"]}
GenuineIntel-6-5CInstructions per FP Arithmetic AVX/SSE 128-bit instruction (lower number means higher occurrence rate)1000 * L2_RQSTS.DEMAND_DATA_RD_MISS / INST_RETIRED.ANYSummary;Power( ( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE ) / 1000000000 ) / duration_timeKernel_UtilizationAverage external Memory Bandwidth Use for reads and writes [GB / sec]( 64 * ( uncore_imc@cas_count_read@ + uncore_imc@cas_count_write@ ) / 1000000000 ) / duration_time1000000000 * ( cbox@event\=0x36\,umask\=0x3\,filter_opc\=0x182@ / cbox@event\=0x35\,umask\=0x3\,filter_opc\=0x182@ ) / ( cbox_0@event\=0x0@ / duration_time )(cstate_core@c3\-residency@ / msr@tsc@) * 100event=0x24,period=200003,umask=0xe7Retired load uops which data sources were L3 hit and cross-core snoop missed in on-pkg core cache. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM100 (Precise event)mem_load_uops_retired.hit_lfbmem_load_uops_retired.l2_missmem_uops_retired.split_loadsoffcore_responseThis event counts the number of cycles 4  uops were  delivered to Instruction Decode Queue (IDQ) from the MITE path. Counting includes uops that may bypass the IDQ. This also means that uops are not being delivered from the Decode Stream Buffer (DSB)This event counts the number of cycles  uops were delivered to Instruction Decode Queue (IDQ) from the MITE path. Counting includes uops that may bypass the IDQ. This also means that uops are not being delivered from the Decode Stream Buffer (DSB)This event counts the total number of uops delivered to Instruction Decode Queue (IDQ) while the Microcode Sequenser (MS) is busy. Counting includes uops that may bypass the IDQ. Uops maybe initiated by Decode Stream Buffer (DSB) or MITEevent=0xc8,period=2000003,umask=0x2Number of times we entered an HLE region; does not count nested transactionsmem_trans_retired.load_latency_gt_64This event counts speculative cache-line split load uops dispatched to the L1 cacheSpeculative cache line split STA uops dispatched to L1 cachertm_retired.startpipelineThis event counts the number of the divide operations executed. Uses edge-detect and a cmask value of 1 on ARITH.FPU_DIV_ACTIVE to get the number of the divide operations executedbr_inst_exec.taken_direct_near_callbr_misp_exec.taken_indirect_jump_non_call_retevent=0x89,period=200003,umask=0x84Taken speculative and retired mispredicted indirect callsNumber of near branch instructions retired that were mispredicted and taken. (Precise Event - PEBS) (Precise event)event=0,period=2000003,umask=0x2cycle_activity.cycles_ldm_pendingevent=0xd,cmask=1,period=2000003,umask=0x3This event counts all not software-prefetch load dispatches that hit the fill buffer (FB) allocated for the software prefetch. It can also be incremented by some lock instructions. So it should only be used with profiling so that the locks can be excluded by asm inspection of the nearby instructionsThis event counts the number of executed Intel AVX masked load operations that refer to an illegal address range with the mask bits set to 0resource_stalls.rsevent=0xa2,period=2000003,umask=0x4event=0xb1,cmask=3,period=2000003,umask=0x2Number of flags-merge uops being allocated. Such uops considered perf sensitive; added by GSR u-archUncore cache clock ticks. Unit: uncore_cbox event=0x34,umask=0x11,filter_state=0x164Bytesunc_c_llc_victims.m_stateLLC misses - demand and prefetch data reads - excludes LLC prefetches. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox llc_misses.rfo_llc_prefetchevent=0x35,umask=0x3,filter_opc=0x192read requests to remote home agent. Unit: uncore_ha event=0x1,umask=0x4unc_m_power_self_refreshunc_p_power_state_occupancy.cores_c0unc_p_freq_max_power_cyclesdtlb_load_misses.miss_causes_a_walkThis event counts the number of flushes of the big or small ITLB pages. Counting include both TLB Flush (covering all sets) and TLB Set Clear (set-specific)event=0x85,period=100003,umask=0x40itlb_misses.stlb_hit_4ktlb_flush.dtlb_threadevent=0xbd,period=100007,umask=0x1Retired store uops that miss the STLB  Supports address when precise (Precise event)Counts all demand & prefetch data readsCounts all prefetch code readsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0090offcore_response.all_pf_rfo.supplier_none.snoop_hitmCounts all demand & prefetch RFOsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0122offcore_response.all_rfo.supplier_none.any_snoopoffcore_response.all_rfo.supplier_none.snoop_missCounts writebacks (modified to exclusive)Counts all demand code readsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020004offcore_response.demand_data_rd.supplier_none.snoop_noneoffcore_response.demand_data_rd.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0002Counts any other requestsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0040offcore_response.pf_l2_code_rd.supplier_none.snoop_not_neededoffcore_response.pf_l2_data_rd.l3_hit.snoop_hit_no_fwdoffcore_response.pf_l2_rfo.supplier_none.snoop_hit_no_fwdoffcore_response.pf_l3_code_rd.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0080020100Randomly selected loads with latency value being above 128  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000091offcore_response.all_pf_code_rd.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000240event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020010offcore_response.pf_l2_rfo.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000200offcore_response.pf_l3_rfo.l3_hit.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000100Number of times RTM abort was triggered  (Precise event)This event counts conditional branch instructions retired (Precise event)This event counts the number of mispredicted ret instructions retired. Non PEBS (Precise event)This event counts resource-related stall cyclesA cross-core snoop initiated by this Cbox due to processor core memory request which hits a non-modified line in some processor core. Unit: uncore_cbox event=0x22,umask=0x48L3 Lookup any request that access cache and found line in I-state. Unit: uncore_cbox This 48-bit fixed counter counts the UCLK cyclesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C8FFFCounts all demand & prefetch RFOs miss in the L3Number of data flits transmitted . Derived from unc_q_txl_flits_g0.data. Unit: uncore_qpi event=0x40,period=200000,umask=0x8event=0x27,period=200000,umask=0x70event=0x30,period=200000,umask=0x41l2_reject_busq.self.demand.mesievent=0x2e,period=200000,umask=0x71event=0x2e,period=200000,umask=0x52SIMD Instructions retiredSIMD packed logical micro-ops retiredx87_comp_ops_exe.fxch.arevent=0x7,period=200000,umask=0x8fL0 DTLB misses due to load operationspage_walks.walksRequires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts requests to the uncore subsystem that have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.any_request.l2_miss.hitm_other_coreoffcore_response.any_request.l2_miss.snoop_miss_or_no_snoop_neededCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that miss the L2 cacheCounts full cache line data writes to uncacheable write combining (USWC) memory region and full cache-line non-temporal writes that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000080Counts partial cache line data writes to uncacheable write combining (USWC) memory region  that true miss for the L2 cache with a snoop miss in the other processor moduleevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000010Counts data cacheline reads generated by hardware L2 cache prefetcher that miss the L2 cacheoffcore_response.pf_l2_data_rd.l2_miss.hitm_other_coreevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200000020cycles_div_busy.fpdivCounts core cycles the floating point divide unit is busyMachine clears due to FP assistsReferences per ICache line that are not available in the ICache (miss). This event counts differently than Intel processors based on Silvermont microarchitectureCounts branch instructions retired for all branch types.  This is an architectural performance event (Must be precise)Counts near relative CALL branch instructions retired (Must be precise)Retired mispredicted near return instructions (Precise event capable) (Must be precise)Retired mispredicted conditional branch instructions that were taken (Precise event capable) (Must be precise)ld_blocks.data_unknownCounts loads blocked because they are unable to find their physical address in the micro TLB (UTLB) (Must be precise)Counts machine clears for any reasonevent=0x9c,period=200003Counts bus lock and split lock requests hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts bus lock and split lock requests outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache line reads generated by hardware L1 data cache prefetcher miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010010offcore_response.pf_l2_rfo.outstandingevent=0xc3,period=20003,umask=0x4Counts page walks completed due to demand data loads (including SW prefetches) whose address translations missed in all TLB levels and were mapped to 2M or 4M pages.  The page walks can end with or without a page faultCounts page walks completed due to demand data stores whose address translations missed in the TLB and were mapped to 4K pages.  The page walks can end with or without a page faultevent=0x85,period=200003,umask=0x10Demand Data Read miss L2, no rejects  Spec update: HSD78, HSM80All requests to L2 cache  Spec update: HSD78, HSM80This event counts retired load uops where the data came from local DRAM. This does not include hardware prefetches  Supports address when precise.  Spec update: HSD74, HSD29, HSD25, HSM30 (Precise event)Counts all demand code reads hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedoffcore_response.pf_l2_rfo.l3_hit.any_responseThis event counts Instruction Cache (ICACHE) missesNumber of times an HLE execution startedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC08FFFCounts all prefetch (that bring data to LLC only) code reads miss in the L3Number of times an HLE transactional execution aborted due to XRELEASE lock not satisfying the address and value requirements in the elision bufferThis event counts the number of uops issued by the Front-end of the pipeline to the Back-end. This event is counted at the allocation stage and will count both retired and non-retired uopsevent=0x34,umask=0x46This event counts load operations from a 2M page that miss the first DTLB level but hit the second and do not cause page walksevent=0xd3,period=100003,umask=0x10Dirty L2 cache lines filling the L2event=0x24,period=200003,umask=0xcevent=0x24,period=200003,umask=0x80event=0x24,period=200003,umask=0x4Cycles when offcore outstanding cacheable Core Data Read transactions are present in SuperQueue (SQ), queue to uncoreCycles when offcore outstanding Demand Data Read transactions are present in SuperQueue (SQ), queue to uncoreCycles with at least 6 offcore outstanding Demand Data Read transactions in uncore queueCounts demand & prefetch code reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x000105B3event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00010122Counts all writebacks from the core to the LLCCounts all demand code reads that hit in the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0001Counts all demand rfo'sCounts number of SSE* or AVX-128 double precision FP scalar uops executed1000 * MEM_LOAD_UOPS_RETIRED.LLC_MISS / INST_RETIRED.ANYLoads with latency value being above 512 (Must be precise)Speculative cache-line split Store-address uops dispatched to L1DCounts all demand & prefetch code reads that miss the LLC  and the data returned from dramoffcore_response.all_reads.llc_miss.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3004003f7Cycles which a Uop is dispatched on port 0Increments each cycle the # of Uops issued by the RAT to RS. Set Cmask = 1, Inv = 1, Any= 1to count stalled cycles of this coreunc_cbo_xsnp_response.inval_mA snoop invalidates a modified line in some processor coreoffcore_response.all_reads.llc_hit.snoop_missCounts all data/code/rfo reads (demand & prefetch) that hit the LLCCounts all demand data writes (RFOs) that miss the LLC and the data is found in M state in remote cache and forwarded from therePCIe allocating writes that miss LLC - DDIO misses. Derived from unc_c_tor_inserts.miss_opcode.ddio_miss. Unit: uncore_cbox LLC misses for PCIe non-snoop reads. Derived from unc_c_tor_inserts.miss_opcode.pcie_read. Unit: uncore_cbox event=0x35,umask=0x1,filter_opc=0x1e6Write requests to memory controller. Derived from unc_m_cas_count.wr. Unit: uncore_imc Counts the number of cycles that the uncore was running at a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band1=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Unit: uncore_pcu freq_band2_cycles %unc_p_freq_ge_3000mhz_cyclesevent=0x51,period=2000003,umask=0x4Retired load uops which data sources were HitM responses from shared LLCThis event counts retired load uops that hit in the last-level cache (L3) and were found in a non-modified state in a neighboring core's private cache (same package).  Since the last level cache is inclusive, hits to the L3 may require snooping the private L2 caches of any cores on the same socket that have the line.  In this case, a snoop was required, and another L2 had the line in a modified state, so the line had to be invalidated in that L2 cache and transferred to the requesting L2UOPS_DISPATCHED.THREAD / UOPS_ISSUED.ANYUOPS_DISPATCHED.THREAD / (( cpu@UOPS_DISPATCHED.CORE\,cmask\=1@ / 2 ) if #SMT_on else cpu@UOPS_DISPATCHED.CORE\,cmask\=1@)Counts all local dram accesses for all demand and L2 prefetches. LLC prefetches are excludedQPI clock ticks. Used to get percentages of QPI cycles events. Unit: uncore_qpi Counts the number of times that the uncore transitioned to a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band1=XXX with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Derived from unc_p_freq_band1_cycles. Unit: uncore_pcu offcore_response.any_data_rd.l2_hit_near_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010003091event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000070event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004008000offcore_response.any_request.l2_hit_this_tile_foffcore_response.any_rfo.l2_hit_far_tileevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800080022event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080022Counts Bus locks and split lock requests that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010002000offcore_response.pf_l2_code_rd.l2_hit_near_tileevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000040event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800401000offcore_response.uc_code_reads.any_responseoffcore_response.uc_code_reads.l2_hit_near_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000044Counts Demand code reads and prefetch code read requests  that accounts for data responses from DRAM FarCounts Demand cacheable data and L1 prefetch data read requests  that accounts for data responses from MCDRAM Far or Other tile L2 hit faroffcore_response.any_read.ddr_farCounts any Read request  that accounts for data responses from DRAM Faroffcore_response.demand_data_rd.ddr_nearevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200001offcore_response.demand_rfo.mcdram_faroffcore_response.partial_reads.ddr_faroffcore_response.partial_writes.mcdram_farCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for data responses from MCDRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080802000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x2000020020Counts Software Prefetches that accounts for data responses from DRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400200no_alloc_cycles.not_deliveredCounts the total number of core cycles for all the I-side page walks. The cycles for page walks started in speculative path will also be includedevent=0x40,period=2000000,umask=0x8l1d_wb_l2.s_statel2_data_rqsts.prefetch.s_statel2_rqsts.ifetch_hitevent=0x24,period=200000,umask=0x1Memory instructions retired above 32 clocks (Precise Event)mem_uncore_retired.other_core_l2_hitmoffcore_response.any_data.any_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4744event=0xb7,period=100000,umask=0x1,offcore_rsp=0x1822offcore_response.data_in.llc_hit_no_other_coreOffcore request = all data, response = local cacheOffcore request = all data, response = local cache or dramoffcore_response.demand_data.llc_hit_no_other_coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x201Offcore demand data reads satisfied by the LLCOffcore demand data reads satisfied by a remote cache or remote DRAMOffcore demand code reads satisfied by a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4702Offcore demand RFO requests that HITM in a remote cacheOffcore other requests satisfied by the LLC  and HITM in a sibling coreoffcore_response.other.remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x830event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F10Offcore prefetch code reads satisfied by any cache or DRAMoffcore_response.pf_ifetch.remote_cache_hitevent=0xfd,period=200000,umask=0x20event=0xfd,period=200000,umask=0x2offcore_response.any_data.remote_dramoffcore_response.any_ifetch.remote_dramOffcore requests that missed the LLCoffcore_response.any_rfo.any_dramOffcore code or data read requests that missed the LLCOffcore request = all data, response = any DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2001offcore_response.other.any_llc_missevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF830event=0xb7,period=100000,umask=0x1,offcore_rsp=0xF870event=0xd5,period=2000000,umask=0x1l1i.cycles_stalledevent=0xb8,period=100000,umask=0x1snoop_response.hitebpu_missed_call_retTaken branches executedevent=0x87,period=2000000,umask=0xfssex_uops_retired.scalar_doubleStack pointer sync operationsevent=0xd1,cmask=1,period=2000000,umask=0x2uops_executed.port4_coreDTLB load miss caused by low part of addressevent=0x8,period=2000000,umask=0x10DTLB miss page walksl2_lines_out.non_silentoffcore_response.demand_code_rd.l3_hit_e.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40400004offcore_response.demand_data_rd.l3_hit_m.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020001offcore_response.demand_rfo.l3_hit.spl_hitoffcore_response.demand_rfo.l3_hit_e.snoop_hit_no_fwdoffcore_response.demand_rfo.l3_hit_s.snoop_noneoffcore_response.demand_rfo.l3_hit_s.spl_hitoffcore_response.other.l3_hit_e.snoop_hitmoffcore_response.other.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80028000event=0x32,period=2000003,umask=0x4Counts once for most SIMD 128-bit packed computational double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 2 computation operations, one for each element.  Applies to packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsCounts the number of times the front-end is resteered when it finds a branch instruction in a fetch line. This occurs for the first time a branch instruction is fetched or when the branch is not tracked by the BPU (Branch Prediction Unit) anymorefrontend_retired.l2_missfrontend_retired.latency_ge_1event=0xc6,period=100007,umask=0x1,frontend=0x400206event=0xc6,period=100007,umask=0x1,frontend=0x100206event=0xc6,period=100007,umask=0x1,frontend=0x300206Counts the number of uops initiated by MITE and delivered to Instruction Decode Queue (IDQ) while the Microcode Sequencer (MS) is busy. Counting includes uops that may 'bypass' the IDQoffcore_response.demand_data_rd.l3_hit_m.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC4008000rtm_retired.aborted_eventsbr_inst_retired.cond_ntakenevent=0x55,period=2000003,umask=0x1event=0xb1,cmask=1,inv=1,period=2000003,umask=0x2Counts the retirement slots usedMispredictions_SMTTotal pipeline cost of Memory Latency related bottlenecks (external memory and off-core caches)100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) * ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , OFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DATA_RD ) / CPU_CLK_UNHALTED.THREAD - (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD)) / #(CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) * ( (( (10 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) - (3.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) ) * MEM_LOAD_RETIRED.L3_HIT * (1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) / 2) / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) + ( (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD)) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * CPU_CLK_UNHALTED.THREAD)) - ( UOPS_ISSUED.ANY + 4 * INT_MISC.RECOVERY_CYCLES ) / (4 * CPU_CLK_UNHALTED.THREAD))) ) )( CPU_CLK_UNHALTED.THREAD_ANY / 2 ) if #SMT_on else CPU_CLK_UNHALTED.THREADUOPS_ISSUED.ANY / cpu@UOPS_ISSUED.ANY\,cmask\=1@BR_INST_RETIRED.NOT_TAKEN / BR_INST_RETIRED.ALL_BRANCHESoffcore_response.demand_data_rd.l2_miss.snoop_missoffcore_response.demand_rfo.l2_miss.snoop_missCounts the number of JCC baclearsCounts the number of times a decode restriction reduced the decode throughput due to wrong instruction length predictionCounts the number of times the MSROM starts a flow of UOPS. It does not count every time a UOP is read from the microcode ROM.  The most common case that this counts is when a micro-coded instruction is encountered by the front end of the machine.  Other cases include when an instruction encounters a fault, trap, or microcode assist of any sort.  The event will count MSROM startups for UOPS that are speculative, and subsequently cleared by branch mispredict or machine clear.  Background: UOPS are produced by two mechanisms.  Either they are generated by hardware that decodes instructions into UOPS, or they are delivered by a ROM (called the MSROM) that holds UOPS associated with a specific instruction.  MSROM UOPS might also be delivered in response to some condition such as a fault or other exceptional condition.  This event is an excellent mechanism for detecting instructions that require the use of MSROM instructionsThis event counts when a data (D) page walk is completed or started.  Since a page walk implies a TLB miss, the number of TLB misses can be counted by counting the number of pagewalksCounts all data/code/rfo references (demand & prefetch) COREWB & ANY_RESPONSEoffcore_requests_outstanding.any.readoffcore_requests_outstanding.demand.read_data_not_emptyREQUEST = ANY_DATA read and RESPONSE = LOCAL_CACHEREQUEST = ANY RFO and RESPONSE = ANY_LOCATIONREQUEST = ANY RFO and RESPONSE = IO_CSR_MMIOREQUEST = DEMAND_DATA_RD and RESPONSE = LOCAL_CACHEoffcore_response.demand_ifetch.all_local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5004REQUEST = PF_DATA and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = PF_DATA_RD and RESPONSE = IO_CSR_MMIOevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3011REQUEST = ANY_DATA read and RESPONSE = ANY_DRAM AND REMOTE_FWDevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3044offcore_response.other.any_dram_and_remote_fwdevent=0xc5,period=20000,umask=0x4DTLB misses casued by low part of addressevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F50DTLB misses caused by low part of address. Count also includes 2M page references because 2M pages do not use the PDEOFFCORE_RESPONSE.ALL_PF_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDCounts demand data reads that hit in the L3Counts all demand data writes (RFOs) that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.pf_l2_rfo.l3_hit.snoop_hit_with_fwdoffcore_response.pf_l3_data_rd.l3_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00122Counts all demand & prefetch RFOs that miss the L3 and the data is returned from local dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00004Counts demand data reads that miss the L3 and the modified data is transferred from remote cacheoffcore_response.pf_l2_rfo.l3_miss.snoop_miss_or_no_fwdoffcore_response.pf_l3_data_rd.l3_miss.remote_hitmCounts all prefetch (that bring data to LLC only) RFOs that miss the L3 and clean or shared data is transferred from remote cacheCORE_POWER.LVL1_TURBO_LICENSE / 2 / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )Counts all CAS (Column Address Select) commands issued to DRAM per memory channel.  CAS commands are issued to specify the address to read or write on DRAM, so this event increments for every read and write. This event counts whether AutoPrecharge (which closes the DRAM Page automatically after a read/write) is enabled or notuncore_upievent=0x33,umask=0x42Counts number of allocations per cycle into the specified Ingress queueevent=0x5c,umask=0x20event=0x5c,umask=0x10unc_iio_comp_buf_occupancy.cmpd.all_partsevent=0xd5,fc_mask=0x04,umask=0x04unc_iio_data_req_by_cpu.mem_write.part1unc_iio_txn_req_by_cpu.peer_read.part3Counts every peer to peer read request for up to a 64 byte transaction of data made by a different IIO unit to the MMIO space of a card on IIO Part3. Does not include requests made by the same IIO unit. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busevent=0x84,ch_mask=0x08,fc_mask=0x07,umask=0x01Counts every peer to peer write request of up to a 64 byte transaction of data made by IIO Part2 to the MMIO space of an IIO target. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busunc_m2m_bypass_m2m_egress.not_takenevent=0x29Inserts into the Memory Controller Prefetch Queue. Unit: uncore_m2m unc_m2m_rxc_ad_insertsevent=0x12,umask=0x1ocr.all_data_rd.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040491ocr.all_data_rd.l3_hit_s.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040490OCR.ALL_PF_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_FWDOCR.ALL_PF_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400100490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200120ocr.all_pf_rfo.l3_hit_f.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C07F7ocr.all_reads.l3_hit.snoop_noneocr.all_reads.l3_hit_e.any_snoopocr.all_reads.l3_hit_s.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080122OCR.ALL_RFO.L3_HIT_F.ANY_SNOOP  OCR.ALL_RFO.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040122OCR.ALL_RFO.L3_HIT_S.ANY_SNOOP  OCR.ALL_RFO.L3_HIT_S.ANY_SNOOPocr.demand_code_rd.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040004Counts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_HIT.SNOOP_NONECounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_E.ANY_SNOOPCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDocr.demand_rfo.l3_hit_s.snoop_noneocr.other.l3_hit.hitm_other_coreocr.other.l3_hit.snoop_missocr.other.l3_hit_m.snoop_noneCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_HIT.HIT_OTHER_CORE_FWD OCR.PF_L1D_AND_SW.L3_HIT.HIT_OTHER_CORE_FWDocr.pf_l1d_and_sw.l3_hit_e.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080010ocr.pf_l2_data_rd.l3_hit_e.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200080ocr.pf_l3_data_rd.l3_hit_m.hit_other_core_fwdocr.pf_l3_data_rd.l3_hit_m.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100080Counts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_E.ANY_SNOOPocr.pf_l3_rfo.l3_hit_f.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80040100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800100100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.all_data_rd.l3_hit_m.snoop_missoffcore_response.all_data_rd.l3_hit_s.hit_other_core_fwdoffcore_response.all_data_rd.l3_hit_s.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020491This event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_M.SNOOP_MISSoffcore_response.all_pf_data_rd.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.SNOOP_MISSoffcore_response.all_pf_rfo.l3_hit_f.hit_other_core_fwdoffcore_response.all_pf_rfo.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_E.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000207F7This event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_E.SNOOP_NONEoffcore_response.all_rfo.l3_hit_m.hit_other_core_no_fwdoffcore_response.demand_code_rd.l3_hit_e.no_snoop_neededoffcore_response.demand_code_rd.l3_hit_m.no_snoop_neededoffcore_response.demand_data_rd.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_FWDoffcore_response.demand_data_rd.l3_hit_s.no_snoop_neededoffcore_response.demand_data_rd.pmm_hit_local_pmm.any_snoopoffcore_response.demand_rfo.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400002offcore_response.demand_rfo.pmm_hit_local_pmm.snoop_not_neededoffcore_response.demand_rfo.supplier_none.hitm_other_coreThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_E.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_S.SNOOP_NONEoffcore_response.pf_l2_data_rd.l3_hit_f.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_data_rd.l3_hit_e.snoop_missoffcore_response.pf_l3_data_rd.l3_hit_f.snoop_noneoffcore_response.pf_l3_data_rd.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_E.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDocr.all_data_rd.l3_miss_remote_hop1_dram.snoop_noneocr.all_pf_data_rd.l3_miss_local_dram.hit_other_core_no_fwdocr.all_pf_rfo.l3_miss.any_snoopocr.all_pf_rfo.l3_miss.no_snoop_neededocr.all_pf_rfo.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C0007F7OCR.ALL_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDCounts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.HIT_OTHER_CORE_FWD OCR.DEMAND_CODE_RD.L3_MISS.HIT_OTHER_CORE_FWDocr.demand_code_rd.l3_miss.no_snoop_neededocr.demand_data_rd.l3_miss_local_dram.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000002ocr.demand_rfo.l3_miss_local_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C008000ocr.other.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC08000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x410008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x110000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000010Counts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000080Counts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.PF_L3_RFO.L3_MISS.HIT_OTHER_CORE_NO_FWDCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.ANY_SNOOPoffcore_response.all_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.HITM_OTHER_COREoffcore_response.demand_code_rd.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.other.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.pf_l2_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.pf_l2_rfo.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.REMOTE_HIT_FORWARDoffcore_response.pf_l2_rfo.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.pf_l3_data_rd.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDocr.all_pf_rfo.pmm_hit_local_pmm.any_snoopOCR.ALL_RFO.ANY_RESPONSE have any response typeOCR.ALL_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONE OCR.ALL_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.all_rfo.supplier_none.hitm_other_coreCounts demand data reads OCR.DEMAND_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.demand_data_rd.supplier_none.any_snoopCounts demand data reads  OCR.DEMAND_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.demand_rfo.supplier_none.no_snoop_neededocr.pf_l2_data_rd.pmm_hit_local_pmm.any_snoopocr.pf_l2_data_rd.supplier_none.snoop_missWrite commands for Intel Optane DC persistent memory. Unit: uncore_imc Counts when a transaction with the opcode type Rsp*Fwd*WB Snoop Response was received which indicates the data was written back to its home socket, and the cacheline was forwarded to the requestor socket.  This snoop response is only used in >= 4 socket systems.  It is used when a snoop HITM's in a remote caching agent and it directly forwards data to a requestor, and simultaneously returns data to its home socket to be written back to memoryevent=0x48,period=1000003,umask=0x4Number of L1D misses that are outstandingNon-modified cache lines that are silently dropped by L2 cache when triggered by an L2 cache fillevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C0001Counts demand data reads that hit a cacheline in the L3 where a snoop was sent but no other cores had the dataocr.hwpf_l2_rfo.l3_hit.snoop_hitmNumber of SSE/AVX computational 256-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0x79,cmask=5,period=2000003,umask=0x4Counts the number of uops delivered to Instruction Decode Queue (IDQ) from the MITE path. This also means that uops are not being delivered from the Decode Stream Buffer (DSB)( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * ( FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE ) + 16 * FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE ) / CPU_CLK_UNHALTED.DISTRIBUTED(cstate_pkg@c10\-residency@ / msr@tsc@) * 100Counts demand data reads that was not supplied by the L3 cacheCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that was not supplied by the L3 cacheevent=0xc5,period=50021,umask=0x10Counts the number of conditional branch instructions retired that were mispredicted and the branch direction was not taken (Precise event)Counts cycles where the Store Buffer was full and no loads caused an execution stallCounts cycles during which the reservation station (RS) is empty for this logical processor. This is usually caused when the front-end pipeline runs into stravation periods (e.g. branch mispredictions or i-cache misses)Uops exclusively fetched by decoder 0event=0xc2,cmask=1,inv=1,period=1000003,umask=0x2Number of page walks outstanding for a store in the PMH each cycleCounts the number of off-core outstanding Demand Data Read transactions every cycle. A transaction is considered to be in the Off-core outstanding state between L2 cache miss and data-return to the coreCounts retired load instructions with local Intel Optane DC persistent memory as the data source and the data request missed L3 (AppDirect or Memory Mode) and DRAM cache(Memory Mode)  Supports address when precise (Precise event)Counts demand instruction fetches and L1 instruction cache prefetches that hit in the L3 or were snooped from another core's caches on the same socketevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x808000002For every cycle, increments by the number of outstanding code read requests pendingUNC_CHA_TOR_OCCUPANCY.IA_MISS_DRD / cha@event\=0x36\,umask\=0xC817FE01\,thresh\=1@Counts responses to snoops indicating the line will now be (I)nvalidated: removed from this core's caches, after the data is forwarded back to the requestor, and indicating the data was found modified(M) in this cores caches cache (aka HitM response).  A single snoop response from the core counts on all hyperthreads of the coreCounts demand instruction fetches and L1 instruction cache prefetches that were supplied by DRAM on a distant memory controller of this socket when the system is in SNC (sub-NUMA cluster) modeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x703C00002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x73C000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000477Counts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by DRAM attached to another socketNumber of DRAM Refreshes Issued. Unit: uncore_imc Local read requests that miss the SF/LLC and are sent to the CHA's home agent. Unit: uncore_cha unc_cha_tor_occupancy.iaevent=0x36,umask=0xC807FE01unc_cha_tor_occupancy.io_missevent=0x35,umask=0xC8178601Cache and Snoop Filter Lookups; Data Read Request. Unit: uncore_cha Data requested of the CPU : CmpD - device sending completion to CPU request. Unit: uncore_iio Number Transactions requested by the CPU : Core writing to Card's MMIO space. Unit: uncore_iio event=0x83,ch_mask=0x20,fc_mask=0x07,umask=0x01unc_iio_data_req_of_cpu.mem_read.part6Multi-socket cacheline Directory Lookups : Found in S state. Unit: uncore_m2m event=0x2c,umask=0x01unc_cha_tor_inserts.io_itomcachenear_localMulti-socket cacheline Directory Updates : From/to any state. Note: event counts are incorrect in 2LM mode. Unit: uncore_m2m Multi-socket cacheline Directory Updates : From/to any state. Note: event counts are incorrect in 2LM modePage walks completed due to a demand data load to a 1G pagel2_reject_xq.anyCounts the number of L2 Cache accesses that resulted in a hit. Counts on a per core basisevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0044Counts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedCounts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedocr.hwpf_l2_rfo.l3_hit.snoop_hit_with_fwdCounts the number of instruction cache hitsocr.corewb_m.l3_miss_localocr.demand_data_and_l1pf_rd.l3_miss_localCounts modified writebacks from L1 cache that miss the L2 cache that were not supplied by the L3 cacheevent=0x3,period=1000003,umask=0x1topdown_fe_bound.otherDRAM Precharge commands. : Precharge due to page table : Counts the number of DRAM Precharge commands sent on this channel. : Prechages from Page TableTOR Inserts : DRd_Opt_Prefs issued by iA Cores : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : ItoMs issued by IO Devices that Hit the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Occupancy : All requests from IO Devices that hit the LLC : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsNumber Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Number Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x8 card plugged in to Lane 6/7, Or x4 card is plugged in to slot 6Counts the number of first level TLB misses but second level hits due to stores that did not start a page walk. Account for all pages sizes. Will result in a DTLB write from STLBCounts the number of page walks completed due to store DTLB misses to a 2M or 4M pageept.epdpe_missevent=0x85,period=2000003,umask=0x20Instruction per taken branch. Unit: cpu_core TOPDOWN_RETIRING.ALL / (5 * CPU_CLK_UNHALTED.CORE)Percentage of total non-speculative loads that are splits. Unit: cpu_atom Number of Instructions per non-speculative Branch Misprediction. Unit: cpu_atom X87_Uop_RatioCounts the number of tagged loads with an instruction latency that exceeds or equals the threshold of 256 cycles as defined in MEC_CR_PEBS_LD_LAT_THRESHOLD (3F6H). Only counts with PEBS enabled  Supports address when precise (Must be precise). Unit: cpu_atom All L2 requests.[This event is alias to L2_RQSTS.REFERENCES]. Unit: cpu_core Retired load instructions whose data sources were L3 and cross-core snoop hits in on-pkg core cache  Supports address when precise (Precise event). Unit: cpu_core Retired load instructions whose data sources were L3 hit and cross-core snoop missed in on-pkg core cache  Supports address when precise (Precise event). Unit: cpu_core event=0x20,cmask=1,period=1000003,umask=0x4Counts number of SSE/AVX computational 256-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. Unit: cpu_core frontend_retired.unknown_branchCycles where a code fetch is stalled due to L1 instruction cache tag miss. Unit: cpu_core event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84400002event=0x47,cmask=3,period=1000003,umask=0x3memory_activity.stalls_l2_missCounts demand read for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that have any type of response. Unit: cpu_core Counts the number of unhalted core clock cycles. (Fixed event). Unit: cpu_atom Counts the total number of issue slots that were not consumed by the backend because allocation is stalled due to a machine clear (nuke) of any kind including memory ordering and memory disambiguation. Unit: cpu_atom Counts the number of issue slots every cycle that were not consumed by the backend due to the physical register file unable to accept an entry (marble stalls). Unit: cpu_atom Counts the number of integer divide uops retired (Precise event). Unit: cpu_atom Counts the number of x87 uops retired, includes those in MS flows (Precise event). Unit: cpu_atom Thread cycles when thread is not in halt state. Unit: cpu_core event=0xe7,period=1000003,umask=0x40Cycles at least 2 micro-op is executed from any thread on physical core. Unit: cpu_core This event is deprecated. Refer to new event UOPS_EXECUTED.STALLS. Unit: cpu_core Incoming VC0 write request. Unit: uncore_imc unc_m_cas_count_rdevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x730004477event=0xc5,period=100003,umask=0x80Counts core clocks when the thread is in the C0.1 or C0.2 power saving optimized states (TPAUSE or UMWAIT instructions) or running the PAUSE instructionevent=0x20,umask=0x0000000001event=0x2,umask=0x000000000fevent=0x83,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000001TOR Inserts for DRd Pref misses from local IA. Unit: uncore_cha event=0x4,umask=0x000000010eevent=0x5,umask=0x000000010eevent=0x84,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000004event=0xc1,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000001event=0x17,umask=0x07event=0x2bunc_cha_llc_lookup.data_rdunc_cha_tor_inserts.allTOR Inserts; RFO hits from local IA. Unit: uncore_cha unc_cha_tor_inserts.ipqevent=0x35,umask=0x0000000010event=0x35,umask=0x00c8a7fd01event=0x35,umask=0x00c803fd04event=0x35,umask=0x00ccc7ff01TOR Inserts; DRd Opt Pref from local IA. Unit: uncore_cha event=0x35,umask=0x00c80fff01event=0x36,umask=0x00ccc7fe01TOR Occupancy; RFO misses from local IA. Unit: uncore_cha event=0x36,umask=0x00c001fd04event=0x36,umask=0x00c8a7fe01unc_cha_tor_occupancy.io_itomTOR Inserts : WCiLFs issued by iA Cores targeting DDR that missed the LLC - HOMed locally. Unit: uncore_cha TOR Occupancy : DRd_Prefs issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00c88f7e01event=0x36,umask=0x00c867fe01unc_cha_tor_occupancy.ia_miss_wcil_pmmTOR Occupancy : WbMtoIs issued by IO Devices. Unit: uncore_cha event=0x8eic_cache_fill_l2Instruction Pipe Stall. IC pipe was stalled during this clock cycle for any reason (nothing valid in pipe ICM1)event=0x8c,umask=0x01event=0x99All L2 Cache Requests (Breakdown 2 - Rare). Data cache read sizedRetired Branch Instructionsx87 instructionsevent=0xd4ex_tagged_ibs_ops.ibs_count_rolloverRemote Link Controller Outbound Packet Types: Data (32B): Remote Link Controller 3dram_channel_data_controller_5Single precision multiply-add FLOPS. Multiply-add counts as 2 FLOPSNumber of SSE Move Opsevent=0xaf,umask=0x40l2_request_g1.all_no_prefetch + l2_pf_hit_l2 + l2_pf_miss_l2_hit_l3 + l2_pf_miss_l2_l3l2_cache_accesses_from_dc_missesL2 Cache Accesses from L1 Data Cache Misses (including prefetch)All L2 Cache Hitsevent=0x45,umask=0xf0Micro-ops DispatchedThe number of instruction fetches that hit in the L1 ITLB. Instruction fetches to a 1GB pageFloating Point Dispatch Faults. XMM fill faultde_dis_uops_from_decoder.decoder_dispatchedbp_l1_tlb_miss_l2_hit + bp_l1_tlb_miss_l2_tlb_missThe number of valid fills into the ITLB originating from the LS Page-Table Walker. Tablewalk requests are issued for L1-ITLB and L2-ITLB misses. Walk for 1G pageevent=0x28f,umask=0x04ex_ret_opsDemand Data Cache Fills by Data Source. From DRAM or IO connected in different NodeDemand Data Cache Fills by Data Source. From Local L2 to the corede_dis_dispatch_token_stalls2.int_sch2_token_stallDynamic indirect predictions (branch used the indirect predictor to make a prediction)Software prefetches that did not fetch data outside of the processor core as the PREFETCH instruction saw a data cache hitHardware prefetch data cache fills from local L2 cacheL2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L1Stream (fetch additional sequential lines into L1 cache)l2_pf_miss_l2_hit_l3.alll2_pf_miss_l2_l3.l1_streamevent=0x72,umask=0x80Op cache missesOp cache accesses of all typesRetired macro-opsRead data beats (64 bytes) for local processor at Coherent Station (CS) 1Read data beats (64 bytes) for local processor at Coherent Station (CS) 6event=0x1f,umask=0x7fflocal_processor_write_data_beats_cs6local_processor_write_data_beats_cs10Write data beats (64 bytes) for remote processor at Coherent Station (CS) 6remote_processor_write_data_beats_cs9local_socket_inf0_inbound_data_beats_ccm5Data beats (32 bytes) at interface 0 for local socket inbound data to CPU Moderator (CCM) 5local_socket_inf1_inbound_data_beats_ccm0local_socket_inf1_inbound_data_beats_ccm5local_socket_inf0_outbound_data_beats_ccm6Data beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 2remote_socket_inf1_inbound_data_beats_ccm5event=0x41e,umask=0xbffData beats (64 bytes) for local socket outbound data from inter-socket xGMI link 5Retired x87 floating-point ops of all typesRetired packed 512-bit floating-point opsfp_ops_retired_by_type.vector_addRetired vector floating-point add opsRetired SSE and AVX integer logical opsevent=0xc,umask=0x05Retired 128-bit packed floating-point convert opsfp_pack_ops_retired.fp128_shufflefp_pack_ops_retired.fp256_cmpRetired 128-bit packed integer logical opsevent=0xd,umask=0x70Number of PRECHARGE commands sentl1_demand_data_cache_fills_from_near_cacheDRAM write data for remote processorlocal_socket_upstream_dma_write_dataLocal socket outbound data from the CPU (e.g. write data)Total cache hitsFR_DISPATCH_STALL_FOR_SEGMENT_LOADMEM_WRITEEVENT_03HEVENT_19HEVENT_84HEVENT_96HEVENT_CBHEVENT_F0HEVENT_F6HEVENT_FFHMAIN_TLB_MISS_STALLDATA_EVICTIONL2D_CACHE_REFILLEVENT_119HEVENT_12AHEVENT_140HEVENT_168HEVENT_1ABHEVENT_1B3HEVENT_1B7HEVENT_1C4HEVENT_1F8HEVENT_22BHEVENT_239HEVENT_23CHEVENT_243HEVENT_280HEVENT_28EHEVENT_29BHEVENT_2BFHEVENT_2C7HEVENT_319HEVENT_357HEVENT_36FHEVENT_395HEVENT_3CDHEVENT_3FDHLDREX_SPECASE_SPECLL_CACHE_RDhni_wdb_occ_cnt_ovflsbsx_rd_reqrnd_s0_rdata_beatscxha_wdb_occclkdiv2_high_qos_depthCOMPLETION_QUEUE_ENTRIES_OVER_THRESHOLDL2_INSTR_CACHE_MISSESBUS_RETRY_DUE_TO_PREVIOUS_ADJACENTLSU_EMPTYCYCLES_IN_SUPERDCACHE_MISS_COMPLETION_STALLFXU_LONG_INSTR_COMPLETION_STALLFINISHED_UNCOND_BRANCHES_MISS_BTBCYCLES_SU1_SCHED_STALLEDLOADS_TRANSLATEDSTORES_TRANSLATEDDATA_L1_CACHE_LOCKSSTORE_TRANSLATE_WHEN_QUEUE_FULLLOAD_MISS_DLFB_FULL_CYCLESLOAD_MISS_LDQ_FULL_CYCLESx87probe-hitdataARMV8pmclog_readpme->table[idx].name	%s
ldlat%s, "pmcid": "0x%08x", "pid": "%d", "base": "0x%016jx", "dyn": "0x%016jx", "pathname": "%s"}
Per-Logical Processor actual clocks when the Logical Processor is active( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )Number of Instructions per non-speculative Branch Misprediction (JEClear)InsTypeIpStoreMeasured Average Frequency for unhalted processors [GHz]l2_rqsts.demand_data_rd_hitThis event counts the number of demand Data Read requests that miss L2 cache. Only not rejected loads are countedl2_trans.l2_fillmem_load_uops_retired.l1_missThis is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were misses in the nearest-level (L1) cache. Counting excludes unknown and UC data source  Supports address when precise (Precise event)Retired load uops that split across a cacheline boundary.(Precise Event - PEBS)  Supports address when precise (Precise event)Uops delivered to Instruction Decode Queue (IDQ) from MITE pathUops delivered to Instruction Decode Queue (IDQ) while Microcode Sequenser (MS) is busyhle_retired.abortedtx_exec.misc4tx_mem.abort_hle_elision_buffer_mismatchSpeculative and retired direct near callsbr_inst_exec.all_indirect_jump_non_call_retbr_inst_exec.all_indirect_near_returnTaken speculative and retired indirect branches with return mnemonicAll (macro) branch instructions retiredConditional branch instructions retired. (Precise Event - PEBS) (Precise event)event=0x89,period=200003,umask=0x41This event counts not taken speculative and retired mispredicted macro conditional branch instructionsevent=0x89,period=200003,umask=0x88br_misp_retired.all_branches_pebsThis is a precise version of BR_MISP_RETIRED.ALL_BRANCHES that counts all mispredicted macro branch instructions retired (Must be precise)cpu_clk_unhalted.thread_p_anyThis event counts stalls occurred due to changing prefix length (66, 67 or REX.W when they change the length of the decoded instruction). Occurrences counting is proportional to the number of prefixes in a 16B-line. This may result in the following penalties: three-cycle penalty for each LCP in a 16-byte chunkThis event counts the number of times that split load operations are temporarily blocked because all resources for handling the split accesses are in useNumber of machine clears (nukes) of any typeuops_executed.core_cycles_ge_1event=0xa1,any=1,period=2000003,umask=0x2event=0xe,period=2000003,umask=0x10llc_misses.code_llc_prefetchevent=0x21,umask=0x4Shared line response from remote cache. Unit: uncore_ha write requests to memory controller. Derived from unc_m_cas_count.wr. Unit: uncore_imc (UNC_M_POWER_SELF_REFRESH / UNC_M_DCLOCKTICKS) * 100.Pre-charge for writes. Unit: uncore_imc event=0x80,occ_sel=1Counts the number of cycles when current is the upper limit on frequency. Unit: uncore_pcu Load miss in all TLB levels causes a page walk that completes. (1G)  Spec update: BDM69This event counts load misses in all DTLB levels that cause a completed page walk (2M and 4M page sizes). The page walk can end with or without a fault  Spec update: BDM69dtlb_store_misses.walk_completed_2m_4mCycle count for an Extended Page table walkevent=0xbc,period=2000003,umask=0x14page_walker_loads.itlb_l3IDQ.DSB_UOPS / (( IDQ.DSB_UOPS + LSD.UOPS + IDQ.MITE_UOPS + IDQ.MS_UOPS ) )MEM_Request_Latencyoffcore_response.all_data_rd.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0240offcore_response.all_pf_code_rd.l3_hit.snoop_not_neededoffcore_response.all_rfo.any_responseoffcore_response.all_rfo.l3_hit.snoop_hit_no_fwdoffcore_response.all_rfo.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020008event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020001offcore_response.other.any_responseoffcore_response.other.l3_hit.any_snoopoffcore_response.other.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x02003C0020offcore_response.pf_l3_code_rd.l3_hit.snoop_not_neededoffcore_response.pf_l3_code_rd.supplier_none.snoop_missoffcore_response.pf_l3_rfo.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0100offcore_response.pf_l3_rfo.supplier_none.snoop_hitmoffcore_response.pf_l3_rfo.supplier_none.snoop_missRandomly selected loads with latency value being above 16  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)Randomly selected loads with latency value being above 4  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)Randomly selected loads with latency value being above 512  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000091offcore_response.all_pf_code_rd.l3_miss.snoop_noneoffcore_response.all_pf_code_rd.l3_miss_local_dram.snoop_hitmoffcore_response.all_pf_code_rd.l3_miss_local_dram.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000090offcore_response.all_pf_data_rd.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000122offcore_response.all_rfo.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000008offcore_response.demand_data_rd.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204008000offcore_response.pf_l2_code_rd.l3_miss.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000040event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000040offcore_response.pf_l3_rfo.l3_miss_local_dram.snoop_hitmunc_cbo_cache_lookup.write_mesiunc_arb_trk_requests.allNumber of entries allocated. Account for Any type: e.g. Snoop, Core aperture, etcevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC007F7offcore_response.pf_llc_rfo.llc_miss.any_responseevent=0x28,period=200000,umask=0x41l2_ld.self.any.m_stateevent=0x2b,period=200000,umask=0x42l2_rqsts.self.prefetch.m_statel2_st.self.s_stateRetired loads that hit the L2 cache (precise event)Retired loads that miss the L2 cachefp_assist.ssimd_assistevent=0xb3,period=2000000,umask=0x2Store splitsevent=0x61,period=200000,umask=0x20event=0x6f,period=200000,umask=0xe0event=0x6b,period=200000,umask=0xe0Explicit writeback bus transactionsext_snoop.all_agents.hitext_snoop.this_agent.hitmRetired branch instructions that were predicted takenMispredicted and taken cond branch instructions retiredevent=0x8,period=200000,umask=0x6event=0x82,period=200000,umask=0x1Number of I-Side page walksL1 Cache evictions for dirty dataLoad uops retired that hit L2 (Precise event capable)  Supports address when precise (Must be precise)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400008000offcore_response.corewb.l2_hitCounts the number of writeback transactions caused by L1 or L2 cache evictions that hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000008offcore_response.demand_code_rd.l2_miss.hit_other_core_no_fwdevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000001Counts demand cacheable data reads of full cache lines that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredoffcore_response.partial_streaming_stores.l2_hitevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000044000offcore_response.pf_l2_data_rd.l2_hitoffcore_response.streaming_stores.l2_miss.anyCounts the number of floating point divide uops retired (Must be precise)BACLEARs asserted for any branch typeevent=0x13,period=200003,umask=0x4Cycles code-fetch stalled due to any reasonbr_inst_retired.callCounts mispredicted retired Jcc (Jump on Conditional Code/Jump if Condition is Met) branch instructions retired, including both when the branch was supposed to be taken and when it was not supposed to be taken (but the processor predicted the opposite condition) (Must be precise)Counts uops retired that are from the complex flows issued by the micro-sequencer (MS).  Counts both the uops from a micro-coded instruction, and the uops that might be generated from a micro-coded assist (Must be precise)ITLB missesCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.corewb.outstandingCounts demand cacheable data reads of full cache lines true miss for the L2 cache with a snoop miss in the other processor moduleCounts demand reads for ownership (RFO) requests generated by a write to full data cache line have any transaction responses from the uncore subsystemCounts data cache line reads generated by hardware L1 data cache prefetcher hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000020Counts reads for ownership (RFO) requests generated by L2 prefetcher outstanding, per cycle, from the time of the L2 miss to when any response is receivedevent=0x85,period=2000003,umask=0x2Counts STLB flushes.  The TLBs are flushed on instructions like INVLPG and MOV to CR3Retired load uops missed L1 cache as data sources  Supports address when precise.  Spec update: HSM30 (Precise event)Offcore outstanding cacheable data read transactions in SQ to uncore. Set Cmask=1 to count cycles  Spec update: HSD62, HSD61, HSM63This event count the number of undelivered (unallocated) uops from the Front-end to the Resource Allocation Table (RAT) while the Back-end of the processor is not stalled. The Front-end can allocate up to 4 uops per cycle so this event can increment 0-4 times per cycle depending on the number of unallocated uops. This event is counted on a per-core basis  Spec update: HSD135Cycles per thread when 4 or more uops are not delivered to Resource Allocation Table (RAT) when backend of the machine is not stalled  Spec update: HSD135Cycles with less than 2 uops delivered by the front end  Spec update: HSD135Randomly selected loads with latency value being above 16  Supports address when precise.  Spec update: HSD76, HSD25, HSM26 (Must be precise)offcore_response.demand_data_rd.l3_miss.any_responseNumber of ITLB page walker hits in the L3 + XSNP  Spec update: HSD25event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0600400091l2_lines_out.dirty_allCore-originated cacheable demand requests that refer to LLCOffcore outstanding code reads transactions in SuperQueue (SQ), queue to uncore, every cycleevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400244Number of cycles waiting for the checkpoints in Resource Allocation Table (RAT) to be recovered after Nuke due to all other cases except JEClear (e.g. whenever a ucode assist is needed like SSE exception, memory disambiguation, etc.)event=0x34,umask=0x06mem_load_uops_llc_miss_retired.remote_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0010Counts prefetch (that bring data to LLC only) data reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x107fc00004LLC misses for PCIe read current. Derived from unc_c_tor_inserts.miss_opcode.pcie_read. Unit: uncore_cbox Counts the number of times that the uncore transitioned to a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band1=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Derived from unc_p_freq_band1_cycles. Unit: uncore_pcu Cycles spent changing Frequency. Unit: uncore_pcu freq_ge_2000mhz_cycles %Counts the number of times that the uncore transitioned to a frequency greater than or equal to 2Ghz. Derived from unc_p_freq_band1_cycles. Unit: uncore_pcu event=0xd,edge=1,filter_band2=30Counts the number of cycles that the uncore transitioned to a frequency greater than or equal to 3Ghz. Derived from unc_p_freq_band2_cycles. Unit: uncore_pcu jkt metricsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC20077offcore_response.all_demand_mlc_pref_reads.llc_miss.remote_hitm_hit_forwardResource stalls out of order resources fullThis event count cycles when Page Miss Handler (PMH) is servicing page walks caused by ITLB missesl2_requests_reject.alloffcore_response.any_code_rd.any_responseoffcore_response.any_code_rd.l2_hit_this_tile_soffcore_response.any_data_rd.l2_hit_near_tileCounts any Prefetch requests that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x18001832f7Counts any Read request  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateoffcore_response.any_request.l2_hit_near_tile_mCounts any request that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateoffcore_response.any_rfo.l2_hit_this_tile_eCounts demand code reads and prefetch code reads that accounts for responses which hit its own tile's L2 with data in F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080080Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses which hit its own tile's L2 with data in E stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000080offcore_response.pf_l1_data_rd.l2_hit_far_tile_mCounts L2 code HW prefetches that accounts for responses which hit its own tile's L2 with data in F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800081000Counts Software Prefetches that accounts for responses which hit its own tile's L2 with data in F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008001000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000000200Counts the number of times the front end resteers for conditional branches as a result of another branch handling mechanism in the front endevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600044offcore_response.any_request.mcdram_farCounts any request that accounts for data responses from MCDRAM Localevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800004Counts demand code reads and prefetch code reads that accounts for data responses from MCDRAM Far or Other tile L2 hit faroffcore_response.pf_l1_data_rd.mcdram_nearoffcore_response.pf_l2_code_rd.ddr_farCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for data responses from DRAM LocalCounts UC code reads (valid only for Outstanding response type)  that accounts for data responses from MCDRAM LocalCounts the number of mispredicted near RET branch instructions retired (Precise event)Counts the number of unhalted core clock cyclesCounts the total D-side page walks that are completed or started. The page walks started in the speculative path will also be countedCycles L1D lockedL2 data demand loads in M stateAll L2 data prefetchesevent=0x26,period=200000,umask=0x20event=0xf2,period=100000,umask=0x1l2_rqsts.ifetchesevent=0x24,period=200000,umask=0x20event=0x24,period=200000,umask=0xcevent=0x27,period=100000,umask=0xf0Memory instructions retired above 256 clocks (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x844All offcore requestsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4722offcore_response.any_rfo.remote_cache_dramOffcore RFO requests satisfied by a remote cache or remote DRAMOffcore writebacks to the LLC or local DRAMOffcore code or data read requests satisfied by the LLC and HIT in a sibling coreoffcore_response.data_ifetch.llc_hit_other_core_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8033Offcore data reads, RFO's and prefetches satisfied by the LLC and HIT in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x703Offcore demand data reads satisfied by any cache or DRAMOffcore demand data reads satisfied by the LLC and not found in a sibling coreOffcore demand data reads satisfied by the LLC and HIT in a sibling coreoffcore_response.demand_data_rd.remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x102Offcore other requests satisfied by the LLC and not found in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x480event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4780event=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F30Offcore prefetch data requests satisfied by the LLC  and HITM in a sibling coreOffcore prefetch data requests satisfied by a remote cacheoffcore_response.pf_data_rd.local_cacheOffcore prefetch data reads satisfied by the LLCOffcore prefetch code reads satisfied by the LLC or local DRAMoffcore_response.pf_ifetch.remote_cache_hitmoffcore_response.pf_rfo.any_cache_dramoffcore_response.pf_rfo.io_csr_mmiooffcore_response.prefetch.local_cacheevent=0x10,period=2000000,umask=0x8SSE* FP double precision UopsSSE FP scalar UopsOffcore writebacks that missed the LLCoffcore_response.data_ifetch.local_dramoffcore_response.demand_data.local_dramOffcore other requests that missed the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4010es_reg_renamesevent=0x88,period=20000,umask=0x8Mispredicted call branches executedevent=0xd2,period=2000000,umask=0x8ssex_uops_retired.packed_doubleSIMD Scalar-Single Uops retired (Precise Event)event=0xb1,any=1,cmask=1,edge=1,inv=1,period=2000000,umask=0x3fuops_executed.core_stall_cycles_no_port5Uops executed on port 4 (core count)event=0xc2,period=2000000,umask=0x2Total cycles using precise uop retired event (Precise Event)dtlb_load_misses.pde_missevent=0x49,period=200000,umask=0x10Counts L2 cache hits when fetching instructions, code readsCore-originated cacheable demand requests missed L3  Spec update: SKL057Counts cycles when offcore outstanding cacheable Core Data Read transactions are present in the super queue. A transaction is considered to be in the Offcore outstanding state between L2 miss and transaction completion sent to requestor (SQ de-allocation). See corresponding Umask under OFFCORE_REQUESTSoffcore_response.demand_code_rd.l3_hit_e.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040004offcore_response.demand_data_rd.l3_hit.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080001offcore_response.demand_data_rd.l3_hit_e.spl_hitoffcore_response.demand_data_rd.l4_hit_local_l4.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000040002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2001C8000offcore_response.other.l3_hit_m.snoop_hit_no_fwdoffcore_response.other.l3_hit_s.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80108000frontend_retired.any_dsb_missInstruction fetch tag lookups that hit in the instruction cache (L1I). Counts at 64-byte cache-line granularityCounts randomly selected loads when the latency from first dispatch to completion is greater than 512 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)Counts randomly selected loads when the latency from first dispatch to completion is greater than 64 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)Counts number of Offcore outstanding Demand Data Read requests that miss L3 cache in the superQ every cycleevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C400001offcore_response.demand_rfo.l3_hit_e.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104008000rob_misc_events.pause_instevent=0xc2,cmask=16,inv=1,period=2000003,umask=0x2Mem;MemoryBW;Offcore_SMTL1MPKI_LoadCounts loads that miss the DTLB (Data TLB) and hit the STLB (Second level TLB)Counts completed page walks  (2M/4M sizes) caused by demand data stores. This implies address translations missed in the DTLB and further levels of TLB. The page walk can end with or without a faultCounts completed page walks (4K page sizes) caused by a code fetch. This implies it missed in the ITLB (Instruction TLB) and further levels of TLB. The page walk can end with or without a faultmem_uops_retired.utlb_missoffcore_response.any_code_rd.l2_miss.hit_other_core_no_fwdCounts any data read (demand & prefetch) that have any response typeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0200008008Counts writeback (modified to exclusive) that miss L2event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080000008Counts demand and DCU prefetch data read that hit in the other module where modified copies were found in other core's L1 cacheThis event counts the number of retired loads that were prohibited from receiving forwarded data from the store because of address mismatch (Precise event)Counts the number of taken JCC branch instructions retired (Precise event)Counts the number of core cycles while the core is not in a halt state. The core enters the halt state when it is running the HLT instruction. This event is a component in many key event ratios.  The core frequency may change from time to time. For this reason this event may have a changing ratio with regards to time. In systems with a constant core frequency, this event can give you a measurement of the elapsed time while the core was not in halt state by dividing the event count by the core frequency. This event is architecturally defined and is a designated fixed counter.  CPU_CLK_UNHALTED.CORE and CPU_CLK_UNHALTED.CORE_P use the core frequency which may change from time to time.  CPU_CLK_UNHALTE.REF_TSC and CPU_CLK_UNHALTED.REF are not affected by core frequency changes but counts as if the core is running at the maximum frequency all the time.  The fixed events are CPU_CLK_UNHALTED.CORE and CPU_CLK_UNHALTED.REF_TSC and the programmable events are CPU_CLK_UNHALTED.CORE_P and CPU_CLK_UNHALTED.REFCounts the number of cycles when no uops are allocated and the alloc pipe is stalled waiting for a mispredicted jump to retire.  After the misprediction is detected, the front end will start immediately but the allocate pipe stalls until the mispredictedDuration of I-side page-walks in core cyclesThis event counts when a data (D) page walk or an instruction (I) page walk is completed or started.  Since a page walk implies a TLB miss, the number of TLB misses can be counted by counting the number of pagewalksevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0120Counts prefetch RFOs that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all demand & prefetch prefetch RFOs event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0020REQUEST = PF_LLC_IFETCH and RESPONSE = ANY_RESPONSEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1f80400010offcore_requests_outstanding.demand.rfoREQUEST = ANY_DATA read and RESPONSE = ANY_CACHE_DRAMREQUEST = ANY_DATA read and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f44offcore_response.data_ifetch.local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f03offcore_response.demand_data_rd.local_dram_and_remote_cache_hitREQUEST = DEMAND_DATA_RD and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = OTHER and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff50event=0xb7,period=100000,umask=0x1,offcore_rsp=0xff40offcore_response.any_ifetch.other_local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf8ffevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3022offcore_response.data_in.other_local_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3003REQUEST = PF_RFO and RESPONSE = OTHER_LOCAL_DRAMoffcore_response.pf_rfo.other_local_dramDTLB miss page walk cyclesevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2710event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5810mem_uncore_retired.local_dram_and_remote_cache_hitoffcore_response.all_pf_rfo.l3_hit.hit_other_core_no_fwdOFFCORE_RESPONSE.ALL_PF_RFO.L3_HIT.SNOOP_HIT_WITH_FWDCounts number of SSE/AVX computational 512-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 16 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT14 RCP14 FM(N)ADD/SUB. FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementoffcore_response.all_data_rd.l3_miss.remote_hitmoffcore_response.all_pf_data_rd.l3_miss.any_snoopoffcore_response.all_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800002Counts number of cache lines that are allocated and written back to L3 with the intention that they are more likely to be reused shortlyFraction of Core cycles where the core was running with power-delivery for license level 1. SMT version; use when SMT is enabled and measuring per logical CPUCORE_POWER.LVL2_TURBO_LICENSE / CPU_CLK_UNHALTED.THREADWrite Pending Queue Allocations. Unit: uncore_imc unc_cha_core_snp.core_gtoneevent=0x33,umask=0x82Counts  transactions that looked into the multi-socket cacheline Directory state, and sent one or more snoops, because the Directory indicated it was neededunc_cha_snoop_resp.rspifwdevent=0xc0,ch_mask=0x04,fc_mask=0x07,umask=0x04Counts every peer to peer write request of 4 bytes of data made to the MMIO space of a card on IIO Part0 by a different IIO unit. Does not include requests made by the same IIO unit.  In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busRead request for up to a 64 byte transaction is made by the CPU to IIO Part0. Unit: uncore_iio Peer to peer read request for up to a 64 byte transaction is made by a different IIO unit to IIO Part2. Unit: uncore_iio event=0x84,ch_mask=0x01,fc_mask=0x07,umask=0x04Read request for up to a 64 byte transaction is made by IIO Part2 to Memory. Unit: uncore_iio Write request of up to a 64 byte transaction is made by IIO Part2 to Memory. Unit: uncore_iio Peer to peer write request of up to a 64 byte transaction is made by IIO Part1 to an IIO target. Unit: uncore_iio uncore_irpunc_m2m_direct2upi_not_taken_creditsevent=0x2e,umask=0x2event=0x2e,umask=0x10Reads to iMC issued at Normal Priority (Non-Isochronous). Unit: uncore_m2m unc_m2m_imc_writes.allWrites to iMC issued. Unit: uncore_m2m M2M Writes Issued to iMC; All, regardless of priorityCounts when the M2M (Mesh to Memory) recieves a prefetch request and inserts it into its outstanding prefetch queue.  Explanatory Side Note: the prefect queue is made from CAM: Content Addressable MemoryPrefetches generated by the flow control queue of the M3UPI unit. Unit: uncore_m3upi Counts Data Response (DRS) packets that attempted to go direct to core bypassing the CHAunc_upi_txl0p_power_cyclesCounts null FLITs (80 bit FLow control unITs) transmitted via any of the 3 Intel Ulra Path Interconnect (UPI) slots on this UPI unitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080491OCR.ALL_DATA_RD.L3_HIT_E.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80080491OCR.ALL_DATA_RD.L3_HIT_F.SNOOP_MISSOCR.ALL_DATA_RD.L3_HIT_M.HITM_OTHER_CORE  OCR.ALL_DATA_RD.L3_HIT_M.HITM_OTHER_COREOCR.ALL_DATA_RD.L3_HIT_S.ANY_SNOOP  OCR.ALL_DATA_RD.L3_HIT_S.ANY_SNOOPocr.all_data_rd.l3_hit_s.snoop_missOCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_NONE OCR.ALL_PF_DATA_RD.L3_HIT.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100490OCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_PF_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDOCR.ALL_PF_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_RFO.L3_HIT_E.HIT_OTHER_CORE_NO_FWDOCR.ALL_PF_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10002007F7ocr.all_rfo.l3_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80200122ocr.demand_code_rd.l3_hit_f.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200004Counts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDocr.demand_code_rd.l3_hit_s.hitm_other_coreocr.demand_data_rd.l3_hit.any_snoopCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDocr.demand_data_rd.l3_hit_s.hit_other_core_no_fwdocr.demand_rfo.l3_hit_e.hit_other_core_no_fwdocr.demand_rfo.l3_hit_e.no_snoop_neededocr.demand_rfo.l3_hit_s.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800108000ocr.pf_l1d_and_sw.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080400ocr.pf_l1d_and_sw.l3_hit_e.hitm_other_coreCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_F.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080020ocr.pf_l2_rfo.l3_hit_f.hitm_other_coreCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_M.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040020ocr.pf_l3_data_rd.l3_hit_f.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDocr.pf_l3_data_rd.l3_hit_s.any_snoopocr.pf_l3_rfo.l3_hit.hit_other_core_no_fwdocr.pf_l3_rfo.l3_hit_e.snoop_noneCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDocr.pf_l3_rfo.l3_hit_s.any_snoopoffcore_response.all_data_rd.l3_hit_f.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_HIT_E.SNOOP_MISSoffcore_response.all_pf_data_rd.l3_hit_f.snoop_missoffcore_response.all_pf_data_rd.l3_hit_m.snoop_missThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDoffcore_response.all_pf_rfo.l3_hit_e.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_E.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020120This event is deprecated. Refer to new event OCR.ALL_PF_RFO.SUPPLIER_NONE.SNOOP_MISSoffcore_response.all_reads.l3_hit_e.hitm_other_coreoffcore_response.all_reads.l3_hit_m.hit_other_core_no_fwdoffcore_response.all_reads.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_F.ANY_SNOOPoffcore_response.demand_code_rd.l3_hit.hit_other_core_fwdoffcore_response.demand_code_rd.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_M.NO_SNOOP_NEEDEDoffcore_response.demand_code_rd.pmm_hit_local_pmm.any_snoopoffcore_response.demand_data_rd.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_F.HIT_OTHER_CORE_FWDoffcore_response.other.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_S.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l1d_and_sw.l3_hit_s.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020010offcore_response.pf_l2_rfo.l3_hit_f.snoop_noneoffcore_response.pf_l2_rfo.l3_hit_m.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.l3_hit_s.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_M.ANY_SNOOPoffcore_response.pf_l3_rfo.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_S.HITM_OTHER_COREoffcore_response.pf_l3_rfo.supplier_none.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020100Intel AVX-512 computational 128-bit packed BFloat16 instructions retiredCounts once for each Intel AVX-512 computational 256-bit packed BFloat16 floating-point instruction retired. Applies to the YMM based VDPBF16PS instruction.  Each count represents 32 computation operations. This event is only supported on products formerly named Cooper Lake and is not supported on products formerly named Cascade Lakeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000490OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDED  OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000120ocr.all_reads.l3_miss.hit_other_core_no_fwdOCR.ALL_READS.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_READS.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDocr.all_reads.l3_miss_remote_hop1_dram.any_snoopocr.all_reads.l3_miss_remote_hop1_dram.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000004ocr.demand_data_rd.l3_miss_local_dram.snoop_missocr.demand_rfo.l3_miss.remote_hitmocr.other.l3_miss.snoop_missocr.other.l3_miss_local_dram.snoop_noneCounts any other requests  OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x110008000ocr.pf_l2_data_rd.l3_miss.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F90000010Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000100offcore_response.all_pf_rfo.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS.ANY_SNOOPoffcore_response.demand_code_rd.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_data_rd.l3_miss.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.HIT_OTHER_CORE_FWDoffcore_response.pf_l1d_and_sw.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.pf_l2_rfo.l3_miss.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPocr.all_data_rd.supplier_none.no_snoop_neededocr.all_pf_data_rd.supplier_none.hit_other_core_fwdOCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWD  OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDOCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDocr.all_pf_data_rd.supplier_none.snoop_missocr.all_pf_rfo.pmm_hit_local_pmm.snoop_noneOCR.ALL_PF_RFO.SUPPLIER_NONE.SNOOP_MISSocr.all_reads.supplier_none.hit_other_core_fwdocr.all_rfo.supplier_none.hit_other_core_no_fwdocr.demand_code_rd.supplier_none.no_snoop_neededCounts demand data reads  OCR.DEMAND_DATA_RD.SUPPLIER_NONE.HITM_OTHER_CORECounts all demand data writes (RFOs) OCR.DEMAND_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONEocr.other.supplier_none.hit_other_core_fwdCounts any other requests  OCR.OTHER.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.pf_l1d_and_sw.supplier_none.no_snoop_neededocr.pf_l2_data_rd.supplier_none.no_snoop_neededCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.pf_l3_data_rd.supplier_none.no_snoop_neededCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDIntel Optane DC persistent memory read latency (ns). Unit: uncore_imc All hits to Near Memory(DRAM cache) in Memory Mode. Unit: uncore_imc unc_cha_tor_occupancy.ia_miss_drdevent=0x38,umask=0x20Write requests to Intel Optane DC persistent memory issued to the iMC from M2M. Unit: uncore_m2m Counts demand instruction fetches and L1 instruction cache prefetches that hit a cacheline in the L3 where a snoop was not needed to satisfy the requestCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit a cacheline in the L3 where a snoop was sent but no other cores had the dataCounts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that hit a cacheline in the L3 where a snoop was sent but no other cores had the dataocr.hwpf_l2_data_rd.l3_hit.anyevent=0x32,period=100003,umask=0x4Number of SSE/AVX computational 256-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0xc6,period=100007,umask=0x1,frontend=0x502006Counts retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 512 cycles which was not interrupted by a back-end stall (Precise event)Counts the number of cycles when no uops were delivered by the Instruction Decode Queue (IDQ) to the back-end of the pipeline when there was no back-end stalls. This event counts for one SMT thread in a given cycleCounts the number of times an RTM execution aborted due to none of the previous 4 categories (e.g. interrupt)event=0x5d,period=100003,umask=0x2tx_mem.abort_capacity_readCounts the number of times a TSX line had a cache conflictCounts the number of times a TSX Abort was triggered due to release/commit but data and address mismatchocr.demand_code_rd.local_dramocr.demand_rfo.local_dramocr.streaming_wr.any_responseocr.streaming_wr.local_dramNumber of instructions retired. General Counter - architectural event (Precise event)Number of uops executed on port 1Counts stores that miss the DTLB (Data TLB) and hit the STLB (2nd Level TLB)Cycles when at least one PMH is busy with a page walk for a storetgl metricsCounts responses to snoops indicating the line may be kept on this core in the (S)hared state, after the data is forwarded back to the requestor, initially the data was found in the cache in the (M)odified state.  A single snoop response from the core counts on all hyperthreads of the coreocr.demand_data_rd.pmmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x700800001ocr.reads_to_core.local_pmmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x708000477event=0x82PMM Commands : Underfill reads. Unit: uncore_imc event=0x37,umask=0x0Fevent=0x35,umask=0xC807FD01TOR Inserts : LLCPrefRFO issued by iA Cores that missed the LLC. Unit: uncore_cha unc_cha_tor_inserts.io_hitTOR Inserts : ItoMs issued by IO Devices that missed the LLC. Unit: uncore_cha unc_cha_tor_inserts.ia_miss_rfo_pref_remoteevent=0x36,umask=0xC8178A01unc_iio_data_req_of_cpu.mem_write.part7unc_iio_data_req_of_cpu.cmpd.part6unc_iio_txn_req_by_cpu.mem_write.part6unc_iio_txn_req_of_cpu.cmpd.part4PCIe Completion Buffer Occupancy of completions with data : Part 6. Unit: uncore_iio event=0x10,umask=0x40Counts the number of cycles the core is stalled due to an instruction cache or Translation Lookaside Buffer (TLB) access which hit in the L2 cacheCounts the number of load uops retired that hit in the L1 data cache  Supports address when precise (Precise event)ocr.demand_data_and_l1pf_rd.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004003C0000event=0x71,period=1000003,umask=0x20Counts the number of missed requests to the instruction cache.  The event only counts new cache line accesses, so that multiple back to back fetches to the exact same cache line and byte chunk count as one.  Specifically, the event counts when accesses from sequential code crosses the cache line boundary, or when a branch target is moved to a new line or to a non-sequential byte chunk of the same lineevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3002184000000ocr.l1wb_m.l3_missThis event is deprecated. Refer to new event MEM_BOUND_STALLS.LOAD_LLC_HITCounts all code reads that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)ocr.uc_rd.any_responseCounts the number of mispredicted taken JCC (Jump on Conditional Code) branch instructions retired (Precise event)Counts the number of retired loads that are blocked because its address exactly matches an older store whose data is not ready (Precise event)topdown_bad_speculation.monuketopdown_be_bound.alloc_restrictionsevent=0x71,period=1000003,umask=0x40Write Pending Queue Occupancy : Accumulates the occupancies of the Write Pending Queue each cycle.  This can then be used to calculate both the average queue occupancy (in conjunction with the number of cycles not empty) and the average latency (in conjunction with the number of allocations).  The WPQ is used to schedule write out to the memory controller and to track the writes.  Requests allocate into the WPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the HA to the iMC.  They deallocate after being issued to DRAM.  Write requests themselves are able to complete (from the perspective of the rest of the system) as soon they have posted to the iMC.  This is not to be confused with actually performing the write to DRAM.  Therefore, the average latency for this queue is actually not useful for deconstruction intermediate write latencies.  So, we provide filtering based on if the request has posted or not.  By using the not posted filter, we can track how long writes spent in the iMC before completions were sent to the HA.  The posted filter, on the other hand, provides information about how much queueing is actually happenning in the iMC for writes before they are actually issued to memory.  High average occupancies will generally coincide with high write major mode countsTOR Inserts : All requests from iA Cores : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsTOR Inserts : DRd_Opts issued by iA Cores. Unit: uncore_cha TOR Inserts : DRd_Opts issued by iA Cores that hit the LLC. Unit: uncore_cha event=0x36,umask=0xC8A7FD01Clockticks of the integrated IO (IIO) traffic controllerPCIe Completion Buffer Inserts of completions with data : Part 2 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 6PCIe Completion Buffer Occupancy : Part 2 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 2Data requested of the CPU : Card writing to DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 5Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 7Total IRP occupancy of inbound read and write requests to coherent memory.  This is effectively the sum of read occupancy and write occupancyCounts the number of page walks due to stores that miss the PDE (Page Directory Entry) cacheCounts the number of page walks completed due to stores whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 2M or 4M pages.  Includes page walks that page faultept.epde_hit64 * LONGEST_LAT_CACHE.MISS / 1000000000 event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2104000002Average external Memory Bandwidth Use for reads and writes [GB / sec]. Unit: cpu_core Uops Per Instruction. Unit: cpu_atom 100 * UOPS_RETIRED.MS / UOPS_RETIRED.ALLPercentage of all uops which are FPDiv uops. Unit: cpu_atom Counts the number of cycles the core is stalled due to an instruction cache or tlb miss which hit in the L2 cache. Unit: cpu_atom Counts the number of load ops retired that hit in DRAM  Supports address when precise (Precise event). Unit: cpu_atom Counts the number of load uops retired  Supports address when precise (Precise event). Unit: cpu_atom event=0x43,period=1000003,umask=0xfdCompleted demand load uops that miss the L1 d-cache. Unit: cpu_core decode.lcpevent=0x79,cmask=6,period=2000003,umask=0x8Uops delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path. Unit: cpu_core event=0x73,period=1000003This event is deprecated. Refer to new event ARITH.FPDIV_ACTIVE. Unit: cpu_core event=0xec,cmask=1,edge=1,period=2000003,umask=0x40Number of all retired NOP instructions. Unit: cpu_core int_misc.unknown_branch_cyclesint_vec_retired.256bitCounts cycles where the pipeline is stalled due to serializing operations. Unit: cpu_core event=0xc2,cmask=1,period=1000003,umask=0x2unc_m_vc0_requests_rdevent=0x12,period=100003,umask=0x8Counts hardware prefetches (which bring data to L2) that have any type of responseamx_ops_retired.int8INST_RETIRED.MACRO_FUSEDPrecise instruction retired with PEBS precise-distribution (Precise event)TOPDOWN.MEMORY_BOUND_SLOTSNumber of uops dispatch to execution  port 6IMC Clockticks at DCLK frequency. Unit: uncore_imc PMM Read Pending Queue occupancy. Unit: uncore_imc event=0x3,umask=0x0000000020event=0xe0,umask=0x0000000010event=0x84,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000002TOR Occupancy for DRd misses from local IA targeting local memory. Unit: uncore_cha Valid Flits Sent : Data. Unit: uncore_upi Valid Flits Received : Slot 1. Unit: uncore_upi event=0x4,umask=0x000000010func_upi_txl_insertsevent=0x83,ch_mask=0x0010,fc_mask=0x07,umask=0x0000000004event=0xc0,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000001event=0x1c,umask=0x03unc_m2m_tracker_inserts.ch0unc_m3upi_cms_clockticksTOR Inserts; Hits from Local IA. Unit: uncore_cha event=0x35,umask=0x00ccc7fe01event=0x35,umask=0x00c001fd04TOR Inserts : IRQ - iA. Unit: uncore_cha unc_cha_tor_inserts.remote_tgtunc_cha_tor_inserts.isocevent=0x35,umask=0x00c897fd01event=0x36,umask=0x00C000FF04event=0x36,umask=0x00c803ff04event=0x35,umask=0x00c837ff01event=0x35,umask=0x00C86F0A01unc_cha_tor_inserts.ia_miss_wcil_ddrunc_cha_tor_occupancy.ia_miss_wilunc_cha_tor_occupancy.ia_miss_remote_wcilf_ddrevent=0x36,umask=0x00cd43fd04ic_cache_fill_sysbp_l1_tlb_miss_l2_hitbp_tlb_relAll L2 Cache Requests (Breakdown 1 - Common). Data cache storesevent=0x60,umask=0x20l2_request_g1.l2_hw_pfAll L2 Cache Requests (Breakdown 2 - Rare). Instruction cache read sizedevent=0x61,umask=0x08l2_cache_req_stat.ic_access_in_l2l2_fill_pending.l2_fill_busyex_ret_mmx_fp_instr.x87_instrex_ret_fus_brnch_instdram_channel_data_controller_2event=0,umask=0xf0The number of x87 floating-point Ops that have retired. The number of events logged per cycle can vary from 0 to 8. Divide and square root Opsfp_ret_sse_avx_ops.dp_add_sub_flopsx87 bottom-executing uOps retiredevent=0x29,umask=0x01ls_mab_alloc.storesL1 DTLB Reload of a page of 1G sizels_l1_d_tlb_miss.tlb_reload_2m_l2_hitCycles where a dispatch group is valid but does not get dispatched due to a token stall. ALSQ 2 Tokens unavailabled_ratio(ex_ret_brn_misp, ex_ret_brn)event=0x60,umask=0xc8L2 Cache Misses from L1 Instruction Cache Missesevent=0xe,umask=0x0eAdd/subtract FLOPS. This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15ls_pref_instr_disp.prefetch_wOps dispatched from either the decoders, OpCache or bothCycles where a dispatch group is valid but does not get dispatched due to a token stall. FP Miscellaneous resource unavailable. Applies to the recovery of mispredicts with FP opsThe number of instruction fetches that hit in the L1 ITLB. L1 Instruction TLB hit (2M page size)Core to L2 cacheable request access status (not including L2 Prefetch). Data cache read hit in L2. ModifiableL2 prefetcher misses in L3. Counts all L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 cachesex_ret_fused_instrDemand Data Cache Fills by Data Source. From DRAM or IO connected in same nodels_any_fills_from_sys.ext_cache_remotels_misal_loads.ma64Core to L2 cache requests (not including L2 prefetch) for instruction cache accessl2_pf_miss_l2_l3.l1_regionInstruction cache hitsl3_xi_sampled_latency_requests.allInterrupts takenWrite data beats (64 bytes) for local processor at Coherent Station (CS) 0event=0x8df,umask=0x7fflocal_socket_inf0_inbound_data_beats_ccm1Data beats (32 bytes) at interface 1 for local socket inbound data to CPU Moderator (CCM) 0Data beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 1local_socket_inf0_outbound_data_beats_ccm7Data beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 0event=0x49f,umask=0xbfeevent=0x55f,umask=0xbferemote_socket_inf0_outbound_data_beats_ccm6Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 1Retired SSE and AVX floating-point multiply opsfp_retired_ser_ops.allRetired scalar floating-point opsevent=0xa,umask=0x01event=0xa,umask=0x60fp_ops_retired_by_type.vector_allRetired MMX integer multiply ops of other typesRetired SSE and AVX integer compare opssse_avx_ops_retired.sse_avx_othersse_avx_ops_retired.allRetired 128-bit packed floating-point square root opsevent=0xd,umask=0x03Number of clocks used by the data bus for writesL1 DTLB misses with L2 DTLB hits for coalesced pages. A coalesced page is a 16k page created from four adjacent 4k pagessmt_contentiond_ratio(ex_ret_ops, total_dispatch_slots)retiringFraction of dispatch slots that remained unused because of stalls not related to the memory subsystemretiring * (1 - d_ratio(ex_ret_ucode_ops, ex_ret_ops))L2 cache misses from L1 data cache missesL1 data cache fills from another CCX cache in any NUMA node1MB/sMemory controller ACTIVATE command rateBUS_ACCESSINSTRUCTIONS_ISSUED_CYCLECYCLES_STALLED_NEON_FULLQEVENT_02HEVENT_13HEVENT_2BHEVENT_4AHEVENT_D6HEVENT_F7HDATA_MICRO_TLB_MISS_STALLEVENT_11BHEVENT_11CHEVENT_126HEVENT_150HEVENT_15FHEVENT_16CHEVENT_181HEVENT_183HEVENT_1A2HEVENT_1BAHEVENT_205HEVENT_223HEVENT_248HEVENT_26CHEVENT_282HEVENT_285HEVENT_29DHEVENT_2E1HEVENT_30BHEVENT_317HEVENT_35BHEVENT_35DHEVENT_38DHEVENT_3ACHEVENT_3BDHEVENT_3BFHEVENT_3C9HEVENT_3D2HEVENT_3FEHISB_SPECdn_rxreq_vivi_dvmophnf_snp_sentrnd_s2_rdata_beatsclkdiv2_waiting_for_wr_dataLSU_LOAD_MISS_LINE_ALIAS_VS_CSQ0LSU_LOAD_VS_STORE_QUEUE_ALIAS_STALLREFETCH_SERIALIZATIONSECOND_SPECULATION_BUFFER_ACTIVESUCCESSFUL_STWCXL2_CACHE_MISSESMARKED_STORE_SENT_TO_STSLSU_MARKED_INSTR_FINISHLOADS_TRANSLATED_ALLOCATED_TO_DLFBDATA_MMU_MISS_CYCLESINSTR_MMU_TLB4K_RELOADSEXT_INPUT_INTR_PENDING_LATENCY_CYCLESPMC0_OVERFLOWINTV_ALLOCATIONSL2_CACHE_DIRTY_REDUNDANT_UPDATESL2_INCOHERENT_LINE_INVALIDATIONSIAC2S_DETECTEDDAC1S_DTECTEDcount=add-pipe-excluding-junk-opslocked-instructionsPRECISEUCFINTEL_ATOM_GOLDMONT_PSStopic: %s
{"type": "pmcallocatedyn"{"type": "thr_exit"v19GenuineIntel-6-[4589]EGenuineIntel-6-37AuthenticAMD-23-[012][0-9A-F]AuthenticAMD-25-[[:xdigit:]]+This category represents fraction of slots where no uops are being delivered due to a lack of required resources for accepting new uops in the Backend. SMT version; use when SMT is enabled and measuring per logical CPUThis category represents fraction of slots utilized by useful work i.e. issued uops that eventually get retiredThis category represents fraction of slots utilized by useful work i.e. issued uops that eventually get retired. SMT version; use when SMT is enabled and measuring per logical CPUCLKSRet;SMT;TmaL1Actual per-core usage of the Floating Point execution units (regardless of the vector width). Values > 1 are possible due to Fused-Multiply Add (FMA) countingBad;BrMispredictsInstructions per Load (lower number means higher occurrence rate)Instructions per FP Arithmetic AVX* 256-bit instruction (lower number means higher occurrence rate)l2_rqsts.all_pfl2_rqsts.l2_pf_missmem_load_uops_retired.l3_hitThis is a precise version (that is, uses PEBS) of the event that counts load uops with locked access retired to the architected path  Supports address when precise.  Spec update: BDM35 (Precise event)event=0xca,period=100003,umask=0x10Number of SIMD Move Elimination candidate uops that were eliminatedNumber of transitions from SSE to AVX-256 when penalty applicable  Spec update: BDM30event=0x80,period=2000003,umask=0x4event=0x79,period=2000003,umask=0x8event=0x79,period=2000003,umask=0x2event=0x9c,cmask=1,inv=1,period=2000003,umask=0x1This is a precise version (that is, uses PEBS) of the event that counts return instructions retired (Precise event)event=0x89,period=200003,umask=0xc4Not taken speculative and retired mispredicted macro conditional branchesThis event counts all mispredicted macro branch instructions retiredCounts number of cycles nothing is executed on any execution port, while there was at least one pending demand load request missing the L1 data cacheCycles 4 Uops delivered by the LSD, but didn't come from the decoderResource-related stall cyclesevent=0xb1,period=2000003,umask=0x1uops_executed_port.port_4Cycles without actually retired uopsAll LLC Misses (code+ data rd + data wr - including demand and prefetch). Unit: uncore_cbox ItoM write misses (as part of fast string memcpy stores) + PCIe full line writes. Derived from unc_c_tor_inserts.miss_opcode. Unit: uncore_cbox write requests to local home agent. Unit: uncore_ha (UNC_P_FREQ_MAX_POWER_CYCLES / UNC_P_CLOCKTICKS) * 100.This event counts load misses in all DTLB levels that cause page walks of any page size (4K/2M/4M/1G)  Spec update: BDM69event=0x8,period=2000003,umask=0x2ept.walk_cyclesitlb.itlb_flushoffcore_response.all_data_rd.supplier_none.snoop_hitmoffcore_response.all_pf_rfo.supplier_none.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0122offcore_response.corewb.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020040offcore_response.pf_l2_data_rd.l3_hit.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010020offcore_response.pf_l3_code_rd.supplier_none.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000020080offcore_response.pf_l3_data_rd.supplier_none.snoop_hit_no_fwdCounts randomly selected loads with latency value being above 256  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)offcore_response.all_data_rd.l3_miss.snoop_noneoffcore_response.all_data_rd.l3_miss_local_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2004000091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000090offcore_response.all_rfo.l3_miss_local_dram.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000008offcore_response.corewb.l3_miss_local_dram.snoop_not_neededoffcore_response.pf_l2_data_rd.l3_miss_local_dram.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000020offcore_response.pf_l3_code_rd.l3_miss.snoop_missoffcore_response.pf_l3_code_rd.l3_miss_local_dram.snoop_missoffcore_response.pf_l3_data_rd.l3_miss_local_dram.snoop_noneoffcore_response.pf_l3_data_rd.l3_miss_local_dram.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000100A cross-core snoop initiated by this Cbox due to processor core memory request which hits a modified line in some processor coreevent=0x34,umask=0x81L3 Lookup any request that access cache and found line in E or S-stateoffcore_response.pf_llc_code_rd.llc_miss.any_responsel2_ifetch.self.e_statel2_ld.self.any.s_statel2_ld.self.prefetch.mesil2_reject_busq.self.prefetch.m_statel2_rqsts.self.any.s_stateevent=0x2e,period=200000,umask=0x48l2_rqsts.self.prefetch.e_stateevent=0x77,period=200000,umask=0x2event=0xe4,period=2000000,umask=0x1Bogus branchesbr_missp_type_retired.returnevent=0x14,period=2000000,umask=0x1event=0x13,period=2000000,umask=0x81event=0x8,period=200000,umask=0xaITLB hitspage_walks.d_side_walksNumber of D-side only page walkscore_reject_l2q.allmem_load_uops_retired.dram_hitCounts the number of load uops retired  Supports address when precise (Must be precise)Counts locked memory uops retired.  This includes regular locks and bus locks. (To specifically count bus locks only, see the Offcore response event.)  A locked access is one with a lock prefix, or an exchange to memory.  See the SDM for a complete description of which memory load accesses are locks  Supports address when precise (Must be precise)offcore_response.any_pf_data_rd.l2_miss.hit_other_core_no_fwdCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests (demand & prefetch) that miss the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000001offcore_response.demand_rfo.l2_miss.snoop_miss_or_no_snoop_neededCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that are outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts reads for ownership (RFO) requests generated by L2 prefetcher that true miss for the L2 cache with a snoop miss in the other processor moduleCounts any data writes to uncacheable write combining (USWC) memory region  that hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600001000Counts data cache lines requests by software prefetch instructions that miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Instructions retired (Precise event capable) (Must be precise)Unfilled issue slots per cycleCounts anytime a load that retires is blocked for any reason (Must be precise)Counts uops issued by the front end and allocated into the back end of the machine.  This event counts uops that retire as well as uops that were speculatively executed but didn't retire. The sort of speculative uops that might be counted includes, but is not limited to those uops issued in the shadow of a miss-predicted branch, those uops that are inserted during an assist (such as for a denormal floating point result), and (previously allocated) uops that might be canceled during a machine clearCounts every core cycle when a Instruction-side (walks due to an instruction fetch) page walk is in progressCounts data reads (demand & prefetch) hit the L2 cacheevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x4000003091Counts data reads generated by L1 or L2 prefetchers have any transaction responses from the uncore subsystemCounts requests to the uncore subsystem miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010008Counts the number of writeback transactions caused by L1 or L2 cache evictions outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand cacheable data reads of full cache lines outstanding, per cycle, from the time of the L2 miss to when any response is received. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache line reads generated by hardware L1 data cache prefetcher true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts reads for ownership (RFO) requests generated by L2 prefetcher miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts any data writes to uncacheable write combining (USWC) memory region  true miss for the L2 cache with a snoop miss in the other processor moduleCounts data cache lines requests by software prefetch instructions have any transaction responses from the uncore subsystemCounts page walks completed due to demand data stores whose address translations missed in the TLB and were mapped to 1GB pages.  The page walks can end with or without a page faultevent=0x49,period=2000003,umask=0x4Counts once per cycle for each page walk occurring due to an instruction fetch. Includes cycles spent traversing the Extended Page Table (EPT). Average cycles per walk can be calculated by dividing by the number of walksoffcore_response.pf_l3_code_rd.l3_hit.any_responseCounts cycles DSB is delivered at least one uops. Set Cmask = 1Speculative cache-line split store-address uops dispatched to L1Devent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100400244offcore_response.all_rfo.l3_miss.local_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00004offcore_response.demand_rfo.l3_miss.any_responseThis event counts the number of cycles spent waiting for a recovery after an event such as a processor nuke, JEClear, assist, hle/rtm abort etcThis event counts cycles when the Reservation Station ( RS ) is empty for the thread. The RS is a structure that buffers allocated micro-ops from the Front-end. If there are many cycles when the RS is empty, it may represent an underflow of instructions delivered from the Front-endCycles which a uop is dispatched on port 0 in this threadevent=0x34,umask=0x4fAn external snoop hits a modified line in some processor coreAn external snoop misses in some processor coreCompleted page walks due to store miss in any TLB levels of any page size (4K/2M/4M/1G)Number of ITLB page walker hits in the L2event=0xbc,period=2000003,umask=0x28offcore_response.pf_l2_rfo.llc_miss.any_responsel2_l1d_wb_rqsts.hit_mevent=0x27,period=200003,umask=0x1mem_load_uops_llc_miss_retired.local_dramOffcore outstanding Demand Data Read transactions in SQ to uncore. Set Cmask=1 to count cyclesOffcore outstanding RFO store transactions in SuperQueue (SQ), queue to uncoreCount issue pipeline slots where no uop was delivered from the front end to the back end when there is no back-end stallCount XClk pulses when this thread is unhalted and the other is haltedNumber of occurences waiting for the checkpoints in Resource Allocation Table (RAT) to be recovered after Nuke due to all other cases except JEClear (e.g. whenever a ucode assist is needed like SSE exception, memory disambiguation, etc.)event=0x22,umask=0x02Filter on any IRQ or IPQ initiated requests including uncacheable, non-coherent requestsevent=0x8,period=2000003,umask=0x84Counts all demand & prefetch code reads that miss the LLC  and the data forwarded from remote cacheCounts all demand code reads that miss the LLC  the data is found in M state in remote cache and forwarded from thereevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x107fc20002llc_references.pcie_ns_writeQPI clock ticks. Use to get percentages for QPI cycles events. Unit: uncore_qpi event=0xb,edge=1,filter_band0=12l1d.all_m_replacementCache lines in M state evicted out of L1D due to Snoop HitM or dirty line replacementMispredicted taken branch instructions retired (Precise event)partial_rat_stalls.flags_merge_uoppartial_rat_stalls.mul_single_uopUops dispatched from any threadThis event counts the number of micro-ops retired (Precise event)unc_m_act_countevent=0x4,umask=0xcCounts the number of core cycles the fetch stalls because of an icache miss. This is a cummulative count of core cycles the fetch stalled for all icache missesThis event counts the number of core cycles the fetch stalls because of an icache miss. This is a cumulative count of cycles the NIP stalled for all icache missesl2_requests.missevent=0x4,period=200003,umask=0x80mem_uops_retired.hitmevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800400044event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080044offcore_response.any_data_rd.l2_hit_this_tile_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x00020032f7event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800408000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800188000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008008000Counts Demand cacheable data write requests  that accounts for any responseoffcore_response.any_rfo.l2_hit_this_tile_sCounts demand code reads and prefetch code reads that are outstanding, per weighted cycle, from the time of the request to when any response is received. The oustanding response should be programmed only on PMC0Counts demand cacheable data and L1 prefetch data reads that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateCounts Demand cacheable data writes that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000002offcore_response.pf_l1_data_rd.l2_hit_far_tileoffcore_response.pf_l2_code_rd.l2_hit_near_tile_e_fCounts L2 code HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateoffcore_response.pf_l2_rfo.l2_hit_far_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080020Counts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses which hit its own tile's L2 with data in E stateoffcore_response.pf_l2_rfo.l2_hit_this_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002001000Counts the number of times the front end resteers for RET branches as a result of another branch handling mechanism in the front endoffcore_response.any_code_rd.mcdram_nearevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400070event=0xb7,period=100007,umask=0x1,offcore_rsp=0x01818032f7Counts any Read request  that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts Bus locks and split lock requests that accounts for data responses from MCDRAM Far or Other tile L2 hit farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800002Counts the number of far branch instructions retired (Precise event)Counts the number of mispredicted far branch instructions retired (Precise event)Counts the number of mispredicted near relative CALL branch instructions retired (Precise event)This event counts the number of core cycles while the thread is not in a halt state. The thread enters the halt state when it is running the HLT instruction. This event is a component in many key event ratios. The core frequency may change from time to time due to transitions associated with Enhanced Intel SpeedStep Technology or TM2. For this reason this event may have a changing ratio with regards to time. When the core frequency is constant, this event can approximate elapsed time while the core was not in the halt state. It is counted on a dedicated fixed counterCycles the number of core cycles when divider is busy.  Does not imply a stall waiting for the dividerevent=0xca,period=200003,umask=0x20recycleq.ld_block_st_forwardrecycleq.ld_splitsevent=0xcb,period=200003,umask=0x1f6.4e-05MiBevent=0x43,period=2000000,umask=0x2L1 data cache read in I state (misses)L1 data cache load locks in S stateL1 data cache stores in E stateevent=0x26,period=200000,umask=0x8l2_lines_out.anyl2_rqsts.ld_missl2_rqsts.prefetch_hitL2 fill transactionsevent=0xf0,period=200000,umask=0x40event=0x27,period=100000,umask=0x8L2 demand store RFOs in M statemem_inst_retired.latency_above_threshold_1024event=0xb,period=3,umask=0x10,ldlat=0x8000Offcore data reads satisfied by the LLCevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1011Offcore code reads satisfied by any cache or DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8044offcore_response.any_rfo.local_cache_dramoffcore_response.any_rfo.remote_cache_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x108event=0xb7,period=100000,umask=0x1,offcore_rsp=0x808event=0xb7,period=100000,umask=0x1,offcore_rsp=0x177offcore_response.demand_data.remote_cacheOffcore demand code reads satisfied by any cache or DRAMoffcore_response.demand_ifetch.io_csr_mmioOffcore other requests satisfied by any cache or DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x210Offcore prefetch code reads satisfied by the IO, CSR, MMIO unitOffcore prefetch code reads satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x740Offcore prefetch code reads satisfied by a remote cacheOffcore prefetch RFO requests satisfied by a remote cache or remote DRAMoffcore_response.prefetch.llc_hit_no_other_coresimd_int_64.shuffle_moveoffcore_response.any_data.any_dramOffcore code reads satisfied by any DRAMOffcore code or data read requests satisfied by a remote DRAMOffcore demand code reads that missed the LLCOffcore demand code reads satisfied by a remote DRAMOffcore prefetch requests satisfied by the local DRAML1I instruction fetch missesevent=0xf6,period=2000000,umask=0x1Late Branch Prediction Unit clearsevent=0x88,period=20000,umask=0x10event=0x89,period=20000,umask=0x7event=0x89,period=2000,umask=0x8Reference base clock (133 Mhz) cycles when thread is not halted (programmable counter)resource_stalls.otherevent=0xc7,period=200000,umask=0x4event=0xc7,period=200000,umask=0x10event=0xb1,any=1,cmask=1,period=2000000,umask=0x1fCycles no Uops issued on ports 0, 1 or 5Uops issued on ports 2, 3 or 4event=0xe,period=2000000,umask=0x2dtlb_misses.stlb_hitBACLEAR asserted, regardless of cause Number of times a request needed a FB entry but there was no entry available for it. That is the FB unavailability was dominant reason for blocking the request. A request includes cacheable/uncacheable demands that is load, store or SW prefetchRetired load instructions with L1 cache hits as data sources  Supports address when precise (Precise event)offcore_response.demand_code_rd.supplier_none.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080001offcore_response.demand_data_rd.l3_hit_e.snoop_not_neededoffcore_response.demand_rfo.l3_hit_e.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40080002offcore_response.demand_rfo.l3_hit_m.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200088000offcore_response.other.l3_hit_s.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200028000frontend_retired.latency_ge_128Counts the number of uops not delivered to Resource Allocation Table (RAT) per thread adding 4  x when Resource Allocation Table (RAT) is not stalled and Instruction Decode Queue (IDQ) delivers x uops to Resource Allocation Table (RAT) (where x belongs to {0,1,2,3}). Counting does not cover cases when: a. IDQ-Resource Allocation Table (RAT) pipe serves the other thread. b. Resource Allocation Table (RAT) is stalled for the thread (including uop drops and clear BE conditions).  c. Instruction Decode Queue (IDQ) delivers four uopsCycles with less than 3 uops delivered by the front-endoffcore_requests_outstanding.cycles_with_l3_miss_demand_data_rdoffcore_response.demand_code_rd.l3_miss.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC4000004offcore_response.demand_data_rd.l3_hit_s.snoop_non_dramoffcore_response.demand_rfo.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204008000event=0x14,cmask=1,period=2000003,umask=0x1Number of cycles using an always true condition applied to  PEBS instructions retired event. (inst_ret< 16)  Spec update: SKL091, SKL044 (Must be precise)Increments when an entry is added to the Last Branch Record (LBR) array (or removed from the array in case of RETURNs in call stack mode). The event requires LBR enable via IA32_DEBUGCTL MSR and branch type selection via MSR_LBR_SELECTINST_RETIRED.ANY / FRONTEND_RETIRED.ANY_DSB_MISSCounts any data read (demand & prefetch) that miss L2offcore_response.any_data_rd.l2_miss.snoop_missCounts demand and DCU prefetch data read that are are outstanding, per cycle, from the time of the L2 miss to when any response is receivedInstruction fetches from IcacheNON_RETURN_IND counts the number of mispredicted near indirect JMP and near indirect CALL branch instructions retired.  This event counts the number of retired branch instructions that were mispredicted by the processor, categorized by type. A branch misprediction occurs when the processor predicts that the branch would be taken, but it is not, or vice-versa.  When the misprediction is discovered, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct path (Precise event)TAKEN_JCC counts the number of mispredicted taken conditional branch (JCC) instructions retired.  This event counts the number of retired branch instructions that were mispredicted by the processor, categorized by type. A branch misprediction occurs when the processor predicts that the branch would be taken, but it is not, or vice-versa.  When the misprediction is discovered, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct path (Precise event)This event counts every cycle when a I-side (walks due to an instruction fetch) page walk is in progress. Page walk duration divided by number of page walks is the average duration of page-walksRetired store uops that split across a cacheline boundary. (Precise Event - PEBS) (Precise event)Counts all prefetch RFOs that hit in the LLCCounts prefetch (that bring data to L2) code reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003c0200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400080Mispredicted not taken branch instructions retired.(Precise Event - PEBS) (Precise event)REQUEST = ANY IFETCH and RESPONSE = LLC_HIT_NO_OTHER_COREevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4ffREQUEST = OTHER and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = PF_DATA_RD and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = PF_RFO and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = PF_IFETCH and RESPONSE = IO_CSR_MMIOREQUEST = PREFETCH and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = ANY IFETCH and RESPONSE = ANY_DRAM AND REMOTE_FWDREQUEST = ANY IFETCH and RESPONSE = REMOTE_DRAMREQUEST = CORE_WB and RESPONSE = ANY_LLC_MISSevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf877event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2050REQUEST = PF_DATA_RD and RESPONSE = OTHER_LOCAL_DRAMoffcore_response.pf_ifetch.other_local_dramSnoop code requestsCycles snoop code requests queueddtlb_misses.pde_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003C0001OFFCORE_RESPONSE.DEMAND_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDoffcore_response.demand_rfo.l3_hit.snoop_hit_with_fwdCounts all prefetch (that bring data to L2) RFOs that hit in the L3offcore_response.all_pf_data_rd.l3_miss.snoop_miss_or_no_fwdCounts all demand & prefetch RFOs that miss in the L3Counts all demand & prefetch RFOs that miss the L3 and clean or shared data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00004offcore_response.demand_code_rd.l3_miss_local_dram.snoop_miss_or_no_fwdoffcore_response.demand_data_rd.l3_miss.remote_hit_forwardoffcore_response.demand_rfo.l3_miss.remote_hitmCounts all demand data writes (RFOs) that miss the L3 and the data is returned from local or remote dramCounts L1 data cache hardware prefetch requests and software prefetch requests that miss the L3 and the data is returned from local or remote dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBC000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00020core_power.lvl0_turbo_license( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE) ) / ( 2 * CPU_CLK_UNHALTED.THREAD )Fraction of Core cycles where the core was running with power-delivery for license level 2 (introduced in SKX)Power_License2_UtilizationCounts DRAM Page Activate commands sent on this channel due to a write request to the iMC (Memory Controller).  Activate commands are issued to open up a page on the DRAM devices so that it can be read or written to with a CAS (Column Access Select) commandevent=0x35,umask=0x21,config1=0x41a33event=0x2,umask=0xfunc_iio_data_req_of_cpu.mem_read.part3Counts when a normal (Non-Isochronous) full line write is issued from the CHA to the any of the memory controller channelsRsp*WB Snoop Responses Received. Unit: uncore_cha event=0xd5,fc_mask=0x04,umask=0x0fPCIe Completion Buffer occupancy of completions with data: Part 1PCIe Completion Buffer occupancy of completions with data: Part 2. Unit: uncore_iio Peer to peer read request for 4 bytes made by a different IIO unit to IIO Part3. Unit: uncore_iio unc_iio_data_req_of_cpu.peer_write.part0Counts every peer to peer read request for up to a 64 byte transaction of data made by a different IIO unit to the MMIO space of a card on IIO Part0. Does not include requests made by the same IIO unit. In the general case, part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busWrite request of up to a 64 byte transaction is made by IIO Part1 to Memory. Unit: uncore_iio Counts every peer to peer write request of up to a 64 byte transaction of data made by IIO Part3 to the MMIO space of an IIO target. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busCounts when the M2M (Mesh to Memory) updates the multi-socket cacheline Directory to a new stateevent=0x3,umask=0x27Idle FLITs transmitted. Unit: uncore_upi event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400040491ocr.all_pf_data_rd.l3_hit.hitm_other_coreOCR.ALL_PF_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDED  OCR.ALL_PF_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDocr.all_pf_rfo.l3_hit_e.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080120OCR.ALL_PF_RFO.L3_HIT_E.SNOOP_MISSocr.all_pf_rfo.l3_hit_s.no_snoop_neededOCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10000807F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800807F7ocr.all_reads.l3_hit_f.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4002007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000407F7OCR.ALL_READS.L3_HIT_M.HIT_OTHER_CORE_NO_FWD  OCR.ALL_READS.L3_HIT_M.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x801007F7ocr.all_rfo.l3_hit.snoop_missocr.demand_code_rd.l3_hit_f.snoop_noneCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040001ocr.demand_rfo.l3_hit.hit_other_core_no_fwdocr.demand_rfo.l3_hit.snoop_missocr.demand_rfo.l3_hit_e.snoop_missCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_F.HITM_OTHER_COREocr.other.l3_hit_e.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100400Counts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.PF_L2_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200010ocr.pf_l2_data_rd.l3_hit_f.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200010ocr.pf_l2_data_rd.l3_hit_m.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040010ocr.pf_l2_data_rd.l3_hit_s.hit_other_core_no_fwdCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_F.HIT_OTHER_CORE_FWDocr.pf_l2_rfo.l3_hit_m.snoop_missCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_HIT.NO_SNOOP_NEEDED OCR.PF_L3_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDocr.pf_l3_data_rd.l3_hit.snoop_hit_with_fwdoffcore_response.all_data_rd.l3_hit_e.snoop_missThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.SUPPLIER_NONE.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT.SNOOP_HIT_WITH_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_E.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_E.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.NO_SNOOP_NEEDEDoffcore_response.all_reads.l3_hit_f.snoop_noneThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT.ANY_SNOOPoffcore_response.all_rfo.l3_hit_m.any_snoopThis event is deprecated. Refer to new event OCR.ALL_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDoffcore_response.other.l3_hit.no_snoop_neededThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.OTHER.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDoffcore_response.pf_l1d_and_sw.l3_hit_m.snoop_noneoffcore_response.pf_l2_data_rd.l3_hit_f.hit_other_core_fwdoffcore_response.pf_l2_data_rd.l3_hit_s.snoop_noneoffcore_response.pf_l2_rfo.l3_hit_e.hitm_other_coreoffcore_response.pf_l2_rfo.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_F.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_M.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020080This event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_S.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_S.SNOOP_MISSevent=0xcf,period=2000003,umask=0x80event=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000490OCR.ALL_PF_DATA_RD.L3_MISS.SNOOP_MISS OCR.ALL_PF_DATA_RD.L3_MISS.SNOOP_MISSOCR.ALL_PF_DATA_RD.L3_MISS.SNOOP_NONE OCR.ALL_PF_DATA_RD.L3_MISS.SNOOP_NONEOCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDocr.all_pf_data_rd.l3_miss_local_dram.snoop_noneocr.all_pf_data_rd.l3_miss_remote_hop1_dram.no_snoop_neededOCR.ALL_PF_RFO.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_PF_RFO.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_PF_RFO.L3_MISS.NO_SNOOP_NEEDEDOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOP  OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPocr.all_pf_rfo.l3_miss_remote_hop1_dram.snoop_missOCR.ALL_READS.L3_MISS.SNOOP_MISS OCR.ALL_READS.L3_MISS.SNOOP_MISSocr.all_reads.l3_miss.snoop_noneocr.all_reads.l3_miss_local_dram.snoop_missOCR.ALL_READS.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWD OCR.ALL_READS.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1100007F7OCR.ALL_RFO.L3_MISS.REMOTE_HIT_FORWARD OCR.ALL_RFO.L3_MISS.REMOTE_HIT_FORWARDCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.ANY_SNOOP OCR.DEMAND_DATA_RD.L3_MISS.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000001Counts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000002Counts any other requests OCR.OTHER.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.OTHER.L3_MISS.HIT_OTHER_CORE_NO_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDocr.pf_l2_data_rd.l3_miss_local_dram.no_snoop_neededocr.pf_l2_data_rd.l3_miss_local_dram.snoop_missCounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.REMOTE_HIT_FORWARDCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDocr.pf_l3_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdocr.pf_l3_data_rd.l3_miss_remote_hop1_dram.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000100offcore_response.all_data_rd.l3_miss_remote_hop1_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDoffcore_response.all_pf_rfo.l3_miss_local_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDoffcore_response.all_reads.l3_miss_remote_hop1_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEoffcore_response.all_rfo.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREoffcore_response.demand_data_rd.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.SNOOP_MISSoffcore_response.demand_rfo.l3_miss_remote_hop1_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS.SNOOP_MISSoffcore_response.other.l3_miss_local_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_LOCAL_DRAM.SNOOP_MISSoffcore_response.pf_l1d_and_sw.l3_miss_remote_hop1_dram.snoop_missoffcore_response.pf_l2_rfo.l3_miss_remote_hop1_dram.hitm_other_coreoffcore_response.pf_l2_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.pf_l3_data_rd.l3_miss.hit_other_core_fwdoffcore_response.pf_l3_data_rd.l3_miss.hit_other_core_no_fwdoffcore_response.pf_l3_data_rd.l3_miss_remote_hop1_dram.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.all_pf_data_rd.supplier_none.snoop_noneOCR.ALL_READS.SUPPLIER_NONE.NO_SNOOP_NEEDED  OCR.ALL_READS.SUPPLIER_NONE.NO_SNOOP_NEEDEDocr.pf_l2_data_rd.pmm_hit_local_pmm.snoop_noneocr.pf_l3_data_rd.supplier_none.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.PMM_HIT_LOCAL_PMM.SNOOP_NONECounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_FWDevent=0xe4,umask=0x1Counts hardware prefetch data reads (which bring data to L2)  that hit a cacheline in the L3 where a snoop hit in another cores caches, data forwarding is required as the data is modifiedCounts all microcode Floating Point assistsCounts the number of uops not delivered to by the Instruction Decode Queue (IDQ) to the back-end of the pipeline when there was no back-end stalls. This event counts for one SMT thread in a given cycleFraction of Physical Core issue-slots utilized by this Logical ProcessorFraction of branches of other types (not individually covered by other metrics in Info.Branches group)Counts demand instruction fetches and L1 instruction cache prefetches that was not supplied by the L3 cacheocr.streaming_wr.l3_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00800Speculatively counts the number of TSX aborts due to a data capacity limitation for transactional writesCounts mispredicted conditional branch instructions retired (Precise event)event=0xa8,cmask=5,period=2000003,umask=0x1Counts the number of machine clears (nukes) of any typeuops_dispatched.port_1Counts cycles when at least 1 micro-op is executed from any thread on physical coreCounts the number of cycles using always true condition (uops_ret &amp;lt; 16) applied to non PEBS uops retired eventCounts the number of X86 instructions retired - an Architectural PerfMon event. Counting continues during hardware interrupts, traps, and inside interrupt handlers. Notes: INST_RETIRED.ANY is counted by a designated fixed counter freeing up programmable counters to count other events. INST_RETIRED.ANY_P is counted by a programmable counter (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80082380ocr.reads_to_core.l3_miss_local_socketCycles where the core is waiting on at least 6 outstanding demand data read requests known to have missed the L3 cacheevent=0xef,period=1000003,umask=0x4Counts responses to snoops indicating the line was kept on this core in the (S)hared state, and that the data was found unmodified but not forwarded back to the requestor, initially the data was found in the cache in the (FSE) Forward, Shared state or Exclusive state.  A single snoop response from the core counts on all hyperthreads of the coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x730000001Counts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were supplied by PMM attached to another socketunc_m_tagchk.nm_rd_hitevent=0xd3,umask=0x08event=0x2,umask=0x08PMM Write Queue Inserts. Unit: uncore_imc TOR Inserts : RFO_Prefs issued by iA Cores that Hit the LLC. Unit: uncore_cha TOR Inserts : ItoMs issued by IO Devices. Unit: uncore_cha TOR Occupancy : RFOs issued by iA Cores. Unit: uncore_cha event=0x35,umask=0xCD43FD04unc_cha_tor_inserts.ia_miss_drd_pmmunc_iio_data_req_of_cpu.mem_read.part5unc_iio_data_req_of_cpu.mem_read.part7event=0xc1,ch_mask=0x20,fc_mask=0x07,umask=0x04unc_iio_txn_req_by_cpu.mem_read.part7unc_iio_txn_req_of_cpu.mem_read.part4unc_iio_txn_req_of_cpu.cmpd.part5event=0xc2,ch_mask=0xff,fc_mask=0x04,umask=0x03Counts the number of cycles the core is stalled due to an instruction cache or TLB miss which hit in the L2, LLC, DRAM or MMIO (Non-DRAM)Counts the total number of store uops retired  Supports address when precise (Precise event)Counts all code reads that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedocr.demand_data_and_l1pf_rd.l3_hit.snoop_hit_with_fwdThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_HIT.SNOOP_NOT_NEEDEDCounts streaming stores which modify a full 64 byte cacheline that were supplied by the L3 cacheCounts L2 cache hardware prefetch code reads (written to the L2 cache only) that were supplied by the L3 cacheocr.hwpf_l2_code_rd.l3_hit.snoop_hitmCounts L2 cache hardware prefetch data reads (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedCounts the number of issue slots every cycle that were not delivered by the frontend due to instruction cache missesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000044Counts streaming stores which modify a full 64 byte cacheline that were not supplied by the L3 cacheocr.hwpf_l2_code_rd.l3_miss_localevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x402184000000Counts uncached memory writes that were not supplied by the L3 cacheThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.OUTSTANDINGCounts uncached memory reads that were supplied by DRAMCounts the number of core cycles while the core is not in a halt state.  The core enters the halt state when it is running the HLT instruction. The core frequency may change from time to time. For this reason this event may have a changing ratio with regards to time. This event uses fixed counter 1Counts the total number of issue slots that were not consumed by the backend because allocation is stalled due to a mispredicted jump or a machine clear. Only issue slots wasted due to fast nukes such as memory ordering nukes are counted. Other nukes are not accounted for. Counts all issue slots blocked during this recovery window including relevant microcode flows and while uops are not yet available in the instruction queue (IQ) even if an FE_bound event occurs during this period. Also includes the issue slots that were consumed by the backend but were thrown away because they were younger than the mispredict or machine clearevent=0x73,period=1000003,umask=0x2Counts the total number of DRAM CAS commands issued on this channelPCIe Completion Buffer Inserts of completions with data : Part 2 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 3PCIe Completion Buffer Occupancy : Part 5 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 5Number Transactions requested of the CPU : Card reading from DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 1Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 1Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Misc Events - Set 1 : Lost Forward : Snoop pulled away ownership before a write was committedCounts the number of page walks completed due to instruction fetches whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 4K pages.  Includes page walks that page faultTotal pipeline cost of branch related instructions (used for program control-flow including function calls). Unit: cpu_core Cycles Per Instruction (per Logical Processor). Unit: cpu_core Instructions Per Cycle across hyper-threads (per physical core). Unit: cpu_core Memory-Level-Parallelism (average number of L1 miss demand load when there is at least one such miss. Per-Logical Processor). Unit: cpu_core L2 cache misses per kilo instruction for all demand loads  (including speculative). Unit: cpu_core Cycles Per Instruction for the Operating System (OS) Kernel mode. Unit: cpu_core Percent of instruction miss cost that hit in the L3. Unit: cpu_atom load ops retired per 1000 instruction. Unit: cpu_atom All retired load instructions  Supports address when precise (Precise event). Unit: cpu_core Cycles MITE is delivering any Uop. Unit: cpu_core Execution stalls while L3 cache miss demand load is outstanding. Unit: cpu_core Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that have any type of response. Unit: cpu_atom This event is deprecated. Refer to new event BR_INST_RETIRED.NEAR_CALL (Precise event). Unit: cpu_atom Counts the number of issue slots not consumed due to a micro-sequencer (MS) scoreboard, which stalls the front-end from issuing uops from the UROM until a specified older uop retires. Unit: cpu_atom arith.fp_divider_activeevent=0xe7,period=1000003,umask=0xacint_vec_retired.add_256event=0x3,period=100003,umask=0x82event=0xb2,period=2000003,umask=0x80Any Rank at Hot state. Unit: uncore_imc uncore_clockevent=0x11,period=100003,umask=0x4OFFCORE_REQUESTS_OUTSTANDING.DATA_RDfrontend_retired.ms_flowsCounts the cycles where the AMX (Advance Matrix Extension) unit is busy performing an operationevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x10070XQ.FULL_CYCLESCounts miss-predicted near indirect branch instructions retired excluding returns. TSX abort is an indirect branch (Precise event)MISC2_RETIRED.LFENCEThis event is deprecated. Refer to new event UOPS_RETIRED.STALLSPrecharge due to read on page miss. Unit: uncore_imc event=0x35,umask=0x00c8f3ff04Valid Flits Received : Idle. Unit: uncore_upi unc_upi_rxl_inserts.slot0RxQ Flit Buffer Bypassed : Slot 0. Unit: uncore_upi event=0x31,umask=0x0000000004event=0x40event=0x83,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000004event=0xc1,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000001event=0x16,umask=0x07Tracker Inserts : Channel 0. Unit: uncore_m2m TOR Inserts; Hits from local IO. Unit: uncore_cha unc_cha_tor_inserts.match_opcTOR Inserts : Just ISOC. Unit: uncore_cha TOR Inserts; DRd Opt misses from local IA. Unit: uncore_cha unc_cha_tor_inserts.ia_crd_prefunc_cha_tor_inserts.ia_itomunc_cha_tor_occupancy.ia_hit_rfoevent=0x36,umask=0x00c001fe04TOR Occupancy : Match the PreMorphed Opcode in b[29:19] of the extended umask field. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_crd_prefunc_cha_tor_inserts.ia_miss_drdpteevent=0x35,umask=0x00ccd7fd01TOR Inserts; LLCPrefData from local IA. Unit: uncore_cha TOR Inserts : WCiLFs issued by iA Cores targeting DDR that missed the LLC - HOMed remotely. Unit: uncore_cha TOR Inserts : WCiLs issued by iA Cores targeting PMM that missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00c8170601unc_cha_tor_occupancy.ia_miss_remote_wcilf_pmmunc_cha_tor_occupancy.ia_miss_wcilf_ddrevent=0x36,umask=0x00c8670601unc_cha_tor_occupancy.ia_miss_local_wcil_ddrbp_l1_tlb_fetch_hitevent=0x60,umask=0x01LS to L2 WCB zero byte store requests. LS (Load/Store unit) to L2 WCB (Write Combining Buffer) zero byte store requestsevent=0x64,umask=0x07Caching: L3 cache accesses. Unit: amd_l3 event=0xd1The number of operations (uOps) and dual-pipe uOps dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one- cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number multi-pipe uOps assigned to pipe 0fp_retx87_fp_ops.add_sub_opsfp_num_mov_elim_scal_op.opt_potentialfp_retired_ser_ops.x87_bot_retBus lock when a locked operations crosses a cache boundary or is done on an uncacheable memory typels_dispatch.ld_st_dispatchL1 DTLB Miss of a page of 1G sizeevent=0x45,umask=0x40event=0x45,umask=0x10event=0x28a,umask=0x01de_dis_dispatch_token_stalls0.alsq3_0_token_stalll2_cache_accesses_from_l2_hwpfl2_cache_misses_from_ic_missl2_cache_req_stat.ic_dc_hit_in_l2 + l2_pf_hit_l2event=0x25,umask=0x08event=0x25,umask=0x02Software Prefetch Data Cache Fills by Data Source. From another cache (home node local)Cycles where a dispatch group is valid but does not get dispatched due to a token stall. Integer Physical Register File resource stall. Applies to all ops that have an integer destination registerThe number of instruction fetches that hit in the L1 ITLB. L1 Instrcution TLB hit (4K or 16K page size)op_cache_hit_miss.op_cache_missAll L3 Request Types. All L3 cache Requests. Unit: amd_l3 L3 Misses by Request Type. Ignores SliceID, EnAllSlices, CoreID, EnAllCores and ThreadMask. Requires unit mask 0xFF to engage event for counting. Unit: amd_l3 Counts the number of SMIs receivedAny Data Cache Fills by Data Source. From DRAM or IO connected in different Nodel3_cache_accessesbp_l1_tlb_miss_l2_tlb_hit + bp_l1_tlb_miss_l2_tlb_missSoftware prefetch data cache fills from L3 cache or different L2 cache in the same CCXls_hw_pf_dc_fills.far_cacheL2 cache requests: data cache reads including hardware and software prefetchl2_pf_hit_l2.l1_streamevent=0x1f,umask=0x7felocal_processor_read_data_beats_cs5local_processor_read_data_beats_cs9local_processor_write_data_beats_cs2Write data beats (64 bytes) for local processor at Coherent Station (CS) 4event=0x15f,umask=0x7fflocal_processor_write_data_beats_cs11Read data beats (64 bytes) for remote processor at Coherent Station (CS) 0event=0xdf,umask=0xbfeevent=0x25f,umask=0xbferemote_processor_write_data_beats_cs10Read data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 2remote_socket_upstream_read_beats_iom1event=0x59f,umask=0x7feevent=0x5df,umask=0x7feData beats (64 bytes) at interface 0 for local socket outbound data from CPU Moderator (CCM) 0event=0x49e,umask=0x7ffData beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 1Data beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 6Data beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 5event=0x5de,umask=0xbffevent=0x59f,umask=0xbffevent=0xd1f,umask=0xf3efp_ops_retired_by_width.pack_256_uops_retiredRetired vector floating-point compare opsevent=0xa,umask=0x80fp_ops_retired_by_type.vector_shuffleevent=0xb,umask=0x60fp_pack_ops_retired.fp128_sqrtRetired 128-bit packed floating-point shuffle ops (may include instructions not necessarily thought of as including shuffles e.g. horizontal add, dot product, and certain MOV instructions)Retired 128-bit packed integer CLM opspacked_int_op_type.int128_allInstruction fetches that miss in both the L1 and L2 ITLBs (page-table walks are requested) for 4k pagesInstruction fetches that hit in the L1 ITLB for 1G pagesPipeline restarts not caused by branch mispredictsNumber of cycles dispatch is stalled for floating-point flush recoveryFraction of dispatch slots that remained unused because the other thread was selectedL2 cache misses from L1 instruction cache missesMacro-ops retiredDRAM read data for local processorremote_processor_write_data_beats_cs0 + remote_processor_write_data_beats_cs1 + remote_processor_write_data_beats_cs2 + remote_processor_write_data_beats_cs3 + remote_processor_write_data_beats_cs4 + remote_processor_write_data_beats_cs5 + remote_processor_write_data_beats_cs6 + remote_processor_write_data_beats_cs7 + remote_processor_write_data_beats_cs8 + remote_processor_write_data_beats_cs9 + remote_processor_write_data_beats_cs10 + remote_processor_write_data_beats_cs11remote_socket_upstream_dma_read_datalocal_socket_outbound_data_from_cpuhisi_sccl,l3cDC_MISSIC_FETCHINSTR_EXECUTEDCYCLES_NO_INSTRUCTIONEVENT_04HEVENT_22HEVENT_27HEVENT_3FHEVENT_CAHEVENT_D2HEVENT_E0HINSTR_RENAMEDBUS_ACCESS_STEXC_FIQEVENT_137HEVENT_14FHEVENT_159HEVENT_17EHEVENT_18EHEVENT_1B5HEVENT_1CAHEVENT_1D9HEVENT_1DAHEVENT_1E9HEVENT_1EEHEVENT_206HEVENT_225HEVENT_24DHEVENT_261HEVENT_26BHEVENT_27AHEVENT_34DHEVENT_34FHEVENT_379HEVENT_37DHEVENT_38FHEVENT_399HEVENT_3ABHEVENT_3CFHLL_CACHE_MISS_RDdn_rxreq_dvmsyncrnd_txreq_flits_totalrni_s0_wdata_beatsrni_wrt_alloccxha_reqtrk_occcxla_avg_rx_tlp_sz_dwsclkdiv2_read_backlogMFVSCR_SYNC_CYCLESCYCLES_NO_COMPLETED_INSTRSL1_DATA_CYCLES_USEDBPU_STALL_ON_LR_DEPENDENCYFPU_INSTR_COMPLETEDCYCLES_THREE_INSTR_COMPLETEDSNOOP_RETRIESSRQ_EMPTYFXU_COMPLETION_STALLBIU_MASTER_REQUESTSBIU_MASTER_RETRIESL2_HIT_CACHE_ACCESSESL2_CACHE_INSTR_ACCESSESinvscrubberread-to-write-turnaroundk8-bu-cpu-clk-unhaltedDMC620_PMU_CD2INTEL_HASWELLARMV8_CORTEX_A57LONGEST_LAT_CACHE.MISS%s, "pmcid": "0x%08x", "pid": "%d", "value": "0x%016jx"}
%s, "pid": "%d", "flags": "0x%08x", "pcomm": "%s"}
GenuineIntel-6-26GenuineIntel-6-3EAuthenticAMD-25-[0245][[:xdigit:]]PipelineActual per-core usage of the Floating Point execution units (regardless of the vector width). SMT version; use when SMT is enabled and measuring per logical CPUInstructions per FP Arithmetic Scalar Double-Precision instruction (lower number means higher occurrence rate). May undercount due to FMA double countingINST_RETIRED.ANY / ( FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE )Mem;MemoryBW;SoCcbox_0@event\=0x0@C3 residency percent per coreevent=0x48,cmask=1,period=2000003,umask=0x1event=0x48,any=1,cmask=1,period=2000003,umask=0x1Requests from L2 hardware prefetchersl2_rqsts.missL2 writebacks that access L2 cacheThis is a precise version (that is, uses PEBS) of the event that counts retired load uops which data source were hits in the nearest-level (L1) cache.
Note: Only two data-sources of L1/FB are applicable for AVX-256bit  even though the corresponding AVX load could be serviced by a deeper level in the memory hierarchy. Data source is reported for the Low-half load. This event also counts SW prefetches independent of the actual data source  Supports address when precise (Precise event)event=0xd1,period=100003,umask=0x8mem_uops_retired.all_storesRetired load uops with locked access. (Precise Event - PEBS)  Supports address when precise.  Spec update: BDM35 (Precise event)offcore_requests_outstanding.all_data_rdevent=0xb7,period=100003,umask=0x1Counts the total number when the front end is resteered, mainly when the BPU cannot provide a correct prediction and this is corrected by other branch handling mechanisms at the front endevent=0x79,cmask=1,period=2000003,umask=0x30Cycles when uops initiated by Decode Stream Buffer (DSB) are being delivered to Instruction Decode Queue (IDQ) while Microcode Sequenser (MS) is busyThis event counts the number of uops initiated by MITE and delivered to Instruction Decode Queue (IDQ) while the Microcode Sequenser (MS) is busy. Counting includes uops that may bypass the IDQrtm_retired.abortedThis event counts both taken and not taken speculative and retired direct near callsevent=0x88,period=200003,umask=0xc4Taken speculative and retired indirect callsbr_inst_retired.near_takenbr_misp_exec.all_conditionalReference cycles when the core is not in halt stateevent=0,any=1,period=2000003,umask=0x2Counts number of cycles the CPU has at least one pending  demand load request missing the L1 data cacheevent=0xd,period=2000003,umask=0x8event=0xa1,period=2000003,umask=0x1event=0xa1,period=2000003,umask=0x4This event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 3This event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 7Counts the number of uops to be executed per-thread each cycleCycles per core when uops are dispatched to port 7uops_retired.retire_slotsPCIe write references (full cache line). Derived from unc_c_tor_inserts.opcode. Unit: uncore_cbox unc_h_requests.reads_remoteunc_m_pre_count.wrDemand load Miss in all translation lookaside buffer (TLB) levels causes a page walk that completes of any page size  Spec update: BDM69This event counts load misses in all DTLB levels that cause a completed page walk (1G  page size). The page walk can end with or without a fault  Spec update: BDM69This event counts the number of cycles while PMH is busy with the page walk  Spec update: BDM69dtlb_store_misses.stlb_hit_4kNumber of ITLB page walker hits in the L3 + XSNP  Spec update: BDM69, BDM98Retired load uops with L2 cache hits as data sources  Supports address when precise.  Spec update: BDM35 (Precise event)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020240offcore_response.all_pf_rfo.supplier_none.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0200020120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x00803C0122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020122offcore_response.corewb.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0040Counts prefetch (that bring data to LLC only) code readsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000008offcore_response.demand_code_rd.l3_hit.snoop_non_dramoffcore_response.demand_data_rd.l3_miss_local_dram.snoop_hitmoffcore_response.demand_data_rd.l3_miss_local_dram.snoop_missoffcore_response.demand_rfo.l3_hit.snoop_non_dramoffcore_response.demand_rfo.l3_miss.snoop_noneoffcore_response.other.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000040offcore_response.pf_l2_rfo.l3_miss_local_dram.any_snoopoffcore_response.pf_l2_rfo.l3_miss_local_dram.snoop_hit_no_fwdoffcore_response.pf_l3_code_rd.l3_miss_local_dram.snoop_hit_no_fwdoffcore_response.pf_l3_data_rd.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0084000100This event counts both direct and indirect macro near call instructions retired (captured in ring 3) (Precise event)L3 Lookup read request that access cache and found line in E or S-state. Unit: uncore_cbox Each cycle count number of all Core outgoing valid entries. Such entry is defined as valid from it's allocation till first of IDI0 or DRS0 messages is sent out. Accounts for Coherent and non-coherent trafficNumber of Core coherent Data Read entries allocated in DirectData mode. Unit: uncore_arb Retired load uop whose Data Source was: Remote cache HITM  Supports address when precise.  Spec update: BDE70 (Precise event)event=0x40,period=2000000,umask=0x83l1d_cache.replmL1 Cacheable Data Writesl2_ld.self.demand.e_statel2_ld_ifetch.self.i_statel2_lines_out.self.anyevent=0x26,period=200000,umask=0x50L2 locked accessesevent=0x2b,period=200000,umask=0x4fevent=0x30,period=200000,umask=0x7fevent=0x30,period=200000,umask=0x51l2_rqsts.self.any.e_stateevent=0xce,period=2000000,umask=0x0Retired Streaming SIMD Extensions 2 (SSE2) vector instructionsmisalign_mem_ref.ld_bubbleevent=0x7,period=200000,umask=0x81bus_io_wait.selfbus_lock_clocks.all_agentsevent=0xc8,period=200000,umask=0x0Branch instructions decodedRetired branch instructions that were mispredicted not-takenbr_missp_type_retired.condbr_missp_type_retired.indevent=0xc0,period=2000000,umask=0x0data_tlb_misses.l0_dtlb_miss_stevent=0xd0,period=200003,umask=0x41offcore_response.any_data_rd.l2_miss.hit_other_core_no_fwdoffcore_response.any_read.l2_miss.hit_other_core_no_fwdCounts reads for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts demand cacheable data reads of full cache lines that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400000002offcore_response.full_streaming_stores.l2_miss.anyevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000800Counts data cache line reads generated by hardware L1 data cache prefetcher that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts reads for ownership (RFO) requests generated by L2 prefetcher that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache lines requests by software prefetch instructions that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts requests to the Instruction Cache (ICache) for one or more bytes in an ICache Line and that cache line is in the ICache (hit).  The event strives to count on a cache line basis, so that multiple accesses which hit in a single cache line count as one ICACHE.HIT.  Specifically, the event counts when straight line code crosses the cache line boundary, or when a branch target is to a new line, and that cache line is in the ICache. This event counts differently than Intel processors based on Silvermont microarchitectureCounts cycles that fetch is stalled due to an outstanding ITLB miss. That is, the decoder queue is able to accept bytes, but the fetch unit is unable to provide bytes due to an ITLB miss.  Note: this event is not the same as page walk cycles to retrieve an instruction translationRetired instructions of near indirect Jmp or call (Precise event capable) (Must be precise)br_inst_retired.returnevent=0xcd,period=200003,umask=0x1issue_slots_not_consumed.recoveryissue_slots_not_consumed.resource_fullCounts a load blocked from using a store forward, but did not occur because the store data was not available at the right time.  The forward might occur subsequently when the data is available (Must be precise)Counts data reads (demand & prefetch) miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data reads generated by L1 or L2 prefetchers true miss for the L2 cache with a snoop miss in the other processor moduleCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts requests to the uncore subsystem true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000010002Counts any data writes to uncacheable write combining (USWC) memory region  have any transaction responses from the uncore subsystemCounts any data writes to uncacheable write combining (USWC) memory region  miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredPage walks outstanding due to a demand data store every cycleevent=0x85,period=2000003,umask=0x8Retired load uops which data sources were hits in L3 without snoops required  Supports address when precise.  Spec update: HSD74, HSD29, HSD25, HSM26, HSM30 (Precise event)Offcore outstanding Demand code Read transactions in SQ to uncore. Set Cmask=1 to count cycles  Spec update: HSD62, HSD61, HSM63Counts all prefetch (that bring data to L2) RFOs hit in the L3Number of times an HLE execution successfully committedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00001Counts demand data reads miss the L3 and the data is returned from local dramoffcore_response.pf_l3_code_rd.l3_miss.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00200This is a non-precise version (that is, does not use PEBS) of the event that counts FP operations retired. For X87 FP operations that have no exceptions counting also includes flows that have several X87, or flows that use X87 uops in the exception handlingloads blocked by overlapping with store buffer that cannot be forwardedFalse dependencies in MOB due to partial compare on addressunc_cbo_cache_lookup.write_iDTLB demand load misses with low part of linear-to-physical address translation missedThis event counts load operations from a 4K page that miss the first DTLB level but hit the second and do not cause page walksDTLB store misses with low part of linear-to-physical address translation missedCompleted page walks due to store misses in one or more TLB levels of 2M/4M page structureevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FBFC00040Demand data read requests sent to uncoreOffcore outstanding RFO store transactions in SQ to uncore. Set Cmask=1 to count cyclesoffcore_response.all_rfo.llc_hit.any_responseCounts demand data reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts non-temporal storesoffcore_response.demand_data_rd.llc_miss.dramLLC lookup request that access cache and found line in E-state or S-stateMisses in all TLB levels that cause a page walk of any page size from demand loadsRetired load uops whose data source was local DRAM (Snoop not needed, Snoop Miss, or Snoop Hit data not forwarded)offcore_response.all_pf_data_rd.llc_hit.hitm_other_coreCounts prefetch data reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23ffc08000Counts prefetch (that bring data to LLC only) data reads that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts all demand code reads that miss the LLC  and the data returned from local dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x67f800001Counts demand data reads that miss the LLC  and the data returned from remote dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3fffc20200txl0p_power_cycles %unc_q_txl_flits_g0.non_datafreq_band3_cycles %event=0x60(UNC_P_FREQ_GE_4000MHZ_CYCLES / UNC_P_CLOCKTICKS) * 100.Counts the number of cycles that the uncore transitioned to a frequency greater than or equal to 4Ghz. Derived from unc_p_freq_band3_cycles. Unit: uncore_pcu This event counts the number of store uops retired (Precise event)event=0xac,period=2000003,umask=0x2event=0x59,cmask=1,period=2000003,umask=0x20Cycles per core when load or STA uops are dispatched to port 2Counts the number of L2 cache missesCounts Demand code reads and prefetch code read requests  that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x18004032f7Counts any request that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in E/F stateCounts any request that accounts for responses which hit its own tile's L2 with data in F stateCounts demand code reads and prefetch code reads that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateoffcore_response.demand_data_rd.l2_hit_near_tile_moffcore_response.demand_rfo.l2_hit_far_tileCounts Demand cacheable data writes that accounts for responses which hit its own tile's L2 with data in S stateCounts Full streaming stores (WC and should be programmed on PMC1) that accounts for any responseevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400080Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateCounts L1 data HW prefetches that accounts for reponses from snoop request hit with data forwarded from it Far(not in the same quadrant as the request)-other tile L2 in E/F/M state. Valid only in SNC4 Cluster modeevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0800082000offcore_response.pf_l1_data_rd.l2_hit_this_tile_eoffcore_response.pf_l2_code_rd.l2_hit_far_tile_e_foffcore_response.pf_l2_rfo.supplier_noneCounts Software Prefetches that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeoffcore_response.uc_code_reads.l2_hit_near_tileoffcore_response.uc_code_reads.l2_hit_this_tile_suops_retired.packed_simdoffcore_response.any_code_rd.ddrCounts Demand cacheable data and L1 prefetch data read requests  that accounts for responses from MCDRAM (local and far)offcore_response.any_rfo.ddrCounts Demand cacheable data write requests  that accounts for data responses from MCDRAM Far or Other tile L2 hit farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200022event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0181800400event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080800400offcore_response.demand_rfo.ddroffcore_response.partial_writes.ddr_farCounts the number of mispredicted branch instructions retired that were conditional jumps and predicted taken (Precise event)Fixed Counter: Counts the number of unhalted reference clock cyclesevent=0xca,period=200003,umask=0x4Counts any retired store that was pushed into the recycle queue for any reasonCounts the number of occurences a retired load that is a cache line split. Each split should be counted only once  Supports address when precise (Precise event)ddr bandwidth read (CPU traffic only) (MB/sec). Unit: uncore_imc Counts the number of load micro-ops retired that cause a DTLB miss  Supports address when precise (Precise event)l1d.m_evictevent=0x53,period=2000000,umask=0x1L1D hardware prefetch requests triggeredevent=0xf1,period=100000,umask=0x2All L2 prefetchesL1D writeback to L2 transactionsLongest latency cache missmem_inst_retired.latency_above_threshold_128event=0xb,period=20000,umask=0x10,ldlat=0x8event=0xb7,period=100000,umask=0x1,offcore_rsp=0x8011event=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF22offcore_response.corewb.llc_hit_no_other_coreoffcore_response.corewb.local_cacheOffcore code or data read requests satisfied by the LLCOffcore code or data read requests satisfied by the LLC or local DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1033Offcore demand data reads satisfied by the IO, CSR, MMIO unitOffcore demand data reads satisfied by a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F04event=0xb7,period=100000,umask=0x1,offcore_rsp=0x4704offcore_response.demand_ifetch.remote_cache_hitoffcore_response.other.any_locationoffcore_response.other.local_cacheOffcore other requests satisfied by a remote cache or remote DRAMoffcore_response.pf_ifetch.remote_cacheOffcore prefetch code reads satisfied by a remote cache or remote DRAMOffcore prefetch RFO requests satisfied by the LLC  and HITM in a sibling coreOffcore prefetch requests satisfied by a remote cacheevent=0xcc,period=2000000,umask=0x1Transitions from MMX to Floating Point instructionsevent=0x12,period=200000,umask=0x10SIMD integer 64 bit unpack operationsTwo Uop instructions decodedoffcore_response.any_ifetch.any_llc_missoffcore_response.any_request.any_llc_missOffcore code or data read requests satisfied by any DRAMOffcore demand data reads satisfied by the local DRAMoffcore_response.demand_data_rd.remote_dramOffcore demand RFO requests satisfied by the local DRAMoffcore_response.pf_ifetch.any_llc_missevent=0x80,period=2000000,umask=0x4event=0x13,period=2000000,umask=0x7BACLEAR asserted, regardless of causeevent=0xc3,period=20000,umask=0x2event=0xa2,period=2000000,umask=0x8ssex_uops_retired.packed_singleUops executed on ports 0-4 (core count)uops_executed.core_stall_cyclesevent=0xdb,period=2000000,umask=0x1event=0x85,period=200000,umask=0x2Counts the demand and prefetch data reads. All Core Data Reads include cacheable 'Demands' and L2 prefetchers (not L3 prefetchers). Counting also covers reads due to page walks resulted from any request typeCounts memory transactions reached the super queue including requests initiated by the core, all L3 prefetches, page walks, etc.event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC01C0004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40080004offcore_response.demand_code_rd.l3_hit_s.snoop_not_neededoffcore_response.demand_code_rd.l4_hit_local_l4.snoop_missoffcore_response.demand_code_rd.l4_hit_local_l4.spl_hitoffcore_response.demand_data_rd.l3_hit_s.spl_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC0400001offcore_response.demand_data_rd.l4_hit_local_l4.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x40020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x40400002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020002offcore_response.other.l3_hit.spl_hitRetired Instructions who experienced a critical DSB miss (Precise event)Retired Instructions who experienced Instruction L1 Cache true miss (Precise event)event=0xc6,period=100007,umask=0x1,frontend=0x400106Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 2 cycles which was not interrupted by a back-end stall (Precise event)Number of times an HLE execution aborted due to unfriendly events (such as interrupts)Counts randomly selected loads when the latency from first dispatch to completion is greater than 4 cycles  Supports address when precise (Must be precise)offcore_response.demand_code_rd.l4_hit_local_l4.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC400001offcore_response.demand_rfo.l3_miss.snoop_non_dramoffcore_response.demand_rfo.l3_miss.spl_hitoffcore_response.demand_rfo.l3_miss_local_dram.snoop_not_neededoffcore_response.other.l3_miss.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C408000arith.divider_activeThis is a precise version of BR_INST_RETIRED.ALL_BRANCHES that counts all (macro) branch instructions retired  Spec update: SKL091 (Must be precise)event=0x3c,any=1,period=25003,umask=0x1event=0xa3,cmask=20,period=2000003,umask=0x14Counts the number of instructions (EOMs) retired. Counting covers macro-fused instructions individually (that is, increments by two)  Spec update: SKL091, SKL044Counts the number of uops that the Resource Allocation Table (RAT) issues to the Reservation Station (RS)Counts cycles during which the Resource Allocation Table (RAT) does not issue any Uops to the reservation station (RS) for the current threadFraction of branches that are non-taken conditionals(BR_INST_RETIRED.NEAR_TAKEN - ( BR_INST_RETIRED.CONDITIONAL - BR_INST_RETIRED.NOT_TAKEN ) - 2 * BR_INST_RETIRED.NEAR_CALL) / BR_INST_RETIRED.ALL_BRANCHESL1D_PEND_MISS.PENDING / ( MEM_LOAD_RETIRED.L1_MISS + MEM_LOAD_RETIRED.FB_HIT )Counts 1 per cycle for each PMH that is busy with a page walk for a load. EPT page walk duration are excluded in SkylakeStores that miss the DTLB (Data TLB) and hit the STLB (2nd Level TLB)Instruction fetch requests that miss the ITLB and hit the STLBThis event counts the number of demand and prefetch transactions that the L2 XQ rejects due to a full or near full condition which likely indicates back pressure from the IDI link. The XQ may reject transactions from the L2Q (non-cacheable requests), BBS (L2 misses) and WOB (L2 write-back victims)L2 cache requests from this coreThis event counts the number of store ops retiredoffcore_response.pf_l2_code_rd.l2_miss.snoop_missThe NO_ALLOC_CYCLES.ALL event counts the number of cycles when the front-end does not provide any instructions to be allocated for any reason. This event indicates the cycles where an allocation stalls occurs, and no UOPS are allocated in that cycleTotal cycles for all the page walks. (I-side and D-side)This event counts retired load uops that hit in the last-level cache (L3) and were found in a non-modified state in a neighboring core's private cache (same package).  Since the last level cache is inclusive, hits to the L3 may require snooping the private L2 caches of any cores on the same socket that have the line.  In this case, a snoop was required, and another L2 had the line in a modified state, so the line had to be invalidated in that L2 cache and transferred to the requesting L2. (Precise Event - PEBS) (Precise event)offcore_response.all_code_rd.llc_hit.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0244Counts demand & prefetch RFOs that hit in the LLC and the snoops to sibling cores hit in either E/S state and the line is not forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003c0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0200offcore_response.any_ifetch.all_local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f22event=0xb7,period=100000,umask=0x1,offcore_rsp=0xff08offcore_response.pf_data.all_local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff10REQUEST = PF_RFO and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = PREFETCH and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = ANY_DATA read and RESPONSE = ANY_LLC_MISSoffcore_response.any_rfo.any_dram_and_remote_fwdREQUEST = ANY RFO and RESPONSE = REMOTE_DRAMREQUEST = PREFETCH and RESPONSE = ANY_LLC_MISSevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2722event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5877offcore_response.all_pf_data_rd.l3_hit.hit_other_core_no_fwdCounts prefetch RFOs that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all demand data writes (RFOs) that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0400Counts all prefetch (that bring data to LLC only) data reads that have any response typeoffcore_response.pf_l3_rfo.l3_hit.hit_other_core_no_fwdOFFCORE_RESPONSE.PF_L3_RFO.L3_HIT.SNOOP_HIT_WITH_FWDCounts all demand code reads that miss the L3 and the data is returned from remote dramCounts L1 data cache hardware prefetch requests and software prefetch requests that miss the L3 and clean or shared data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B800080Core cycles where the core was running in a manner where Turbo may be clipped to the AVX512 turbo scheduleidi_misc.wb_downgrade( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * ( FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE ) + 16 * FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE ) / CPU_CLK_UNHALTED.THREADFraction of Core cycles where the core was running with power-delivery for license level 2 (introduced in SKX). SMT version; use when SMT is enabled and measuring per logical CPULLC misses - Uncacheable reads (from cpu) . Derived from unc_cha_tor_inserts.ia_miss. Unit: uncore_cha upi_data_bandwidth_txunc_cha_hitme_hit.ex_rdsunc_cha_llc_victims.total_mevent=0x3d,umask=0x01unc_iio_data_req_of_cpu.peer_read.part2event=0x83,ch_mask=0x08,fc_mask=0x07,umask=0x08unc_iio_data_req_of_cpu.peer_write.part2unc_iio_txn_req_by_cpu.peer_read.part2event=0xc1,ch_mask=0x08,fc_mask=0x07,umask=0x08unc_iio_txn_req_by_cpu.peer_write.part3unc_iio_txn_req_of_cpu.peer_write.part2Total IRP occupancy of inbound read and write requests.  This is effectively the sum of read occupancy and write occupancyunc_m2m_directory_update.a2sunc_m2m_directory_update.i2sunc_m2m_imc_reads.allunc_upi_clockticksProtocol header and credit FLITs received from any slot. Unit: uncore_upi Cycles in which the Tx of the Intel Ultra Path Interconnect (UPI) is in L0p power mode. Unit: uncore_upi OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.ALL_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200491OCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.ALL_PF_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080490ocr.all_pf_rfo.l3_hit.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800080120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100120OCR.ALL_READS.L3_HIT.HITM_OTHER_CORE OCR.ALL_READS.L3_HIT.HITM_OTHER_CORE OCR.ALL_READS.L3_HIT.HITM_OTHER_COREocr.all_reads.l3_hit_f.hitm_other_coreOCR.ALL_READS.L3_HIT_S.HITM_OTHER_CORE  OCR.ALL_READS.L3_HIT_S.HITM_OTHER_COREocr.all_reads.l3_hit_s.no_snoop_neededOCR.ALL_RFO.L3_HIT.HITM_OTHER_CORE OCR.ALL_RFO.L3_HIT.HITM_OTHER_CORE OCR.ALL_RFO.L3_HIT.HITM_OTHER_COREOCR.ALL_RFO.L3_HIT_F.HITM_OTHER_CORE  OCR.ALL_RFO.L3_HIT_F.HITM_OTHER_COREevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040122OCR.ALL_RFO.L3_HIT_S.SNOOP_NONECounts all demand code reads  OCR.DEMAND_CODE_RD.L3_HIT_S.NO_SNOOP_NEEDEDCounts demand data reads OCR.DEMAND_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.DEMAND_DATA_RD.L3_HIT.HIT_OTHER_CORE_NO_FWDCounts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_HIT.ANY_SNOOP OCR.DEMAND_RFO.L3_HIT.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0002ocr.other.l3_hit_e.snoop_missCounts any other requests  OCR.OTHER.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200208000ocr.other.l3_hit_m.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80108000Counts L1 data cache hardware prefetch requests and software prefetch requestsCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_E.ANY_SNOOPocr.pf_l1d_and_sw.l3_hit_e.hit_other_core_no_fwdocr.pf_l1d_and_sw.l3_hit_m.snoop_missCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_HIT_S.ANY_SNOOPocr.pf_l1d_and_sw.l3_hit_s.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_HIT.HITM_OTHER_CORE OCR.PF_L2_DATA_RD.L3_HIT.HITM_OTHER_COREocr.pf_l2_rfo.l3_hit_e.hit_other_core_no_fwdocr.pf_l2_rfo.l3_hit_f.snoop_noneCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_HIT.HITM_OTHER_CORE OCR.PF_L3_DATA_RD.L3_HIT.HITM_OTHER_CORECounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100080ocr.pf_l3_rfo.l3_hit_m.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200040100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100100offcore_response.all_data_rd.l3_hit_f.any_snoopThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_S.SNOOP_MISSoffcore_response.all_pf_data_rd.l3_hit_m.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_F.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_S.HITM_OTHER_COREoffcore_response.all_pf_rfo.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_S.NO_SNOOP_NEEDEDoffcore_response.all_reads.l3_hit_f.no_snoop_neededThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_M.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004007F7offcore_response.all_rfo.pmm_hit_local_pmm.snoop_noneoffcore_response.demand_code_rd.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_F.SNOOP_MISSoffcore_response.demand_data_rd.l3_hit_m.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_rfo.l3_hit.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_E.ANY_SNOOPThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_S.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.OTHER.SUPPLIER_NONE.HITM_OTHER_COREoffcore_response.pf_l1d_and_sw.l3_hit.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_E.SNOOP_MISSoffcore_response.pf_l1d_and_sw.l3_hit_f.hitm_other_coreoffcore_response.pf_l1d_and_sw.l3_hit_m.hitm_other_coreoffcore_response.pf_l1d_and_sw.supplier_none.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_data_rd.l3_hit_f.snoop_missThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_E.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDoffcore_response.pf_l2_rfo.pmm_hit_local_pmm.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020020offcore_response.pf_l3_data_rd.l3_hit_e.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_F.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_M.HIT_OTHER_CORE_FWDoffcore_response.pf_l3_rfo.l3_hit_m.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_M.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_RFO.SUPPLIER_NONE.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.SUPPLIER_NONE.SNOOP_NONEOCR.ALL_DATA_RD.L3_MISS.ANY_SNOOP OCR.ALL_DATA_RD.L3_MISS.ANY_SNOOP OCR.ALL_DATA_RD.L3_MISS.ANY_SNOOPOCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDOCR.ALL_PF_DATA_RD.L3_MISS.REMOTE_HIT_FORWARD OCR.ALL_PF_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x840007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8100007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000122Counts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.ANY_SNOOP OCR.DEMAND_CODE_RD.L3_MISS.ANY_SNOOPCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDCounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.HITM_OTHER_CORECounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDCounts demand data reads OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.REMOTE_HIT_FORWARDocr.demand_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDocr.other.l3_miss.hit_other_core_fwdocr.other.l3_miss_remote_hop1_dram.hitm_other_coreCounts L1 data cache hardware prefetch requests and software prefetch requests  OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.pf_l2_data_rd.l3_miss.hit_other_core_no_fwdCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.L3_MISS.SNOOP_NONEocr.pf_l2_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x410000010ocr.pf_l2_rfo.l3_miss_local_dram.snoop_noneocr.pf_l3_rfo.l3_miss.hitm_other_coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000100ocr.pf_l3_rfo.l3_miss_remote_hop1_dram.any_snoopCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDoffcore_response.all_reads.l3_miss_local_dram.snoop_noneoffcore_response.all_rfo.l3_miss_remote_hop1_dram.any_snoopoffcore_response.all_rfo.l3_miss_remote_hop1_dram.hit_other_core_fwdoffcore_response.demand_code_rd.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.NO_SNOOP_NEEDEDoffcore_response.pf_l2_data_rd.l3_miss_local_dram.no_snoop_neededThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_LOCAL_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.SNOOP_MISSoffcore_response.pf_l2_rfo.l3_miss_remote_hop1_dram.any_snoopThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_rfo.l3_miss.hit_other_core_no_fwdocr.all_data_rd.pmm_hit_local_pmm.snoop_noneocr.all_data_rd.supplier_none.hit_other_core_fwdocr.all_reads.supplier_none.no_snoop_neededocr.all_rfo.pmm_hit_local_pmm.snoop_not_neededocr.demand_code_rd.supplier_none.snoop_noneCounts any other requests  OCR.OTHER.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDocr.pf_l2_rfo.pmm_hit_local_pmm.snoop_not_neededCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDocr.pf_l3_rfo.supplier_none.no_snoop_neededTOR Inserts : DRds issued by iA Cores that Missed the LLC. Unit: uncore_cha TOR Inserts : DRds issued by iA Cores that Missed the LLC : Counts the number of entries successfuly inserted into the TOR that match qualifications specified by the subevent.   Does not include addressless requests such as locks and interruptsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C0004Counts demand data reads that hit a cacheline in the L3 where a snoop was sent or notevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FC03C0010Counts hardware prefetch data reads (which bring data to L2)  that hit a cacheline in the L3 where a snoop was sentStore Read transactions pending for off-core. Highly correlatedCounts the cycles for which the thread is active and the queue waiting for responses from the uncore cannot take any more entriesCounts number of SSE/AVX computational 256-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementevent=0xc6,period=100007,umask=0x1,frontend=0x508006event=0xa3,cmask=6,period=1000003,umask=0x6event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC00400ocr.hwpf_l2_rfo.l3_missCounts the number of times an RTM execution aborted due to incompatible memory typeevent=0x54,period=100003,umask=0x4Counts hardware prefetch data reads (which bring data to L2)  that have any type of responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000020br_inst_retired.cond_takenTaken conditional branch instructions retired (Precise event)br_misp_retired.condNumber of instructions retired. Fixed Counter - architectural event (Precise event)Counts the number of instructions retired - an Architectural PerfMon event. Counting continues during hardware interrupts, traps, and inside interrupt handlers. Notes: INST_RETIRED.ANY is counted by a designated fixed counter freeing up programmable counters to count other events. INST_RETIRED.ANY_P is counted by a programmable counter (Precise event)Counts cycles after recovery from a branch misprediction or machine clear till the first uop is issued from the resteered pathCounts the number of times that split load operations are temporarily blocked because all resources for handling the split accesses are in useTBD  Supports address when precise (Precise event)SMT;TmaL1Counts demand data reads that hit a modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that hit in the L3 or were snooped from another core's caches on the same socketocr.reads_to_core.l3_hit.snoop_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0477For every cycle, increments by the number of demand data read requests pending that are known to have missed the L3 cacheevent=0xef,period=1000003,umask=0x10Counts L1 data cache prefetch requests and software prefetches (except PREFETCHW) that were supplied by DRAM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those DRAM accesses that are controlled by the close SNC Clusterevent=0xd3,umask=0x04event=0x83DRAM Activate Count : All Activates. Unit: uncore_imc PMM Write Pending Queue Occupancy. Unit: uncore_imc Local write requests that miss the SF/LLC and are sent to the CHA's home agent. Unit: uncore_cha event=0x35,umask=0xC001FD04event=0x36,umask=0xC001FF01event=0x36,umask=0xC80FFE01event=0x35,umask=0xC807FF01unc_cha_tor_inserts.ia_crdevent=0x36,umask=0xC817FF01TOR Inserts : DRds issued by iA Cores targeting PMM Mem that Missed the LLC. Unit: uncore_cha TOR Inserts : DRds issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha TOR Inserts : DRds issued by iA Cores targeting DDR Mem that Missed the LLC. Unit: uncore_cha unc_iio_txn_req_of_cpu.mem_write.part6PCIe Completion Buffer Occupancy of completions with data : Part 4. Unit: uncore_iio event=0x12,umask=0x78event=0x2,umask=0x0Fmem_bound_stalls.load_l2_hitocr.demand_data_and_l1pf_rd.l3_hit.snoop_hit_no_fwdocr.hwpf_l2_code_rd.l3_hit.snoop_hit_no_fwdbus_lock.block_cyclesbus_lock.lock_cyclesCounts all code reads that were supplied by DRAMocr.l1wb_m.any_responseCounts the total number of mispredicted branch instructions retired for all branch types (Precise event)event=0x74,period=1000003,umask=0x1topdown_be_bound.store_buffertopdown_fe_bound.decodeevent=0x71,period=1000003,umask=0x10TOR Occupancy : All requests from IO Devices : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsPCIe Completion Buffer Occupancy : Part 1 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 1PCIe Completion Buffer Occupancy : Part 3 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 3Number Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Number Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 3Number Transactions requested of the CPU : CmpD - device sending completion to CPU request : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x8 card plugged in to Lane 2/3, Or x4 card is plugged in to slot 2Number Transactions requested of the CPU : Card writing to DRAM : Also known as Inbound.  Number of 64B cache line requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 3Instruction-Level-Parallelism (average number of uops executed when there is at least 1 uop executed). Unit: cpu_core TOPDOWN_BAD_SPECULATION.ALL / (5 * CPU_CLK_UNHALTED.CORE)Instruction per (near) call (lower number means higher occurrence rate). Unit: cpu_atom CPU_CLK_UNHALTED.CORE / CPU_CLK_UNHALTED.REF_TSCTBD. Unit: cpu_core Retired store instructions that split across a cacheline boundary  Supports address when precise (Precise event). Unit: cpu_core event=0x21,period=100003,umask=0x80event=0xb3,period=2000003,umask=0x1Counts the number of instruction cache misses. Unit: cpu_atom event=0xc6,period=100007,umask=0x1,frontend=0x600206Cycles DSB is delivering optimal number of Uops. Unit: cpu_core Cycles when optimal number of uops was delivered to the back-end when the back-end is not stalled. Unit: cpu_core Counts the number of cycles that the head (oldest load) of the load buffer is stalled due to a pagewalk when load subsequently retires. Unit: cpu_atom Counts demand data reads that were not supplied by the L3 cache. Unit: cpu_atom Counts the number of issue slots every cycle that were not consumed by the backend due to a machine clear (nuke). Unit: cpu_atom Counts the total number of uops retired (Precise event). Unit: cpu_atom Cycles total of 2 uops are executed on all ports and Reservation Station was not empty. Unit: cpu_core event=0xe7,period=1000003,umask=0x10event=0xb2,period=2000003,umask=0x2unc_m_act_count_rdevent=0x12,cmask=1,period=100003,umask=0x10event=0x12,period=100003,umask=0x2Number of all Vector (also called packed) Half-Precision FP arithmetic instructions(1) retiredCounts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The decode pipeline works at a 32 Byte granularityevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x733004477CPU_CLK_UNHALTED.PAUSENumber of retired integer ADD/SUB (regular or horizontal), SAD 256-bit vector instructionsINT_VEC_RETIRED.VNNI_256Number of TMA slots that were wasted due to incorrect speculation by (any type of) branch mispredictions. This event estimates number of specualtive operations that were issued but not retired as well as the out-of-order engine recovery past a branch mispredictionDRAM underfill read CAS commands issued. Unit: uncore_imc event=0x3,umask=0x0000000088event=0xe0,umask=0x0000000008unc_m_pre_count.rd_pch0event=0xe0,umask=0x0000000020event=0x35,umask=0x00cd43ff04event=0x2,umask=0x0000000080unc_upi_rxl_flits.slot2unc_upi_rxl_occupancy.slot1Tx Flit Buffer Allocations. Unit: uncore_upi Read request for 4 bytes made by IIO Part0 to Memory. Unit: uncore_iio event=0x83,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000004event=0x84,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000001event=0xc1,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000001event=0xc1,ch_mask=0x0008,fc_mask=0x07,umask=0x0000000004event=0xc2,ch_mask=0x10,fc_mask=0x07,umask=0x0000000004Tracker Inserts : Channel 1. Unit: uncore_m2m unc_cha_tor_inserts.io_miss_rfoevent=0x35,umask=0x0000000001unc_cha_tor_inserts.prq_iosfTOR Inserts : PRQ - IOSF. Unit: uncore_cha TOR Inserts; RFO from local IO. Unit: uncore_cha TOR Occupancy; All from local IA. Unit: uncore_cha TOR Occupancy; LLCPrefRFO misses from local IA. Unit: uncore_cha TOR Occupancy; All from local IO. Unit: uncore_cha TOR Occupancy : All from Remote. Unit: uncore_cha TOR Occupancy; DRd Pref hits from local IA. Unit: uncore_cha TOR Occupancy; DRd Opt Pref misses from local IA. Unit: uncore_cha TOR Occupancy; RFO prefetch from local IA. Unit: uncore_cha unc_cha_tor_occupancy.ia_miss_drd_pref_localunc_cha_tor_occupancy.ia_miss_drd_pref_remoteevent=0x35,umask=0x00cd43fd04unc_cha_tor_inserts.ia_miss_crd_pref_localevent=0x35,umask=0x00C877DE01unc_cha_tor_inserts.ia_wcilfevent=0x35,umask=0x00C86FFE01unc_cha_tor_occupancy.ia_miss_drd_pref_remote_pmmTOR Occupancy : DRd_Prefs issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed remotely. Unit: uncore_cha TOR Inserts : DDR Access. Unit: uncore_cha l2_request_g1.rd_blk_xl2_request_g2.bus_locks_originatorevent=0x64,umask=0x20l2_cache_req_stat.ls_rd_blk_cevent=0x64,umask=0x04l2_cache_req_stat.ic_fill_missl2_pf_miss_l2_l3L2 prefetcher misses in L3. All L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 cachesevent=0x6,umask=0xfeThe number of MMX, SSE or x87 instructions retired. The UnitMask allows the selection of the individual classes of instructions as given in the table. Each increment represents one complete instruction. Since this event includes non-numeric instructions it is not suitable for measuring MFLOPS. SSE instructions (SSE, SSE2, SSE3, SSSE3, SSE4A, SSE41, SSE42, AVX)The number of MMX, SSE or x87 instructions retired. The UnitMask allows the selection of the individual classes of instructions as given in the table. Each increment represents one complete instruction. Since this event includes non-numeric instructions it is not suitable for measuring MFLOPS. x87 instructionsTagged IBS Ops. Number of Ops tagged by IBS that retiredevent=0x47,umask=0x38dram_channel_data_controller_7event=0,umask=0x10fp_retx87_fp_ops.mul_opsevent=0x3,umask=0x10Double precision add/subtract FLOPSNumber of SSE Move Ops eliminatedls_dispatch.ld_dispatchThe number of accesses to the data cache for load and store references. This may include certain microcode scratchpad accesses, although these are generally rare. Each increment represents an eight-byte access, although the instruction may only be accessing a portion of that. This event is a speculative eventls_l1_d_tlb_miss.tlb_reload_32k_l2_missTotal Page Table Walks IC Type 0Total Page Table Walks DC Type 1event=0x4b,umask=0x02L2 Cache Accesses from L2 HWPFL2 Cache Hits from L2 HWPFbp_l1_tlb_miss_l2_hit + bp_l1_tlb_miss_l2_missRetired lock instructions. Non-speculative lock succeededA non-cacheable store and the non-cacheable commit buffer is fullls_tlb_flushde_dis_dispatch_token_stalls1.int_phy_reg_file_token_stallAll Op Cache accesses. Counts Op Cache micro-tag hit/miss eventsThe number of CPUID instructions retiredls_hw_pf_dc_fills.ext_cache_remoteevent=0xab,umask=0x04de_dis_dispatch_token_stalls2.agsq_token_stallCycles where a dispatch group is valid but does not get dispatched due to a token stall. No tokens for Integer Scheduler Queue 1 availablel1_data_cache_fills_from_remote_nodeDemand data cache fills from cache of another CCX when the address was in a different NUMA nodels_any_fills_from_sys.local_l2Any data cache fills from local L2 cache or L3 cache or different L2 cache in the same CCXHardware prefetch data cache fills from all types of data sourcesCore to L2 cache requests (not including L2 prefetch) with status: data cache read hit modifiable line in L2Core to L2 cache requests (not including L2 prefetch) with status: data cache shared read hit in L2Core to L2 cache requests (not including L2 prefetch) for data cache hitsCore to L2 cache requests (not including L2 prefetch) for data and instruction cache accessL2 prefetches accepted by the L2 pipeline which hit in the L2 cache of type L2NextLine (fetch the next line into L2 cache)L2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L2NextLine (fetch the next line into L2 cache)event=0x72,umask=0x10L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L2Stride (fetch additional lines into L2 cache when each access is a constant distance from the previous)l3_xi_sampled_latency.dram_farevent=0xd6,umask=0x08event=0xd6,umask=0x10event=0xd6,umask=0x1bevent=0x1df,umask=0x7feremote_processor_read_data_beats_cs11event=0x15f,umask=0xbffremote_socket_upstream_read_beats_iom2local_socket_inf0_inbound_data_beats_ccm4event=0x41e,umask=0x7ffData beats (64 bytes) at interface 1 for local socket outbound data from CPU Moderator (CCM) 0remote_socket_inf1_inbound_data_beats_ccm2Data beats (64 bytes) at interface 1 for remote socket outbound data from CPU Moderator (CCM) 3Retired SSE and AVX floating-point ops of all typesevent=0xa,umask=0x02event=0xa,umask=0xb0event=0xb,umask=0x0esse_avx_ops_retired.sse_avx_macevent=0xb,umask=0xd0fp_pack_ops_retired.fp128_cmpevent=0xc,umask=0x0eevent=0xc,umask=0x20Retired 256-bit packed floating-point ops of other typespacked_int_op_type.int128_addpacked_int_op_type.int128_shuffleRetired 256-bit packed integer ops of all typesStore-to-load conflicts (load unable to complete due to a non-forwardable conflict with an older store)event=0xaa,umask=0x07Fraction of dispatch slots that remained unused because of backend stallsL3 misses (including cacheline state change requests)Mixed SSE/AVX stallslocal_socket_upstream_write_beats_iom0 + local_socket_upstream_write_beats_iom1 + local_socket_upstream_write_beats_iom2 + local_socket_upstream_write_beats_iom3hisi_sccl,ddrcBU_INTERNAL_L2_REQUESTPC_BRANCH_PREDMEM_ACCESSAXI_READOPS_ISSUEDEVENT_17HEVENT_1BHEVENT_32HEVENT_36HEVENT_71HEVENT_D4HSECOND_EXECUTION_UNIT_PIPEPLD_STALLDATA_MAIN_TLB_MISS_STALLEVENT_146HEVENT_15CHEVENT_185HEVENT_186HEVENT_191HEVENT_1BDHEVENT_1C3HEVENT_1C7HEVENT_1EBHEVENT_209HEVENT_22CHEVENT_240HEVENT_25DHEVENT_2C9HEVENT_2D0HEVENT_2F9HEVENT_345HEVENT_34BHEVENT_34CHEVENT_359HEVENT_39CHEVENT_3F2HEVENT_3F6HL1D_CACHE_WB_CLEANL2D_CACHE_LDrnd_rrt_allocrni_rrt_occ_ovflcxha_rddatbypcxha_snphaz_occclkdiv2_medium_qos_depthclkdiv2_rdwrclk_cycle_countSPEC_BUFFER_CYCLESBRANCH_UNIT_STALL_CYCLESDATA_BKPT_MATCHESDTLB_MISSESL3_CACHE_HITSFXU0_IDLE_FXU1_BUSYLSU_COMPLETION_STALLINSTR_FETCHEDCQ_REDIRECTSINSTR_L1_CACHE_RELOADSL2_CACHE_INSTR_ALLOCATIONSL2_CACHE_CLEAN_UPDATESL2_CACHE_LOCKSL2_CLEAN_LINE_INVALIDATIONSDAC2S_DTECTEDTSC-add-pipe-junk-opsssPOWER8BR_MISP_RETIRED.ALL_BRANCHESunrecognized kvpair: %s:%s
%s, "pmcid": "0x%08x", "pid": "%d", "tid": "%d", "cpuflags": "0x%08x", "cpuflags2": "0x%08x", "pc": [ GenuineIntel-6-2EThis category represents fraction of slots where the processor's Frontend undersupplies its Backend. Frontend denotes the first part of the processor core responsible to fetch operations that are executed later on by the Backend part. Within the Frontend; a branch predictor predicts the next address to fetch; cache-lines are fetched from the memory subsystem; parsed into instructions; and lastly decoded into micro-operations (uops). Ideally the Frontend can issue Machine_Width uops every cycle to the Backend. Frontend Bound denotes unutilized issue-slots when there is no Backend stall; i.e. bubbles where Frontend delivered no uops while Backend could have accepted them. For example; stalls due to instruction-cache misses would be categorized under Frontend Bound. SMT version; use when SMT is enabled and measuring per logical CPUCoreIPCRet;FlopsRet;Flops_SMTActual per-core usage of the Floating Point execution units (regardless of the vector width)Instructions per (near) call (lower number means higher occurrence rate)INST_RETIRED.ANY / BR_INST_RETIRED.NEAR_TAKENFraction of Uops delivered by the DSB (aka Decoded ICache; or Uop Cache)Mem;Backend;CacheMissesC7_Pkg_ResidencyThis event counts L1D data line replacements including opportunistic replacements, and replacements that require stall-for-replace or block-for-replaceevent=0xf1,period=100003,umask=0x1event=0x24,period=200003,umask=0x50event=0xd2,period=20011,umask=0x1This event counts the demand and prefetch data reads. All Core Data Reads include cacheable Demands and L2 prefetchers (not L3 prefetchers). Counting also covers reads due to page walks resulted from any request typeThis event counts both cacheable and noncachaeble code read requestssq_misc.split_lockThis event counts the number of split locks in the super queuefp_assist.x87_outputdsb2mite_switches.penalty_cyclesevent=0x9c,cmask=1,period=2000003,umask=0x1Number of times an HLE execution aborted due to none of the previous 4 categories (e.g. interrupts)This event counts the number of memory ordering Machine Clears detected. Memory Ordering Machine Clears can result from one of the following:
1. memory disambiguation,
2. external snoop, or
3. cross SMT-HW-thread snoop (stores) hitting load bufferNumber of times an RTM execution aborted due to HLE-unfriendly instructionsevent=0x5d,period=2000003,umask=0x4tx_exec.misc5cpl_cycles.ring0_transevent=0x88,period=200003,umask=0x84event=0xc4,period=400009,umask=0x1Direct and indirect near call instructions retired. (Precise Event - PEBS) (Precise event)This is a precise version (that is, uses PEBS) of the event that counts both direct and indirect macro near call instructions retired (captured in ring 3) (Precise event)event=0x89,period=200003,umask=0xffThread cycles when thread is not in halt statecycle_activity.cycles_l2_misscycle_activity.stalls_ldm_pendingCore cycles the allocator was stalled due to recovery from earlier clear event for any thread running on the physical core (e.g. misprediction or memory nuke)lsd.cycles_4_uopsThis event counts stall cycles caused by absence of eligible entries in the reservation station (RS). This may result from RS overflow, or from RS deallocation because of the RS array Write Port allocation scheme (each RS entry has two write ports instead of four. As a result, empty entries could not be used, although RS is not really full). This counts cycles that the pipeline backend blocked uop delivery from the front enduops_executed_port.port_0_coreuops_executed_port.port_1uops_executed_port.port_2_corePCU clock ticks. Use to get percentages of PCU cycles events. Unit: uncore_pcu Cycles when PMH is busy with page walks  Spec update: BDM69Store misses in all DTLB levels that cause page walks  Spec update: BDM69Code miss in all TLB levels causes a page walk that completes. (2M/4M)  Spec update: BDM69event=0x85,period=100003,umask=0x10This event counts retired load uops which data sources were L3 hit and a cross-core snoop hit in the on-pkg core cache  Supports address when precise.  Spec update: BDM100 (Precise event)This event counts retired load uops which data sources were hits in the last-level (L3) cache without snoops required  Supports address when precise.  Spec update: BDM100 (Precise event)This event counts retired load uops which data sources were hits in the mid-level (L2) cache  Supports address when precise.  Spec update: BDM35 (Precise event)offcore_response.all_data_rd.l3_hit.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010120offcore_response.all_pf_rfo.l3_hit.any_snoopoffcore_response.all_pf_rfo.supplier_none.any_snoopoffcore_response.all_rfo.l3_hit.snoop_hitmoffcore_response.all_rfo.l3_hit.snoop_not_neededCounts writebacks (modified to exclusive) have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x01003C0004offcore_response.demand_rfo.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C8000offcore_response.pf_l2_code_rd.l3_hit.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0000010080Counts all prefetch (that bring data to LLC only) data reads have any response typeevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0404000091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0204000091event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0122offcore_response.all_rfo.l3_miss_local_dram.snoop_hit_no_fwdoffcore_response.corewb.l3_miss_local_dram.snoop_missoffcore_response.demand_data_rd.l3_miss_local_dram.snoop_non_dramoffcore_response.other.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020020offcore_response.pf_l3_code_rd.l3_miss_local_dram.snoop_non_dramThis event counts both direct and indirect near call instructions retired (Precise event)uncore_ncuCounts all demand & prefetch code reads hit in the L3 and the snoops to sibling cores hit in either E/S state and the line is not forwardedoffcore_response.all_data_rd.llc_hit.hit_other_core_no_fwdCounts all requests hit in the L3offcore_response.all_reads.llc_miss.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x06040007F7Counts prefetch (that bring data to LLC only) code reads miss in the L3l2_ld.self.any.e_stateevent=0x29,period=200000,umask=0x44l2_reject_busq.self.any.i_statel2_reject_busq.self.any.m_stateevent=0x2e,period=200000,umask=0x44simd_sat_instr_retiredDecode stall due to IQ fullevent=0x80,period=200000,umask=0x2event=0xa9,cmask=1,period=2000000,umask=0x1event=0x7,period=200000,umask=0x82bus_hitm_drv.all_agentsevent=0x7a,period=200000,umask=0x20event=0x60,period=200000,umask=0x40event=0x70,period=200000,umask=0x40event=0x69,period=200000,umask=0x40Bus stalled for snoopsRetired mispredicted branch instructions (precise event) (Precise event)event=0x88,period=2000000,umask=0x2Bus cycles when core is not haltedCycles the divider is busyevent=0xc3,period=200000,umask=0x1Memory accesses that missed the DTLBevent=0xd1,period=200003,umask=0x20Counts data reads (demand & prefetch) that true miss for the L2 cache with a snoop miss in the other processor moduleCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredCounts reads for ownership (RFO) requests (demand & prefetch) that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand cacheable data reads of full cache lines that hit the L2 cacheCounts demand reads for ownership (RFO) requests generated by a write to full data cache line that miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts demand data partial reads, including data in uncacheable (UC) or uncacheable write combining (USWC) memory types that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache lines requests by software prefetch instructions that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)baclears.allMachine clears due to memory ordering issueevent=0x13,period=200003,umask=0x2event=0xc4,period=200003,umask=0xf7Counts mispredicted branch instructions retired including all branch types (Must be precise)ld_blocks.4k_aliasUops issued to the back end per cycleevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000013010Counts reads for ownership (RFO) requests (demand & prefetch) miss the L2 cache with a snoop hit in the other processor module, data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cache lines requests by software prefetch instructions miss the L2 cache with a snoop hit in the other processor module, data forwarding is requiredCounts the number of instructions that retire execution. For instructions that consist of multiple uops, this event counts the retirement of the last uop of the instruction. The counter continues counting during hardware interrupts, traps, and inside interrupt handlers.  This event uses fixed counter 0.  You cannot collect a PEBs record for this event (Must be precise)dtlb_store_misses.walk_completed_1gbIncrements the number of outstanding L1D misses every cycle. Set Cmask = 1 and Edge =1 to count occurrencesRetired load uops which data sources were L3 and cross-core snoop hits in on-pkg core cache  Supports address when precise.  Spec update: HSD29, HSD25, HSM26, HSM30 (Precise event)Retired load uops which data sources were HitM responses from shared L3  Supports address when precise.  Spec update: HSD29, HSD25, HSM26, HSM30 (Precise event)Retired load uops which data sources were data hits in L3 without snoops required  Supports address when precise.  Spec update: HSD74, HSD29, HSD25, HSM26, HSM30 (Precise event)offcore_response.all_rfo.l3_hit.hit_other_core_no_fwdoffcore_response.pf_l2_data_rd.l3_miss.any_responseCounts all near executed branches (not necessarily retired)Stall cycles due to IQ is fullCycles at least 1 micro-op is executed from any thread on physical core  Spec update: HSD30, HSM31event=0xe,any=1,cmask=1,inv=1,period=2000003,umask=0x1Number of slow LEA or similar uops allocated. Such uop has 3 sources (for example, 2 sources + immediate) regardless of whether it is a result of LEA instruction or notevent=0x22,umask=0x84unc_cbo_xsnp_response.hit_externalunc_arb_coh_trk_occupancy.allMisses in ITLB that causes a page walk of any page sizeoffcore_response.pf_llc_data_rd.llc_hit.any_responseoffcore_response.demand_data_rd.llc_miss.local_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0600400002Dirty L2 cache lines evicted by L2 prefetchevent=0x27,period=200003,umask=0xfevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3f803c0244Counts all data/code/rfo references (demand & prefetch)offcore_response.all_rfo.llc_hit.no_snoop_neededCounts all demand data readsevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10800fp_comp_ops_exe.sse_packed_doublefp_comp_ops_exe.sse_scalar_singleCycles with at least half of the requests outstanding are waiting for data return from memory controller. Account for coherent and non-coherent requests initiated by IA Cores, Processor Graphics Unit, or LLC. Unit: uncore_arb Cycles with at least half of the requests outstanding are waiting for data return from memory controller. Account for coherent and non-coherent requests initiated by IA Cores, Processor Graphics Unit, or LLCunc_cbo_cache_lookup.eCounts all data/code/rfo reads (demand & prefetch) that hit in the LLC and sibling core snoop returned a clean responseCounts prefetch (that bring data to L2) data reads that hit in the LLCCounts prefetch (that bring data to L2) data reads that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all data/code/rfo reads (demand & prefetch) that miss the LLC  the data is found in M state in remote cache and forwarded from thereLLC misses - demand and prefetch data reads - excludes LLC prefetches. Derived from unc_c_tor_inserts.miss_opcode.demand. Unit: uncore_cbox (UNC_Q_TxL0P_POWER_CYCLES / UNC_Q_CLOCKTICKS) * 100.freq_ge_4000mhz_cycles %Counts the number of load micro-ops retired that miss in L1 D cacheCounts Demand code reads and prefetch code read requests  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateCounts Demand cacheable data and L1 prefetch data read requests  that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts Demand cacheable data and L1 prefetch data read requests  that accounts for responses which hit its own tile's L2 with data in S stateCounts any Read request  that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateCounts any request that accounts for responses which hit its own tile's L2 with data in M stateoffcore_response.demand_code_rd.l2_hit_far_tileoffcore_response.demand_code_rd.l2_hit_near_tile_e_fevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000004event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000001offcore_response.demand_rfo.l2_hit_far_tile_moffcore_response.demand_rfo.l2_hit_this_tile_eCounts L1 data HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses from a snoop request hit with data forwarded from its Near-other tile's L2 in M stateoffcore_response.pf_software.any_responseoffcore_response.pf_software.l2_hit_this_tile_eevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004001000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180608000event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080208000Counts demand cacheable data and L1 prefetch data reads that accounts for responses from MCDRAM (local and far)offcore_response.partial_writes.mcdramCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses from DDR (local and far)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0180600020unc_e_rpq_insertsl2_transactions.wbl2_write.rfo.s_statemem_inst_retired.latency_above_threshold_8mem_load_retired.llc_unshared_hitOffcore L1 data cache writebacksoffcore_response.any_data.remote_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x244Offcore code reads satisfied by the LLC and HIT in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x18FFOffcore RFO requests satisfied by the IO, CSR, MMIO unitoffcore_response.corewb.any_locationoffcore_response.data_in.any_locationOffcore request = all data, response = remote cacheOffcore demand code reads satisfied by the LLC  and HITM in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8002Offcore demand RFO requests satisfied by the LLCoffcore_response.other.llc_hit_other_core_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x130event=0xb7,period=100000,umask=0x1,offcore_rsp=0x710event=0xb7,period=100000,umask=0x1,offcore_rsp=0x140Offcore prefetch RFO requests satisfied by the LLC and not found in a sibling coreOffcore prefetch RFO requests satisfied by the LLC or local DRAMsimd_int_128.packed_arithevent=0x12,period=200000,umask=0x40event=0xfd,period=200000,umask=0x1SIMD integer 64 bit shuffle/move operationsevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF811offcore_response.any_request.any_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x60FFoffcore_response.any_rfo.remote_dramoffcore_response.demand_data_rd.local_dramoffcore_response.demand_ifetch.local_dramOffcore other requests satisfied by a remote DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6010arith.mulevent=0xa7,period=2000000,umask=0x1event=0xe5,period=2000000,umask=0x1br_inst_exec.non_callsevent=0xc5,period=2000,umask=0x2event=0xa2,period=2000000,umask=0x1SIMD Vector Integer Uops retired (Precise Event)Cycles no Uops are decodedThis event is deprecated. Refer to new event L2_LINES_OUT.USELESS_HWPFmem_inst_retired.stlb_miss_storesmem_load_misc_retired.ucoffcore_response.demand_code_rd.l3_hit_e.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200400004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000080001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x100020001offcore_response.demand_rfo.l3_hit_m.spl_hitRetired instructions after front-end starvation of at least 1 cycle (Must be precise)event=0xc6,period=100007,umask=0x1,frontend=0x410006frontend_retired.latency_ge_2_bubbles_ge_3Counts randomly selected loads when the latency from first dispatch to completion is greater than 128 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)offcore_requests_outstanding.l3_miss_demand_data_rd_ge_6offcore_response.demand_code_rd.l3_miss_local_dram.spl_hitoffcore_response.demand_data_rd.l3_hit_e.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x203C400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000001rtm_retired.aborted_memtypeexe_activity.3_ports_utilexe_activity.bound_on_stores100 * ((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) * ( ( (CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (min( CPU_CLK_UNHALTED.THREAD , OFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DATA_RD ) / CPU_CLK_UNHALTED.THREAD - (min( CPU_CLK_UNHALTED.THREAD , cpu@OFFCORE_REQUESTS_OUTSTANDING.ALL_DATA_RD\,cmask\=4@ ) / CPU_CLK_UNHALTED.THREAD)) / #(CYCLE_ACTIVITY.STALLS_L3_MISS / CPU_CLK_UNHALTED.THREAD + (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD) - (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD))) ) + ( (( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) * ( (( (10 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) - (3.5 * ((CPU_CLK_UNHALTED.THREAD / CPU_CLK_UNHALTED.REF_TSC) * msr@tsc@ / 1000000000 / duration_time)) ) * MEM_LOAD_RETIRED.L3_HIT * (1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) / 2) / CPU_CLK_UNHALTED.THREAD) / #(( CYCLE_ACTIVITY.STALLS_L2_MISS - CYCLE_ACTIVITY.STALLS_L3_MISS ) / CPU_CLK_UNHALTED.THREAD) ) + ( (( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) / ( (MEM_LOAD_RETIRED.L2_HIT * ( 1 + (MEM_LOAD_RETIRED.FB_HIT / MEM_LOAD_RETIRED.L1_MISS) )) + cpu@L1D_PEND_MISS.FB_FULL\,cmask\=1@ ) ) * (( CYCLE_ACTIVITY.STALLS_L1D_MISS - CYCLE_ACTIVITY.STALLS_L2_MISS ) / CPU_CLK_UNHALTED.THREAD)) / #((( CYCLE_ACTIVITY.STALLS_MEM_ANY + EXE_ACTIVITY.BOUND_ON_STORES ) / (CYCLE_ACTIVITY.STALLS_TOTAL + (EXE_ACTIVITY.1_PORTS_UTIL + (UOPS_RETIRED.RETIRE_SLOTS / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * EXE_ACTIVITY.2_PORTS_UTIL) + EXE_ACTIVITY.BOUND_ON_STORES)) * (1 - (IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - ( UOPS_ISSUED.ANY + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) ) )Total pipeline cost of instruction fetch bandwidth related bottlenecksDSBmiss;FedCond_NTBad;BranchesCycles when at least one PMH is busy with a page walk for a load. EPT page walk duration are excluded in SkylakeCounts cycles when at least one PMH (Page Miss Handler) is busy with a page walk for a loadCounts completed page walks  (1G sizes) caused by demand data stores. This implies address translations missed in the DTLB and further levels of TLB. The page walk can end with or without a faultCounts the number of DTLB flush attempts of the thread-specific entriesCounts any data read (demand & prefetch) that miss L2 with a snoop miss responseCounts demand and DCU prefetch RFOs that hit in the other module where modified copies were found in other core's L1 cacherehabq.lockStalls due to FP assistsThis event counts all instruction fetches, not including most uncacheable
fetchesIND_CALL counts the number of near indirect CALL branch instructions retired.  Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)RETURN counts the number of mispredicted near RET branch instructions retired.  This event counts the number of retired branch instructions that were mispredicted by the processor, categorized by type. A branch misprediction occurs when the processor predicts that the branch would be taken, but it is not, or vice-versa.  When the misprediction is discovered, all the instructions executed in the wrong (speculative) path must be discarded, and the processor must start fetching from the correct path (Precise event)Counts prefetch RFOs that hit in the LLC and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresoffcore_response.demand_code_rd.llc_hit.snoop_missoffcore_response.demand_rfo.llc_hit_m.hitmREQUEST = DEMAND_RFO and RESPONSE = LLC_HIT_M and SNOOP = HITMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2380408000Counts prefetch (that bring data to LLC only) code reads that hit in the LLC and the snoops sent to sibling cores return clean responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400040Direct and indirect mispredicted near call instructions retired. (Precise Event - PEBS) (Precise event)This event counts the number of micro-ops retired. (Precise Event) (Precise event)event=0x60,period=2000000,umask=0x2REQUEST = DEMAND_DATA and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = DEMAND_IFETCH and RESPONSE = REMOTE_CACHE_HITMoffcore_response.demand_rfo.all_local_dram_and_remote_cache_hitREQUEST = DEMAND_RFO and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7f40offcore_response.prefetch.local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xf802REQUEST = PF_DATA and RESPONSE = OTHER_LOCAL_DRAMevent=0x85,period=2000000,umask=0x4event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5844Counts the number of lines that are silently dropped by L2 cache when triggered by an L2 cache fill. These lines are typically in Shared state. A non-threaded eventevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0491OFFCORE_RESPONSE.ALL_DATA_RD.L3_HIT.SNOOP_HIT_WITH_FWDCounts prefetch (that bring data to L2) data reads that hit in the L3 and sibling core snoops are not needed as either the core-valid bit is not set or the shared line is present in multiple coresCounts all prefetch (that bring data to L2) RFOs that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C0020OFFCORE_RESPONSE.PF_L2_RFO.L3_HIT.SNOOP_HIT_WITH_FWDCounts all prefetch (that bring data to LLC only) data reads that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedCounts all demand & prefetch data reads that miss the L3 and clean or shared data is transferred from remote cacheCounts prefetch (that bring data to L2) data reads that miss the L3 and the data is returned from local or remote dramCounts all prefetch (that bring data to L2) RFOs that miss the L3 and the data is returned from local or remote dramCounts all prefetch (that bring data to LLC only) data reads that miss the L3 and clean or shared data is transferred from remote cacheoffcore_response.pf_l3_rfo.l3_miss_remote_dram.snoop_miss_or_no_fwdFraction of Core cycles where the core was running with power-delivery for baseline license level 0. SMT version; use when SMT is enabled and measuring per logical CPUFraction of Core cycles where the core was running with power-delivery for license level 1Counts the number of entries in the Write Pending Queue (WPQ) at each cycle.  This can then be used to calculate both the average queue occupancy (in conjunction with the number of cycles not empty) and the average latency (in conjunction with the number of allocations).  The WPQ is used to schedule writes out to the memory controller and to track the requests.  Requests allocate into the WPQ soon after they enter the memory controller, and need credits for an entry in this buffer before being sent from the CHA to the iMC (memory controller).  They deallocate after being issued to DRAM.  Write requests themselves are able to complete (from the perspective of the rest of the system) as soon they have 'posted' to the iMC.  This is not to be confused with actually performing the write to DRAM.  Therefore, the average latency for this queue is actually not useful for deconstruction intermediate write latencies.  So, we provide filtering based on if the request has posted or not.  By using the 'not posted' filter, we can track how long writes spent in the iMC before completions were sent to the HA.  The 'posted' filter, on the other hand, provides information about how much queueing is actually happenning in the iMC for writes before they are actually issued to memory.  High average occupancies will generally coincide with high write major mode counts. Is there a filter of sorts???unc_cha_requests.reads_remote7.11E-06Bytesunc_iio_data_req_of_cpu.mem_write.part0event=0x53,umask=0x02Snoop filter capacity evictions for M-state entries. Unit: uncore_cha unc_cha_snoop_resp.rsp_wbwbPCIe Completion Buffer occupancy of completions with data: Part 0. Unit: uncore_iio event=0xc0,ch_mask=0x02,fc_mask=0x07,umask=0x04Counts every read request for 4 bytes of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part2. In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busevent=0xc0,ch_mask=0x04,fc_mask=0x07,umask=0x01Counts ever peer to peer read request for 4 bytes of data made by a different IIO unit to the MMIO space of a card on IIO Part0. Does not include requests made by the same IIO unit. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busWrite request of up to a 64 byte transaction is made to IIO Part3 by the CPU. Unit: uncore_iio unc_iio_txn_req_by_cpu.peer_write.part0unc_i_coherent_ops.pcitomRFO request issued by the IRP unit to the mesh with the intention of writing a partial cacheline to coherent memory.  RFO is a Read For Ownership command that requests ownership of the cacheline and moves data from the mesh to IRP cacheTraffic in which the M2M to iMC Bypass was not taken. Unit: uncore_m2m Counts when the M2M (Mesh to Memory) updates the multi-socket cacheline Directory state from from S (Shared) to A (SnoopAll)BL Ingress (from CMS) AllocationsBL Egress (to CMS) Occupancy; All. Unit: uncore_m2m Counts incoming FLITs (FLow control unITs) which bypassed the slot1 RxQ buffer  (Receive Queue) and passed directly across the BGF and into the Egress.  This is a latency optimization, and should generally be the common case.  If this value is less than the number of FLITs transfered, it implies that there was queueing getting onto the ring, and thus the transactions saw higher latencyRetired load instructions whose data sources was forwarded from a remote cache  Supports address when preciseocr.all_data_rd.l3_hit_m.hit_other_core_no_fwdOCR.ALL_DATA_RD.L3_HIT_M.SNOOP_MISSocr.all_data_rd.l3_hit_s.snoop_noneOCR.ALL_PF_DATA_RD.L3_HIT_F.SNOOP_MISSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80080120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200120ocr.all_pf_rfo.l3_hit_f.snoop_noneocr.all_pf_rfo.l3_hit_s.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8007C07F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000807F7OCR.ALL_READS.L3_HIT_E.SNOOP_NONEocr.all_reads.l3_hit_f.hit_other_core_fwdOCR.ALL_READS.L3_HIT_M.ANY_SNOOP  OCR.ALL_READS.L3_HIT_M.ANY_SNOOPOCR.ALL_READS.L3_HIT_S.NO_SNOOP_NEEDED  OCR.ALL_READS.L3_HIT_S.NO_SNOOP_NEEDEDocr.all_reads.l3_hit_s.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0122OCR.ALL_RFO.L3_HIT_E.NO_SNOOP_NEEDED  OCR.ALL_RFO.L3_HIT_E.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040122ocr.all_rfo.l3_hit_m.hit_other_core_fwdocr.all_rfo.l3_hit_m.snoop_missocr.demand_data_rd.l3_hit_e.no_snoop_neededocr.demand_rfo.l3_hit.hitm_other_coreocr.demand_rfo.l3_hit_f.snoop_missocr.demand_rfo.l3_hit_s.hit_other_core_no_fwdCounts any other requests  OCR.OTHER.L3_HIT_F.NO_SNOOP_NEEDEDCounts any other requests  OCR.OTHER.L3_HIT_S.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400080400ocr.pf_l1d_and_sw.l3_hit_m.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040010Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800200020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200020ocr.pf_l2_rfo.l3_hit_m.hitm_other_coreocr.pf_l2_rfo.l3_hit_m.snoop_noneocr.pf_l3_data_rd.l3_hit_e.any_snoopCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDocr.pf_l3_data_rd.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040100offcore_response.all_data_rd.l3_hit_e.no_snoop_neededoffcore_response.all_data_rd.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NONEoffcore_response.all_data_rd.supplier_none.no_snoop_neededoffcore_response.all_pf_data_rd.l3_hit_m.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100400490offcore_response.all_pf_rfo.l3_hit_s.hitm_other_coreoffcore_response.all_reads.l3_hit.snoop_hit_with_fwdThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT_F.NO_SNOOP_NEEDEDoffcore_response.all_reads.l3_hit_f.snoop_missoffcore_response.all_reads.l3_hit_s.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_E.HITM_OTHER_COREoffcore_response.demand_code_rd.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDoffcore_response.demand_data_rd.l3_hit_f.hitm_other_coreoffcore_response.demand_data_rd.pmm_hit_local_pmm.snoop_not_neededThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_M.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_RFO.SUPPLIER_NONE.SNOOP_NONEThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_E.HITM_OTHER_COREoffcore_response.other.l3_hit_e.hit_other_core_fwdThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_HIT_S.HITM_OTHER_COREoffcore_response.pf_l1d_and_sw.supplier_none.snoop_noneThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_S.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400010This event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.SNOOP_HIT_WITH_FWDThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.PF_L2_RFO.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPoffcore_response.pf_l3_data_rd.supplier_none.hitm_other_coreoffcore_response.pf_l3_rfo.l3_hit_e.no_snoop_neededoffcore_response.pf_l3_rfo.l3_hit_f.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_M.HITM_OTHER_COREoffcore_response.pf_l3_rfo.l3_hit_m.snoop_missocr.all_data_rd.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F84000491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x804000491ocr.all_data_rd.l3_miss_remote_dram.snoop_miss_or_no_fwdOCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOP  OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPOCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDED  OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x90000490ocr.all_pf_rfo.l3_miss.hitm_other_coreOCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORE  OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREOCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_READS.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2100007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000122ocr.all_rfo.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000122OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDocr.demand_code_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x810000004ocr.demand_data_rd.l3_miss.hitm_other_coreocr.demand_data_rd.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000001ocr.demand_data_rd.l3_miss_local_dram.any_snoopCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_MISS.HITM_OTHER_CORE OCR.DEMAND_RFO.L3_MISS.HITM_OTHER_COREocr.demand_rfo.l3_miss.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000002ocr.demand_rfo.l3_miss.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000002ocr.demand_rfo.l3_miss_local_dram.hit_other_core_fwdCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDCounts any other requests OCR.OTHER.L3_MISS.SNOOP_MISSocr.other.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63B808000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x84000400ocr.pf_l2_data_rd.l3_miss_local_dram.hit_other_core_fwdocr.pf_l2_rfo.l3_miss_local_dram.any_snoopCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103C000080Counts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_CORECounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.REMOTE_HITMCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.REMOTE_HIT_FORWARDocr.pf_l3_rfo.l3_miss_local_dram.no_snoop_neededocr.pf_l3_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS.REMOTE_HIT_FORWARDoffcore_response.all_reads.l3_miss_local_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.ALL_READS.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDoffcore_response.demand_data_rd.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEoffcore_response.demand_rfo.l3_miss_local_dram.hit_other_core_fwdoffcore_response.demand_rfo.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_NONEOCR.ALL_PF_DATA_RD.SUPPLIER_NONE.SNOOP_NONECounts all demand code reads  OCR.DEMAND_CODE_RD.SUPPLIER_NONE.ANY_SNOOPocr.demand_data_rd.supplier_none.hit_other_core_fwdCounts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.SUPPLIER_NONE.ANY_SNOOPCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.PMM_HIT_LOCAL_PMM.ANY_SNOOPevent=0xea,umask=0x4unc_m2m_tag_hit.nm_rd_hit_dirtyCounts demand instruction fetches and L1 instruction cache prefetches that hit a cacheline in the L3 where a snoop hit in another cores caches, data forwarding is required as the data is modifiedocr.demand_rfo.l3_hit.snoop_not_neededocr.hwpf_l2_data_rd.l3_hit.snoop_sentCounts hardware prefetch RFOs (which bring data to L2) that hit a cacheline in the L3 where a snoop hit in another cores caches, data forwarding is required as the data is modifiedCycles where at least 1 outstanding Demand RFO request is pendingRetired instructions that are fetched after an interval where the front-end delivered no uops for a period of at least 2 cycles which was not interrupted by a back-end stall (Precise event)( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.512B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.512B_PACKED_SINGLE) ) / ( 2 * CPU_CLK_UNHALTED.DISTRIBUTED )CORE_POWER.LVL0_TURBO_LICENSE / CPU_CLK_UNHALTED.DISTRIBUTEDCounts the number of times an HLE execution aborted due to HLE-unfriendly instructions and certain unfriendly events (such as AD assists etc.)Counts miscellaneous requests, such as I/O and un-cacheable accesses that was not supplied by the L3 cacheCycles where at least one demand data read request known to have missed the L3 cache is pending.  Note that this does not capture all elapsed cycles while requests are outstanding - only cycles from when the requests were known to have missed the L3 cacheCounts the number of times we entered an RTM region. Does not count nested transactionsCounts the number of times a TSX Abort was triggered due to a non-release/commit store to lockocr.demand_data_rd.local_dramCounts both direct and indirect near call instructions retired (Precise event)This event distributes cycle counts between active hyperthreads, i.e., those in C0.  A hyperthread becomes inactive when it executes the HLT or MWAIT instructions.  If all other hyperthreads are inactive (or disabled or do not exist), all counts are attributed to this hyperthread. To obtain the full count when the Core is active, sum the counts from each hyperthreadCycles where the Store Buffer was full and no loads caused an execution stallCycles without actually retired instructionsCounts end of periods where the Reservation Station (RS) was empty. Could be useful to closely sample on front-end latency issues (see the FRONTEND_RETIRED event of designated precise events)Counts, on the per-thread basis, cycles during which at least one uop is dispatched from the Reservation Station (RS) to ports 2 and 3Flops;FpArith;InsTypeCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that resulted in a snoop hit in another core's caches which forwarded the unmodified data to the requesting coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1008000477ocr.hwpf_l3.l3_missevent=0xef,period=1000003,umask=0x8event=0xb7,period=100003,umask=0x1,offcore_rsp=0x73C000001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x90002380All DRAM read CAS commands issued (including underfills). Unit: uncore_imc event=0x45,umask=0x02unc_m_wpq_inserts.pch1event=0x36,umask=0xC001FE04TOR Inserts : RFOs issued by iA Cores. Unit: uncore_cha unc_cha_tor_occupancy.ia_drdunc_cha_tor_inserts.ia_miss_drd_pref_localTOR Inserts : RFOs issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha unc_cha_tor_inserts.ia_clflushunc_cha_tor_inserts.io_itomcachenearunc_cha_tor_inserts.ia_miss_full_streaming_wrTOR Inserts : PCIRdCurs issued by IO Devices that hit the LLC. Unit: uncore_cha Four byte data request of the CPU : Card writing to DRAM. Unit: uncore_iio unc_iio_data_req_by_cpu.mem_read.part6event=0x84,ch_mask=0x80,fc_mask=0x07,umask=0x80event=0x16Cycles in L1. Unit: uncore_upi TOR Inserts : ItoMs issued by IO Devices to locally HOMed memory. Unit: uncore_cha event=0x34,period=200003,umask=0x2ocr.all_code_rd.l3_hit.snoop_missCounts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were supplied by the L3 cache where a snoop was sent, the snoop hit, but no data was forwardedocr.hwpf_l2_code_rd.l3_hit.snoop_not_neededocr.partial_streaming_wr.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1008003C0000Counts the total number of BACLEARS due to all branch types including conditional and unconditional jumps, returns, and indirect branchesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000004Counts demand instruction fetches and L1 instruction cache prefetches that were not supplied by the L3 cacheCounts L2 cache hardware prefetch data reads (written to the L2 cache only) that were not supplied by the L3 cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x102184000000This event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.LOCAL_DRAMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000000000000002Counts demand reads for ownership (RFO) and software prefetches for exclusive ownership (PREFETCHW) that have an outstanding request. Returns the number of cycles until the response is received (i.e. XQ to XQ latency)Counts streaming stores which modify only part of a 64 byte cacheline that have any type of responseCounts uncached memory reads that have any type of responseCounts the number of issue slots every cycle that were not consumed by the backend due to scoreboards from the instruction queue (IQ), jump execution unit (JEU), or microcode sequencer (MS)event=0x71,period=1000003,umask=0x72event=0x71,period=1000003,umask=0x4unc_cha_tor_inserts.ia_miss_wcilNumber Transactions requested by the CPU : Core writing to Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 1Clockticks of the IO coherency tracker (IRP)Counts the number of page walks completed due to loads (including SW prefetches) whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 2M or 4M pages. Includes page walks that page faultCounts the number of page walks completed due to load DTLB misses to a 4K pageCounts the number Extended Page Directory Pointer Entry hits.  The Extended Page Directory cache is used by Virtual Machine operating systems while the guest operating systems use the standard TLB cachesCounts the number of page walks due to an instruction fetch that miss the PDE (Page Directory Entry) cacheCounts the total number of BACLEARSBranch instructions per taken branch. Unit: cpu_core Fraction of Uops delivered by the DSB (aka Decoded ICache; or Uop Cache). Unit: cpu_core Ratio between Mispredicted branches and unknown branches. Unit: cpu_atom Counts the number of load ops retired that hit in the L3 cache (Precise event). Unit: cpu_atom mem_uops_retired.load_latency_gt_32Counts the number of cache lines replaced in L1 data cache. Unit: cpu_core Demand requests that miss L2 cache. Unit: cpu_core Retired load instructions missed L2 cache as data sources  Supports address when precise (Precise event). Unit: cpu_core event=0x2a,period=100003,umask=0x1,offcore_rsp=0x10003C0002Demand Data Read requests sent to uncore. Unit: cpu_core Number of PREFETCHW instructions executed. Unit: cpu_core event=0xb3,period=2000003,umask=0x2event=0xc6,period=100007,umask=0x1,frontend=0x608006Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 16 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core Number of machine clears due to memory ordering conflicts. Unit: cpu_core event=0xc1,period=100003,umask=0x1fevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x10002Counts the total number of branch instructions retired for all branch types (Precise event). Unit: cpu_atom event=0x73,period=1000003,umask=0x1event=0xb0,cmask=1,period=1000003,umask=0x9All branch instructions retired (Precise event). Unit: cpu_core Cycles total of 3 uops are executed on all ports and Reservation Station was not empty. Unit: cpu_core Execution stalls while memory subsystem has an outstanding load. Unit: cpu_core Loads blocked due to overlapping with a preceding store that cannot be forwarded. Unit: cpu_core Cycles Uops delivered by the LSD, but didn't come from the decoder. Unit: cpu_core event=0xb2,period=2000003,umask=0x20incoming read request page status is Page Hit. Unit: uncore_imc event=0x1dCounts every 64B read request entering the Memory Controller 1 to DRAM (sum of all channels). Unit: uncore_imc unc_mc1_wrcas_count_freerunevent=0x12,period=100003,umask=0x20event=0x2a,period=100003,umask=0x1,offcore_rsp=0x1008000002event=0x2a,period=100003,umask=0x1,offcore_rsp=0x80082380FP_ARITH_INST_RETIRED2.512B_PACKED_HALFFRONTEND_RETIRED.UNKNOWN_BRANCH (Precise event)Core clocks when the thread is in the C0.1 or C0.2 or running a PAUSE in C0 ACPI stateint_misc.mba_stallsINT_VEC_RETIRED.SHUFFLESINT_VEC_RETIRED.VNNI_128PMM Write Pending Queue inserts. Unit: uncore_imc Activate due to read, write, underfill, or bypass. Unit: uncore_imc event=0x83,ch_mask=0x0001,fc_mask=0x07,umask=0x0000000001Write request of 4 bytes made by IIO Part2 to Memory. Unit: uncore_iio unc_iio_txn_req_of_cpu.peer_write.part6event=0x50,umask=0x0000000004unc_upi_txl_flits.slot1unc_upi_txl_flits.nullValid Flits Sent : Idle. Unit: uncore_upi unc_upi_rxl_basic_hdr_match.ncsunc_upi_rxl_inserts.slot1event=0xc1,ch_mask=0x0004,fc_mask=0x07,umask=0x0000000004event=0x83,ch_mask=0x0040,fc_mask=0x07,umask=0x0000000080unc_m2m_prefcam_inserts.upi_allchevent=0x35,umask=0x00c807fd01TOR Inserts; LLCPrefRFO misses from local IA. Unit: uncore_cha unc_cha_tor_inserts.wbqevent=0x36,umask=0x00ccc7fd01event=0x36,umask=0x00cc43fe04event=0x36,umask=0x0000000040event=0x36,umask=0x00c827fe01TOR Occupancy; ItoM from local IO. Unit: uncore_cha TOR Inserts : ItoMCacheNears issued by iA Cores. Unit: uncore_cha TOR Occupancy : DRds issued by iA Cores targeting PMM Mem that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00c867ff01event=0x36,umask=0x00c8678601TOR Occupancy : WCiLFs issued by iA Cores targeting DDR that missed the LLC. Unit: uncore_cha event=0x36,umask=0x00c86e8a01bp_dyn_ind_predic_fw32_missThe number of 64 byte instruction cache line was fulfilled from the L2 cachel2_request_g1.prefetch_l2_cmdevent=0x61,umask=0x04event=0x63,umask=0x40LS to L2 WCB close requests. LS (Load/Store unit) to L2 WCB (Write Combining Buffer) close requestsevent=0x72,umask=0xffevent=0xc7Retired Indirect Branch Instructions Mispredictedevent=0,umask=0x80Total number of fp uOps on pipe 1This is a speculative event. The number of cycles in which the FPU scheduler is empty. Note that some Ops like FP loads bypass the schedulerThis is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Single-precision add/subtract FLOPSThis is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemes. Number of SSE Move Ops eliminatedfp_retired_ser_ops.x87_ctrl_retL1 DTLB Miss of a page of 4K sizels_tablewalker.ic_type1ls_tablewalker.ic_type0l2_cache_hits_from_dc_missesuops_dispatchedApproximate: Outbound data bytes for all Remote Links for a node (die)event=0x85,umask=0x04fp_disp_faults.xmm_fill_faultRetired lock instructions. Low speculative cacheable lock speculation succeededls_rdtscNumber of reads of the TSC (RDTSC instructions). The count is speculativeLS MAB Allocates by Type. Loadsls_refills_from_sys.ls_mabresp_lcl_cacheL1 DTLB Miss. DTLB reload hit a coalesced pagels_pref_instr_displs_sw_pf_dc_fill.ls_mabresp_rmt_dramde_dis_dispatch_token_stalls1.fp_misc_rsrc_stallde_dis_dispatch_token_stalls1.taken_branch_buffer_rsrc_stallevent=0xae,umask=0x10The number of times a branch used the indirect predictor to make a predictionThe number of instruction fetches that hit in the L1 ITLB. L1 Instruction TLB hit (1G page size)The number of indirect branches retired that were not correctly predicted. Each such mispredict incurs the same penalty as a mispredicted conditional branch instruction. Note that only EX mispredicts are countedDivide/square root FLOPs. This is a retire-based event. The number of retired SSE/AVX FLOPs. The number of events logged per cycle can vary from 0 to 64. This event requires the use of the MergeEvent since it can count above 15 events per cycle. See 2.1.17.3 [Large Increment per Cycle Events]. It does not provide a useful count without the use of the MergeEventls_alloc_mab_countMacro-ops RetiredRetired branch instructions mispredictedevent=0x1c8ls_dmnd_fills_from_sys.near_cachels_hw_pf_dc_fills.local_l2Core to L2 cache requests (not including L2 prefetch) with status: data cache read hit non-modifiable line in L2l2_pf_miss_l2_hit_l3.l2_stridel2_pf_miss_l2_l3.l2_streamevent=0x72,umask=0x01event=0xac,umask=0x10Retired Lock instructions which caused a bus lockDivide ops executedex_no_retire.not_completeRetired microcode opsevent=0x19f,umask=0x7feevent=0x2df,umask=0x7feWrite data beats (64 bytes) for local processor at Coherent Station (CS) 1event=0x11f,umask=0x7ffRead data beats (64 bytes) for remote processor at Coherent Station (CS) 9Write data beats (64 bytes) for remote processor at Coherent Station (CS) 1remote_processor_write_data_beats_cs8Write data beats (64 bytes) for local socket upstream DMA at IO Moderator (IOM) 3event=0x4df,umask=0x7feevent=0x51f,umask=0x7felocal_socket_inf0_outbound_data_beats_ccm5Data beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 4Data beats (64 bytes) at interface 0 for remote socket outbound data from CPU Moderator (CCM) 7event=0x45f,umask=0xbffData beats (64 bytes) for local socket outbound data from inter-socket xGMI link 7Retired scalar floating-point multiply opssse_avx_ops_retired.mmx_addsse_avx_ops_retired.mmx_logicalRetired 128-bit packed floating-point subtract opsRetired 128-bit packed floating-point blend opsevent=0xc,umask=0x30event=0xc,umask=0xe0Retired 128-bit packed integer shift opspacked_int_op_type.int256_shiftFloating-point dispatch faults of all types for SSE and AVX opsls_l1_d_tlb_miss.all_l2_missde_src_op_disp.decoderNumber of cycles dispatch is stalled for floating-point register file tokensPipelineL2;frontend_bound_group6.103515625e-5MiBdram_read_data_for_local_processorevent-two-hyphconfig=0x2cDTLB_REFILLPC_WRITEMEM_UNALIGNED_ACCESSEVENT_08HEVENT_11HEVENT_38HEVENT_53HEVENT_62HEVENT_9DHEVENT_A6HEVENT_F4HPLE_CACHE_LINE_REQ_COMPLETEDBR_INDIRECT_SPECEVENT_103HEVENT_154HEVENT_179HEVENT_189HEVENT_219HEVENT_222HEVENT_25BHEVENT_2A4HEVENT_2D5HEVENT_306HEVENT_307HEVENT_356HEVENT_369HEVENT_39AHEVENT_3C5HEVENT_3C7HEVENT_3D5HEXC_TRAP_DABORTL3D_CACHE_REFILLdn_rxreq_tlbi_dvmoprnd_txdat_flitscxha_chidat_up_stallcxra_snp_trk_occclkdiv2_enqueueclkdiv2_write_depthclkdiv2_low_qos_depthclkdiv2_t_mac_trackerREFRESHED_DSTSFPU_RENORMALIZATIONLSU_ALIAS_VS_FSQ_WB0_WB1FP_LOAD_INSTR_COMPLETED_IN_LSUITLB_NON_SPECULATIVE_MISSESGPR_RENAME_BUFFER_ENTRIES_OVER_THRESHOLDTHIRD_SPECULATIVE_BRANCH_BUFFER_RESOLVED_CORRECTLYL2_CACHE_CASTOUTSOVERFLOWUOPS_COMPLETEDLOAD_UOPS_COMPLETEDTAKEN_BRANCHES_FINISHEDCYCLES_BU_SCHED_STALLEDSTORES_COMPLETEDDATA_L1_CACHE_RELOADSDATA_MMU_BUSY_CYCLESBIU_MASTER_INSTR_SIDE_REQUESTSBIU_MASTER_DATA_SIDE_REQUESTSPMC3_OVERFLOWSTASH_HIT_DLFBosstore-pipe-junk-opsinvalidk8-fr-retired-taken-branches-mispredictedk8-ic-missINVERTIAPINTEL_NEHALEM_EXINTEL_BROADWELL_XEONbad magicex_ret_brn_mispls_int_takenumaskl3cache%s, "pid": "%d"}
UOPS_RETIRED.RETIRE_SLOTS / (4 * CPU_CLK_UNHALTED.THREAD)Uops Per InstructionFloating Point Operations Per CycleBackend;Cor;Pipeline;PortsUtilINST_RETIRED.ANY / MEM_UOPS_RETIRED.ALL_LOADSINST_RETIRED.ANY / ( 1 * ( FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE ) + 2 * FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + 4 * ( FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE ) + 8 * FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE )IDQ.DSB_UOPS / ( IDQ.DSB_UOPS + LSD.UOPS + IDQ.MITE_UOPS + IDQ.MS_UOPS )MLPAverage_FrequencyThis event counts the number of L2 cache lines in the Invalidate state filling the L2. Counting does not cover rejectsl2_rqsts.all_demand_missl2_rqsts.code_rd_missevent=0x24,period=200003,umask=0x41Demand Data Read requests that access L2 cachel2_trans.l1d_wbl2_trans.l2_wbmem_load_uops_l3_hit_retired.xsnp_missThis is a precise version (that is, uses PEBS) of the event that counts retired load uops which data sources were hits in the last-level (L3) cache without snoops required  Supports address when precise.  Spec update: BDM100 (Precise event)event=0xd1,period=50021,umask=0x4mem_uops_retired.all_loadsThis is a precise version (that is, uses PEBS) of the event that counts store uops true STLB miss retired to the architected path. True STLB miss is an uop triggering page walk that gets completed without blocks, and later gets retired. This page walk can end up with or without a fault  Supports address when precise (Precise event)offcore_requests_outstanding.cycles_with_demand_rfofp_arith_inst_retired.doubleNumber of SSE/AVX computational scalar double precision floating-point instructions retired.  Each count represents 1 computation. Applies to SSE* and AVX* scalar double precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementNumber of SIMD FP assists due to Output valuesNumber of X87 assists due to output valueevent=0xa0,period=2000003,umask=0x3event=0x79,period=2000003,umask=0x3cidq.mite_cyclesUops initiated by Decode Stream Buffer (DSB) that are being delivered to Instruction Decode Queue (IDQ) while Microcode Sequenser (MS) is busyevent=0xc9,period=2000003,umask=0x4rtm_retired.aborted_misc3event=0x5d,period=2000003,umask=0x8tx_mem.abort_capacity_writeSpeculative and retired indirect return branchesTaken speculative and retired mispredicted macro conditional branchesCounts number of cycles nothing is executed on any execution port, while there was at least one pending demand* load request missing the L2 cache.(as a footprint) * includes also L1 HW prefetch requests that may or may not be required by demandsevent=0xc0,period=2000003This event counts how many times the load operation got the true Block-on-Store blocking code preventing store forwarding. This includes cases when:
 - preceding store conflicts with the load (incomplete overlap);
 - store forwarding is impossible due to u-arch limitations;
 - preceding lock RMW operations are not forwarded;
 - store has the no-forward bit set (uncacheable/page-split/masked stores);
 - all-blocking stores are used (mostly, fences and port I/O);
and others.
The most common case is a load blocked due to its address range overlapping with a preceding smaller uncompleted store. Note: This event does not take into account cases of out-of-SW-control (for example, SbTailHit), unknown physical STA, and cases of blocking loads on store due to being non-WB memory type or a lock. These cases are covered by other events.
See the table of not supported store forwards in the Optimization Guidemachine_clears.smcmove_elimination.int_eliminatedresource_stalls.sbNumber of uops to be executed per-thread each cycleevent=0xa1,any=1,period=2000003,umask=0x4event=0xc2,period=2000003,umask=0x2llc_misses.uncacheablepower_self_refresh %dtlb_load_misses.walk_completed_1gevent=0x8,period=2000003,umask=0x8event=0x49,period=100003,umask=0x8itlb_misses.miss_causes_a_walktlb_flush.stlb_anyUNC_ARB_TRK_OCCUPANCY.ALL / arb@event\=0x81\,umask\=0x1@This event counts store uops retired to the architected path with a filter on bits 0 and 1 applied.
Note: This event counts AVX-256bit load/store double-pump memory uops as a single uop at retirement  Supports address when precise (Precise event)Retired load uops with locked access  Supports address when precise.  Spec update: BDM35 (Precise event)Any memory transaction that reached the SQoffcore_response.all_pf_code_rd.any_responseoffcore_response.all_pf_code_rd.l3_hit.snoop_noneoffcore_response.all_pf_data_rd.supplier_none.any_snoopoffcore_response.corewb.l3_hit.snoop_not_neededoffcore_response.demand_rfo.l3_hit.snoop_noneoffcore_response.other.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020010event=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020200offcore_response.pf_l3_data_rd.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0400020080Number of SSE/AVX computational 128-bit packed single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 4 calculations per elementCounts randomly selected loads with latency value being above 128  Supports address when precise.  Spec update: BDM100, BDM35 (Must be precise)offcore_response.all_data_rd.l3_miss.snoop_hit_no_fwdoffcore_response.all_pf_code_rd.l3_miss_local_dram.snoop_hit_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x00BC000090offcore_response.all_pf_data_rd.l3_miss_local_dram.snoop_not_neededoffcore_response.all_rfo.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x20003C0008offcore_response.demand_code_rd.l3_miss_local_dram.snoop_non_dramoffcore_response.demand_data_rd.l3_miss.snoop_missoffcore_response.demand_data_rd.l3_miss_local_dram.any_snoopoffcore_response.pf_l2_code_rd.l3_miss.snoop_hit_no_fwdoffcore_response.pf_l2_rfo.l3_miss.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0104000020offcore_response.pf_l3_rfo.l3_miss.snoop_hit_no_fwdoffcore_response.pf_l3_rfo.l3_miss.snoop_missL3 Lookup any request that access cache and found line in MESI-state. Unit: uncore_cbox L3 Lookup any request that access cache and found line in E or S-state. Unit: uncore_cbox Total number of Core outgoing entries allocated. Accounts for Coherent and non-coherent traffic. Unit: uncore_arb event=0x80,cmask=1,umask=0x01Counts all data/code/rfo reads (demand & prefetch) miss the L3 and the data is returned from remote dramqpi_data_bandwidth_txunc_m_clockticksevent=0x40,period=200000,umask=0x10l2_ads.selfl2_data_rqsts.self.mesievent=0x2c,period=200000,umask=0x4fl2_ifetch.self.m_stateevent=0x29,period=200000,umask=0x71event=0x2b,period=200000,umask=0x44l2_reject_busq.self.prefetch.i_statesimd_inst_retired.scalar_doublesimd_uop_type_exec.logical.sSIMD unpacked micro-ops executedevent=0x86,period=2000000,umask=0x1event=0x5,period=200000,umask=0x8fprefetch.prefetchntaprefetch.prefetcht1Number of Bus Not Ready signals assertedevent=0x61,period=200000,umask=0x0bus_hit_drv.all_agentsbus_hit_drv.this_agentevent=0x6d,period=200000,umask=0x40event=0x6c,period=200000,umask=0x40Cycles during which interrupts are disabledCycles during which interrupts are pending and disabledext_snoop.all_agents.hitmevent=0x3b,period=200000,umask=0xc0event=0x89,period=200000,umask=0x1cpu_clk_unhalted.busevent=0x3c,period=2000000,umask=0x0L0 DTLB misses due to store operationsCounts the number of demand and L1 prefetcher requests rejected by the L2Q due to a full or nearly full condition which likely indicates back pressure from L2Q. It also counts requests that would have gone directly to the XQ, but are rejected due to a full or nearly full condition, indicating back pressure from the IDI link. The L2Q may also reject transactions from a core to ensure fairness between cores, or to delay a core's dirty eviction when the address conflicts with incoming external snoopsdl1.dirty_evictionevent=0xd0,period=200003,umask=0x43Counts store uops retired where the data requested spans a 64 byte cache line boundary  Supports address when precise (Must be precise)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000043091offcore_response.any_data_rd.l2_miss.anyevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600003091Counts requests to the uncore subsystem that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is requiredoffcore_response.any_rfo.l2_miss.anyoffcore_response.demand_code_rd.outstandingCounts demand cacheable data reads of full cache lines that true miss for the L2 cache with a snoop miss in the other processor moduleoffcore_response.demand_rfo.outstandingCounts partial cache line data writes to uncacheable write combining (USWC) memory region  that miss the L2 cache with a snoop hit in the other processor module, no data forwarding is required. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600000100Counts the number of demand write requests (RFO) generated by a write to partial data cache line, including the writes to uncacheable (UC) and write through (WT), and write protected (WP) types of memory that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts data cacheline reads generated by hardware L2 cache prefetcher that miss the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.pf_l2_rfo.l2_miss.snoop_miss_or_no_snoop_neededevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x3600004800Counts data cache lines requests by software prefetch instructions that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xe6,period=200003,umask=0x10Counts machine clears due to memory ordering issues.  This occurs when a snoop request happens and the machine is uncertain if memory ordering will be preserved as another core is in the process of modifying the dataCounts when a memory store of a uop spans a page boundary (a split) is retired (Must be precise)event=0xcb,period=200003,umask=0x4Retired near return instructions (Precise event capable) (Must be precise)Retired mispredicted instructions of near indirect Jmp or near indirect call. (Precise event capable) (Must be precise)Cycles the integer divide unit is busyevent=0x3,period=200003,umask=0x10Machine clears due to memory disambiguationoffcore_response.any_data_rd.outstandingoffcore_response.any_rfo.outstandingevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000400Counts the number of writeback transactions caused by L1 or L2 cache evictions hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.pf_l2_data_rd.outstandingevent=0xc3,period=20003,umask=0x2Retired mispredicted instructions of near indirect Jmp or near indirect call (Precise event capable) (Must be precise)Machines clear due to a page faultCounts page walks completed due to demand data loads (including SW prefetches) whose address translations missed in all TLB levels and were mapped to 4K pages.  The page walks can end with or without a page faultPage walks outstanding due to an instruction fetch every cycleCounts all L2 HW prefetcher requests that hit L2Miss in mid-level (L2) cache. Excludes Unknown data-source  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)Retired load uops that miss the STLB  Supports address when precise.  Spec update: HSD29, HSM30 (Precise event)Cycles with at least 6 offcore outstanding Demand Data Read transactions in uncore queue  Spec update: HSD78, HSD62, HSD61, HSM63, HSM80offcore_response.demand_code_rd.l3_miss.any_responseNumber of times HLE lock could not be elided due to ElisionBufferAvailable being zeroevent=0x5c,cmask=1,edge=1,period=100003,umask=0x1Increments at the frequency of XCLK (100 MHz) when not haltedExecution stalls due to L2 cache misses  Spec update: HSM63, HSM80Non-SW-prefetch load dispatches that hit fill buffer allocated for H/W prefetchNumber of uops executed on the core  Spec update: HSD30, HSM31unc_cbo_cache_lookup.extsnp_esL3 Lookup external snoop request that access cache and found line in I-state. Unit: uncore_cbox L3 Lookup external snoop request that access cache and found line in I-stateAn external snoop misses in some processor core. Unit: uncore_cbox dtlb_load_misses.pde_cache_missdtlb_store_misses.pde_cache_missNumber of DTLB page walker hits in the L1+FBNumber of DTLB page walker hits in Memory  Spec update: HSD25page_walker_loads.ept_itlb_l2event=0xbd,period=100003,umask=0x1l2_l1d_wb_rqsts.missRetired load uops whose data source was local memory (cross-socket snoop not needed or missed)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0004event=0xac,period=2000003,umask=0x8Loads with latency value being above 64 (Must be precise)Cycles per thread when uops are dispatched to port 0Cycles which a Uop is dispatched on port 3Counts total number of uops to be executed per-thread each cycle. Set Cmask = 1, INV =1 to count stall cyclesCycle PMH is busy with a walkCounts all demand & prefetch code reads that miss the LLCevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x87f800244offcore_response.demand_data_rd.llc_miss.remote_hitmoffcore_response.pf_l2_data_rd.llc_miss.any_dramllc_misses.pcie_non_snoop_readunc_p_freq_band2_transitionsCounts the number of cycles that the uncore transitioned to a frequency greater than or equal to the frequency that is configured in the filter.  (filter_band3=XXX, with XXX in 100Mhz units). One can also use inversion (filter_inv=1) to track cycles when we were less than the configured frequency. Derived from unc_p_freq_band3_cycles. Unit: uncore_pcu (UNC_P_FREQ_GE_1200MHZ_CYCLES / UNC_P_CLOCKTICKS) * 100.event=0xe,filter_band3=40l2_store_lock_rqsts.hit_eRetired store uops that miss the STLB (Precise event)Each cycle there was a miss-pending demand load this thread, increment by 1. Note this is in DCU and connected to Umask 1. Miss Pending demand load should be deduced by OR-ing increment bits of DCACHE_MISS_PEND.PENDINGevent=0xa2,period=2000003,umask=0x2offcore_response.any_code_rd.outstandingCounts any Prefetch requests that accounts for responses which hit its own tile's L2 with data in M stateoffcore_response.bus_locks.l2_hit_near_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000400offcore_response.demand_code_rd.l2_hit_this_tile_sevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0002000001offcore_response.demand_data_rd.l2_hit_this_tile_sevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000080002event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1800400100Counts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateCounts Partial writes (UC or WT or WP and should be programmed on PMC1) that accounts for responses which hit its own tile's L2 with data in S stateCounts L2 code HW prefetches that accounts for any responseCounts L2 code HW prefetches that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0010000020offcore_response.pf_l2_rfo.l2_hit_this_tile_sevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0100400044offcore_response.any_data_rd.ddroffcore_response.any_data_rd.mcdramCounts any Prefetch requests that accounts for data responses from MCDRAM Far or Other tile L2 hit farCounts any request that accounts for data responses from MCDRAM Far or Other tile L2 hit faroffcore_response.demand_data_rd.ddr_farevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0080200002event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0101000040offcore_response.pf_software.ddrevent=0x3,umask=0x02l1d_wb_l2.e_stateevent=0x26,period=200000,umask=0x4event=0x26,period=200000,umask=0x80All L2 missesl2_transactions.anyAll L2 transactionsevent=0x27,period=100000,umask=0x40event=0x27,period=100000,umask=0xemem_inst_retired.latency_above_threshold_16event=0xb,period=50000,umask=0x10,ldlat=0x4mem_inst_retired.latency_above_threshold_512mem_load_retired.hit_lfbevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF11event=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF44event=0xb7,period=100000,umask=0x1,offcore_rsp=0x444offcore_response.any_ifetch.remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x7F22event=0xb7,period=100000,umask=0x1,offcore_rsp=0xFF77offcore_response.data_ifetch.local_cacheOffcore code or data read requests that HITM in a remote cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x103offcore_response.demand_data.local_cacheoffcore_response.demand_data.remote_cache_hitoffcore_response.demand_data_rd.io_csr_mmioevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x8001event=0xb7,period=100000,umask=0x1,offcore_rsp=0x1001event=0xb7,period=100000,umask=0x1,offcore_rsp=0x702Offcore prefetch data requests satisfied by the LLCOffcore prefetch data reads satisfied by a remote cache or remote DRAMOffcore prefetch code reads satisfied by the LLC and HIT in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x420offcore_response.pf_rfo.local_cache_dramoffcore_response.pf_rfo.remote_cache_hitOffcore prefetch requests satisfied by a remote cache or remote DRAMevent=0xf7,period=20000,umask=0x2simd_int_128.packsimd_int_128.packed_logicalevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x6022event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2004event=0xb7,period=100000,umask=0x1,offcore_rsp=0x2070Loads dispatched from stage 305bpu_clears.lateUnconditional call branches executedAll non call branches executedMispredicted conditional branches executedMispredicted indirect call branches executedild_stall.mruTotal cycles (Precise Event)Uops issued on ports 0, 1 or 5event=0xb1,period=2000000,umask=0x20event=0xe,period=2000000,umask=0x1uops_retired.active_cyclesdtlb_load_misses.anyitlb_miss_retiredmem_load_l3_hit_retired.xsnp_nonemem_load_retired.l1_hitCounts the number of cases when the offcore requests buffer cannot take more entries for the core. This can happen when the superqueue does not contain eligible entries, or when L1D writeback pending FIFO requests is full.Note: Writeback pending FIFO has six entriesCounts the number of offcore outstanding Code Reads transactions in the super queue every cycle. The 'Offcore outstanding' state of the transaction lasts from the L2 miss until the sending transaction completion to requestor (SQ deallocation). See the corresponding Umask under OFFCORE_REQUESTSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4001C0004offcore_response.demand_code_rd.l3_hit_e.snoop_not_neededoffcore_response.demand_code_rd.l3_hit_s.any_snoopoffcore_response.demand_data_rd.l4_hit_local_l4.spl_hitoffcore_response.demand_data_rd.supplier_none.spl_hitoffcore_response.demand_rfo.l3_hit_e.snoop_not_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x801C8000Counts once for most SIMD 256-bit packed single computational precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 8 computation operations, one for each element.  Applies to packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsCounts once for most SIMD scalar computational single precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 1 computational operation. Applies to SIMD scalar single precision floating-point instructions: ADD SUB MUL DIV MIN MAX SQRT RSQRT RCP FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0xc6,period=100007,umask=0x1,frontend=0x401006frontend_retired.latency_ge_4Instruction fetch tag lookups that miss in the instruction cache (L1I). Counts at 64-byte cache-line granularityCounts, on the per-thread basis, cycles when no uops are delivered to Resource Allocation Table (RAT). IDQ_Uops_Not_Delivered.core =4Counts randomly selected loads when the latency from first dispatch to completion is greater than 8 cycles.  Reported latency may be longer than just the memory latency  Supports address when precise (Must be precise)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x20001C0004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000040001event=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC400002Unfriendly TSX abort triggered by a vzeroupper instructionevent=0xc0,cmask=10,inv=1,period=2000003,umask=0x1Counts the number of times where store forwarding was prevented for a load operation. The most common case is a load blocked due to the address of memory access (partially) overlapping with a preceding uncompleted store. Note: See the table of not supported store forwards in the Optimization GuideCounts end of periods where the Reservation Station (RS) was empty. Could be useful to precisely locate front-end Latency Bound issuesRetdtlb_load_misses.walk_activeCounts cycles when at least one PMH (Page Miss Handler) is busy with a page walk for a storeLoads missed L1offcore_response.any_code_rd.l2_miss.hitm_other_coreCounts any rfo reads (demand & prefetch) that miss L2 and the snoops to sibling cores hit in either E/S state and the line is not forwardedCounts any rfo reads (demand & prefetch) that miss L2 with a snoop miss responseoffcore_response.pf_l2_data_rd.l2_miss.snoop_missLoads blocked due to store data not readyThis event counts the number of times that pipeline stalled due to FP operations needing assistsTAKEN_JCC counts the number of taken conditional branch (JCC) instructions retired. Branch prediction predicts the branch target and enables the processor to begin executing instructions long before the branch true execution path is known. All branches utilize the branch prediction unit (BPU) for prediction. This unit predicts the target address not only based on the EIP of the branch but also based on the execution path through which execution reached this EIP. The BPU can efficiently predict the following branch types: conditional branches, direct calls and jumps, indirect calls and jumps, returns (Precise event)Counts the number of cycles when no uops are allocated and a RATstall is assertedThis event counts when an instruction (I) page walk is completed or started.  Since a page walk implies a TLB miss, the number of TLB misses can be counted by counting the number of pagewalksRetired load uops which data sources were HitM responses from shared LLC. (Precise Event - PEBS) (Precise event)Counts prefetch data reads that hit in the LLC and the snoops sent to sibling cores return clean responseCounts prefetch (that bring data to L2) RFOs that hit in the LLC and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x17004001b3REQUEST = ANY_DATA read and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = ANY_REQUEST and RESPONSE = ALL_LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = ANY_REQUEST and RESPONSE = IO_CSR_MMIOREQUEST = ANY_REQUEST and RESPONSE = LOCAL_CACHEevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff22REQUEST = ANY RFO and RESPONSE = LLC_HIT_OTHER_CORE_HITMREQUEST = CORE_WB and RESPONSE = ANY_LOCATIONoffcore_response.corewb.local_dram_and_remote_cache_hitevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xff33event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5080event=0xb7,period=100000,umask=0x1,offcore_rsp=0x250REQUEST = PF_DATA_RD and RESPONSE = LOCAL_CACHEREQUEST = PF_DATA_RD and RESPONSE = REMOTE_CACHE_HITMoffcore_response.pf_rfo.all_local_dram_and_remote_cache_hitREQUEST = PREFETCH and RESPONSE = ANY_LOCATIONevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x40ffREQUEST = DEMAND_DATA and RESPONSE = OTHER_LOCAL_DRAMREQUEST = DEMAND_RFO and RESPONSE = REMOTE_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4080event=0xb4,period=100000,umask=0x2event=0xb3,cmask=1,period=2000000,umask=0x4snoopq_requests_outstanding.invalidateOffcore data reads, RFOs, and prefetches satisfied by the LLC and HIT in a sibling coreevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x2701event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5804event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5880mem_uncore_retired.local_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0120Counts demand data reads that have any response typeCounts all prefetch data reads that miss the L3 and the modified data is transferred from remote cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00004event=0xb7,period=100003,umask=0x1,offcore_rsp=0x103FC00001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00001offcore_response.demand_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x83FC00400core_power.lvl2_turbo_licenseRate of silent evictions from the L2 cache per Kilo instruction where the evicted lines are dropped (no writeback to L3 or memory)DRAM Page Activate commands sent due to a write request. Unit: uncore_imc Streaming stores (partial cache line). Unit: uncore_cha LLC_MISSES.PCIE_READunc_cha_snoop_resp.rspcnflctsCounts when a transaction with the opcode type Rsp*Fwd*WB Snoop Response was received which indicates the data was written back to its home socket, and the cacheline was forwarded to the requestor socket.  This snoop response is only used in &gt;= 4 socket systems.  It is used when a snoop HITM's in a remote caching agent and it directly forwards data to a requestor, and simultaneously returns data to its home socket to be written back to memoryunc_iio_data_req_by_cpu.mem_read.part1Counts every read request for 4 bytes of data made by a unit on the main die (generally a core) or by another IIO unit to the MMIO space of a card on IIO Part3. In the general case, Part3 refers to a x4 PCIe card plugged into the fourth slot of a PCIe riser card, but it could brefer to  any device attached to the IIO unit using the lanes starting at lane 12 of the 16 lanes supported by the busPeer to peer read request for 4 bytes made by IIO Part2 to an IIO target. Unit: uncore_iio unc_iio_data_req_of_cpu.peer_read.part3Counts every peer to peer write request of 4 bytes of data made by IIO Part0 to the MMIO space of an IIO target. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busevent=0xc1,ch_mask=0x08,fc_mask=0x07,umask=0x04Write request of up to a 64 byte transaction is made to IIO Part0 by the CPU. Unit: uncore_iio event=0xc1,ch_mask=0x04,fc_mask=0x07,umask=0x01event=0x84,ch_mask=0x08,fc_mask=0x07,umask=0x04Read request for up to a 64 byte transaction is made by IIO Part3 to Memory. Unit: uncore_iio PCIITOM request issued by the IRP unit to the mesh with the intention of writing a full cacheline. Unit: uncore_irp Counts when the M2M (Mesh to Memory) looks into the multi-socket cacheline Directory state , and found the cacheline marked in the S (Shared) state indicating the cacheline is either stored in another socket in the S(hared) state , and so there is no need to snoop the other sockets for the latest data.  The data may be stored in any state in the local socketevent=0x2e,umask=0x4Multi-socket cacheline Directory update from S to A. Unit: uncore_m2m M2M Writes Issued to iMC; All, regardless of priority. Unit: uncore_m2m BL Egress (to CMS) Occupancy; Allunc_upi_txl_flits.all_nullevent=0x2,umask=0x47event=0x2,umask=0x97ocr.all_data_rd.l3_hit_m.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100491event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100491OCR.ALL_DATA_RD.L3_HIT_S.SNOOP_NONEOCR.ALL_PF_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWD  OCR.ALL_PF_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200490ocr.all_pf_data_rd.l3_hit_m.no_snoop_neededOCR.ALL_PF_DATA_RD.L3_HIT_M.SNOOP_MISSocr.all_pf_rfo.l3_hit_e.hitm_other_coreOCR.ALL_PF_RFO.L3_HIT_E.HITM_OTHER_CORE  OCR.ALL_PF_RFO.L3_HIT_E.HITM_OTHER_COREOCR.ALL_PF_RFO.L3_HIT_F.NO_SNOOP_NEEDED  OCR.ALL_PF_RFO.L3_HIT_F.NO_SNOOP_NEEDEDOCR.ALL_PF_RFO.L3_HIT_S.ANY_SNOOP  OCR.ALL_PF_RFO.L3_HIT_S.ANY_SNOOPOCR.ALL_PF_RFO.L3_HIT_S.SNOOP_NONEocr.all_reads.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2002007F7OCR.ALL_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWD  OCR.ALL_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200100122Counts all demand code reads OCR.DEMAND_CODE_RD.L3_HIT.HIT_OTHER_CORE_FWD OCR.DEMAND_CODE_RD.L3_HIT.HIT_OTHER_CORE_FWDocr.demand_code_rd.l3_hit_e.no_snoop_neededCounts demand data reads OCR.DEMAND_DATA_RD.L3_HIT.HITM_OTHER_CORE OCR.DEMAND_DATA_RD.L3_HIT.HITM_OTHER_COREocr.demand_data_rd.l3_hit_e.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x800040001Counts demand data reads  OCR.DEMAND_DATA_RD.L3_HIT_S.ANY_SNOOPCounts all demand data writes (RFOs) OCR.DEMAND_RFO.L3_HIT.NO_SNOOP_NEEDED OCR.DEMAND_RFO.L3_HIT.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003C0002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200002event=0xb7,period=100003,umask=0x1,offcore_rsp=0x8003C8000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80048000Counts any other requests  OCR.OTHER.L3_HIT_M.HIT_OTHER_CORE_NO_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_HIT.HIT_OTHER_CORE_NO_FWD OCR.PF_L1D_AND_SW.L3_HIT.HIT_OTHER_CORE_NO_FWDocr.pf_l1d_and_sw.l3_hit_e.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400200400Counts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_M.NO_SNOOP_NEEDEDocr.pf_l2_data_rd.l3_hit_s.no_snoop_neededocr.pf_l2_rfo.l3_hit_e.no_snoop_neededCounts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_F.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100040020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80100020ocr.pf_l2_rfo.l3_hit_s.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80100020ocr.pf_l3_data_rd.l3_hit.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100100080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000100100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_E.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_F.SNOOP_MISSoffcore_response.all_data_rd.l3_hit_m.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80020491offcore_response.all_pf_data_rd.l3_hit_f.hit_other_core_no_fwdoffcore_response.all_pf_data_rd.l3_hit_s.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020490This event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_M.HIT_OTHER_CORE_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_READS.L3_HIT.HIT_OTHER_CORE_FWDoffcore_response.all_reads.l3_hit_f.hit_other_core_fwdoffcore_response.all_reads.l3_hit_s.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_M.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDoffcore_response.demand_data_rd.supplier_none.no_snoop_neededThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.SUPPLIER_NONE.SNOOP_NONEoffcore_response.demand_rfo.l3_hit_f.hitm_other_coreThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_F.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_S.HIT_OTHER_CORE_NO_FWDoffcore_response.other.l3_hit_f.snoop_noneoffcore_response.pf_l1d_and_sw.l3_hit.snoop_missoffcore_response.pf_l1d_and_sw.l3_hit_s.snoop_missoffcore_response.pf_l1d_and_sw.pmm_hit_local_pmm.snoop_not_neededoffcore_response.pf_l1d_and_sw.supplier_none.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT.SNOOP_MISSoffcore_response.pf_l2_data_rd.l3_hit_e.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_HIT_E.SNOOP_NONEoffcore_response.pf_l2_data_rd.l3_hit_f.snoop_noneoffcore_response.pf_l2_data_rd.l3_hit_s.hitm_other_coreoffcore_response.pf_l2_data_rd.l3_hit_s.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020010This event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_HIT.NO_SNOOP_NEEDEDoffcore_response.pf_l2_rfo.l3_hit_m.any_snoopoffcore_response.pf_l2_rfo.l3_hit_m.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_HIT_F.ANY_SNOOPoffcore_response.pf_l3_data_rd.l3_hit_s.hit_other_core_fwdoffcore_response.pf_l3_rfo.l3_hit_f.hit_other_core_fwdoffcore_response.pf_l3_rfo.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_S.SNOOP_NONEThis event is deprecated. Refer to new event OCR.PF_L3_RFO.SUPPLIER_NONE.SNOOP_MISSevent=0xcf,period=2000003,umask=0x20OCR.ALL_DATA_RD.L3_MISS.SNOOP_NONE OCR.ALL_DATA_RD.L3_MISS.SNOOP_NONEocr.all_data_rd.l3_miss_local_dram.hit_other_core_no_fwdOCR.ALL_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.SNOOP_MISSOCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWD  OCR.ALL_PF_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x13C0007F7event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2040007F7ocr.all_reads.l3_miss_remote_hop1_dram.snoop_noneocr.all_rfo.l3_miss.hit_other_core_no_fwdOCR.ALL_RFO.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_RFO.L3_MISS.NO_SNOOP_NEEDED OCR.ALL_RFO.L3_MISS.NO_SNOOP_NEEDEDOCR.ALL_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISSocr.all_rfo.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1010000122event=0xb7,period=100003,umask=0x1,offcore_rsp=0x83C000004Counts all demand code reads OCR.DEMAND_CODE_RD.L3_MISS.HIT_OTHER_CORE_NO_FWD OCR.DEMAND_CODE_RD.L3_MISS.HIT_OTHER_CORE_NO_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000004ocr.demand_rfo.l3_miss.snoop_noneCounts any other requests  OCR.OTHER.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDCounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.L3_MISS.NO_SNOOP_NEEDED OCR.PF_L1D_AND_SW.L3_MISS.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x23C000400event=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000400ocr.pf_l2_data_rd.l3_miss.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000020Counts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_MISS.SNOOP_NONEevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x210000020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x404000080Counts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS.HITM_OTHER_CORE OCR.PF_L3_RFO.L3_MISS.HITM_OTHER_COREocr.pf_l3_rfo.l3_miss.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) RFOs OCR.PF_L3_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISS_OR_NO_FWDThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.all_pf_rfo.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.all_rfo.l3_miss_local_dram.no_snoop_neededoffcore_response.all_rfo.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.REMOTE_HIT_FORWARDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS.SNOOP_NONEThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_REMOTE_HOP1_DRAM.NO_SNOOP_NEEDEDoffcore_response.other.l3_miss_local_dram.snoop_miss_or_no_fwdThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.ANY_SNOOPoffcore_response.pf_l2_data_rd.l3_miss.no_snoop_neededoffcore_response.pf_l2_data_rd.l3_miss_remote_hop1_dram.any_snoopThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS_LOCAL_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.PF_L3_DATA_RD.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_NO_FWDoffcore_response.pf_l3_rfo.l3_miss_local_dram.hit_other_core_no_fwdocr.all_data_rd.pmm_hit_local_pmm.any_snoopocr.all_reads.supplier_none.hitm_other_coreocr.all_rfo.pmm_hit_local_pmm.any_snoopOCR.ALL_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOP OCR.ALL_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPCounts all demand code reads  OCR.DEMAND_CODE_RD.SUPPLIER_NONE.HIT_OTHER_CORE_NO_FWDocr.pf_l1d_and_sw.supplier_none.hit_other_core_fwdocr.pf_l1d_and_sw.supplier_none.snoop_missCounts prefetch (that bring data to L2) data reads OCR.PF_L2_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.pf_l2_data_rd.supplier_none.hitm_other_coreocr.pf_l2_rfo.supplier_none.hit_other_core_no_fwdCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDocr.pf_l3_rfo.supplier_none.snoop_noneDirty line underfill read hits to Near Memory(DRAM cache) in Memory Mode. Unit: uncore_m2m Counts Software prefetch requests that hit the L2 cache. Accounts for PREFETCHNTA and PREFETCHT0/1/2 instructions when FB is not fullevent=0xd1,period=50021,umask=0x20Counts hardware prefetch RFOs (which bring data to L2) that hit a cacheline in the L3 where a snoop was sentCounts miscellaneous requests, such as I/O and un-cacheable accesses that hit a cacheline in the L3 where a snoop was sent but no other cores had the dataevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1E003C8000Counts streaming stores that hit a cacheline in the L3 where a snoop was sent or notNumber of SSE/AVX computational 128-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 2 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform 2 calculations per element. The DAZ and FTZ flags in the MXCSR register need to be set when using these eventsevent=0xc6,period=100007,umask=0x1,frontend=0x500406ocr.demand_rfo.dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000800assists.anyevent=0xc5,period=50021,umask=0x2A version of INST_RETIRED that allows for a more unbiased distribution of samples across instructions retired. It utilizes the Precise Distribution of Instructions Retired (PDIR) feature to mitigate some bias in how retired instructions get sampled. Use on Fixed Counter 0 (Precise event)SummaryLSDocr.demand_rfo.snc_cache.hitmocr.hwpf_l1d_and_swpf.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0477ocr.demand_code_rd.l3_miss_localevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F844027F0event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F3FC00477event=0xb7,period=100003,umask=0x1,offcore_rsp=0x94000800core_snoop_response.misscore_snoop_response.s_fwd_mCounts demand reads for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were supplied by DRAM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those DRAM accesses that are controlled by the close SNC Clusterocr.reads_to_core.any_responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F330004772LM Tag Check : Miss, no data in this line. Unit: uncore_imc DRAM Precharge commands. : Precharge due to read. Unit: uncore_imc event=0x4,umask=0x0fLocal INVITOE requests (exclusive ownership of a cache line without receiving data) that miss the SF/LLC and are sent to the CHA's home agent. Unit: uncore_cha unc_cha_tor_inserts.ia_hit_rfoevent=0x35,umask=0xC001FE01unc_cha_tor_inserts.ia_drd_prefevent=0x35,umask=0xC8077E01event=0x35,umask=0xC8168A01unc_cha_tor_occupancy.ia_miss_drd_pmmData requested by the CPU : Core writing to Card's MMIO space. Unit: uncore_iio unc_iio_data_req_by_cpu.mem_read.part4unc_iio_data_req_of_cpu.mem_read.part4unc_iio_num_req_of_cpu.commit.allunc_iio_comp_buf_occupancy.cmpd.part7PCIe Completion Buffer Occupancy of completions with data : Part 0-7. Unit: uncore_iio Valid Flits Received : All Data. Unit: uncore_upi Valid Flits Sent : All Data. Unit: uncore_upi Valid Flits Sent : Null FLITs transmitted to any slot. Unit: uncore_upi unc_cha_tor_inserts.io_itom_localCounts the number of L2 Cache accesses that resulted in a hit from a front door request only (does not include rejects or recycles), Counts on a per core basisCounts the number of memory uops retired.  A single uop that performs both a load AND a store will be counted as 1, not 2 (e.g. ADD [mem], CONST)  Supports address when precise (Precise event)Counts all code reads that were supplied by the L3 cacheThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_HIT.SNOOP_HIT_WITH_FWDocr.hwpf_l1d_and_swpf.l3_hit.snoop_hitmocr.hwpf_l2_code_rd.l3_hitevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2001F803C0000ocr.uc_rd.l3_hit.snoop_hit_no_fwdCounts the number of BACLEARS due to a conditional jumpCounts the number of requests that hit in the instruction cache.  The event only counts new cache line accesses, so that multiple back to back fetches to the exact same cache line and byte chunk count as one.  Specifically, the event counts when accesses from sequential code crosses the cache line boundary, or when a branch target is moved to a new line or to a non-sequential byte chunk of the same lineCounts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were not supplied by the L3 cacheevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000800Counts the number of core cycles during which there are pending interrupts while interrupts are masked (disabled). Increments by 1 each core cycle that both EFLAGS.IF is 0 and an INTR is pending (which means the APIC is telling the ROB to cause an INTR). This event does not increment if EFLAGS.IF is 0 but all interrupt in the APICs Interrupt Request Register (IRR) are inhibited by the PPR (thus either by ISRV or TPR)  because in these cases the interrupts would be held up in the APIC and would not be pended to the ROB. This event does count when an interrupt is only inhibited by MOV/POP SS state machines or the STI state machine. These extra inhibits only last for a single instructions and would not be importantocr.all_code_rd.outstandingevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x184000040ocr.uc_rd.local_drammachine_clears.anytopdown_be_bound.serializationCounts the number of issue slots every cycle that were not delivered by the frontend due to wrong predecodesCounts the number of uops issued by the front end every cycleTOR Occupancy : DRd_Opts issued by iA Cores : For each cycle, this event accumulates the number of valid entries in the TOR that match qualifications specified by the subevent.     Does not include addressless requests such as locks and interruptsTOR Occupancy : DRd_Opt_Prefs issued by iA Cores that hit the LLC. Unit: uncore_cha PCIe Completion Buffer Inserts of completions with data : Part 1 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 1PCIe Completion Buffer Occupancy : Part 4 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 4Data requested of the CPU : Card reading from DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 7Data requested of the CPU : Card writing to DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Data requested of the CPU : Card writing to DRAM : Number of DWs (4 bytes) the card requests of the main die.    Includes all requests initiated by the Card, including reads and writes. : x4 card is plugged in to slot 7Number Transactions requested by the CPU : Core reading from Card's MMIO space : Also known as Outbound.  Number of requests initiated by the main die, including reads and writes. : x4 card is plugged in to slot 3Clockticks of the mesh to memory (M2M)Counts the number of Extended Page Directory Entry hits.  The Extended Page Directory cache is used by Virtual Machine operating systems while the guest operating systems use the standard TLB cachesCounts the number of Extended Page Directory Pointer Entry missesInstructions per FP Arithmetic Scalar Double-Precision instruction (lower number means higher occurrence rate). Unit: cpu_core Counts the total number of issue slots  that were not consumed by the backend due to backend stalls.  Note that uops must be available for consumption in order for this event to count.  If a uop is not available (IQ is empty), this event will not count.   The rest of these subevents count backend stalls, in cycles, due to an outstanding request which is memory bound vs core bound.   The subevents are not slot based events and therefore can not be precisely added or subtracted from the Backend_Bound_Aux subevents which are slot basedevent=0x4,period=20003,umask=0x7mem_uops_retired.load_latency_gt_8Counts all the retired split loads  Supports address when precise (Precise event). Unit: cpu_atom Number of L1D misses that are outstanding. Unit: cpu_core Retired store instructions that miss the STLB  Supports address when precise (Precise event). Unit: cpu_core Retired load instructions which data sources missed L3 but serviced from local dram  Supports address when precise. Unit: cpu_core event=0x40,period=100003,umask=0x2event=0xb3,period=2000003,umask=0x4Counts the number of requests to the instruction cache for one or more bytes of a cache line. Unit: cpu_atom event=0xc6,period=100007,umask=0x1,frontend=0x600406icache_tag.stallsld_head.l1_bound_at_retevent=0x5,period=1000003,umask=0xc0Execution stalls while L1 cache miss demand load is outstanding. Unit: cpu_core Counts demand read for ownership (RFO) requests and software prefetches for exclusive ownership (PREFETCHW) that were not supplied by the L3 cache. Unit: cpu_core Counts the number of far branch instructions retired, includes far jump, far call and return, and Interrupt call and return (Precise event). Unit: cpu_atom serialization.non_c01_ms_scbIndirect near branch instructions retired (excluding returns) (Precise event). Unit: cpu_core TMA slots where uops got dropped. Unit: cpu_core integer ADD, SUB, SAD 128-bit vector instructions. Unit: cpu_core event=0xa4,period=10000003,umask=0x10TMA slots available for an unhalted logical processor. Fixed counter - architectural event. Unit: cpu_core uops_dispatched.port_5_11Cycles at least 1 micro-op is executed from any thread on physical core. Unit: cpu_core PRE command sent to DRAM due to page table idle timer expiration. Unit: uncore_imc event=0x12,period=100003,umask=0xeevent=0x12,period=100003,umask=0x10event=0x13,period=100003,umask=0x4event=0x11,period=100003,umask=0x2Counts retired load instructions with remote Intel Optane DC persistent memory as the data source and the data request missed L3Number of retired micro-operations (uops) for load or store memory accessesCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that resulted in a snoop hit in another core's caches which forwarded the unmodified data to the requesting coreCounts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by a cache on a remote socket where a snoop was sent and data was returned (Modified or Not Modified)event=0xcf,period=100003,umask=0x4fp_arith_inst_retired2.vectorMEMORY_ACTIVITY.STALLS_L2_MISSASSISTS.PAGE_FAULTevent=0xb0,period=1000003,umask=0x8Number of retired integer ADD/SUB (regular or horizontal), SAD 128-bit vector instructionsUops executed on port 6IMC Clockticks at HCLK frequency. Unit: uncore_imc event=0x84,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000002event=0x84,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000002event=0x50,umask=0x0000000008Requests for exclusive ownership of a cache line without receiving data. Unit: uncore_cha event=0x2,umask=0x0000000020unc_upi_rxl_occupancy.slot2event=0x84,ch_mask=0x0002,fc_mask=0x07,umask=0x0000000004event=0x84,ch_mask=0x0020,fc_mask=0x07,umask=0x0000000001event=0xc2,ch_mask=0x08,fc_mask=0x07,umask=0x0000000004event=0x1b,umask=0x07event=0x21,umask=0x0302event=0x2aTOR Inserts; RFO misses from local IO. Unit: uncore_cha TOR Inserts; RFO Pref hits from local IA. Unit: uncore_cha event=0x35,umask=0x00c886fe01event=0x35,umask=0x00c8877e01unc_cha_tor_occupancy.ia_hit_crdevent=0x36,umask=0x00c80ffd01TOR Occupancy : PRQ - Non IOSF. Unit: uncore_cha TOR Occupancy : Just Local Targets. Unit: uncore_cha event=0x36,umask=0x00c88ffd01unc_cha_tor_occupancy.ia_miss_drd_prefevent=0x36,umask=0x00cc43fd04unc_cha_tor_inserts.ia_wbeftoeevent=0x36,umask=0x00c8f3fd04unc_cha_tor_inserts.ia_miss_wcil_pmmunc_cha_tor_inserts.ia_miss_local_wcil_pmmunc_cha_tor_occupancy.ia_miss_local_wcilf_ddrTOR Occupancy : WCiLs issued by iA Cores targeting DDR that missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0x00c8c3ff04event=0x87,umask=0x01event=0x61,umask=0x02l2_cache_req_stat.ic_fill_hit_xl2_cache_req_stat.ic_fill_hit_sevent=0x71,umask=0xffThe number of retired taken branch instructions that were mispredictedex_ret_near_ret_mispreddram_channel_data_controller_4The number of operations (uOps) dispatched to each of the 4 FPU execution pipelines. This event reflects how busy the FPU pipelines are and may be used for workload characterization. This includes all operations performed by x87, MMX, and SSE instructions, including moves. Each increment represents a one-cycle dispatch event. This event is a speculative event. Since this event includes non-numeric operations it is not suitable for measuring MFLOPS. Total number uOps assigned to pipe 3Double precision multiply FLOPSThis is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15. Single-precision multiply FLOPSevent=0x46,umask=0x0cSoftware Prefetch Instructions (PREFETCHNTA instruction) Dispatchedls_pref_instr_disp.store_prefetch_wevent=0x52,umask=0x02de_dis_dispatch_token_stalls0.alu_token_stallL2 ITLB Misses & Instruction page walksevent=0x43,umask=0x40Software Prefetch Instructions Dispatched (Speculative). Prefetch_T0_T1_T2. PrefetchT0, T1 and T2 instructions. See docAPM3 PREFETCHlevelde_dis_dispatch_token_stalls1.load_queue_token_stallRetired Ops. Use macro_ops_retired insteadevent=0x60,umask=0xe8ls_dmnd_fills_from_sys.allIn-flight L1 data cache misses i.e. Miss Address Buffer (MAB) allocations each cyclel2_pf_hit_l2.l1_regionL2 prefetches accepted by the L2 pipeline which miss the L2 cache and hit in the L3 cache of type L2Stride (fetch additional lines into L2 cache when each access is a constant distance from the previous)L2 prefetches accepted by the L2 pipeline which miss the L2 and the L3 caches of type L2Burst (aggressively fetch additional sequential lines into L2 cache)L3 cache hits. Unit: amd_l3 l3_xi_sampled_latency_requests.near_cachel3_xi_sampled_latency_requests.far_cacheNumber of cycles the divider is busyCycles with no retire while the oldest op is waiting for load dataremote_processor_read_data_beats_cs0event=0x2df,umask=0xbfeWrite data beats (64 bytes) for remote socket upstream DMA at IO Moderator (IOM) 0event=0x89f,umask=0xbffevent=0x45e,umask=0x7felocal_socket_inf1_inbound_data_beats_ccm2event=0x49f,umask=0x7felocal_socket_inf1_outbound_data_beats_ccm7Data beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 2event=0x55e,umask=0xbfeData beats (32 bytes) at interface 0 for remote socket inbound data to CPU Moderator (CCM) 5Data beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 4remote_socket_inf1_outbound_data_beats_ccm7event=0x8,umask=0x3ffp_ops_retired_by_type.vector_divevent=0xb,umask=0x0bsse_avx_ops_retired.mmx_allsse_avx_ops_retired.sse_avx_packevent=0xb,umask=0xc0event=0xb,umask=0xf0Retired SSE and AVX integer ops of all typesfp_pack_ops_retired.fp128_divRetired 128-bit packed floating-point divide opsfp_pack_ops_retired.fp128_blendevent=0xc,umask=0x09Retired 128-bit packed floating-point ops of all typesRetired 256-bit packed floating-point subtract opsfp_pack_ops_retired.fp256_divRetired 256-bit packed floating-point shuffle ops (may include instructions not necessarily thought of as including shuffles e.g. horizontal add, dot product, and certain MOV instructions)event=0xd,umask=0x40packed_int_op_type.int256_allumc_act_cmd.rdumc_pchg_cmd.allOps fetched from instruction cache and dispatchedIn each cycle counts dispatch slots left empty because the front-end did not supply opsl1_data_cache_fills_from_different_ccxAll TLBs flushedLocal socket upstream DMA write dataDC_DCACHE_ACCESSES_BY_LOCKSFR_RETIRED_NEAR_RETURNS_MISPREDICTEDPMUEXTIN0_EVTEVENT_0BHEVENT_16HEVENT_20HEVENT_45HEVENT_46HEVENT_4CHEVENT_57HEVENT_6EHEVENT_7FHEVENT_9CHEVENT_DEHEVENT_DFHDATA_CACHE_DEPENDENT_STALLEVENT_11FHEVENT_134HEVENT_148HEVENT_16EHEVENT_19FHEVENT_1D7HEVENT_1E4HEVENT_1E5HEVENT_1F7HEVENT_1FBHEVENT_20CHEVENT_214HEVENT_21EHEVENT_22DHEVENT_275HEVENT_291HEVENT_2AFHEVENT_2BAHEVENT_2DDHEVENT_2EFHEVENT_2F4HEVENT_2FFHEVENT_310HEVENT_314HEVENT_318HEVENT_31FHEVENT_327HEVENT_32DHEVENT_35EHEVENT_363HEVENT_368HEVENT_391HEVENT_3A5HEVENT_3A9HEVENT_3FAHL2D_CACHE_REFILL_LDL2D_CACHE_INVALMEM_ACCESS_STCRYPTO_SPECL2D_TLB_REFILL_RDhnf_mc_retrieshnf_txdat_stallhni_txrsp_retryackhni_awvalid_no_awreadyrnd_s2_wdata_beatscxla_tx_tlp_buffer_fullVIU1_INSTR_COMPLETEDINSTR_BKPT_MATCHESLOAD_MISS_ALIAS_ON_TOUCHL1_DATA_TOUCH_MISS_CYCLESL1_DATA_TOTAL_MISSTOUCHES_TRANSLATEDlow-op-pos-1rdszdwordk8-fr-retired-taken-branchesINTEL_CORE2kern.hwpmc.cpuid%s, "tsc": "%jd"%s, "tid": "%d", "pid": "%d", "flags": "0x%08x", "tdname": "%s"}
v5GenuineIntel-6-7AThis category represents fraction of slots wasted due to incorrect speculations. This include slots used to issue uops that do not eventually get retired and slots for which the issue-pipeline was blocked due to recovery from earlier incorrect speculation. For example; wasted work due to miss-predicted branches are categorized under Bad Speculation category. Incorrect data speculation followed by Memory Ordering Nukes is another examplePipeline;Ret;Retire1 / (INST_RETIRED.ANY / CPU_CLK_UNHALTED.THREAD)( (FP_ARITH_INST_RETIRED.SCALAR_SINGLE + FP_ARITH_INST_RETIRED.SCALAR_DOUBLE) + (FP_ARITH_INST_RETIRED.128B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.128B_PACKED_SINGLE + FP_ARITH_INST_RETIRED.256B_PACKED_DOUBLE + FP_ARITH_INST_RETIRED.256B_PACKED_SINGLE) ) / ( 2 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ) )( ((BR_MISP_RETIRED.ALL_BRANCHES / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT )) * (( UOPS_ISSUED.ANY - UOPS_RETIRED.RETIRE_SLOTS + 4 * ( INT_MISC.RECOVERY_CYCLES_ANY / 2 ) ) / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) + (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * (BR_MISP_RETIRED.ALL_BRANCHES * (12 * ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY ) / CPU_CLK_UNHALTED.THREAD) / ( BR_MISP_RETIRED.ALL_BRANCHES + MACHINE_CLEARS.COUNT + BACLEARS.ANY )) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) ) * (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )) / BR_MISP_RETIRED.ALL_BRANCHESIpLoadL2_Cache_Fill_BWL2MPKI1000 * MEM_LOAD_UOPS_RETIRED.L3_MISS / INST_RETIRED.ANYUtilization of the core's Page Walker(s) serving STLB misses triggered by instruction/Load/Store accessesl2_rqsts.all_code_rdevent=0x24,period=200003,umask=0xe4Demand requests that miss L2 cacheevent=0x24,period=200003,umask=0xe2L1D writebacks that access L2 cacheevent=0x2e,period=100003,umask=0x41(null)mem_uops_retired.lock_loadsevent=0xb0,period=100003,umask=0x2Number of SIMD FP assists due to input valuesCycles Decode Stream Buffer (DSB) is delivering any UopThis counts the number of cycles that the instruction decoder queue is empty and can indicate that the application may be bound in the front end.  It does not determine whether there are uops being delivered to the Alloc stage since uops can be delivered by bypass skipping the Instruction Decode Queue (IDQ) when it is emptyidq.ms_dsb_occurUops initiated by MITE and delivered to Instruction Decode Queue (IDQ) while Microcode Sequenser (MS) is busyCycles per thread when 3 or more uops are not delivered to Resource Allocation Table (RAT) when backend of the machine is not stalledmemoryevent=0xc3,period=100003,umask=0x2This event counts loads with latency value being above 128  Spec update: BDM100, BDM35 (Must be precise)Number of times an RTM execution aborted due to incompatible memory typeevent=0xc9,period=2000003,umask=0x2tx_mem.abort_conflictbr_inst_exec.taken_indirect_near_callbr_inst_retired.not_takenbr_misp_retired.retcpu_clk_unhalted.thread_anyStalls caused by changing prefix length of the instructionCases when loads get true Block-on-Store blocking code preventing store forwardingCycles per thread when uops are executed in port 2This event counts, on the per-thread basis, cycles during which uops are dispatched from the Reservation Station (RS) to port 2event=0xb1,cmask=1,period=2000003,umask=0x1Counts number of cycles no uops were dispatched to be executed on this threadCycles per core when uops are exectuted in port 0uops_executed_port.port_5uops_issued.stall_cyclesevent=0x35,umask=0x3,filter_opc=0x1c8unc_h_snoop_resp.rspsevent=0x21,umask=0x2unc_m_dclockticksfreq_max_limit_thermal_cycles %(UNC_P_FREQ_MAX_OS_CYCLES / UNC_P_CLOCKTICKS) * 100.Demand load Miss in all translation lookaside buffer (TLB) levels causes a page walk that completes (4K)  Spec update: BDM69This event counts store misses in all DTLB levels that cause a completed page walk (4K page size). The page walk can end with or without a fault  Spec update: BDM69Number of DTLB page walker hits in the L2  Spec update: BDM69, BDM98page_walker_loads.itlb_l1event=0x24,period=200003,umask=0xc2This event counts memory transactions reached the super queue including requests initiated by the core, all L3 prefetches, page walks, and so onoffcore_response.all_pf_code_rd.l3_hit.snoop_missoffcore_response.all_pf_rfo.l3_hit.snoop_hit_no_fwdoffcore_response.all_rfo.supplier_none.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x0100020122offcore_response.demand_data_rd.any_responseoffcore_response.demand_data_rd.l3_hit.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80020001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x04003C0010offcore_response.pf_l2_data_rd.supplier_none.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F803C0020event=0xb7,period=100003,umask=0x1,offcore_rsp=0x10003C0080offcore_response.pf_l3_rfo.l3_hit.snoop_hit_no_fwdNumber of SSE/AVX computational single precision floating-point instructions retired; some instructions will count twice as noted below. Applies to SSE* and AVX* scalar and packed single precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT RSQRT RCP SQRT DPP FM(N)ADD/SUB.  DPP and FM(N)ADD/SUB instructions count twice as they perform multiple calculations per elementoffcore_response.all_data_rd.l3_miss_local_dram.snoop_not_neededoffcore_response.all_pf_data_rd.supplier_none.snoop_non_dramevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000020090event=0xb7,period=100003,umask=0x1,offcore_rsp=0x023C000120event=0xb7,period=100003,umask=0x1,offcore_rsp=0x013C000122offcore_response.corewb.l3_miss_local_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004000001offcore_response.demand_data_rd.l3_miss_local_dram.snoop_noneevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x043C008000event=0xb7,period=100003,umask=0x1,offcore_rsp=0x1004008000offcore_response.other.l3_miss_local_dram.snoop_non_dramunc_cbo_xsnp_response.hit_xcoreNumber of Writes allocated - any write transactions: full/partials writes and evictions. Unit: uncore_arb unc_arb_coh_trk_requests.allCycles with at least one request outstanding is waiting for data return from memory controller. Account for coherent and non-coherent requests initiated by IA Cores, Processor Graphics Unit, or LLC.;. Unit: uncore_arb offcore_response.all_data_rd.llc_hit.hitm_other_coreoffcore_response.demand_rfo.llc_miss.any_responseuncore interconnectevent=0x40,period=2000000,umask=0xa1event=0x2d,period=200000,umask=0x44L2 cache line modificationsl2_no_req.selfevent=0x30,period=200000,umask=0x58event=0x2e,period=200000,umask=0x74SIMD packed logical micro-ops executedsimd_uop_type_exec.mul.sevent=0xb3,period=2000000,umask=0x4simd_uop_type_exec.unpack.arSIMD unpacked micro-ops retiredmacro_insts.all_decodedAll Instructions decodedNon-CISC nacro instructions decodedMemory references that cross an 8-byte boundary (At Retirement)Store splits (Ar Retirement)event=0x62,period=200000,umask=0x0IO requests waiting in the bus queuebus_trans_rfo.selfext_snoop.this_agent.hitRetired branch instructionsevent=0xa,period=2000000,umask=0x0event=0x2,period=200000,umask=0x81event=0xc2,period=2000000,umask=0x10data_tlb_misses.l0_dtlb_miss_ldITLB misses (Must be precise)Counts data reads generated by L1 or L2 prefetchers that miss the L2 cacheCounts data read, code read, and read for ownership (RFO) requests (demand & prefetch) that true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x0000018000offcore_response.bus_locks.any_responseCounts demand instruction cacheline and I-side prefetch requests that miss the instruction cache that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000000800Counts partial cache line data writes to uncacheable write combining (USWC) memory region  that hit the L2 cache. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.sw_prefetch.l2_miss.hit_other_core_no_fwdevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0400001000uops_retired.fpdivevent=0xc4,period=200003,umask=0x80Retired conditional branch instructions (Precise event capable) (Must be precise)Counts mispredicted near RET branch instructions retired, where the return address taken was not what the processor predicted (Must be precise)Counts bus lock and split lock requests outstanding, per cycle, from the time of the L2 miss to when any response is receivedCounts the number of writeback transactions caused by L1 or L2 cache evictions have any transaction responses from the uncore subsystem. Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)Counts demand reads for ownership (RFO) requests generated by a write to full data cache line true miss for the L2 cache with a snoop miss in the other processor module.  Requires MSR_OFFCORE_RESP[0,1] to specify request type and response. (duplicated for both MSRs)offcore_response.pf_l1_data_rd.outstandingoffcore_response.streaming_stores.any_responsePage walk completed due to a demand load to a 1GB pagedtlb_load_misses.walk_pendingevent=0xf2,period=100003,umask=0x6Cycles in which the L1D is lockedCounts all demand data writes (RFOs) miss the L3 and the data is returned from local dramoffcore_response.pf_l3_data_rd.l3_miss.any_responseAny uop executed by the Divider. (This includes all divide uops, sqrt, ...)Mispredicted branch instructions at retirementCycles with pending L1 data cache miss loads. Set Cmask=8 to count cycleCycles at least 4 micro-op is executed from any thread on physical core  Spec update: HSD30, HSM31Cycles which a uop is dispatched on port 5 in this threadCycles which a uop is dispatched on port 7 in this threadNumber of flags-merge uops allocated. Such uops add delayuops_retired.core_stall_cyclesunc_cbo_xsnp_response.hitm_evictionevent=0x83,umask=0x01Demand load Miss in all translation lookaside buffer (TLB) levels causes a page walk that completes of any page sizeNumber of DTLB page walker loads from memory  Spec update: HSD25page_walker_loads.ept_dtlb_l3Number of ITLB page walker loads that hit in the L2offcore_response.pf_l2_data_rd.llc_hit.any_responseCount the number of modified Lines evicted from L1 and missed L2. (Non-rejected WBs from the DCU.)event=0x24,period=200003,umask=0x8All retired load uops. (Precise Event)offcore_response.demand_rfo.llc_hit.no_snoop_neededCounts miscellaneous accesses that include port i/o, MMIO and uncacheable memory accesses. It also includes L2 hints sent to LLC to keep a line from being evicted out of the core cachesevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x10400event=0x10,period=2000003,umask=0x80dsb2mite_switches.countCounts cycles MITE is delivered at least one uops. Set Cmask = 1Cycles while L2 cache miss load* is outstandingCycles the RS is empty for the threadCycles per core when load or STA uops are dispatched to port 3Counts the number of LLC evictions allocated. Unit: uncore_arb Counts the number of allocated write entries, include full, partial, and LLC evictions. Unit: uncore_arb LLC lookup request that access cache and found line in S-stateFilter on processor core initiated cacheable write requests. Unit: uncore_cbox Filter on external snoop requestsPage walk for a large page completed for Demand loadMiss in all TLB levels causes a page walk that completes of any page size (4K/2M/4M/1G)offcore_response.all_pf_data_rd.llc_hit.hit_other_core_no_fwdCounts prefetch (that bring data to LLC only) data reads that hit in the LLCoffcore_response.pf_llc_data_rd.llc_hit.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1003c0080event=0xb7,period=100003,umask=0x1,offcore_rsp=0x107fc00001event=0x35,umask=0x3,filter_opc=0x1e4Cycles where transmitting QPI link is in half-width mode. Unit: uncore_qpi freq_band0_cycles %unc_p_freq_ge_1200mhz_cyclesevent=0x28,period=200003,umask=0x2Data from local DRAM either Snoop not needed or Snoop Miss (RspI)This event counts the number of load uops retired (Precise event)Sample stores and collect precise store operation via PEBS record. PMC3 only. (Precise Event - PEBS) (Must be precise)Resource stalls due to memory buffers or Reservation Station (RS) being fully utilizedresource_stalls2.all_fl_emptyCycles when Allocator is stalled if BOB is full and new branch needs itoffcore_response.any_code_rd.l2_hit_far_tileoffcore_response.any_code_rd.l2_hit_far_tile_mCounts Demand cacheable data and L1 prefetch data read requests  that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0004000070offcore_response.any_pf_l2.l2_hit_this_tile_mevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x00080032f7offcore_response.any_request.l2_hit_near_tileoffcore_response.any_rfo.l2_hit_near_tile_mCounts Bus locks and split lock requests that accounts for responses which hit its own tile's L2 with data in E stateCounts Bus locks and split lock requests that accounts for responses which hit its own tile's L2 with data in M stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400004Counts demand code reads and prefetch code reads that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateCounts demand cacheable data and L1 prefetch data reads that accounts for responses which hit its own tile's L2 with data in F stateevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x0008000001Counts Partial reads (UC or WC and is valid only for Outstanding response type).  that accounts for any responseoffcore_response.partial_reads.l2_hit_this_tile_eevent=0xb7,period=100007,umask=0x1,offcore_rsp=0x1000400100Counts L2 data RFO prefetches (includes PREFETCHW instruction) that accounts for responses from a snoop request hit with data forwarded from its Far(not in the same quadrant as the request)-other tile's L2 in E/F state. Valid only for SNC4 cluster modeCounts UC code reads (valid only for Outstanding response type)  that accounts for reponses from snoop request hit with data forwarded from its Near-other tile L2 in E/F/M stateevent=0xc2,period=200003,umask=0x40Counts Demand cacheable data and L1 prefetch data read requests  that accounts for data responses from DRAM Faroffcore_response.any_data_rd.ddr_nearoffcore_response.any_pf_l2.mcdram_nearoffcore_response.any_request.ddrCounts any request that accounts for responses from MCDRAM (local and far)offcore_response.bus_locks.mcdram_farCounts Demand cacheable data writes that accounts for data responses from MCDRAM Far or Other tile L2 hit faroffcore_response.partial_writes.mcdram_nearCounts L1 data HW prefetches that accounts for data responses from DRAM FarCounts L1 data HW prefetches that accounts for data responses from MCDRAM Localoffcore_response.pf_l2_code_rd.ddr_nearCounts L2 code HW prefetches that accounts for data responses from MCDRAM Localoffcore_response.pf_l2_rfo.mcdram_nearoffcore_response.pf_l2_rfo.non_dramoffcore_response.pf_software.mcdramCounts the number of mispredicted near CALL branch instructions retired (Precise event)no_alloc_cycles.mispredictsevent=0x3,period=200003,umask=0x40Counts the total number of core cycles for all the D-side page walks. The cycles for page walks started in speculative path will also be includedl1d_cache_prefetch_lock_fb_hitevent=0x28,period=100000,umask=0x4All L1 writebacks to L2L2 data demand requestsevent=0x24,period=200000,umask=0x3event=0xf0,period=200000,umask=0x8l2_write.lock.e_stateevent=0xb,period=100,umask=0x10,ldlat=0x400Memory instructions retired above 2048 clocks (Precise Event)Retired loads that hit the L1 data cache (Precise Event)event=0xb7,period=100000,umask=0x1,offcore_rsp=0x711offcore_response.any_rfo.llc_hit_other_core_hitoffcore_response.corewb.remote_cache_hitmevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x203Offcore demand RFO requests satisfied by the IO, CSR, MMIO unitoffcore_response.demand_rfo.local_cacheevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x1880offcore_response.pf_data.local_cache_dramevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x110Offcore prefetch data reads satisfied by the LLC and HIT in a sibling coreoffcore_response.pf_data_rd.remote_cache_dramoffcore_response.pf_ifetch.remote_cache_dramoffcore_response.prefetch.llc_hit_other_core_hitmOffcore prefetch requests that HITM in a remote cacheSuper Queue lock splits across a cache linefp_comp_ops_exe.sse_single_precisionSSE* FP single precision Uopsoffcore_response.any_data.any_llc_missevent=0xb7,period=100000,umask=0x1,offcore_rsp=0xF810offcore_response.prefetch.any_llc_missEarly Branch Prediciton Unit clearsevent=0x89,period=20000,umask=0x2Mispredicted unconditional branches executedevent=0x89,period=2000,umask=0x30Mispredicted near retired calls (Precise Event)event=0x87,period=2000000,umask=0x8inst_queue_write_cyclesevent=0xc0,period=2000000,umask=0x4inst_retired.total_cyclesinst_retired.total_cycles_psmachine_clears.mem_orderresource_stalls.fpcwevent=0x8,period=200000,umask=0x1dtlb_misses.anydtlb_misses.walk_completedCounts the total number of L2 code requestsCounts requests from the L1/L2/L3 hardware prefetchers or Load software prefetches that hit L2 cacheAll retired store instructions  Supports address when precise (Precise event)mem_load_l3_hit_retired.xsnp_hitmCounts retired load instructions with at least one uop was load missed in L1 but hit FB (Fill Buffers) due to preceding miss to the same cache line with data not ready  Supports address when precise (Precise event)offcore_response.demand_code_rd.l4_hit_local_l4.snoop_hit_no_fwdoffcore_response.demand_data_rd.l3_hit_e.any_snoopoffcore_response.demand_data_rd.l3_hit_m.snoop_missCounts any other requestshave any response typeoffcore_response.other.l3_hit_m.snoop_missfrontend_retired.latency_ge_8Counts the total number of uops delivered by the Microcode Sequencer (MS). Any instruction over 4 uops will be delivered by the MS. Some instructions such as transcendentals may additionally generate uops from the MSevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x7C400001event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2000400002Far branch instructions retired  Spec update: SKL091 (Precise event)Taken branch instructions retired  Spec update: SKL091 (Precise event)event=0xa3,cmask=16,period=2000003,umask=0x10MispredictionsFetch_UpC(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) * (DSB2MITE_SWITCHES.PENALTY_CYCLES / CPU_CLK_UNHALTED.THREAD) / #(4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) + ((IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) )))) * (( IDQ.ALL_MITE_CYCLES_ANY_UOPS - IDQ.ALL_MITE_CYCLES_4_UOPS ) / ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ) / 2) / #((IDQ_UOPS_NOT_DELIVERED.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))) - (4 * IDQ_UOPS_NOT_DELIVERED.CYCLES_0_UOPS_DELIV.CORE / (4 * ( ( CPU_CLK_UNHALTED.THREAD / 2 ) * ( 1 + CPU_CLK_UNHALTED.ONE_THREAD_ACTIVE / CPU_CLK_UNHALTED.REF_XCLK ) ))))DSB_Misses_Cost_SMT( BR_INST_RETIRED.NEAR_CALL + BR_INST_RETIRED.NEAR_RETURN ) / BR_INST_RETIRED.ALL_BRANCHESPage walk completed due to a demand data load to a 4K pageCounts 1 per cycle for each PMH (Page Miss Handler) that is busy with a page walk for an instruction fetch request. EPT page walk duration are excluded in Skylake michroarchitectureCounts any code reads (demand & prefetch) that miss L2event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000001event=0xb7,period=100007,umask=0x1,offcore_rsp=0x1680000040Counts the number of branch instructions retired.. (Precise event)Counts the number of JCC branch instructions retired (Precise event)Counts the number of near indirect JMP and near indirect CALL branch instructions retired (Precise event)Cycles the divider is busy.This event counts the cycles when the divide unit is unable to accept a new divide UOP because it is busy processing a previously dispatched UOP. The cycles will be counted irrespective of whether or not another divide UOP is waiting to enter the divide unit (from the RS). This event might count cycles while a divide is in progress even if the RS is empty.  The divide instruction is one of the longest latency instructions in the machine.  Hence, it has a special event associated with it to help determine if divides are delaying the retirement of instructionsMachine clears happen when something happens in the machine that causes the hardware to need to take special care to get the right answer. When such a condition is signaled on an instruction, the front end of the machine is notified that it must restart, so no more instructions will be decoded from the current path.  All instructions "older" than this one will be allowed to finish.  This instruction and all "younger" instructions must be cleared, since they must not be allowed to complete.  Essentially, the hardware waits until the problematic instruction is the oldest instruction in the machine.  This means all older instructions are retired, and all pending stores (from older instructions) are completed.  Then the new path of instructions from the front end are allowed to start into the machine.  There are many conditions that might cause a machine clear (including the receipt of an interrupt, or a trap or a fault).  All those conditions (including but not limited to MACHINE_CLEARS.MEMORY_ORDERING, MACHINE_CLEARS.SMC, and MACHINE_CLEARS.FP_ASSIST) are captured in the ANY event. In addition, some conditions can be specifically counted (i.e. SMC, MEMORY_ORDERING, FP_ASSIST).  However, the sum of SMC, MEMORY_ORDERING, and FP_ASSIST machine clears will not necessarily equal the number of ANYTotal page walks that are completed (I-side and D-side)Retired load uops which data sources were LLC and cross-core snoop hits in on-pkg core cache. (Precise Event - PEBS) (Precise event)offcore_response.all_pf_rfo.llc_hit.hitm_other_coreCounts demand data writes (RFOs) that hit in the LLC and the snoops sent to sibling cores return clean responseevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0200event=0xb7,period=100003,umask=0x1,offcore_rsp=0x2003c0100event=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400120This event counts any requests that miss the LLC where the data was returned from local DRAMevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x300400100REQUEST = PF_LLC_IFETCH and RESPONSE = LLC_MISS_LOCAL and SNOOP = DRAMREQUEST = ANY RFO and RESPONSE = LOCAL_DRAM AND REMOTE_CACHE_HITREQUEST = DEMAND_DATA and RESPONSE = LOCAL_CACHEREQUEST = DEMAND_RFO and RESPONSE = ANY_CACHE_DRAMREQUEST = DEMAND_RFO and RESPONSE = LOCAL_CACHEREQUEST = PF_DATA_RD and RESPONSE = LLC_HIT_OTHER_CORE_HITREQUEST = PF_IFETCH and RESPONSE = ANY_LOCATIONREQUEST = ANY RFO and RESPONSE = ANY_LLC_MISSevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x3008REQUEST = DEMAND_DATA_RD and RESPONSE = REMOTE_DRAMevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x4050REQUEST = PF_DATA_RD and RESPONSE = ANY_LLC_MISSevent=0x3,period=200000,umask=0x2snoopq_requests_outstanding.data_not_emptyOutstanding snoop invalidate requestsdtlb_load_misses.large_walk_completedDTLB load miss large page walksevent=0xb7,period=100000,umask=0x1,offcore_rsp=0x5808event=0xb7,period=100000,umask=0x1,offcore_rsp=0x1850event=0xb7,period=100000,umask=0x1,offcore_rsp=0x5820offcore_response.pf_l1d_and_sw.l3_hit.hitm_other_coreCounts prefetch (that bring data to L2) data reads that hit in the L3 and the snoop to one of the sibling cores hits the line in M state and the line is forwardedevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x4003C0100offcore_response.all_pf_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00120offcore_response.all_rfo.l3_miss.remote_hitmevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x63FC00002offcore_response.pf_l2_data_rd.l3_miss_local_dram.snoop_miss_or_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x604000010Counts all prefetch (that bring data to L2) RFOs that miss the L3 and the modified data is transferred from remote cacheoffcore_response.pf_l2_rfo.l3_miss_local_dram.snoop_miss_or_no_fwdcore_power.throttleStreaming stores (full cache line). Unit: uncore_cha Ingress (from CMS) Request Queue Rejects; PhyAddr Match. Unit: uncore_cha Ingress (from CMS) Request Queue Rejects; PhyAddr MatchCounts snoop filter capacity evictions for entries tracking modified lines in the cores cache. Snoop filter capacity evictions occur when the snoop filter is full and evicts an existing entry to track a new entry. Does not count clean evictions such as when a cores cache replaces a tracked cacheline with a new cachelineevent=0x5c,umask=0x04unc_iio_comp_buf_occupancy.cmpd.part2Counts every write request of 4 bytes of data made to the MMIO space of a card on IIO Part0 by a unit on the main die (generally a core) or by another IIO unit. In the general case, Part0 refers to a standard PCIe card of any size (x16,x8,x4) that is plugged directly into one of the PCIe slots. Part0 could also refer to any device plugged into the first slot of a PCIe riser card or to a device attached to the IIO unit which starts its use of the bus using lane 0 of the 16 lanes supported by the busCounts every peer to peer write request of 4 bytes of data made to the MMIO space of a card on IIO Part1 by a different IIO unit. Does not include requests made by the same IIO unit. In the general case, Part1 refers to a x4 PCIe card plugged into the second slot of a PCIe riser card, but it could refer to any x4 device attached to the IIO unit using lanes starting at lane 4 of the 16 lanes supported by the busevent=0x83,ch_mask=0x02,fc_mask=0x07,umask=0x08event=0xc1,ch_mask=0x08,fc_mask=0x07,umask=0x02Counts every write request of up to a 64 byte transaction of data made by IIO Part2 to a unit on the main die (generally memory). In the general case, Part2 refers to a x4 or x8 PCIe card plugged into the third slot of a PCIe riser card, but it could refer to any x4 or x8 device attached to the IIO unit and using lanes starting at lane 8 of the 16 lanes supported by the busPeer to peer read request of up to a 64 byte transaction is made by IIO Part0 to an IIO target. Unit: uncore_iio event=0x84,ch_mask=0x02,fc_mask=0x07,umask=0x02Multi-socket cacheline Directory lookups (any state found). Unit: uncore_m2m Multi-socket cacheline Directory update from A to I. Unit: uncore_m2m AD Egress (to CMS) Allocations. Unit: uncore_m2m unc_upi_l1_power_cyclesunc_upi_txl_flits.idleProtocol header and credit FLITs transmitted across any slot. Unit: uncore_upi Counts protocol header and credit FLITs (80 bit FLow control unITs) transmitted across any of the 3 UPI (Ultra Path Interconnect) slots on this UPI unitocr.all_data_rd.l3_hit_m.snoop_missOCR.ALL_PF_DATA_RD.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_PF_DATA_RD.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_PF_DATA_RD.L3_HIT.NO_SNOOP_NEEDEDocr.all_pf_data_rd.l3_hit_e.hit_other_core_fwdOCR.ALL_PF_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDED  OCR.ALL_PF_DATA_RD.L3_HIT_E.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1000200490event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040490OCR.ALL_PF_RFO.L3_HIT.SNOOP_MISS OCR.ALL_PF_RFO.L3_HIT.SNOOP_MISSOCR.ALL_PF_RFO.L3_HIT_E.ANY_SNOOP  OCR.ALL_PF_RFO.L3_HIT_E.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100200120OCR.ALL_READS.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_READS.L3_HIT.NO_SNOOP_NEEDED OCR.ALL_READS.L3_HIT.NO_SNOOP_NEEDEDocr.all_reads.l3_hit_e.snoop_noneocr.all_reads.l3_hit_s.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x80200122ocr.demand_code_rd.l3_hit.hitm_other_coreocr.demand_data_rd.l3_hit.snoop_noneocr.demand_data_rd.l3_hit_e.snoop_missocr.demand_data_rd.l3_hit_m.hitm_other_coreCounts all demand data writes (RFOs)  OCR.DEMAND_RFO.L3_HIT_E.NO_SNOOP_NEEDEDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80040002ocr.pf_l1d_and_sw.l3_hit.snoop_hit_with_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080400ocr.pf_l1d_and_sw.l3_hit_e.snoop_noneocr.pf_l1d_and_sw.l3_hit_f.snoop_missocr.pf_l2_data_rd.l3_hit.hitm_other_coreocr.pf_l2_data_rd.l3_hit_e.snoop_missevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200080010Counts prefetch (that bring data to L2) data reads  OCR.PF_L2_DATA_RD.L3_HIT_F.ANY_SNOOPocr.pf_l2_data_rd.l3_hit_m.any_snoopCounts all prefetch (that bring data to L2) RFOs OCR.PF_L2_RFO.L3_HIT.ANY_SNOOP OCR.PF_L2_RFO.L3_HIT.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x803C0020Counts all prefetch (that bring data to L2) RFOs  OCR.PF_L2_RFO.L3_HIT_S.HIT_OTHER_CORE_FWDCounts all prefetch (that bring data to LLC only) data reads OCR.PF_L3_DATA_RD.L3_HIT.ANY_SNOOP OCR.PF_L3_DATA_RD.L3_HIT.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x100080080ocr.pf_l3_data_rd.l3_hit_e.snoop_noneocr.pf_l3_data_rd.l3_hit_f.no_snoop_neededocr.pf_l3_rfo.l3_hit.no_snoop_neededCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_HIT_E.HIT_OTHER_CORE_FWDevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200200100This event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_HIT_F.HIT_OTHER_CORE_NO_FWDoffcore_response.all_data_rd.l3_hit_m.snoop_noneoffcore_response.all_data_rd.supplier_none.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x200020491offcore_response.all_pf_data_rd.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_HIT_E.ANY_SNOOPoffcore_response.all_reads.l3_hit_s.snoop_missThis event is deprecated. Refer to new event OCR.ALL_RFO.ANY_RESPONSEThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_HIT_E.HITM_OTHER_COREoffcore_response.all_rfo.l3_hit_m.hit_other_core_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x3F80400122This event is deprecated. Refer to new event OCR.ALL_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x400020122This event is deprecated. Refer to new event OCR.DEMAND_CODE_RD.L3_HIT_M.SNOOP_MISSThis event is deprecated. Refer to new event OCR.DEMAND_DATA_RD.L3_HIT_F.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_HIT_S.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.DEMAND_RFO.SUPPLIER_NONE.ANY_SNOOPThis event is deprecated. Refer to new event OCR.OTHER.L3_HIT_E.HIT_OTHER_CORE_FWDThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.ANY_RESPONSEoffcore_response.pf_l2_data_rd.l3_hit_e.any_snoopoffcore_response.pf_l2_data_rd.l3_hit_s.no_snoop_neededoffcore_response.pf_l2_rfo.l3_hit_m.no_snoop_neededoffcore_response.pf_l3_data_rd.l3_hit.hit_other_core_fwdoffcore_response.pf_l3_data_rd.l3_hit_s.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT.NO_SNOOP_NEEDEDThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_HIT_E.HITM_OTHER_COREoffcore_response.pf_l3_rfo.l3_hit_f.hitm_other_coreoffcore_response.pf_l3_rfo.l3_hit_m.no_snoop_neededevent=0xcf,period=2000003,umask=0x40ocr.all_data_rd.l3_miss.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x204000491OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOP  OCR.ALL_PF_DATA_RD.L3_MISS_LOCAL_DRAM.ANY_SNOOPocr.all_pf_data_rd.l3_miss_local_dram.no_snoop_neededevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x104000490ocr.all_pf_data_rd.l3_miss_remote_hop1_dram.any_snoopevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x43C000120ocr.all_pf_rfo.l3_miss_remote_hop1_dram.hitm_other_coreOCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_PF_RFO.L3_MISS_REMOTE_HOP1_DRAM.HIT_OTHER_CORE_FWDOCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWD  OCR.ALL_RFO.L3_MISS_LOCAL_DRAM.HIT_OTHER_CORE_FWDocr.demand_code_rd.l3_miss.hitm_other_coreCounts all demand code reads  OCR.DEMAND_CODE_RD.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREocr.demand_data_rd.l3_miss_local_dram.hit_other_core_no_fwdCounts demand data reads  OCR.DEMAND_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPocr.other.l3_miss_remote_hop1_dram.snoop_missocr.pf_l1d_and_sw.l3_miss_local_dram.hitm_other_coreocr.pf_l2_rfo.l3_miss.hit_other_core_no_fwdevent=0xb7,period=100003,umask=0x1,offcore_rsp=0xBC000020ocr.pf_l3_data_rd.l3_miss.hit_other_core_fwdocr.pf_l3_rfo.l3_miss.any_snoopCounts all prefetch (that bring data to LLC only) RFOs  OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_DATA_RD.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.ALL_PF_DATA_RD.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_PF_RFO.L3_MISS.ANY_SNOOPoffcore_response.all_pf_rfo.l3_miss.no_snoop_neededoffcore_response.all_reads.l3_miss_local_dram.hit_other_core_no_fwdThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS.ANY_SNOOPThis event is deprecated. Refer to new event OCR.ALL_RFO.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.DEMAND_RFO.L3_MISS_LOCAL_DRAM.SNOOP_NONEoffcore_response.demand_rfo.l3_miss_remote_hop1_dram.no_snoop_neededoffcore_response.other.l3_miss_local_dram.hit_other_core_no_fwdoffcore_response.other.l3_miss_remote_hop1_dram.hitm_other_coreThis event is deprecated. Refer to new event OCR.PF_L1D_AND_SW.L3_MISS_REMOTE_HOP1_DRAM.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L2_DATA_RD.L3_MISS.REMOTE_HIT_FORWARDoffcore_response.pf_l2_data_rd.l3_miss_remote_hop1_dram.snoop_missThis event is deprecated. Refer to new event OCR.PF_L2_RFO.L3_MISS.HITM_OTHER_COREThis event is deprecated. Refer to new event OCR.PF_L3_RFO.L3_MISS_REMOTE_HOP1_DRAM.ANY_SNOOPoffcore_response.pf_l3_rfo.l3_miss_remote_hop1_dram.snoop_noneocr.all_pf_data_rd.pmm_hit_local_pmm.snoop_not_neededOCR.ALL_PF_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDED  OCR.ALL_PF_DATA_RD.SUPPLIER_NONE.NO_SNOOP_NEEDEDOCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOP OCR.ALL_PF_RFO.PMM_HIT_LOCAL_PMM.ANY_SNOOPocr.all_reads.pmm_hit_local_pmm.snoop_not_neededOCR.ALL_RFO.SUPPLIER_NONE.HITM_OTHER_CORE  OCR.ALL_RFO.SUPPLIER_NONE.HITM_OTHER_COREocr.demand_code_rd.supplier_none.snoop_missocr.demand_data_rd.supplier_none.snoop_missCounts any other requests  OCR.OTHER.SUPPLIER_NONE.HITM_OTHER_CORECounts L1 data cache hardware prefetch requests and software prefetch requests OCR.PF_L1D_AND_SW.PMM_HIT_LOCAL_PMM.SNOOP_NOT_NEEDEDCounts all prefetch (that bring data to LLC only) data reads  OCR.PF_L3_DATA_RD.SUPPLIER_NONE.HIT_OTHER_CORE_FWDocr.pf_l3_data_rd.supplier_none.snoop_noneUNC_M_PMM_READ_LATENCYTag Hit; Underfill Rd Hit from NearMem, Clean Lineocr.hwpf_l3.l3_hit.anyevent=0x60,period=1000003,umask=0x8Counts all microcode FP assistsCounts number of SSE/AVX computational 256-bit packed double precision floating-point instructions retired; some instructions will count twice as noted below.  Each count represents 4 computation operations, one for each element.  Applies to SSE* and AVX* packed double precision floating-point instructions: ADD SUB HADD HSUB SUBADD MUL DIV MIN MAX SQRT FM(N)ADD/SUB.  FM(N)ADD/SUB instructions count twice as they perform 2 calculations per elementUops not delivered by IDQ when backend of the machine is not stalledTOPDOWN.SLOTSINST_RETIRED.ANY / CPU_CLK_UNHALTED.DISTRIBUTEDCounts the number of times an HLE execution aborted due to various memory events (e.g., read/write capacity and conflicts)event=0xb7,period=100003,umask=0x1,offcore_rsp=0x3FFFC08000event=0xc9,period=100003,umask=0x20event=0xc9,period=100003,umask=0x1ocr.hwpf_l1d_and_swpf.dramThis event distributes Core crystal clock cycle counts between active hyperthreads, i.e., those in C0 sleep-state. A hyperthread becomes inactive when it executes the HLT or MWAIT instructions. If one thread is active in a core, all counts are attributed to this hyperthread. To obtain the full count when the Core is active, sum the counts from each hyperthreadevent=0xa3,cmask=8,period=1000003,umask=0x8event=0x5e,period=1000003,umask=0x1event=0xa4,period=10000003,umask=0x2Counts the number of uops executed from any threaduops_executed.cycles_ge_2uops_executed.cycles_ge_3Cycles when the memory subsystem has an outstanding load. Increments by 4 for every such cycleMemoryBW;OffcoreCounts retired load instructions with remote Intel Optane DC persistent memory as the data source and the data request missed L3 (AppDirect or Memory Mode) and DRAM cache(Memory Mode)  Supports address when precise (Precise event)Retired instructions with at least 1 uncacheable load or Bus Lock  Supports address when precise (Precise event)Counts demand data reads that hit in the L3 or were snooped from another core's caches on the same socketCounts demand data reads that resulted in a snoop hit in another core's caches which forwarded the unmodified data to the requesting coreCounts responses to snoops indicating that the data was not found (IHitI) in this core's caches. A single snoop response from the core counts on all hyperthreads of the Coreevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x708000004Counts demand data reads that were supplied by PMM attached to this socket, unless in Sub NUMA Cluster(SNC) Mode.  In SNC Mode counts only those PMM accesses that are controlled by the close SNC Clusterocr.reads_to_core.remote_pmmevent=0xe0,umask=0x01event=0x35,umask=0xC001FF04unc_cha_tor_inserts.io_missTOR Inserts : CRd_Prefs issued by iA Cores that Missed the LLC. Unit: uncore_cha event=0x35,umask=0xC8977E01TOR Inserts : RFO_Prefs issued by iA Cores that Missed the LLC - HOMed locally. Unit: uncore_cha event=0x36,umask=0xc8f3fe04unc_cha_tor_inserts.ia_miss_drd_remote_ddrunc_iio_data_req_of_cpu.cmpd.part0unc_iio_txn_req_of_cpu.cmpd.part2event=0xc1,ch_mask=0x80,fc_mask=0x07,umask=0x01unc_iio_txn_req_of_cpu.mem_write.part5unc_iio_comp_buf_occupancy.cmpd.part6unc_iio_comp_buf_occupancy.cmpd.part4event=0x34,period=200003,umask=0x7Counts the number of cycles the core is stalled due to a demand load miss which hit in the L2, LLC, DRAM or MMIO (Non-DRAM)Counts the number of load uops retired that miss in the L1 data cache  Supports address when precise (Precise event)Counts all code reads that were supplied by the L3 cache where a snoop was sent, the snoop hit, and modified data was forwardedCounts cacheable demand data reads, L1 data cache hardware prefetches and software prefetches (except PREFETCHW) that were supplied by the L3 cache where no snoop was needed to satisfy the requestevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x1F803C0002Counts L2 cache hardware prefetch RFOs (written to the L2 cache only) that were supplied by the L3 cache where a snoop was sent but the snoop missedCounts all data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that were supplied by the L3 cache where a snoop was sent, the snoop hit, and non-modified data was forwardedCounts uncached memory writes that were supplied by the L3 cacheThis event is deprecated. Refer to new event OCR.DEMAND_DATA_AND_L1PF_RD.L3_MISS_LOCALevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x2184000002ocr.uc_rd.l3_miss_localocr.uc_wr.l3_miss_localCounts the number of bus locks a core issued its self (e.g. lock to UC or Split Lock) and does not include cache locksocr.corewb_m.outstandingevent=0xb7,period=100003,umask=0x1,offcore_rsp=0x8000000000000020ocr.uc_rd.outstandingCounts the number of retired loads that are blocked for any of the following reasons:  DTLB miss, address alias, store forward or data unknown (includes memory disambiguation blocks and ESP consuming load blocks) (Precise event)Counts the total number of machine clears for any reason including, but not limited to, memory ordering, memory disambiguation, SMC, and FP assistCounts the total number of issue slots that were not consumed by the backend because allocation is stalled due to a mispredicted jump or a machine clearCounts the number of issue slots every cycle that were not consumed by the backend due to memory reservation stalls in which a scheduler is not able to accept uopsevent=0x74,period=1000003,umask=0x10event=0xc2,period=1000003Counts the total number of uops retired (Precise event)Number of DRAM Refreshes Issued : Counts the number of refreshes issuedevent=0x35,umask=0xC001FE01,config1=0x40040e33unc_cha_tor_inserts.ia_hit_drd_opt_prefevent=0x36,umask=0xC827FE01PCIe Completion Buffer Occupancy : Part 0 : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 0/1/2/3, Or x8 card plugged in to Lane 0/1, Or x4 card is plugged in to slot 0Data requested by the CPU : Core writing to Card's MMIO space : Number of DWs (4 bytes) requested by the main die.  Includes all requests initiated by the main die, including reads and writes. : x16 card plugged in to Lane 4/5/6/7, Or x8 card plugged in to Lane 4/5, Or x4 card is plugged in to slot 4Counts the number of page walks completed due to stores whose address translations missed in all Translation Lookaside Buffer (TLB) levels and were mapped to 1G pages.  Includes page walks that page faultCounts the number of page walks completed due to instruction fetch misses to a 4K pageFloating Point Operations Per Cycle. Unit: cpu_core Counts the number of cycles the core is stalled due to an instruction cache or tlb miss which hit in DRAM or MMIO (Non-DRAM). Unit: cpu_atom event=0x61,period=100003,umask=0x2Retired instructions that are fetched after an interval where the front-end had at least 1 bubble-slot for a period of 2 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core Retired instructions that are fetched after an interval where the front-end delivered no uops for a period of 8 cycles which was not interrupted by a back-end stall (Precise event). Unit: cpu_core Cycles Decode Stream Buffer (DSB) is delivering any Uop. Unit: cpu_core ld_head.st_addr_at_retCounts the number of machine clears due to memory ordering caused by a snoop from an external agent. Does not count internally generated machine clears such as those due to memory disambiguation. Unit: cpu_atom memory_activity.cycles_l1d_missevent=0x47,cmask=5,period=1000003,umask=0x5Counts randomly selected loads when the latency from first dispatch to completion is greater than 32 cycles  Supports address when precise (Must be precise). Unit: cpu_core Far branch instructions retired (Precise event). Unit: cpu_core event=0xc5,period=400009,umask=0x11Mispredicted non-taken conditional branch instructions retired (Precise event). Unit: cpu_core Core crystal clock cycles when this thread is unhalted and the other thread is halted. Unit: cpu_core Execution stalls while L2 cache miss demand load is outstanding. Unit: cpu_core Uops executed on port 0. Unit: cpu_core uops_executed.stallsunc_m_vc1_requests_wrevent=0x1eunc_m_pre_count_idleThis 48-bit fixed counter counts the UCLK cycles. Unit: uncore_clock Counts the number of page walks completed due to load DTLB misses to any page size. Unit: cpu_atom This event is deprecated. Refer to new event L1D_PEND_MISS.L2_STALLSevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x10003C0004event=0x2a,period=100003,umask=0x1,offcore_rsp=0x3F803C0002Counts all (cacheable) data read, code read and RFO requests including demands and prefetches to the core caches (L1 or L2) that hit a modified line in a distant L3 Cache or were snooped from a distant core's L1/L2 caches on this socket when the system is in SNC (sub-NUMA cluster) modeevent=0x2a,period=100003,umask=0x1,offcore_rsp=0x808004477OFFCORE_REQUESTS_OUTSTANDING.CYCLES_WITH_DATA_RDCounts core clocks when the thread is in the C0.2 light-weight faster wakeup time but less power saving optimized state.  This state can be entered via the TPAUSE or UMWAIT instructionsNumber of uops dispatch to execution  port 1event=0xe0,umask=0x0000000002event=0x3,umask=0x0000000011unc_m_cas_count.wr_preunc_m_pmm_rpq_occupancy.gnt_wait_sch0event=0x83,ch_mask=0x0080,fc_mask=0x07,umask=0x0000000001TOR Inserts for CRd misses from local IA. Unit: uncore_cha Matches on Transmit path of a UPI Port : Non-Coherent Standard. Unit: uncore_upi Prefetch CAM Inserts : UPI - All Channels. Unit: uncore_m2m unc_m2m_prefcam_demand_merge.xpt_allchOSB Snoop Broadcast : Local Rd. Unit: uncore_cha event=0x34,umask=0x00001bc1ffevent=0x35,umask=0x0000000004event=0x35,umask=0x00C000FF04event=0x36,umask=0x0000000001TOR Occupancy : IRQ - iA. Unit: uncore_cha unc_cha_tor_occupancy.noncohunc_cha_tor_occupancy.io_hit_itomTOR Occupancy; RFO from local IA. Unit: uncore_cha event=0x36,umask=0x00C8977E01unc_cha_tor_occupancy.ia_llcprefdataTOR Occupancy; LLCPrefData from local IA. Unit: uncore_cha TOR Occupancy; LLCPrefData misses from local IA. Unit: uncore_cha event=0x35,umask=0x00C8970601event=0x35,umask=0x00C80F7E01event=0x35,umask=0x00C88EFE01unc_cha_tor_inserts.ia_miss_remote_wcil_ddrevent=0x36,umask=0x00c8170a01event=0x36,umask=0x00c8970a01TOR Occupancy : DRd_Prefs issued by iA Cores targeting DDR Mem that Missed the LLC - HOMed locally. Unit: uncore_cha unc_cha_tor_occupancy.ia_clflushTOR Occupancy : ItoMCacheNears issued by iA Cores. Unit: uncore_cha event=0x36,umask=0x00cc57ff01event=0x36,umask=0x00c87fde01TOR Occupancy : WCiLFs issued by iA Cores targeting PMM that missed the LLC. Unit: uncore_cha event=0x36,umask=0x00cd43fe04bp_l2_btb_correctic_cache_inval.fill_invalidatedevent=0xc8Retired Near Returns Mispredictedevent=0x7c7,umask=0x02Single-precision multiply FLOPSls_locks.bus_lockevent=0x41,umask=0x01event=0x45,umask=0x80L1 DTLB Miss of a page of 2M sizels_l1_d_tlb_miss.tlb_reload_32k_l2_hitl2_cache_misses_from_dc_missesl2_cache_misses_from_l2_hwpfL1 ITLB Missesl2_itlb_missesL1 Branch Prediction Overrides Existing Prediction (speculative)All FLOPS. This is a retire-based event. The number of retired SSE/AVX FLOPS. The number of events logged per cycle can vary from 0 to 64. This event can count above 15fp_ret_sse_avx_ops.mac_flopsNumber of Scalar Ops optimized. This is a dispatch based speculative event, and is useful for measuring the effectiveness of the Move elimination and Scalar code optimization schemesSSE bottom-executing uOps retired. The number of serializing Ops retiredx87 control word mispredict traps due to mispredictions in RC or PC, or changes in mask bits. The number of serializing Ops retiredfp_disp_faults.ymm_fill_faultRetired lock instructions. Bus lock when a locked operations crosses a cache boundary or is done on an uncacheable memory type. Comparable to legacy bus lockDispatch of a single op that performs a load from and store to the same memory address. Number of single ops that do load/store to an addressls_l1_d_tlb_miss.tlb_reload_coalesced_page_missevent=0x59,umask=0x10event=0x5a,umask=0x02de_dis_uops_from_decoder.opcache_dispatchedThe number of valid fills into the ITLB originating from the LS Page-Table Walker. Tablewalk requests are issued for L1-ITLB and L2-ITLB misses. Walk for 2M pageThe number of MMX, SSE or x87 instructions retired. The UnitMask allows the selection of the individual classes of instructions as given in the table. Each increment represents one complete instruction. Since this event includes non-numeric instructions it is not suitable for measuring MFLOPSDemand Data Cache Fills by Data Source. From cache of different CCX in same nodeAny Data Cache Fills by Data Source. From DRAM or IO connected in same nodels_hw_pf_dc_fills.mem_io_remotels_hw_pf_dc_fills.ext_cache_localInstruction Cache (32B) Fetch Miss RatioRetired branch instructions (all types of architectural control flow changes, including exceptions and interrupts)Demand data cache fills from L3 cache or different L2 cache in the same CCXSoftware prefetch instructions dispatched (speculative) of all typesSoftware prefetch data cache fills from cache of another CCX in the same NUMA nodels_hw_pf_dc_fills.near_cacheL2 cache requests: from hardware prefetchers to prefetch directly into L2 (hit or miss)L2 cache requests: data cache state change to writable, check L2 for current statel2_cache_req_stat.dc_hit_in_l2event=0x70,umask=0x08event=0x72,umask=0x20event=0x72,umask=0x40event=0xd6,umask=0x02event=0x9f,umask=0x7feRead data beats (64 bytes) for local processor at Coherent Station (CS) 7local_processor_write_data_beats_cs4local_processor_write_data_beats_cs7Write data beats (64 bytes) for local processor at Coherent Station (CS) 9event=0x1f,umask=0xbfeWrite data beats (64 bytes) for remote processor at Coherent Station (CS) 2event=0x85f,umask=0x7felocal_socket_inf0_inbound_data_beats_ccm0local_socket_inf1_outbound_data_beats_ccm6event=0x4de,umask=0xbfeData beats (32 bytes) at interface 1 for remote socket inbound data to CPU Moderator (CCM) 1event=0x59e,umask=0xbffRetired SSE and AVX control word mispredict trapsevent=0xa,umask=0x90event=0xb,umask=0x02sse_avx_ops_retired.mmx_cmpsse_avx_ops_retired.mmx_movsse_avx_ops_retired.mmx_otherfp_pack_ops_retired.fp128_macRetired 256-bit packed floating-point compare opsevent=0xd,umask=0x0cumc_act_cmd.alll2_cache_req_stat.ic_dc_miss_in_l2 + l2_pf_miss_l2_hit_l3.all + l2_pf_miss_l2_l3.allL3 cache accesseslocal_socket_inf0_outbound_data_beats_ccm0 + local_socket_inf1_outbound_data_beats_ccm0 + local_socket_inf0_outbound_data_beats_ccm1 + local_socket_inf1_outbound_data_beats_ccm1 + local_socket_inf0_outbound_data_beats_ccm2 + local_socket_inf1_outbound_data_beats_ccm2 + local_socket_inf0_outbound_data_beats_ccm3 + local_socket_inf1_outbound_data_beats_ccm3 + local_socket_inf0_outbound_data_beats_ccm4 + local_socket_inf1_outbound_data_beats_ccm4 + local_socket_inf0_outbound_data_beats_ccm5 + local_socket_inf1_outbound_data_beats_ccm5 + local_socket_inf0_outbound_data_beats_ccm6 + local_socket_inf1_outbound_data_beats_ccm6 + local_socket_inf0_outbound_data_beats_ccm7 + local_socket_inf1_outbound_data_beats_ccm7Estimated memory read bandwidthl3_cache_rduncore_hisi_l3c.rd_hit_cpipe���������c���O�4����5�������
��
mm�y_����������6����������6����������6������n���$����������6����������6����������6������n���$������n���$������n���$����������6����������6��KO
��������������o���������g����������6����������6������n���$����������6��!!
?+%;�P�!�"�@"��$&<p&d�&�0'�p'��'�0( �(H`*p�*��*�0+��+p/(�/L@0l�0��0�p2��2��2038p3X�3|`4��4�05��5  6Hp6p�6� 7��7��78,@8T@<��B�0C��C� D 	�DH	�Et	�E�	G�	�G�	@L
pL<
PMh
�M�
pN�
P�
�T U�V(Xx�Y��[
]\
�^�
�`�
�b@d��e��f$�hp`j�oD�p��r�t(�u4�v`�vt@w|pw�zR|� EA�B
B��@04A�B
A�o(`P�A�B
I����A$��!+A�B
F���!$��"RA�B
E��DA �#WA�B
B��Q$@#VA�B
E��HA(x#;A�B
A�v$H�#RA�B
E��DA$p�#YA�B
E��KA$�$YA�B
E��KA$�@$�A�B
D������%%A�B
A�`�%%A�B
A�`$(&jA�B
H��YA$PP&aA�B
E��SA$x�&�A�B
L���|A �@*zA�B
A�u��*BA�B
A�}��*?A�B
A�z�*1A�B
A�l$$+�A�B
B��qA L�,;A�B
B��up�,1A�B
A�l��,1A�B
A�l��,4A�B
A�o �-PA�B
A�K(�<-�A�B
C����A$ �-zA�B
E��lA$H.LA�B
E��~A$p0.xA�B
E��jA$��.aA�B
H��PA$��.LA�B
E��~A$��.LA�B
E��~A  /WA�B
B��Q$4\/_A�B
E��QA$\�/LA�B
E��~A��/"A�B
^$��/7A�B
[AW(��/�A�B
F���uA(��3�A�B
F����A$$<:9A�B
YA[$LT:�A�B
C���� t�:4A�B
B��n$��:uA�B
D���m(�P;�A�B
F����A �<NA�B
A�I(@<(A�B
I���A(<D=�A�B
I����A(h�=WA�B
F���CA�B#A�B
A�^(�,B�A�B
F����A$��BrA�B
D���j$8C�A�B
F����$0�C�A�B
F����(XE�A�B
I���hA$��IMA�B
B��nC(��jA�B
I����A�hkA�B
A� ��k/A�B
B��i 	�k5A�B
B��ozPLR|�y�@$4I���A�B
F���k.y.R.[CB.R.F.P.K.Lh�Ig�A�B
I���`.�.4I.m.G.].mCB.Q.V.T.H��K�7�A�B
I���`.g.$v.m.G.].mCB.Q.V.T.H$M��A�B
I���`.O.e.m.G.].mCB.Q.V.T.HP�Ne��A�B
I���`.F.^.m.u.W.hCB.K.P.N.H��O���A�B
I���`.g.$v.m.G.].mCB.Q.V.T.H�@Q���A�B
I���`.g.,z.m.G.].mCB.Q.V.T.H4�R�[�A�B
I���`.^. p.m.G.].mCB.Q.V.T.H�XTe/�A�B
I���`.F.^.m.u.W.hCB.K.P.N.H�|Ua�A�B
I���Z.C.].s.u.W.hCB.K.P.N.H�Va��A�B
I���Z.C.].s.u.W.hCB.K.P.N.Hd�W���A�B
I���`.O. g.m.G.].mCB.Q.V.T.H�(Y��A�B
I���`.U.$n.m.G.].mCB.Q.V.T.���Z�S�A�B
I���C.Z.$p.t.O.Y.k.a.X.`.}.B.L.X.Y.B.x.T.W.gCB.O.O.O.P.U.[.V.K.H��^��A�B
I���`.^. p.m.G.].mCB.Q.V.T.H�8`���A�B
I���`.^. p.m.G.].mCB.Q.V.T.H�aa��A�B
I���Z.C.].s.u.W.hCB.K.P.N.Hh�b���A�B
I���`.O.e.m.G.].mCB.Q.V.T.0�tePk�A�B
C���O.L.J.V.I.U��SV�[�Ú��X�t�������V�����a����H��r�Ѓ���^[]Ð����������U��S�[��K��\�t���a�t���a�P��U��[]�������������U��SWV����p�[���E��`���L$h�����t$�MW�)D$ )D$0)D$@)D$P�D$`�L$$�M�D$4�L$0�M�L$<�M �L$@�u�D$ �u�9U���ljD$���{��L$�D$PQ�+U�����1U��t�D$ PV�2U����tYW�5U��������}t[���tU��������������PW�U����t'�����u��-�U�������_����v��T���ƅ�uW�T���Ɖt$�t$�D$ P�}T���D$��������t$1���l�D$�|$���F�|$9�t6�D$����t�xW�0�t$�D$�T�T$����u΋L$����>���l�D$���D$�T���҉T$t�L$���t	��1���u9��vrӅ��t$tS�@�D$,�T$�J�L$(�L$ Q�t$ P�R����x-�D$ Pj�������T�����������u�E�L$8�1����S��������t
V���S����t$�;L$hu�e�^_[]��S��������������U��SWV�� �[��f�B�E؅����z��E�d���E�΍h��{S�E䋃p���E�F�u�1��!������9���E�@�}��;E����E��U�҉}�tԋE�����������������9�u�A�@��t����t����W�w
�`��r��_u�S�w
�`��ră�_t�9}�9]�~���U���DU苼�49]�~��U���DU苜�4�1���E�� ^_[]Ð����U��SV���[��7�E�M��`���U�M�E�E�Pj�������4R���;M�u��^[]��.R��������������U��SV�[��������t �M�u	�����������9r�t��Hu��Q���������1�^[]Ð��������U��SV���[��w�E��`���M��E��E�E�Pj�������sQ���;M�u
����^[]��jQ����������U��S�[���������t�E����1���Q������[]����U��SV���[����E�M��`���U�M�E�E�Pj	��������P���;M�u��^[]���P��������������U��SV���[��w�E�M��`���U�M�E��E��E�Pj�������mP���;M�u��^[]��gP�������U��SV���[���E�M��`���U�M�E��E��E�Pj�������
P���;M�u��^[]��P�������U��SWVP�[�ø�E�������a���������a�����L�������a�P����p��t���������1���������8��,n��Dd�D��}���=tX=u^�����T�\��d����O��D��2�B��|���5�������(��N�������k����w�"��P��N�������M�U��9��tp��	rR�����E�)ύ��U��1Ґ������L���L�0T� �ш�T���9�uՋu�؉�1ɐ������������Ή�A9�u�1���������^_[]Ð���U��S�[���jj�������N��[]Ð����������U��S�[�û�jj��������M��[]Ð����������U��SV����0�[�Ä���`���D$,��Pj�������M����x�E($(D$@1��������;L$,u�e�^[]��fM������U��SV���[����E��`���M�E�E�Pj
�������M����x�E�M�1��������;M�u��^[]��L���������������U��SWV������[�à���`����$��������t1��;�$����e�^_[]Í�{�P��L�����������x�DŽ$���$�PQ�L��������$��������D$<
�L$<QjP�4L�������D$?��=
��f�f�$�f�$�f�$�f�$�f�$�fD$pfD$`fD$PfD$@�D$@Pj�������K�����5�t$foD$@f��fp��f~Dž�t�D$P�������QPR��K��jW��K����l���#ǃp���p�Vj�������@K��������t��$���1��<$���$��fo�	��������fo�L	�fn�fp�PfT$fo�,	�fT$ fo�	�fo�<	���fo�f��fp��fp��f�fp��fb�f�t$fo�f�|$ fnWDfnfb�fb�fn��fn��fb�f��fb��z��2f�Ń� �����u�9$t.�$)���������t�����z��x��:�8����DIu�$��������p��������������D$�����l�$�����L$ �������L$�����L$�4$� ����������L$�|$��G�|$��Ht9�:���L$ t�tۃ�u�L$���aI�Kǃ����������������������t$t+=����������ǃ��������������t$�������������o����'I�������U��S�[����E= wP�����UUUU��)щʁ�3333����3333щ��ʁ�i�����u�������Eȋ������
�H�1�[]�����U��S�[��[��E��w
������r�IH�1����������(��[]Ð�������������U��S�[��������*�U��9P�t��Iu���G�1���[]ÐU��S�[����E��w	������
�G�1�[]Ð��������������U��WV�Z�Š��M�������Ps!���a��Dd�����9Ht<��9�u�1��3�}��%���=@u&��tf1���u��,n���w�뿋^_]Í����=�w�����1���s⋴���������돍����=�s"������|���q�����Dd��,n��`����������s��|���L���D���������s��L���d���(�������=�s��d���D���
���������2s��D�������������=�s���������������u���a���a�����΁��1��������t����������p��p����U��SV�[�������u�qF����u����E�����^[]Ð����U��S�[�Û��E��w	������
�E�1�[]��������������U��S�[��[��E��w	���$���
�JE�1�[]��������������U��S�[����������t�����
E������[]Ð�����������U��S�[����������t�E��x9��~�����D���D������[]�U��SWV�[�É����������tP�u��xV9��~N����������xOk�d��Pj��D����t5�0��PjW�cD����x+�E�01���.D���!D������^_[]�V��C���됐����U��SV���[�����E��`���M�E��E� �E���E����E�Pj��������C����x�E�E��1��������;M�u��^[]��C������U��SV���[��g��E��`���M�E�E�Pj�������jC���;M�u��^[]��dC����U��SV���[�����`���E�E�E�E�E�0�E�E�E�E�Pj�������C����x�E�E��1��������;M�u��^[]���B��������U��SV����[�Ô���`���D$�E�D$�E�D$�E�$��Pj
�������B���;L$u
���e�^[]��B���������������U��SV���[��'��E��`���M�E�E�Pj�������*B���;M�u��^[]��$B����U��SV���[�����E��`���M�E�E�Pj��������A���;M�u��^[]���A����U��SV�[�Ê������t �M�u	�����������9r�t��Hu��`A���������1�^[]Ð��������U��SV���[��'���`���E�E�E�E�E��E�E�E�E�Pj�������A���;M�u��^[]��A�U��SV���[�����E��`���M�E�E�Pj��������@���;M�u��^[]���@����U���M��������r�E�`1�]Ð�������������U������}u�M��t�9t]ËE�@(�@$� 1�]Ð��������U��SWV��<�[����}�U��`���E�`�G(�G$�E��€���E���O��E�MĉU������`��1�� �����\���������~��|���v��X���n�� ���f�����^��h���V������N��<���F�����>������6������.�����&��$�����������(�����T���������E��E��Mԍ��{��M�E�PQ�>�������ƋEԃ��E��E���/��E����k�E��E��*����������������u��EP�Q>���ƅ���j�u�V�>�����~�u�V�e>����t���Nr�PV�Q>����������jPV�W>��������y*�PV�>�����'���J�PV�>�����N��W�������j=V�>��������x�-��Gj�E�PW�>���M�9���9�}����	G$����������}���j=V�E>���H�M�x��������E�P�E�P�=����t_�ƋEԋ�����}̐�����PV�9=����t�����u��y��������E�G��E܋E�G��E�uЍE�P��<���ƅ�u��}�O�P�����H����E��Eȃ�Awq�������1��Gt�E�����G$�G(�������Eċ;E������<^_[]Ëu܉��U؉у�!�!Ѿ����	�uˋE�	E�u���>�E�	E����M���u�1��Mԃ9t��1Ґ��Q�A��9�Iu��	1��-���1҉U܀O�P����V<������U��SWV���[����}�M�1��u����P��t������@�����uK��u��}���^��P�{<���?�e��@����V��P��h��DP�R<��������P��D��H�6�����}��H��t(�W�U싗<�|�}�tV1������L���LJ<�W���L�W�M�Q�U�R����;������u��<s0���4��aPQ�u���;���u�<���u���ҋM�������9����ʉ�)�QRV���m;���u��<���)�WQ�u�E���@;�E����M�Q �U�)���<���<��u��6���V����M��}�)�1�9�L���ҋu�W�u�u���:���<}����)�PQV����:���u��<�H)�P����}�������t&�����tb�9��������9#t\��T�F�G�F��������uC�N$�O�����ɉOI��w;�Ћ�������F,�G$���������^_[]Í��*�P��9��������F,�G$�F0�G(���F,�G$�F0�G(�F4�N<�G,�O4�FD�G<�����=������B�N,�O$�N4�U�V<�O(�O,�!����=�������B�N,�O$�N4�U�V0�W(�W,PQR�B�F,�G8�F0�G0�N4�O4�F<�G(���V=��i�U�vP��8�I�F,�G$�F0�G(�F4�G,�F<�G4�FD�w<j@PV�U��9���E�@D�HL�G<�OD�@T�GL�@\�GT�@d�G\�@l�Gd�@t�Gl�@|�GtV�u�C7�U����L�G$�F�G(�F�F�T�F,�G0�F4�G,�F8�G$�F<�G(�3�F,�G$�F4�G(�F<�G,��F,�G$�F0�G(�F4�G,�F8�G0�J����G81���r&��������D�D�D�@�A��;O8r�=���@�L�<���U�)�RjQ��7��F,�G(�F0�G$�F4�G,������=�������B�N,�O$�N0�O(�N4�O,�N<�U�VD�O0�O4PRQ�M�F,�G$�F0�G(�F4�G,�F<�ND�G0�O8�FL�G@�*�F,�O$�Gl�F0�Gd�F4�Gh�F<j@PQ�U��17�U���F�Nу��N�F�G�O�F@�F�G�W �1������U�P��L�7���G$���U�u���G$��0��P�6�����{���������������Ph
����J�����s�Ph
QR�6������������U��M�������x�U��Dt]Ã�Pu��E��H��P1�]Ð������U��SWV�[���hT�d5��1���t|����@������LJLh4jP��5���E��@LJHLJDLJP���th�4����D��t���	W�r4����^_[]Ð�������U��SV�[��j��u��D��t	P�<4��V�34��^[]�������������U��SWVP�[��(��u�M�b�E����������t:�p��1��������u�P��3����tG��<�F�u�>u�~�u��������^_[]Ð����������U��SWV��L���[�ä���`���E��E�@�E��U����jjPRQ�4�����tf�����P�E�P�4�����}�u���?�P�E�P�4����tF�������t&���������PV�3����t�����u���w��}��;E�u)���L^_[]Í�e��P�E�P�@4����t؍�0����3������������U��S�[�û��M�8��t-�@kM<���u!��)������Q/�Ph*QR�31�[]Ð�U��SWV���x�[��c���`���D$p�D$@��t�D$ j@QP�C3���&�D$�L$ ���jjPQR�d3��������<$��t��>�����D$�����L$� �����W�t$�F3���~�����D$t]jWP�<3����u�j�D$dPj�D$,P�|$ W�-3���D$W�03���|$u��D$dD$`u��D$h�D ��t	��-u��1��<$�;D$pu��e�^_[]�1����1��������U��SWV���H�[��3��M��`���D$@�d�����1������t>�|$�x��������������VP��0����t"��<�G�u�?u��u�|$1ɸ�1�W��ҋ|$��t�L$�21Ʌ�DL$�DD$�;T$@u
�ʍe�^_[]��0��������U��SWV��<���[��d���`���E�R�/���E���E܉u�W��G0�G(�G �G�G�G�G��}�f�G0�����{��M�E�PQ�/���E��������E���%+�E���G�Eč�-
��Eȍ�G��E̍�����EЍ�S��Eԍ�Nr�E��0��jjV�V0���M�f�A0�u��E�P�0/���E���u��E�P�/�����(�Nju�u�P��0����t��u�W��0�����r�u�W��0�����v�u�W�0������u�W�0�������u�W�0�������u�W�y0���������k�PW�a0��������͕�PW�I0��������C��PW�10�����������PW�0���������Q�PW�0��������L��PW��/���������Q�PW��/���������k�PW�/��������\��P�/�����k�������QP�/�����S������K������G�VWP�)/���5����jjV�v.���M�A ����j
jV�^.���M����A���jjV�A.���M�A���A���j
jV�#.���M�A3����j
jV�.���M�A2���j
jV��-���M�A4���j
jV��-���M�A5���jjV��-���M�A$�i���jjV�-���M�A(�Q���jjV�-���M�A6�9���jjV�{-���M�A7�!���jjV�c-���M�A,�	���jjV�K-���M�A���A���jjV�--���M�A���A�����u��1,��1��u����;M�u
��<^_[]���-�Y,���������U��S�[���1����1Ʌ�����[]Ð������������U��SWV��H�[������`��E���E�\��P�q-����t����QP�M-�����E���E�1�������t]�p����>��E��
���>t?��<�~�t�}t�uW�,����t�W�u��,���}�u΋V��M�����~�u��E��;E�u��H^_[]��O+���������������U��SWVP�[���1��v�����tL�x����z���E��	��?t/��<�w�t�uV�,����t���t�PV�u��,���у�u˃�^_[]Ð�������������U��SWV���[��v�1�����ti�x����z���E���E����������t=��<�w�t�uV�+����t���tPV�u���G���t�PV�u��r+���Ã�u���^_[]Ð��������������U��SWV��(�[����1��T������`�x��0�� ��E̍�
��Eԍ�j/��E؍�;��E܍�
��E���E䍃���E荃���E썃����E�>��E������������������<�wЅ�t�uV�*����t�V�u��*���G�u�G܅�tP�u��y*���G؅�tP�u��f*���G�tP�u��S*���G�tP�u��@*���G�tP�u��-*���G�tP�u��*���G���tP�u��*���G���tP�u���)������1���P�u���)��� �����������(^_[]Ð���U��SWV���[��C���`���d�����E��E�@�E��u����jjPVQ�X)������������PV�O)����u���?�PV�;)������Džh����MW��E�@<�@4�@,�@$�`����p���1�1��"������w�M��$��l����p�����������p���P�d'����tG��<�F�u�>u�~�u��/�E�x�V������t�����������h����6�����P��p���W�^(��9��ut��ʹ�PW�G(�������F	���p�����M���~$�E�����1�f��Džh���N����M���	��	�N(�N(�V$��|����F,���p������
����l�������p��������������e��PV�'�������h����M����E�@��@���d����	;M��\�Đ^_[]��E�f���}~
�O����	G$�G1��D�p������������������<���t���p���P�%����u�v��++�VP�$'����tu�����VP�'�����E���������G(O$�O$�G(�G,�1����p�����t^��tY���p�����uI���u?�}��l����F�E������G(�G,�U��u�������	�	�	O$�W(�	�	���l�����F�}�}�t���F��U���p���u�����΁����D�����p���D΋�l����B�
���l����h���	���N�}�}�t���N��}�t�� �N��E���p���u��ti�ց����D��
��p���DƋ�l����J��>�M���tB�F���p�������������	�%	O$	G(�G,�O1�����F���p���1�������#�����U��SV�[��z��M�A�P���s�uQV��������^[]��P���vፋԴ�PQj@�%���U��SWV���[��&��u�U��`���E�W���F�M��d �E�F�E���E��E卋���9�w�U�9�w:f�E�}
�E�jPV�$���E�t�u��$���;E�u���^_[]���"��E�u�uW�$�u��q$���t�v�a$��W�h$��������U��SWV��<�[��C��}��`��������E�������������O8�w4�G,�������G0������G$�������G(�������������������������������������t����������������<�������l�P�����QV�����������W�����������������Rh���V�u#��4V�|#�����������
w��ȋ}��G�/�΃�F�����V������S#���}�G���7������O����ˍ49���9�w��uS��t�ً����QRP�`!�������t����������"��������;E�u"����<^_[]�������u��������� �����t����������W"�������V�X"��������U��SWV��0�[��3��u��`��������E����������q�F(������F,�N4������V0�N8��������������������������t����������������b���P����������R�����VQhW�!��$W�!�����������
w��ȋu��F�/�σ�G�����W������x!���u�F���>������N����ˍ<9���9�w��uS��t�ً����QRP��������t����������� ��������;E�u"���0^_[]�������u�&��������������t����������| �������V�} �������������U��SWV��(�[��S��}��`��������E�������������G$�O(���������������������t�����������,�������a��WQPRVh���V����V�������������
w��ȋu��F�/�σ�G�����W���������u�F���>������N����ˍ<9���9�w��uS��t�ً����QRP���������t������������������;E�u"���(^_[]�������u�o��������!�����t������������������V��������U��SWV���[�ã��u��`���T�����E�X��������F$�N(��`�����l�����X�����d�����d���t��l������e��������p���QPRWh�V�C��V�J��������ƒ�
w��ЋM��A�'�׃�GW���+���M�A���9�q��p����<9���9�w��uF��tRVP�F�����d���t��l��������T����;E�u�E�Ġ^_[]��u��������d���t��l����T��V�[�����������U��SWV���[��3��u��`���L�����E�X�����T����q�F0��P����V4�F$��D����F(�N,��H�����`�����l�����X�����d�����d���t��l������e�����T�����Y���p���R��P���P��D�����H���VQh�W���$W������������
w��ȋu��F�/�σ�G��T���W��P����x���u�F���>��P����N��p����ˍ<9���9�w��uS��t�ً�T���QRP������d���t��T�����l���������L����;E�u"��İ^_[]���T����u�&����T��������d���t��T�����l����|����T���V�}�������������U��SWV��0�[��S��}��`��������E�������������W0�G,������G$������G(��������������������������t�����������4�������I����WjRj���������������PQhV���,V������������
w��ȋ}��G�/�΃�F�����V���������}�G���7������O����ˍ49���9�w��uS��t�ً����QRP��������t�������������������;E�u"����0^_[]�������u�B��������������t�����������������V����������U��SWV���[��s��u��`���L�����E�X�����T�����F,��P����N0�F$��H����V(��`�����l�����X�����d�����d���t��l������e�����T������r��p���Q��P�����H���RVPh�W���� W�������������
w��ȋu��F�/�σ�G��T���W��P��������u�F���>��P����N��p����ˍ<9���9�w��uS��t�ً�T���QRP�������d���t��T�����l����$����L����;E�u"��Ĭ^_[]���T����u�u����T����'���d���t��T�����l���������T���V��������������U��SWV���[�ã��u��`���T�����E�X��������F$�N(��`�����l�����X�����d�����d���t��l������e���������p���QPRWh�V�C��V�J��������ƒ�
w��ЋM��A�'�׃�GW���+���M�A���9�q��p����<9���9�w��uF��tRVP�F�����d���t��l��������T����;E�u�E�Ġ^_[]��u��������d���t��l����T��V�[�����������U��SWV���[��3��u��`���E�X������}�F$��`�����l�����X�����d�����d���t��l������e�����3+��p���PQRh�V����V����������‰�T�����
w��ЋM��A�'�׃�GW������M�A���9�q��p����<9���9�w��uF��tRVP�������d���t��l����0����T����;E�u�E�Ġ^_[]��u����>���d���t��l�������V�����������������U��SWV���[�����u��`���E�X������
�F$��`�����l�����X�����d�����d���t��l������e���������p���PQRh�V�m��V�t��������‰�T�����
w��ЋM��A�'�׃�GW���O���M�A���9�q��p����<9���9�w��uF��tRVP�j�����d���t��l��������T����;E�u�E�Ġ^_[]��u��������d���t��l����x��V����������������U��SWV��(�[��S��}��`��������E�������������G$�O(���������������������t�����������,�������}Y�WjQPRVh���V���� V�������������
w��ȋu��F�/�σ�G�����W���������u�F���>������N����ˍ<9���9�w��uS��t�ً����QRP���������t������������������;E�u"���(^_[]�������u�m�������������t������������������V������U��SWV��(�[�ã��u��`��������E������������N,�F$������V(���������������������t���������������������jQjR�����PVhW�$��$W�+�����������
w��ȋu��F�/�σ�G�����W���������u�F���>������N����ˍ<9���9�w��uS��t�ً����QRP��������t����������_��������;E�u"���(^_[]�������u���������b�����t�����������������V��������U��SWV��,�[����M�u��`��������E�W���A����Dž�����������F4������V0�F$������~(�F,������������������������������E����G��������RW����������PQhV�:��$V�A������*����
w��Ȉ����������u�+�΃�FV����������������������u��������<���9���9�w�����������tQ���QP�
����t�v�s�������F������1��u�~8uS�����j�t�<Ph���V�R��V�Y������Q����
�Q��Ȉ����������d1���'s��������������������G�F8H9�s�j�t�<�����h���V��
��V��
����������������
w��Ȉ����������.�������΃�FV���
���������������������ύ4���9���9�w���P��tW���QP����������t����������������������荍����U�u�����PQR�����������������������΃�FV�������������������������ύ4���9���9�w������tW���QP�
���������t�����������荍����uPQV�)�������t�����"������t�������������;E�u>���,^_[]������P�^�������P�O�������P�@���	������t'��!���������t�������������t�����r���E�t�p�_��W�f������U��SWV��,�[��C��u��`��������E�����������F$������Fd������Nh�Vl���������������������t�����������������?��������Q�����RVPhW�
�� W�
�����������
w��ȋu��F�/�σ�G�����W������
���u�F���>������N����ˍ<9���9�w��uS��t�ً����QRP��������t�����������	��������;E�u"���,^_[]�������u�E�������������t����������	�������V�	������������U��SWV��,�[��s��}��`��������E�������������G,������G$������W(���������������������t�����������0�������A���W�����R�����QPhV���� V�������������
w��ȋu��F�/�σ�G�����W����������u�F���>������N����ˍ<9���9�w��uS��t�ً����QRP���������t����������$��������;E�u"���,^_[]�������u�u��������'�����t������������������V��������������U��SWV�� �[�ã��u��`���E���������F$���������������������t����������������PQRhV�M��V�T��������‰������
w��ЋM��A�'�׃�GW���/���M�A���9�q����<9���9�w��uF��tRVP�J�������t�������������;E�u�E�� ^_[]��u���������t�����X��V�_���������������U��SWV��(�[��3��}��`��������E������������q�G$�O(���������������������t�����������,�������
s�WQPRVh���V����V�������������
w��ȋu��F�/�σ�G�����W���������u�F���>������N����ˍ<9���9�w��uS��t�ً����QRP��������t������������������;E�u"���(^_[]�������u�O�������������t�����������������V�������U��SWV����l����[��}���`���d�����E�B�J�����0��p���RP���,��Wh�V�U��V�\�����������
��h���w��ȋ�l�����F�(�σ��GP�4��h�������l����F���>�N��p����<9���9�w��u2��h�����tQRP�B�����d����;E�u��Đ^_[]���l�����U��S�[��{���T��P�����U��SWV�[��Y��}j�����WP�(����x���|�PQV���V���W� U��SV�[��
��u�uV�s���������^[]��U��SV�[��ڱ���a�����������H��r�Ѓ���^[]����U��������]�U��������]�����������h�����h������h�����h�����h ����� h(�����$h0�����(h8�p�����,h@�`�����0hH�P�����4hP�@�����8hX�0�����<h`� �����@hh������Dhp������Hhx����Lh������Ph�������Th������Xh������\h������`h������dh������hh��p�����lh��`�����ph��P�����th��@�����xh��0�����|h�� ������h��������h��������h������h������h�������h������h������h ������h(������h0������h8�p������h@�`������hH�P������hP�@������hX�0������h`� ������hh�������hp�������hx����������p�%i
 �
6�
j,��� `�� �� ��� ��� ��� �@� ϒ� �@� �� {
� �t
� �2"� �p� �� d3� N�� �>� �� ,�� �� �>� ��� ��$� ��� �#� �� N�� �2"� ?� J� n�� �� =�� j� ��� �
� �t
� �p� �� ��� q,� ��� ^�� A� ��� U� �$� ?� �
� Y?� B�� ��� U?� 4U� ��� K�
� ��� � � ^�� ��� ��� #�� ��� 
�� EU� �
� �� �� �?� 1�� \� }�� �t
� �
� ��� 0A� r?� ^�� �?� ��@�@F�@C\@��@H�#@��@<� @ī	@R�
@@T�#@T\
@]�#@�p@@�#@�2"@E�@#@�p@Ϋ@��@�@�?@`!@T�@��@X�@p3@��@�@@(A@�?B@s�C@&�D@#E@_�F@8G@i�H@�I@4�J@�?K@۫L@��M@�?N@�O@@AP@��Q@{3R@�S@}�T@#U@�2"V@`!W@�X@)`!Y@�?Z@<�$p@�q@�t
r@G@^�@C`!@F� @�2"@��@�@b\@r�#@Q	@��
@J�$@�,@�
@�@c�@�t
@|�#@��@M`!@��@u
@T�$@#@��@P� @y�
@)#@h�@�@WU@�@^�$ @m�!@�2""@��#@��$@ñ%@l\&@3"'@Ͼ(@�?)@�*@W`!+@��,@�,-@v\.@H�/@
�0@[1@3#2@�?3@�?4@u
5@=#6@r�7@��#8@��9@�:@�?;@R�<@�,=@aU>@3"?@\�@@�3A@��B@�C@�pD@h�$E@r�$F@�G@kUH@��I@a`!J@�?K@|�$L@�M@��
N@�O@|�P@�?Q@ͱR@��#S@�?T@U@�,V@��$W@�X@��Y@��Z@��[@�\@��]@^@��_@0�`@��a@��#b@w�c@uUd@��
e@�f@�pg@�?h@��i@Ŋj@��k@�3l@f�m@��$n@ױo@WAp@G#q@�r@پs@Ut@��u@��v@�,w@x@��y@��z@��{@e|@p�}@ ~@��$@��@��@:��@%��@Z� �@���@�?�@���@u
�@@�@�U�@���@�,�@�\�@��@(��@��@���@��
�@�\�@��@o�@d� �@��@��@�p�@@�@�3�@��$�@��#�@��@ u
�@D��@���@���@��
�@���@��@��#�@*�@4�@�,�@y�@ϊ�@@�@�\�@��@���@��@�,�@��@�U�@��@q�@���@�\�@aA�@��@&@�@ي�@2��@
q�@/��@��@���@>�@���@�,�@��@�,�@���@���@��@���@3"�@n� �@���@<��@���@q�@H�@���@"3"�@�,�@Q#�@!q�@k`!�@��@	��@R�@��@�?�@��@��
�@��$�@��$�@,3"�@*u
�@+q�@���@@�@��@���@9��@�\�@��@4u
�@��@�U�@0@�@��@���@x� �@-�@\�@kA�@��#�@@�@�� �@u`!�@�U�@uA�@F��@5q�@��@@�@�\�@�� �@��@�@F�@C\@��@H�#@��@<� @ī	@R�
@@T�#@T\
@]�#@�p@@�#@A@@��
A@�B@�UP@��
Q@f`@��$a@�� b@��
c@#@d@�� e@0@f@)�g@63"h@��n@8�p@[#q@��r@�s@>u
t@v#�@
-�@�A�@�#�@��@`!�@���@�@��@��@���@N@�@?q�@��#�@�A�@Rq�@���@z��@N��@G@^�@C`!@F� @�2"@��@�@b\@r�#@Q	@��
@J�$@�,@�
@�@c�@�t
@|�#@��@M`!@��@u
@T�$@#@��@P� @y�
@)#@h�@�@WU@�@^�$ @m�!@�2""@��#@��$@ñ%@l\&@3"'@Ͼ(@�?)@�*@W`!+@��,@�,-@v\.@H�/@
�0@[1@3#2@�?3@�?4@u
5@=#6@r�7@��#8@��9@�:@�?;@R�<@�,=@aU>@3"?@\�@@�3A@��B@�C@�pD@h�$E@r�$F@�G@kUH@��I@a`!J@�?K@|�$L@�M@��
N@�O@|�P@�?Q@ͱR@��#S@�?T@U@�,V@��$W@�X@��Y@��Z@��[@�\@��]@^@��_@0�`@��a@��#b@w�c@uUd@��
e@�f@�pg@�?h@��i@Ŋj@��k@�3l@f�m@��$n@ױo@WAp@G#q@�r@پs@Ut@��u@��v@�,w@x@��y@��z@��{@e|@p�}@ ~@��$@��@��@:��@%��@Z� �@���@�?�@���@u
�@@�@�U�@���@�,�@�\�@��@(��@��@���@��
�@�\�@��@o�@d� �@��@��@�p�@@�@�3�@��$�@��#�@��@ u
�@D��@���@���@��
�@���@��@��#�@*�@4�@�,�@y�@ϊ�@@�@�\�@��@���@��@�,�@��@�U�@��@q�@���@�\�@aA�@��@&@�@ي�@2��@
q�@/��@��@���@>�@���@�,�@��@�,�@���@���@��@���@3"�@n� �@���@<��@���@q�@H�@���@"3"�@�,�@Q#�@!q�@k`!�@��@	��@R�@��@�?�@��@��
�@��$�@��$�@,3"�@*u
�@+q�@���@@�@��@���@9��@�\�@��@4u
�@��@�U�@0@�@��@���@x� �@-�@\�@kA�@��#�@@�@�� �@u`!�@�U�@uA�@F��@5q�@��@@�@�\�@�� �@:@Ae�A��ARu
A�A�Acu
Aī	AP�
AQ�A�AA�Av�A#Apu
A�A-A�� A�AA`!Ac�Ap3A�A�U`AD3"aA��#zA�A�AR3"�A#�Ap�A�A�
A -A2-AGA^�AC`!AF� A�2"A��A�Ab\Ar�#AQ	A��
AJ�$A�,A�
A�Ac�A�t
A|�#A��AM`!A��Au
AT�$A#A��AP� Ay�
A)#Ah�A�AWUA�A^�$ Am�!A�2""A��#A��$Añ%Al\&A3"'AϾ(A�?)A�*AW`!+A��,A�,-Av\.AH�/A
�0A[1A3#2A�?3A�?4Au
5A=#6Ar�7A��#8A��9A�:A�?;AR�<A�,=AaU>A3"?A\�@A�3AA��BA�CA�pDAh�$EAr�$FA�GAkUHA��IAa`!JA�?KA|�$LA�MA��
NA�OA|�PA�?QAͱRA��#SA�?TAUA�,VA��$WA�XA��YA��ZA��[A�\A��]A^A��_A0�`A��aA��#bAw�cAuUdA��
eA�fA�pgA�?hA��iAŊjA��kA�3lAf�mA��$nAױoAWApAG#qA�rAپsAUtA��uA��vA�,wAxA��yA��zA��{Ae|Ap�}A ~A��$A��A��A:��A%��AZ� �A���A�?�A���Au
�A@�A�U�A���A�,�A�\�A��A(��A��A���A��
�A�\�A��Ao�Ad� �A��A��A�p�A@�A�3�A��$�A��#�A��A u
�AD��A���A���A��
�A���A��A��#�A*�A4�A�,�Ay�Aϊ�A@�A�\�A��A���A��A�,�A��A�U�A��Aq�A���A�\�AaA�A��A&@�Aي�A2��A
q�A/��A��A���A>�A���A�,�A��A�,�A���A���A��A���A3"�An� �A���A<��A���Aq�AH�A���A"3"�A�,�AQ#�A!q�Ak`!�A��A	��AR�A��A�?�A��A��
�A��$�A��$�A,3"�A*u
�A+q�A���A@�A��A���A9��A�\�A��A4u
�A��A�U�A0@�A��A���Ax� �A-�A\�AkA�A��#�A@�A�� �Au`!�A�U�AuA�AF��A5q�A��A@�A�\�A�� �Azu
BR@B"B��#B[�BB.�B��B�UBÓ	BB@
B~�BaqBC�
B]@Bf�BB0�B�UBN�BI-B�3Bq�B|�B�Bɏ BM@B�`!B�`!BΓB�B��$B�\ B9�!B��"BD�#Blq$B�3%B�`!&B-'B�(B�u
)Bԏ *BT-+Bٓ,B�\-Bh@.B�\/B��0B��1B�2B�3B��$4B_-5B��6BZ3"7B��
8Bs@9B�A:B��;B�<B'�=B�>B�\?Bߏ @B)�AB�u
BBCBO�DB�u
EB�#FB{�GB��$HB��IB;�JB�KB%LB4�MB8NBe3"OB�`!PB�u
QBY�RB��SB��#TB�\UB�u
VB2�WB=�XBp3"YB�AZBZ�[B�#\BF�]B�^B�`!_B��
`B�3aB��bBj-cBu-dB�\eB��fBwqgB� hB�AiBH�jB��kB�`!lBCmB��$nB��oB��pB��qB�-rB��sB�3tB��uBNvB0wB�u
xB�#yB?�zB��{B��|BS�}B{3"~B�qB��B�`!�B���B�`!�B�U�B�#�B�#�B���Bd��B�#�B���B^��Bi��B���B�3"�B�U�Bt��B�#�B���B�
�B��Bo��B;�B��B�
�B�u
�B��B�u
�B���B���BJ��B	�$�BX@�B��B�`!�B��BQ��B���B�B���B���B�\�BV�B�� �B���B�q�B���B�-�B�B���Bc@�B� �B�u
�B�3"�B���B� �B~@�B���B�`!�B���B���B�#�B�-�B��B�q�Bë�Bz��B�#�B� �B�@�B��B�#�B�@�B��B�3"�B�3�B�3�B�
�B�B���Bn@�BF�B�3�B���B�3�B��BV�B�$�BV�B�3"�B�3"�Be��B$V�B4�B��B\��B�@�B��B��B�u
�B�$�B*�$�B���B4�B�B�3"�B�u
�B�#�B)�BB�B�3"�B٫�B���B�u
�B4�B���By@�B'�
�B/V�B5�$�B!� �B��B��B@�$�B���B�@�B:V�BY�B2�
C�@C�C��CdCa!C�3"CoC=�
C�#	C��
CɯCK�$Cp�
CEVCQC(�C{�C3�C?CV�$C	v
C�C��C��C�#C>�C��C��C�Ca�$CJC� CU!C'�#"Ca!#C�-$C�3"%C��&Cԯ'CI�(C��)C��*C,� +C	#,Cl�$-C��.C&�/C�0C\1C�q2Cg�3Cg4CU�5Cz6CPV7C[V8C7� 9CH�
:C"�;CB� <C�=C�@>CB?C	#@C1�AC��BCM� CC<�DC-�ECfVFC`GCa!HC��IC��JC߯KCT�LC�3"MC�-NC�OC��PC�@QC��RC�-SC��TC�qUC�VCrWC��XCv
YC�ZC2�#[C��\C	#]C_�^CS�
_CǬ`C�3"aCҬbC^�
cC�dCqVeCv
fC}gC|VhCG�iC�VjC�3"kC(a!lC	mCi�
nC8�oCC�pC�@qC�rCN�sC��tCw�$uCkvC�wC�VxCt�
yC
4"zC�{C|C�}C`�~C�@CX� �Cj��C3a!�Ck��C�q�C>a!�C
��Cv�C�q�C���C�
�Cݬ�C��C	]�Cc� �C�@�C�V�C��$�C��C�@�C&��C�V�C��Cu��C�V�C���C1��Cn� �C*v
�CIa!�C��C��C��C���Cv��C���C=�#�C<��C��CY��C��C���C5v
�Cr��C��C#��C���C��$�C}��C�C�@�C���C��C	��C�@�C�@�C��C��
�C��$�Cd��C]�C�@�C���Cy� �C4�Co��C(4�CB�C��
�C@v
�C���C�� �C���C(	#�CKv
�Cz��C��C��
�C%B�C���C3	#�C��C�-�C���C0B�CH�#�CVv
�Cav
�C���C���C���C���C���C��$�C;B�C��C]�CTa!�C34�C���C���C�-�CFB�CȲ�C*�C�q�C��C�-�C�V�C.��C�V�C��$�C���C��C���C5�C��$�C9��C�@�C��CӲ�C>	#�C
��C>4�CG��C��C�-�C��$�C޲D�@D�qD�D�qD��
DS�#D^�#D�D��
	D��
D_a!DËD�
D�@D�D��$D�D�DR�D��$DQBD�qDja!D��$D�� DI4D�@D*]D�-Dlv
D��$DD� DR�!D\B"D�-#DT4$D��%D5]&D��$'D��(D@)D�*D�+D�q,D�$-D�.D��
/D.0D�V1D��2DgB3D�4D�5D�6DO�7D]�8D�9DrB:D��;D΋<Dً=D$�>D �?D��@D_4AD.BD�CD@]DDI	#ED]�FD*�GDj4HD�VID�@JDT	#KD_	#LD4"MDwv
ND 4"OD�PDh�QDh�RDZ�SD��TDs�UDi�#VD�� WDK]XDj	#YD��ZDua![D�\D�a!]D�$^D�_Dr`D��aD
�bD�$cD�@dD�eDAfDe�gD'�$hDt�#iD+�jDAkD�@lDrmD�nD�� oD5�pD��qDs�rD��
sD~�tDV]uD�v
vDp�wD��xD+4"yD6�zDA{D��|D64"}Dr~D�@D��D��D~��D'r�D ��D���D2r�D���D/��DA��D��Da]�D{��D�a!�D��DA4"�D��D2�$�D��D:��D���D�� �D��D��
�D)A�DL4"�D�#�D��Du	#�D���D�V�D���DK�D@��DE��DL��D4A�D=�$�D�V�D?A�D��DH�$�D��DW4"�D�a!�D���D�D���D��DV�DW�D���Du4�D�v
�Dʿ�D���D�DA�Da�Dl�DP��D�a!�D!��D�a!�D��D��
�D���D���D��D��#�D.�D��#�D���D�a!�D}B�DK��D	�D�� �D�B�Db4"�D���D���D�a!�D
�Dw�D��#�D=r�D���DV��D�4�D��
�Dտ�Dl]�D�D���D��Da��D[��D���DW��D�B�DHr�Df��D�DJA�D��
�D��Dw]�DW�D�B�D
�D
�Db��D���D�	#�D��D�4�D�D�	#�D���D%.�D�]�DS�$�D�
�D*�DƐ �D�a!�D���D:@Ae�A��ARu
A�A�Acu
Aī	AP�
AQ�A�AA�Av�A#Apu
A�A-A�� A�AA`!Ac�Ap3A�A�U`AD3"aA��#zA�A�AR3"�A,�A�
A	�@A0.AA+�BASrCAq�FA�	#GA��HA&
LA�v
MA�	#PA�QA^�$RA�v
SA��VAДWAr�$XAUAfA��$gA8
hA�4iA��jAѐ lA�4mA�BnA!�
pA=.qA)�
rA�4sAܐ tA5uA�BvA��$wA��xAm�yA�a!|A��}A3�
~AE.�AW�A6��A&��A�4�A��Al��A��#�A��A�v
�A"�A���A%��A
AbAJ
cAcAdAgreAGA^�AC`!AF� A�2"A��A�Ab\Ar�#AQ	A��
AJ�$A�,A�
A�Ac�A�t
A|�#A��AM`!A��Au
AT�$A#A��AP� Ay�
A)#Ah�A�AWUA�A^�$ Am�!A�2""A��#A��$Añ%Al\&A3"'AϾ(A�?)A�*AW`!+A��,A�,-Av\.AH�/A
�0A[1A3#2A�?3A�?4Au
5A=#6Ar�7A��#8A��9A�:A�?;AR�<A�,=AaU>A3"?A\�@A�3AA��BA�CA�pDAh�$EAr�$FA�GAkUHA��IAa`!JA�?KA|�$LA�MA��
NA�OA|�PA�?QAͱRA��#SA�?TAUA�,VA��$WA�XA��YA��ZA��[A�\A��]A^A��_A0�`A��aA��#bAw�cAuUdA��
eA�fA�pgA�?hA��iAŊjA��kA�3lAf�mA��$nAױoAWApAG#qA�rAپsAUtA��uA��vA�,wAxA��yA��zA��{Ae|Ap�}A ~A��$A��A��A:��A%��AZ� �A���A�?�A���Au
�A@�A�U�A���A�,�A�\�A��A(��A��A���A��
�A�\�A��Ao�Ad� �A��A��A�p�A@�A�3�A��$�A��#�A��A u
�AD��A���A���A��
�A���A��A��#�A*�A4�A�,�Ay�Aϊ�A@�A�\�A��A���A��A�,�A��A�U�A��Aq�A���A�\�AaA�A��A&@�Aي�A2��A
q�A/��A��A���A>�A���A�,�A��A�,�A���A���A��A���A3"�An� �A���A<��A���Aq�AH�A���A"3"�A�,�AQ#�A!q�Ak`!�A��A	��AR�A��A�?�A��A��
�A��$�A��$�A,3"�A*u
�A+q�A���A@�A��A���A9��A�\�A��A4u
�A��A�U�A0@�A��A���Ax� �A-�A\�AkA�A��#�A@�A�� �Au`!�A�U�AuA�AF��A5q�A��A@�A�\�A�� �Azu
BR@B"B��#B[�BB.�B��B�UBÓ	BB@
B~�BaqBC�
B]@Bf�BB0�B�UBN�BI-B�3Bq�B|�B�Bɏ BM@B�`!B�`!BΓB�B��$B�\ B9�!B��"BD�#Blq$B�3%B�`!&B-'B�(B�u
)Bԏ *BT-+Bٓ,B�\-Bh@.B�\/B��0B��1B�2B�3B��$4B_-5B��6BZ3"7B��
8Bs@9B�A:B��;B�<B'�=B�>B�\?Bߏ @B)�AB�u
BBCBO�DB�u
EB�#FB{�GB��$HB��IB;�JB�KB%LB4�MB8NBe3"OB�`!PB�u
QBY�RB��SB��#TB�\UB�u
VB2�WB=�XBp3"YB�AZBZ�[B�#\BF�]B�^B�`!_B��
`B�3aB��bBj-cBu-dB�\eB��fBwqgB� hB�AiBH�jB��kB�`!lBCmB��$nB��oB��pB��qB�-rB��sB�3tB��uBNvB0wB�u
xB�#yB?�zB��{B��|BS�}B{3"~B�qB��B�`!�B���B�`!�B�U�B�#�B�#�B���Bd��B�#�B���B^��Bi��B���B�3"�B�U�Bt��B�#�B���B�
�B��Bo��B;�B��B�
�B�u
�B��B�u
�B���B���BJ��B	�$�BX@�B��B�`!�B��BQ��B���B�B���B���B�\�BV�B�� �B���B�q�B���B�-�B�B���Bc@�B� �B�u
�B�3"�B���B� �B~@�B���B�`!�B���B���B�#�B�-�B��B�q�Bë�Bz��B�#�B� �B�@�B��B�#�B�@�B��B�3"�B�3�B�3�B�
�B�B���Bn@�BF�B�3�B���B�3�B��BV�B�$�BV�B�3"�B�3"�Be��B$V�B4�B��B\��B�@�B��B��B�u
�B�$�B*�$�B���B4�B�B�3"�B�u
�B�#�B)�BB�B�3"�B٫�B���B�u
�B4�B���By@�B'�
�B/V�B5�$�B!� �B��B��B@�$�B���B�@�B:V�BY�B2�
C�@C�C��CdCa!C�3"CoC=�
C�#	C��
CɯCK�$Cp�
CEVCQC(�C{�C3�C?CV�$C	v
C�C��C��C�#C>�C��C��C�Ca�$CJC� CU!C'�#"Ca!#C�-$C�3"%C��&Cԯ'CI�(C��)C��*C,� +C	#,Cl�$-C��.C&�/C�0C\1C�q2Cg�3Cg4CU�5Cz6CPV7C[V8C7� 9CH�
:C"�;CB� <C�=C�@>CB?C	#@C1�AC��BCM� CC<�DC-�ECfVFC`GCa!HC��IC��JC߯KCT�LC�3"MC�-NC�OC��PC�@QC��RC�-SC��TC�qUC�VCrWC��XCv
YC�ZC2�#[C��\C	#]C_�^CS�
_CǬ`C�3"aCҬbC^�
cC�dCqVeCv
fC}gC|VhCG�iC�VjC�3"kC(a!lC	mCi�
nC8�oCC�pC�@qC�rCN�sC��tCw�$uCkvC�wC�VxCt�
yC
4"zC�{C|C�}C`�~C�@CX� �Cj��C3a!�Ck��C�q�C>a!�C
��Cv�C�q�C���C�
�Cݬ�C��C	]�Cc� �C�@�C�V�C��$�C��C�@�C&��C�V�C��Cu��C�V�C���C1��Cn� �C*v
�CIa!�C��C��C��C���Cv��C���C=�#�C<��C��CY��C��C���C5v
�Cr��C��C#��C���C��$�C}��C�C�@�C���C��C	��C�@�C�@�C��C��
�C��$�Cd��C]�C�@�C���Cy� �C4�Co��C(4�CB�C��
�C@v
�C���C�� �C���C(	#�CKv
�Cz��C��C��
�C%B�C���C3	#�C��C�-�C���C0B�CH�#�CVv
�Cav
�C���C���C���C���C���C��$�C;B�C��C]�CTa!�C34�C���C���C�-�CFB�CȲ�C*�C�q�C��C�-�C�V�C.��C�V�C��$�C���C��C���C5�C��$�C9��C�@�C��CӲ�C>	#�C
��C>4�CG��C��C�-�C��$�C޲D�@D�qD�D�qD��
DS�#D^�#D�D��
	D��
D_a!DËD�
D�@D�D��$D�D�DR�D��$DQBD�qDja!D��$D�� DI4D�@D*]D�-Dlv
D��$DD� DR�!D\B"D�-#DT4$D��%D5]&D��$'D��(D@)D�*D�+D�q,D�$-D�.D��
/D.0D�V1D��2DgB3D�4D�5D�6DO�7D]�8D�9DrB:D��;D΋<Dً=D$�>D �?D��@D_4AD.BD�CD@]DDI	#ED]�FD*�GDj4HD�VID�@JDT	#KD_	#LD4"MDwv
ND 4"OD�PDh�QDh�RDZ�SD��TDs�UDi�#VD�� WDK]XDj	#YD��ZDua![D�\D�a!]D�$^D�_Dr`D��aD
�bD�$cD�@dD�eDAfDe�gD'�$hDt�#iD+�jDAkD�@lDrmD�nD�� oD5�pD��qDs�rD��
sD~�tDV]uD�v
vDp�wD��xD+4"yD6�zDA{D��|D64"}Dr~D�@D��D��D~��D'r�D ��D���D2r�D���D/��DA��D��Da]�D{��D�a!�D��DA4"�D��D2�$�D��D:��D���D�� �D��D��
�D)A�DL4"�D�#�D��Du	#�D���D�V�D���DK�D@��DE��DL��D4A�D=�$�D�V�D?A�D��DH�$�D��DW4"�D�a!�D���D�D���D��DV�DW�D���Du4�D�v
�Dʿ�D���D�DA�Da�Dl�DP��D�a!�D!��D�a!�D��D��
�D���D���D��D��#�D.�D��#�D���D�a!�D}B�DK��D	�D�� �D�B�Db4"�D���D���D�a!�D
�Dw�D��#�D=r�D���DV��D�4�D��
�Dտ�Dl]�D�D���D��Da��D[��D���DW��D�B�DHr�Df��D�DJA�D��
�D��Dw]�DW�D�B�D
�D
�Db��D���D�	#�D��D�4�D�D�	#�D���D%.�D�]�DS�$�D�
�D*�DƐ �D�a!�D���D:@Ae�A��ARu
A�A�Acu
Aī	AP�
AQ�A�AA�Av�A#Apu
A�A-A�� A�AA`!Ac�Ap3A�A�U`AD3"aA��#zA�A�AR3"�A,�A�
A	�@A0.AA+�BASrCAq�FA�	#GA��HA&
LA�v
MA�	#PA�QA^�$RA�v
SA��VAДWAr�$XAUAfA��$gA8
hA�4iA��jAѐ lA�4mA�BnA!�
pA=.qA)�
rA�4sAܐ tA5uA�BvA��$wA��xAm�yA�a!|A��}A3�
~AE.�AW�A6��A&��A�4�A��Al��A��#�A��A�v
�A"�A���A%��A|� AuA!A�]"A/#A>$AA�%A>&A�)A��#*A��+AW-A��/AA1A1�4A��5A� 6Am4"7AFDA]EA�NA?�OA��$\A�v
]A�A^A�]_A;�oA���AGA^�AC`!AF� A�2"A��A�Ab\Ar�#AQ	A��
AJ�$A�,A�
A�Ac�A�t
A|�#A��AM`!A��Au
AT�$A#A��AP� Ay�
A)#Ah�A�AWUA�A^�$ Am�!A�2""A��#A��$Añ%Al\&A3"'AϾ(A�?)A�*AW`!+A��,A�,-Av\.AH�/A
�0A[1A3#2A�?3A�?4Au
5A=#6Ar�7A��#8A��9A�:A�?;AR�<A�,=AaU>A3"?A\�@A�3AA��BA�CA�pDAh�$EAr�$FA�GAkUHA��IAa`!JA�?KA|�$LA�MA��
NA�OA|�PA�?QAͱRA��#SA�?TAUA�,VA��$WA�XA��YA��ZA��[A�\A��]A^A��_A0�`A��aA��#bAw�cAuUdA��
eA�fA�pgA�?hA��iAŊjA��kA�3lAf�mA��$nAױoAWApAG#qA�rAپsAUtA��uA��vA�,wAxA��yA��zA��{Ae|Ap�}A ~A��$A��A��A:��A%��AZ� �A���A�?�A���Au
�A@�A�U�A���A�,�A�\�A��A(��A��A���A��
�A�\�A��Ao�Ad� �A��A��A�p�A@�A�3�A��$�A��#�A��A u
�AD��A���A���A��
�A���A��A��#�A*�A4�A�,�Ay�Aϊ�A@�A�\�A��A���A��A�,�A��A�U�A��Aq�A���A�\�AaA�A��A&@�Aي�A2��A
q�A/��A��A���A>�A���A�,�A��A�,�A���A���A��A���A3"�An� �A���A<��A���Aq�AH�A���A"3"�A�,�AQ#�A!q�Ak`!�A��A	��AR�A��A�?�A��A��
�A��$�A��$�A,3"�A*u
�A+q�A���A@�A��A���A9��A�\�A��A4u
�A��A�U�A0@�A��A���Ax� �A-�A\�AkA�A��#�A@�A�� �Au`!�A�U�AuA�AF��A5q�A��A@�A�\�A�� �Azu
BR@B"B��#B[�BB.�B��B�UBÓ	BB@
B~�BaqBC�
B]@Bf�BB0�B�UBN�BI-B�3Bq�B|�B�Bɏ BM@B�`!B�`!BΓB�B��$B�\ B9�!B��"BD�#Blq$B�3%B�`!&B-'B�(B�u
)Bԏ *BT-+Bٓ,B�\-Bh@.B�\/B��0B��1B�2B�3B��$4B_-5B��6BZ3"7B��
8Bs@9B�A:B��;B�<B'�=B�>B�\?Bߏ @B)�AB�u
BBCBO�DB�u
EB�#FB{�GB��$HB��IB;�JB�KB%LB4�MB8NBe3"OB�`!PB�u
QBY�RB��SB��#TB�\UB�u
VB2�WB=�XBp3"YB�AZBZ�[B�#\BF�]B�^B�`!_B��
`B�3aB��bBj-cBu-dB�\eB��fBwqgB� hB�AiBH�jB��kB�`!lBCmB��$nB��oB��pB��qB�-rB��sB�3tB��uBNvB0wB�u
xB�#yB?�zB��{B��|BS�}B{3"~B�qB��B�`!�B���B�`!�B�U�B�#�B�#�B���Bd��B�#�B���B^��Bi��B���B�3"�B�U�Bt��B�#�B���B�
�B��Bo��B;�B��B�
�B�u
�B��B�u
�B���B���BJ��B	�$�BX@�B��B�`!�B��BQ��B���B�B���B���B�\�BV�B�� �B���B�q�B���B�-�B�B���Bc@�B� �B�u
�B�3"�B���B� �B~@�B���B�`!�B���B���B�#�B�-�B��B�q�Bë�Bz��B�#�B� �B�@�B��B�#�B�@�B��B�3"�B�3�B�3�B�
�B�B���Bn@�BF�B�3�B���B�3�B��BV�B�$�BV�B�3"�B�3"�Be��B$V�B4�B��B\��B�@�B��B��B�u
�B�$�B*�$�B���B4�B�B�3"�B�u
�B�#�B)�BB�B�3"�B٫�B���B�u
�B4�B���By@�B'�
�B/V�B5�$�B!� �B��B��B@�$�B���B�@�B:V�BY�B2�
C�@C�C��CdCa!C�3"CoC=�
C�#	C��
CɯCK�$Cp�
CEVCQC(�C{�C3�C?CV�$C	v
C�C��C��C�#C>�C��C��C�Ca�$CJC� CU!C'�#"Ca!#C�-$C�3"%C��&Cԯ'CI�(C��)C��*C,� +C	#,Cl�$-C��.C&�/C�0C\1C�q2Cg�3Cg4CU�5Cz6CPV7C[V8C7� 9CH�
:C"�;CB� <C�=C�@>CB?C	#@C1�AC��BCM� CC<�DC-�ECfVFC`GCa!HC��IC��JC߯KCT�LC�3"MC�-NC�OC��PC�@QC��RC�-SC��TC�qUC�VCrWC��XCv
YC�ZC2�#[C��\C	#]C_�^CS�
_CǬ`C�3"aCҬbC^�
cC�dCqVeCv
fC}gC|VhCG�iC�VjC�3"kC(a!lC	mCi�
nC8�oCC�pC�@qC�rCN�sC��tCw�$uCkvC�wC�VxCt�
yC
4"zC�{C|C�}C`�~C�@CX� �Cj��C3a!�Ck��C�q�C>a!�C
��Cv�C�q�C���C�
�Cݬ�C��C	]�Cc� �C�@�C�V�C��$�C��C�@�C&��C�V�C��Cu��C�V�C���C1��Cn� �C*v
�CIa!�C��C��C��C���Cv��C���C=�#�C<��C��CY��C��C���C5v
�Cr��C��C#��C���C��$�C}��C�C�@�C���C��C	��C�@�C�@�C��C��
�C��$�Cd��C]�C�@�C���Cy� �C4�Co��C(4�CB�C��
�C@v
�C���C�� �C���C(	#�CKv
�Cz��C��C��
�C%B�C���C3	#�C��C�-�C���C0B�CH�#�CVv
�Cav
�C���C���C���C���C���C��$�C;B�C��C]�CTa!�C34�C���C���C�-�CFB�CȲ�C*�C�q�C��C�-�C�V�C.��C�V�C��$�C���C��C���C5�C��$�C9��C�@�C��CӲ�C>	#�C
��C>4�CG��C��C�-�C��$�C޲D�@D�qD�D�qD��
DS�#D^�#D�D��
	D��
D_a!DËD�
D�@D�D��$D�D�DR�D��$DQBD�qDja!D��$D�� DI4D�@D*]D�-Dlv
D��$DD� DR�!D\B"D�-#DT4$D��%D5]&D��$'D��(D@)D�*D�+D�q,D�$-D�.D��
/D.0D�V1D��2DgB3D�4D�5D�6DO�7D]�8D�9DrB:D��;D΋<Dً=D$�>D �?D��@D_4AD.BD�CD@]DDI	#ED]�FD*�GDj4HD�VID�@JDT	#KD_	#LD4"MDwv
ND 4"OD�PDh�QDh�RDZ�SD��TDs�UDi�#VD�� WDK]XDj	#YD��ZDua![D�\D�a!]D�$^D�_Dr`D��aD
�bD�$cD�@dD�eDAfDe�gD'�$hDt�#iD+�jDAkD�@lDrmD�nD�� oD5�pD��qDs�rD��
sD~�tDV]uD�v
vDp�wD��xD+4"yD6�zDA{D��|D64"}Dr~D�@D��D��D~��D'r�D ��D���D2r�D���D/��DA��D��Da]�D{��D�a!�D��DA4"�D��D2�$�D��D:��D���D�� �D��D��
�D)A�DL4"�D�#�D��Du	#�D���D�V�D���DK�D@��DE��DL��D4A�D=�$�D�V�D?A�D��DH�$�D��DW4"�D�a!�D���D�D���D��DV�DW�D���Du4�D�v
�Dʿ�D���D�DA�Da�Dl�DP��D�a!�D!��D�a!�D��D��
�D���D���D��D��#�D.�D��#�D���D�a!�D}B�DK��D	�D�� �D�B�Db4"�D���D���D�a!�D
�Dw�D��#�D=r�D���DV��D�4�D��
�Dտ�Dl]�D�D���D��Da��D[��D���DW��D�B�DHr�Df��D�DJA�D��
�D��Dw]�DW�D�B�D
�D
�Db��D���D�	#�D��D�4�D�D�	#�D���D%.�D�]�DS�$�D�
�D*�DƐ �D�a!�D���D��F~4"F�FyrF�4F��#FLF��F�a!F<�
	F|�
F�BF-AF�]
F��F0�F?AF�F��F-WF_F�AF�F�]F��$FJ�F��FNAFF�FV�F1�F�rF��$ FW�!Ft"F�A#F�a!$F�4%F��&F�B'F?�(FH�)F��*F[A+Fs,F�v
-F�.Fg�/F� 0F�1FI�2F`
3F��4F�]5F��$6F��7F��8F��$9F�v
:FO.;FZ�<FN�=F��>F�?F�@F�AAF� BF�CFj�DF�EFq
FF��GF�AHF�BIFw�JF�KF>WLF\WMF{WNF�]OF��PF��QF
CRF� SF"�TFb!UF�WVF��#WF�4"XF�WYF��ZFȭ[F�\F�A]FiA^F�4_F��$`F�	#aF|AbF��cF��dFCeF(�fF��gF[�hF�riFn�jF~�kFʬlF5�mF�	#nF�roF�pF��qF�4"rF5sF�AtF�ruF�4"vF�wFw
xF�WyF��zF�	#{F�
|F��#}Fh�~F�AFw
�F�4"�F��F-C�F%� �FZ�
�F٭�F�	#�F�W�F�#�F�W�F��F��F���F���F���F5�F��F3w
�F�W�F���F��F�A�F���F���F��F��F��FM��Fj�
�F65�F�]�F_��F��F���Fd.�F��F���F�4"�F|�
�F�A�F��
�F���F	X�F=C�F�A�F
�$�FMw
�Fmw
�F�EBEn�Eb!E�4"E4�E�w
E�#E�AEbC	E،
E�E&�#EH5
E2� E�	#E:�#E��E
#Ec5Eq�EP�#E�E��E�E��E
# E2�!E�r"E��0��0�r0��0B0�w
0�0B0"�$0s5	0>�
0U�04B0��

0��0�05"0��
0K�0а0��0��0)
#0<
#0��0+�0�50г0�0�05"0�0� 0B!0��"0�#0�r$0F�%0�&0s'07�$(0��)0�*0n�+0J�$,0b�-0�
.0�/0��00^10�20��30~�40��50$X60�70f�#807X90JB:0^B;0B<0�=0s>0��?0u�#@0��A02�B0��C0��D0��#E0F0�5G01b!H0!^I0.�J0n�K0��#L0%M0�5N07BO0Rb!P0��Q0~BR0O�S0�5T0KU0UXV0hXW0�XX0+�Y0'�Z0��[0��\0��]0E^^0�X_0�`0�5a0��b0�Cc0vb!d0I� e03sf0�g0��#h0�i0b�j0��k0=�l0c�$m005"n0P�o0Osp0��q0osr0�s0v.t0�.u0
�v0{�w0�x0hy0U
#z0JB{0jB|0�b!}0D5"~0'0�w
�0���0���0���0��0��0��#�0���0_5"�0���0a��0t��0g
#�0��0P��0���0��
�0}�$�0
�0�
�0�s�0��0�B�0�0���0Y^�0s5"�0e��0�B�0�.�0�C�0"��01�0 ��0�
�06�0�5"�0�b!�05��0�#�0�.�0x^�0�B�0�B�0s
#�0q� �0�.�0��0´�0=�09��0G��0��
�0���0E��0�
�0A��0J�0�b!�0�C�09�#�0���0�^�0��0��0'��0�C�0�s�0���07��0��
�0U��0Ǖ�0S��0M��0T��0�
�0�s�0^��0W�0�X�0��0�� �0C�0ו�0�s�0�w
�0�C�01��0״�0]��0�
�0��1�
1*61+�
1�.1K�#1�.1n1I1�C	1T�
1�w
1�� 1j�
1�� 1��1�
#1�5"1�1�1�w
1p�1�X1j�1N�1�1�b!1��1X1��1�^15�
1� 1?q!1P�
"1�b!#1l�$1�
#%1�5"&1đ '1�B(1� )1 C*1�+1{�,1l-19�.1�./1��01��11�.3��3T�#3�
#3�3{�3�3�.3c�#3/	3�
#
3��3w�#3� 
3D3��34�3؍3�3�3#� 3I�3�3&/3��#3��3<� 3M� 3��$3�3�32�3(� 3�!38C"3�b!#3H�$3t�%3��#&3_� '3��#(3_�)3K�*3Z�+3_�,3s� -3.3��/3PC03b�13�� 23�� 33��43A653m�63"c!73��#83��93k�:3�
#;3��<37c!=3�>3��?3��@3��A3�5"B3��#C3�#D3��E3�5"F3�sG3�XH3x
I3�5"J3�XK3q�L3v�M3^CN3Oc!O3��P3uc!Q3��R3��S3 �#T3�U3V3?/W3 DX3�rY32DZ3+[3ED\3�c!]3��^3��_3�`3f�
a3Yb3��c3t�
d3��e3.�#f3x
g3��h3i3}Cj3��k3�5"l3��m3έn3�5"o3��p3��q32r3��s3�
#t3��
u3�
#v3�Bw3h6x3�c!y3#z39�#{3U/|3��
}3��~3#3�c!�3"Y�3��3�6�3�c!�3�c!�39#�3�^�3WD�3���3�B�3�^�3���3���3���3���3��3n/�3��3AY�3L�3���3��3���3�3
��3���3���3�B�3G#�%&p�%��%&P��%�6G����/"3����#3�*d!�/�/,7��D=�pY����#2d!	�^
T�
[����^��ے ��a#O6"tsY����?����$��6�|�����L��~Y�c��8�^6"��^���#���^���^���#��/�4��C�G��C�t�6d!�+tZx
C@�C�gx
��C���D���D�J���	Cl6"������q�Ld!�C���Yy�����Mc�
��c!���L#��@�# �B��
^#��
Ʊ� ��@d!��~/T�#��ծ)�6"�6�B�_Y6��B���s�C2������s�C��sȒ eD'x
?���$�oD��<�L�ۮ����6"�s�ɱ�D���cY�h��$ ,�@�̒ 7x
�D�֒ ���C�\&`&��&�;&�\&`&|x
����U���6��b�������_@�_.���#�D��#�6�ɖ�C4���# _q��#|x
`��`U}6"�6}6"b[���[���
��?���_x��x�6�@�6�.�h#�Dh#�6`ɖ��4�� w�q�D���%��%p�%��%`�%�%��%`�%@�%�%��%�%`&&�&p	&@&
&�&�����
C����4�*��C�����-=��DZd!�Ytd!�/���x
�C���{
�#��\tot}t�\�$�Ep���C����V_ot���������u��d!�Zj_�ot�e!��t	�#��<f!���|

7otF�E���C�#�7w���$la�-}
��p��)�$�8�
k� ��D�f!I7"�\���7d}
�\��LO��/�ǵP���D���E�c#�f!��� 700X�#�dw��k#X�#��� ��u#�#�g!�x5FJFR7"�w�Z�$"DZ�Z��XF��|�#�a����\�g!�a���$]�a5� ���E�
�7V��j}
ٶ�g!���#��$� ��w��� � ����7t}
;D�#�Y�)�Dla�)#R�
U���0
��D}}
���#����V��Djr�b��ݶ��7"��y�|���F�bwl��g!�W�
�8"�l��E�/�@�W#���#	�/0����*����� ]�N�7�#�]����
���y��y��$��`G���
�]��?0�FL�H��}
�����$�#c��טLF+z��*H�����1�
H��]��0 H������$�H��$��;^<�
u��$�PD3�J�
�7�dF\�
���� ��F;�#��G�(�:�JF��jc~
S����A��Eٶ,�l�
[�8����c��8n�����cN�������~z6��cn�8"X��	9"�c���
�V�
�z.9"�����S���-~
�ES������
7�S��F�R^�S�[~
�^�0S���g��GS�|~
����#S��:)�ͺ�#��Kd��ͺ����Y�ͺ�E�z:9"8ͺ\�wdf9"fHͺ�0���ͺ1���6�ͺE�D��z�Gͺ-8��a#�ͺM�#�0��~
ͺ�8ۙ�d��ͺ��$��$�~
ͺӺ��^Hͺ�d��#&���$ͺ���R�ͺZh!���9"ͺG9��$��ͺ�^�~
��oͺ��#��ͺȕ ��#�~
ͺ�H
�J��ͺ� �z�#P�ͺ'9jh!n�w�ͺ���9"
seͺA�U�R_ͺ������ͺ�� I��ͺI����ͺ�0y�����
ͺn�?���ïͺDI8
��#ͺ�eF�#�e�$ͺ�V� �Hͺ�V�#�z�9"ͺ��b_P�ͺ{P
D�'�ͺ�9p
*�$�ͺ��_��
ͺ��
\{�I�ͺ�I>1@H��ͺ��e�#�#�ͺe1����ͺ��#�N�$ͺE�J
A�ͺ��
7��
ͺbH����ͺƳZ2�Jͺ��
���_ͺ�9"g� �;"~2ͺ�� ��
�K�
ͺf�[�h�ͺ��~h!}�#�2ͺ���r���ͺ��#VG" ͺ���;"��3ͺ��
U�$�<"ͺ�h!�`{� ͺ�A`���3ͺ�Y��0ͺך��K ͺ��#��0�ͺ�#4q�$z�ͺ�#��
��O�ͺ)�x�&4_`ͺُ)�~��Hͺ�`�<"	�ͺ.���K� ͺ�f�{��I�ͺ�H�#1h&�ͺ:I4Z��aͺ�
v!��̜ͺb8	��"�ͺu4�4�Iͺ˻��<"�ͺ�#	��!ͺ�#������a�
����a�
�{�c�{a�
�4�!i	a�
<�#QJ~�a�
��[�M}a�
�
����a�
-��Y�#a�
���~a�
����a�
�{G<�a�
��di!��$a�
��vJh���#a�
��r�j�U�a�
��:#~��#a�
�p�
��
�i!a�
��
���a�
�|�ŀ
a�
�c���G�i!a�
Bd���#�4a�
��Q�]h�<"�YO#��.��Y�h"�
-�
l��Y���"j!�
�YL�dE�Q�
�Y���L�G���Y��[�C5��$�Yh�f�h;x��Y0�ؠ�G��
�Y+�������
�Yp5�!Fj!���Y%H@jj!�H�Y��$!�
+�#�� �YſP�#��d�Y.�w����� �Yſ6��="���Y�L����="�YJ@�$�!��Ye�3�
`�#�Y.��MQ�$�Yb>"���e/l�Y�M�:� �Y#� f"��YX�
>"��n"�YD���j!���Y��=���$�Y���*��Y��p#���Y�� ؁
/�E�$/���@�VE�$�J=� ��"E�$��}�
�сE�$n���
�J�E�$�E�A�#E�$Ց^��d�E�$d��H��5�E�$���L�$�E�$#q�*�
IE�$o�$=���
�lE�$dI�M�
E�$*��l!K�IE�$�;�A�mE�$"JW�<f8#E�$nm��h
b�E�$*6�6�#�mE�$�;���MNE�$������
��E�$V�$?"�#h<E�$h<�JfE�$n�QNi���E�$���#�J�#E�$OK�#�<��$E�$���
�<hfE�$���<!�$��E�$��8���KE�$�#����?�E�$����
E�$����D#�E�$}��� -�#kNE�$9�
h#��ނE�$Q�#,n=E�$=E�$o�ƲE�$Ʋ�� "$��E�$������E�$�Pn,���E�$���N��E�E�$E���*E�$*�K���%�nv#f=Q�%��.��n��
%�J+A�W�%���NS3�J�S�_��KG$J��
Lm�
�6J��6��
�=�J���)�K�ݘ J�8?"� �?"��J�)L3� R7m�#J��=wX�
S�J�q�k��L��J�v7��O��J�y����B�J��ǒ�#�fJ�->[�$���?"J�go�|�
Y� J���w�\�� J��L*O���J���o�#ƙ J��MM�J�u����#J�!K�>��LOJ�/#��i$�J��j!�?"�7�KJ�r���$���KJ����
�#�J�*o�?"��J��>��4k!��J���
?M� Xk!J�:� �o�7��#J���+�Y���
J��
�$��}�J�!L�� ��J�����J��k!Ś @LS�J�� "��$=�J�8�S�e�J�����$�$��J��cLx�:�J��gU��$J��w�>%�oJ���x�:�J�>�y�@"J����LU��$J��$�
>%�oJ�`+�W�J�&?��$J@"�OJ�"�/�#J���r� h%�OJ�b���J�P��J�q@"-#���J�����J��%L����J�1�
J����J�~�
��Ӄ
J���J��gIJ�i��gIJ��k!3P�L��J��g�L��J��#K#�L
�J��L)�
�L
�J�8���aMJ���
����$J��� ��J���J��gh��#T�J��8�����jJ�k�����J���<��@"K�J���V�n���J�8��A��m#J�{M��Ȝ J���
��
�$J��#��
���J���,�{�J���%�s�?�J���#E9}l!J�q9kpJ�Xp���AJ�QPG�
b� J�U��J��o�y�
ƻJ�����#����J��%��#����J�%��

J�fP8���J�xMB�l!J���
�ش�PJ���S�f��MJ�#��#�
4BJ��M���"�J���˓{�)�J�G�@�X/�J����@"��J�T�IIN��J�e�M A"Y�$J���$���\�J�DA"��mNf�J���f�
�MJ��B	C?&;kJ�ap��k��J��A"��
��+�
J���� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&����J�ҥ�$%CJ������J���N��J���&�pJ�X"�8�$J��9�9�l!DB"J��#��@"��J�T��#YC��$J��#IN��J�e�&�� :J�� A"Y�$J���$#��#��J�n�
��\�J�DA"nk�k��J�m!mNf�J��(����J���$f�
�MJ��B��
�N�J��N?&;kJ�ap
��k�J�,�k��J��A"<NG��B"J��N�P�NJ�.�؝ ��J�[N6���sJ�F�14OJ���$�y�J���T�M�5:J����B"&�#[�J�r��:��
m!J���q��
J����q���P���:y�H�Cm!�P��o��:w�e�z��P��o��:�k����P��o��:J�#����P��o��:$�����P��o��:��4��P�P��o��:�f��k�P��o��:�� Y�)Q�P��o��:�3��O�P��o��:��Sl9��P��o��:�C��$�m!�P��o��:�:%�C�P��o��:��
U���P���:��f����P��o��:����;��P��o��:�Q���
�P��o��:$��N��
�P��o��:�N���B"�P��o��:����
�P���:��
�:���P�C�:����#��P�C�:3C"=�Y��P�C�:��
��Q�P�C�:���1n!�P�C�:"�����P�C�:��
�
�P�C�:KP�&1q�P�Co��:y�
� �q�P�Co��:�$'�$.� �P�Co��:D�����P�Co��:�%�
���o��:-�9�
g� ��o��:<�$q�����:�QgPM�
���:6�Q���
�qB����:yl!�����
�&���:�� a�#W��lq����:��
������:OC"V�� ���:�q�##�Q��
�:��� ���Q��
�:��B���O�Q��
�:?�LrP�qj�Q��
�:�q7r2�e���Q��
�:��R���:�:�Q��
�:XrN�$�r��L;�Q��
�:n�$��
�k�o��Q��
�:fn!#'��g�3� �Q��
�:r� D����R��n!u��+��m��
P�����
Ԉ��l�
bC"�v�#��#�� ��C"�r���P�� NPo!��$�dRqP��s#��D"�;[�#�˶+�
����.�F�F���D"�l)��4D���r����#`��;'�� S������;a�}�� �$~Qs#��D";o!m�� �;QKo!����� ��#����˶�;P�j�����<��+st��-�`R;<�����
���;'UDvD�#���Q��H�� �$+ST#s#��D"�����R��8m��$��
� ���Os����
�$W��R�L�ms���� ��E"����D���P��#|�ES�����x
�C���{
�#��\tot}t�\�$�Ep���C����V_ot���������u��d!�Zj_�ot�e!��t	�#��<f!���|

7otF�E���C�#�7w���$la�-}
��p��)�$�8�
k� ��D�f!I7"�\���7d}
�\��LO��/�ǵP���D���E�c#�f!��� 700X�#�dw��k#X�#��� ��u#�#�g!�x5FJFR7"�w�Z�$"DZ�Z��XF��|�#�a��D�\�g!�a�f�]�a5� ���E�
�7V��j}
ٶ�g!���#��$� ��w��� � ����7t}
;D�#�Y�)�Dla�)#R�
U���0
��D}}
���#����V��Djr�b��ݶ��7"��y�|���F�bwl��g!�W�
�8"�l��E�/�@�W#�%� 	�/0����*����� ]�N�7�#�]����
���y��y��$��`G���
�]��?0�FL�H��}
�����$�#c��טLF+z��*H�����1�
H��]��0 H������$�H��$��;^<�
u��$�PD3�J�
�7�dF\�
���� ��F;�#��G�(�:�JF��jc~
S����A��Eٶ,�l�
[�8����c��8n���
�cN�*�
��#q� ������#w�����
�V�
�z.9"�����S���-~
�ES������
7�S��F�R^�S�[~
�^�0S���g��GS�|~
����#S��:)�ͺ�#��Kd��ͺ����Y�ͺ�E�z:9"8ͺ\�wdf9"fHͺ�0���ͺ1���6�ͺE�D��z�Gͺ-8��a#�ͺM�#�0��~
ͺ�8ۙ�d��ͺ��$��$�~
ͺӺ��^Hͺ�d��#&���$ͺ���R�ͺZh!���9"ͺG9��$��ͺ�^�~
�oͺ��#��ͺȕ 0G�~
ͺǷ
�J��ͺ� �z)�P�ͺ'9jh!n�w�ͺ���9"
seͺA�U�R_ͺ��$�$��ͺ�� I��ͺI����ͺ�0y�����
ͺn�?���ïͺDI8
��#ͺ�eF�#�e�$ͺ�V� �Hͺ�V�#�z�9"ͺ��b_P�ͺ{P
D�'�ͺ�9p
*�$�ͺ��_��
ͺ��
\{�I�ͺx#>1@Hm�
ͺޕe�#�##�ͺ�'���TGͺ4#��(ͺ��ƳZ2
�ͺ]m����ͺ]�g� �;"oͺ���
�KZE"ͺ�#�[���ͺo~h!}�#^�ͺYS�r��sͺ��#VG��ͺp�
�;"���ͺJ�#U�$�A$ͺ4��`�oͺM�A`����ͺ�GY��.)ͺ�)��K2�ͺ����0�ͺ�#t4p�$ͺH�$4q�$z�ͺ�#��
��O�ͺ)�x�&4_`ͺُ)�~��Hͺ�`�<"	�ͺ.���K� ͺ�f�{��I�ͺ�H�#1h&�ͺ:I4Z��aͺ�
v!��̜ͺb8	��"�ͺu4�4�Iͺ˻��<"�ͺ{��ͺ�� ���ͺ������
��ͺ��J����ͺ�������ͺ��6�
��
��ͺ���#Xp��ͺ���8t��ͺ��[o!����ͺ��u���
��ͺ���T���ͺ��������ͺ���tt��ͺ���$e�
�ͺ�SL���ͺ����	���ͺ��A���
��ͺ����$����ͺ��$�� ��ͺ��� �<��ͺ���P�E"��ͺ����*��ͺ���t����ͺ�����<��ͺ���
�H��ͺ�������ͺ��V���=�
ͺ=�
���E�ͺE�DS�tE�ͺE�+uE�ͺE�E݊
E�ͺE�ws�
E�ͺE���
�E�ͺE�4$�E�ͺE��H��E�ͺE���
�E�ͺE�T��E�ͺE�&�
��E�ͺE��p��E�ͺE����#IͺI#�
=ͺ=vSL�=ͺ=�$�=ͺ=$�d�=ͺ=&%=ͺ=��:�=ͺ=.#Q=ͺ=
���=ͺ=�o!WQ=ͺ=F"DF"=ͺ=�U�
=ͺ=R���=ͺ='� RT��ͺ��D�o!?�ͺ?�b#a�?�ͺ?�M� a?�ͺ?�3�7�?�ͺ?�&��F"?�ͺ?��#_�
?�ͺ?����F"?�ͺ?���
s�?�ͺ?�~� ��?�ͺ?��=?�ͺ?�?�$Ԗ?�ͺ?��Sq�$?�ͺ?���
T�#ͺ�#�F"^��ͺ��:��ͺ��T���ͺ�����ͺ�p!�S�ͺ�l$n�ͺ�R=���ͺ�cu�Q�ͺ�T�Q�ͺ�II��ͺ��y��ͺ�OT�� �ͺ��p���ͺ��ď*�ͺ*��=*�ͺ*��L*�ͺ*�v�Q*�ͺ*�Q���*�ͺ*���#*�ͺ*���� *�ͺ*�����
*�ͺ*��R*�ͺ*��=:�*�ͺ*���G�*�ͺ*��u�**�ͺ*���$���ͺ����T�uͺ�u�
��uͺ�u�T��
�uͺ�u��$�T�uͺ�u�2p!�uͺ�u��u�uͺ�u�=�$�uͺ�u�(v�uͺ�uv����uͺ�uv���uͺ�u��9�uͺ�u��.� �uͺ�u;#H�.�ͺ.�q�T+ͺ+����+ͺ+"��+ͺ+GR��+ͺ+�$\�+ͺ+��;++ͺ+j� uRzIͺzI�� �R7�ͺ7�dvu7�ͺ7���d#7�ͺ7��$�I7�ͺ7��Inp!7�ͺ7��-q7�ͺ7�Viq7�ͺ7�$�
��7�ͺ7��2U7�ͺ7�Jw+7�ͺ7��/>7�ͺ7��� j�7�ͺ7����
ͺ
k>Q��vͺ�v�!G"�vͺ�v�����vͺ�v�R@J�vͺ�v�#��vͺ�v�>T�
�vͺ�v"����vͺ�v-���vͺ�v��e��vͺ�v��p!�vͺ�vY��>�vͺ�v��7��vͺ�v	�
SnUͺnU�2�vͺ�vL����vͺ�v��H�$�vͺ�v�+ZS�vͺ�v����vͺ�v�p!�U�vͺ�v��$���vͺ�v-���vͺ�v��|J�vͺ�vs�n�vͺ�vC��q�vͺ�v�$�vͺ�v�q!��ͺ���+��$�ͺ�O{��ͺ��J?�ͺ���� �ͺ�1����ͺ��v��ͺ�J?}��ͺ�|^��ͺ�����ͺ��}?�ͺ�5�
�S�ͺ��J,�ͺ� ���ͺ��?]G"ͺ]G"<��]G"ͺ]G"��L,]G"ͺ]G"H ��]G"ͺ]G"��֫]G"ͺ]G"(� 0w]G"ͺ]G"�q�U]G"ͺ]G"y i�
]G"ͺ]G"e�/$]G"ͺ]G"_� ��]G"ͺ]G"&���]G"ͺ]G"Yq!� ]G"ͺ]G"0V�#
 #ͺ
 #k$ ��,ͺ�,
���$�,ͺ�,r��,ͺ�,alw�,ͺ�,��
� �,ͺ�,��
4�,ͺ�,)!�?�,ͺ�,p�q!�,ͺ�,�q!�$�,ͺ�,1@\V�,ͺ�,`!��
�,ͺ�,�w��
�,ͺ�,�� �� Q�ͺQ���
G;�ͺ;�&�S�;�ͺ;�3�$�� ;�ͺ;�N�^�;�ͺ;���;�ͺ;�#�m�;�ͺ;�)�
��;�ͺ;�7� �,;�ͺ;�V���;�ͺ;�k� �w;�ͺ;�N���;�ͺ;����;�ͺ;��#	��!ͺ�#����i@a�
���$a�
�{�c�Ta�
�4�!�Va�
<�#QJ�a�
��[��Sa�
�
��{�a�
-����a�
���Va�
���f�$a�
�{G<�a�
��di!��$a�
��vJh���#a�
��r�j�U�a�
��:#~��#a�
�p�
��
�i!a�
��
���a�
�|�ŀ
a�
�c���G�i!a�
Bd���#�4a�
��Q�]h�<"�YO#��.��Y�h"�
-�
l��Y���"j!�
�YL�dE�Q�
�Y���L�G���Y��[�C5��$�Yh�f�h;x��Y0�ؠ�G��
�Y+�������
�Yp5�!Fj!���Y%H@jj!�H�Y��$!�
+�#�� �YſP�#��d�Y.�w����� �Yſ6��="���Y�L����="�YJ@�$�!��Ye�3�
`�#�Y.��MQ�$�Yb>"���e/l�Y�M�:� �Y#� f"��YX�
>"��n"�YD���j!���Y��=���$�Y���*��Y��p#���Y�� ؁
XXE�$XX��@�VE�$�J=� ��"E�$��}�
�сE�$n���
�J�E�$�E�A�#E�$Ց^��d�E�$d��H��5�E�$���L�$�E�$#q�*���E�$�$=���
�� E�$ʼ�M���E�$r!�l!K�!E�$Y��A�'� E�$�XW�<f�� E�$����h
G�E�$��6�#��E�$-���MNE�$������
��E�$V��UU��E�$��9$%K��E�$��\�
�
��E�$���r!���E�$��R�jX��E�$���7� ��E�$�����-��E�$��!Ya #��E�$���r!� #��E�$��s�n���E�$��Ors!��E�$����$Hs!��E�$����
 "��E�$�������E�$��bY��E�$��k��U��E�$��s� r���E�$����X��E�$��"x����E�$���� 1B��E�$��o$mB��E�$��[��X��E�$������E�$�����U��E�$��� \"��E�$��������E�$��՘(� E�E�$E��"�YE�E�$E�T�
d� E�E�$E��Y�$E�E�$E��� �rE�E�$E��-�E�E�$E�M���E�E�$E�����E�E�$E�_x��E�E�$E��&�E�E�$E�\�	VE�E�$E��$�s!E�E�$E�9�$v�$E�E�$E�����=E�$=�x�r=E�$=�"��$=E�$=��
�X=E�$=#��=E�$=aK��=E�$=Ys=E�$=��=E�$=
Z��=E�$=���G"=E�$=GZXY=E�$=�Y2�=E�$=EV�K=E�$=�Y� #?�E�$?�Ǯ�B?�E�$?��BZ?�E�$?��Z.?�E�$?��x��$?�E�$?�3$٨ ?�E�$?�b�~V?�E�$?�!#	�?�E�$?��s!A?�E�$?�� �?�E�$?��?Z?�E�$?��K��?�E�$?�H.L� ?�E�$?��i$�E�$�����E�$�C���E�$�E��s!�E�$���
<C�E�$�*�$���E�$�n��V�E�$����� �E�$�R!#�Z�E�$���xC�E�$��Z}�E�$�3t!�G"�E�$���[��E�$�H"y*�E�$*�'[`[*�E�$*�W��*�E�$*��C�*�E�$*��[��
*�E�$*���*�E�$*�{ZL*�E�$*�5#-*�E�$*�'��*�E�$*�L���*�E�$*��$�[*�E�$*��ƍ
*�E�$*��ZOL*�E�$*�\���uE�$�u@y���uE�$�u�$ �uE�$�up�Bs�uE�$�uĩ ���uE�$�u$#�
�uE�$�uDH"_�$�uE�$�u��yy�uE�$�u�H"~s�uE�$�u��$�L�uE�$�u�!#��uE�$�u���V�uE�$�uW�i�uE�$�u�H"�s+E�$+_�
[�+E�$+G\y#+E�$+�H"�+E�$+e��+E�$+
�~.+E�$+>��7�E�$7�����$7�E�$7�Q��y7�E�$7�C��C7�E�$7�����7�E�$7�!I"��7�E�$7�{��$7�E�$7�5�9�7�E�$7�k��� 7�E�$7�J ��7�E�$7�Q�$97�E�$7� ��7�E�$7��!#v\7�E�$7�m��vE�$�vW$�.�vE�$�v���Z�vE�$�v2[UI"�vE�$�v9� 2W�vE�$�v�s���vE�$�vunW�vE�$�vY��vE�$�v֒���vE�$�v��q� �vE�$�v�.�� �vE�$�v��y�vE�$�v�� �vE�$�vd[)��vE�$�vվ�\�vE�$�v�#D�vE�$�v ��vE�$�v���L�vE�$�v-z���vE�$�v��
�\�vE�$�vR��vE�$�v��
�vE�$�v>�
��vE�$�vot!�vE�$�v_DW�vE�$�v����vE�$�v
u��E�$�p���
�E�$�L��!#�E�$��$>�E�$�iz�t!�E�$��I"M�E�$��e��E�$��I"z�E�$�z���E�$��7/�E�$���?��E�$�P��$�E�$���:"#�E�$�{�
��
]G"E�$]G"�
9�]G"E�$]G"�t!N�]G"E�$]G"{�?M]G"E�$]G"�zs/]G"E�$]G"�D��]G"E�$]G"�D�]G"E�$]G"J"
�
]G"E�$]G"u!�[]G"E�$]G"�W�[]G"E�$]G"v"#�W]G"E�$]G"#XH�]G"E�$]G"���]G"E�$]G"��\�,E�$�,2t�z�,E�$�,��fX�,E�$�,�u��,E�$�,�I�
�,E�$�,KJ"M\�,E�$�,� *]�,E�$�,��
{M�,E�$�,
����,E�$�,[u!���,E�$�,�u!%��,E�$�,$�
f]�,E�$�,�M���,E�$�,��g�
;�E�$;��$a�;�E�$;�-$��;�E�$;����/;�E�$;�{�u!;�E�$;�E{��;�E�$;�� �;�E�$;���M;�E�$;��ٚ;�E�$;����J";�E�$;���X;�E�$;�/N��;�E�$;��##�;�E�$;�$?"�#T�E�$���JfE�$n�QNi���E�$���#�J�#E�$OK�#�<��$E�$���
�<hfE�$���<!�$��E�$��8���KE�$�#����?�E�$����
E�$����D#�E�$}��� -�#kNE�$9�
h#��ނE�$Q�#,n=E�$=E�$o�ƲE�$Ʋ�� "$��E�$������E�$�Pn,���E�$���N��E�E�$E���*E�$*�K���%�nv#f=Q�%��.��n��
%�J+A�W�%���NS3�J�S�_��KG$J��
Lm�
�6J��6��
�=�J���)�K�ݘ J�8?"� �?"��J�)L3� R7m�#J��=wX�
S�J�q�k��L��J�v7��O��J�y����B�J��ǒ�#�fJ�->[�$���?"J�go�|�
Y� J���w�\�� J��L*O���J���o�#��
J�*�MM�J�u���nNJ��"#�>��jtJ��J"��i$˓J��t�?"�7�J�)���$���KJ����
�#�J�*o�?"��J��>��4k!��J���
?M� Xk!J�:� �o�7��#J���+�Y���
J��
�$��}�J�!L�� ��J�����J��k!Ś @LS�J�� "��$�\J���S�EJ��]��$�$t�J��XcLJ�:�J��g�
�$J��w�<�oJ���J�:�J�>�y�@"J����L�
�$J��$�
<�oJ�`+�W�J�&?��$J@"�OJ�"�/�#J���r� h%�OJ�b���J�P��J�q@"-#���J�����J��%L����J�1�
J����J�~�
��Ӄ
J���J��gIJ�i��gIJ��k!3P�L��J��g�L��J��#K#�L
�J��L)�
�L
�J�8���aMJ���
����$J��� ��J���J��gh��#T�J��8�����jJ�k�����J���<��@"K�J���V�n���J�8��A��m#J�{M��Ȝ J���
��
�$J��#��
���J���,�{�J���%�s�?�J���#E9}l!J�q9kpJ�Xp���AJ�QPG�
b� J�U��J��o�y�
ƻJ�����#����J��%��#����J�%��

J�fP8���J�xMB�l!J����ش�PJ���S�f��MJ�#��#�
4BJ��M���"�J���˓{�)�J�G�@�X/�J����@"��J�T�IIN��J�e�M A"Y�$J���$���\�J�DA"��mNf�J���f�
�MJ��B	C?&;kJ�ap��k��J��A"��
��+�
J���� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&����J�ҥ�$%CJ������J���N��J���&�pJ�X"�8�$J��9�9�l!DB"J��#��@"��J�T��#YC��$J��#IN��J�e�&�� :J�� A"Y�$J���$#��#��J�n�
��\�J�DA"nk�k��J�m!mNf�J��(����J���$f�
�MJ��B��
�N�J��N?&;kJ�ap
��k�J�,�k��J��A"<NG��B"J��N�P�NJ�.�؝ ��J�[N6���sJ�F�14OJ���$�y�J���T�M�3!J�)��B"&�#��J�)Y�:��
m!J���q��
J������N�N�x
}{����
NO��x
/����$G���x
���������x
v!��uoY�Y�x
����"uz����x
s���m��v!�\�x
dO���O�O�{�x
]���6�����x
�������]��x
����"� �YǏ
�x
?u��m�#�
[$�x
wE��e9��$�x
�v!����^8K"�x
P���O�_���x
����v�Y!��x
�K"Qm����x
o!QB� ��	$�x
"�
Q���OdL"�x
�uQi����$�x
��Q�$��*^�x
[� Q���	$2�$�x
G|Qҿ0��/�x
	�##����R��n!u��+��m��
P�����
Ԉ��l�
bC"�v�#��#�� ��C"�r���P�� NPo!��$�dRqP��s#��D"�;[�#�˶+�
����.�F�F���D"�l)��4D���r����#`��;'�� S������;a�}�� �$~Qs#��D";o!m�� �;QKo!����� ��#����˶�;P�j�����<��+st��-�`R;<�����
���;'UDvD�#���Q��H�� �$+ST#s#��D"�����R��8m��$��
� ���Os����
�$W��R�L�ms���� ��E"����D���P��#|�ES���
}�x
�C���{

}#��\tot}t
}\�$�Ep���C
}���V_ot��
}������u��d!
}Zj_�ot�e!
}�t	�#��<f!
}��|

7otF
}E���C�#
}7w���$la
}-}
��p�
})�$�8�
k� 
}�D�f!I7"�\
}��7d}
�\
}�LO��/
}ǵP���D��
}E�c#�f!��
} 700X�#
}dw��k#X�#
}�� ��u#�#
}g!�x5FJFR7"
}w�Z�$"DZ�Z�
}XF��|�#�a
}�D�\�g!�a
}f�]�a5� 
}��E�
�7V
}�j}
ٶ�g!
}��#��$� �
}w��� � ��
}�7t}
;D�#
}Y�)�Dla
})#R�
U��
}0
��D}}

}��#���
}V��Djr
}b��ݶ
}�7"��y�|�
}�F�bwl��g!
}W�
�8"�l��E
}/�@�W#
}%� 	�/0��
}�*����� ]
}N�7�#�]
}���
���y
}�y��$��`G
}��
�]��?0
}FL�H��}

}����$�#c
}�טLF+z
}�*H��
}��1�
H��]
}�0 H���
}��$�H��$
}L�;^<�
u��$
}\3�J�
�7
}dF\�
���� 
}�F;�#��G
}(�:�JF�
}jc~
S���
}A��Eٶ,
}l�
[�8��
}�c��8n�
}���cN���
}���~z6
}�cn�8"X�
}	9"�c���

}V�
�z.9"
}����S��
}-~
�ES���
}��
7�S��F
}R^�S�[~

}^�0S��
}g��GS�|~

}���#S��:)�ͺ�#��Kd��ͺ����Y�ͺ�E�z:9"8ͺ\�wdf9"fHͺ�0���ͺ1���6�ͺE�D��z�Gͺ-8��a#�ͺM�#�0��~
ͺ�8ۙ�d��ͺ��$��$�~
ͺӺ��^Hͺ�d��#&���$ͺ���R�ͺZh!���9"ͺG9��$��ͺ�^�~
�oͺ��#��ͺȕ 0G�~
ͺǷ
�J��ͺ� �z)�P�ͺ'9jh!n�w�ͺ���9"
seͺA�U�R_ͺ��$�$��ͺ�� I��ͺI����ͺ�0y�����
ͺn�?���ïͺDI8
��#ͺ�eF�#�e�$ͺ�V� �Hͺ�V�#�z�9"ͺ��b_P�ͺ{P
D�'�ͺ�9p
*�$�ͺ��_��
ͺ��
\{�I�ͺx#>1@Hm�
ͺޕe�#�##�ͺ�'���TGͺ4#��(ͺ���J
��
ͺ��
7���ͺbH���L"ͺƳZ2
�ͺ]m����ͺ]�g� �;"oͺ���
�KZE"ͺ�#�[���ͺo~h!}�#^�ͺYS�r��sͺ��#VG��ͺp�
�;"���ͺJ�#U�$�A$ͺ4��`�oͺM�A`����ͺ�GY��.)ͺ�)��K2�ͺ����0�ͺ�#t4p�$ͺH�$4q�$z�ͺ�#��
��O�ͺ)�x�&4_`ͺُ)�~��Hͺ�`�<"	�ͺ.���K� ͺ�f�{��I�ͺ�H�#1h&�ͺ:I4Z��aͺ�
v!��̜ͺb8	��"�ͺu4�4�Iͺ˻��<"�ͺ�#��##ͺ##
�$��
�ͺ��##��ͺ�/0���Eͺ�E�Y}9�
ͺ9�
�O:��##ͺ�##1a��ͺ�+ZaZFͺZFu��T�Fͺ�F����'�
ͺ'�
r��?!Pͺ!P�uG($ͺ($�#	��!ͺ�#����i@a�
���$a�
�{�c�Ta�
�4�!�Va�
<�#QJ�a�
��[��Sa�
�
��{�a�
-����a�
���Va�
���f�$a�
�{G<�a�
��di!��$a�
��vJh���#a�
��r�j�U�a�
��:#~��#a�
�p�
��
�i!a�
��
���a�
�|�ŀ
a�
�c���G�i!a�
Bd���#�4a�
��Q�]h�<"�YO#��.��Y�h"�
-�
l��Y���"j!�
�YL�dE�Q�
�Y���L�G���Y��[�C5��$�Yh�f�h;x��Y0�ؠ�G��
�Y+�������
�Yp5�!Fj!���Y%H@jj!�H�Y��$!�
+�#�� �YſP�#��d�Y.�w����� �Yſ6��="���Y�L����="�YJ@�$�!��Ye�3�
`�#�Y.��MQ�$�Yb>"���e/l�Y�M�:� �Y#� f"��YX�
>"��n"�YD���j!���Y��=���$�Y���*��Y��p#���Y�� ؁
XXE�$XX��@�VE�$�J=� ��"E�$��}�
�сE�$n���
�J�E�$�E�A�#E�$Ց^��d�E�$d��H��5�E�$���L�$�E�$#q�*���E�$�$=���
�� E�$ʼ�M���E�$r!�l!K�!E�$Y��A�'� E�$�XW�<f�� E�$����h
G�E�$��6�#��E�$-���MNE�$������
��E�$V���b��E�$���$]hPE�$hP���^E�$�^������
E�$��
��
�^b0E�$b0�
9��E�$���ر_E�$_$#��E�$�`]6$#U}E�$U}["�}�	$E�$�	$���� %�E�$%��0�u��E�$����}m$E�$m$<:vv�E�$v��
 �kE�$kB�$�$�E�$���vvR�E�$R��v!bZr$#E�$r$#� u�#�E�$#�$?"�#T�E�$���JfE�$n�QNi���E�$���#�J�#E�$OK�#�<��$E�$���
�<hfE�$���<!�$��E�$��8���KE�$�#����?�E�$����
E�$����D#�E�$}��� -�#kNE�$9�
h#��ނE�$Q�#,n=E�$=E�$o�ƲE�$Ʋ�� "$��E�$������E�$�Pn,���E�$���N��E�E�$E���*E�$*�K���%�nv#f=Q�%��.��n��
%�J+A�W�%���NS3�J�S�_��KG$J��
Lm�
�6J��6��
�=�J���)�K�ݘ J�8?"� �?"��J�)L3� R7m�#J��=wX�
S�J�q�k��L��J�v7��O��J�y����B�J��ǒ�#�fJ�->[�$���?"J�go�|�
Y� J���w�\�� J��L*O���J���o�#��
J�*�MM�J�u���nNJ��"#�>��jtJ��J"��i$˓J��t�?"�7�J�)���$���KJ����
�#�J�*o�?"��J��>��4k!��J���
?M� Xk!J�:� �o�7��#J���+�Y���
J��
�$��}�J�!L�� ��J�����J��k!Ś @LS�J�� "��$�\J���S�EJ��]��$�$t�J��XcLJ�:�J��g�
�$J��w�<�oJ���J�:�J�>�y�@"J����L�
�$J��$�
<�oJ�`+�W�J�&?��$J@"�OJ�"�/�#J���r� h%�OJ�b���J�P��J�q@"-#���J�����J��%L����J�1�
J����J�~�
��Ӄ
J���J��gIJ�i��gIJ��k!3P�L��J��g�L��J��#K#�L
�J��L)�
�L
�J�8���aMJ���
����$J��� ��J���J��gh��#T�J��8�����jJ�k�����J���<��@"K�J���V�n���J�8��A��m#J�{M��Ȝ J���
��
�$J��#��
���J���,�{�J���%�s�?�J���#E9}l!J�q9kpJ�Xp���AJ�QPG�
b� J�U��J��o�y�
ƻJ�����#����J��%��#����J�%��

J�fP8���J�xMB�l!J����ش�PJ���S�f��MJ�#��#�
4BJ��M���"�J���˓{�)�J�G�@�X/�J����@"��J�T�IIN��J�e�M A"Y�$J���$���\�J�DA"��mNf�J���f�
�MJ��B	C?&;kJ�ap��k��J��A"��
��+�
J���� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&����J�ҥ�$%CJ������J���N��J���&�pJ�X"�8�$J��9�9�l!DB"J��#��@"��J�T��#YC��$J��#IN��J�e�&�� :J�� A"Y�$J���$#��#��J�n�
��\�J�DA"nk�k��J�m!mNf�J��(����J���$f�
�MJ��B��
�N�J��N?&;kJ�ap
��k�J�,�k��J��A"<NG��B"J��N�P�NJ�.�؝ ��J�[N6���sJ�F�14OJ���$�y�J���T�M�3!J�)��B"&�#��J�)Y�:��
m!J���q��
J����q���P���:y�H�Cm!�P��o��:w�e�z��P��o��:�k����P��o��:J�#����P��o��:$�����P��o��:��4��P�P��o��:�f��k�P��o��:�� Y�)Q�P��o��:�3��O�P��o��:��Sl9��P��o��:�C��$�m!�P��o��:�:%�C�P��o��:��
U���P���:��f����P��o��:����;��P��o��:�Q���
�P��o��:$��N��
�P��o��:�N���B"�P��o��:����
�P���:��
�:���P�C�:����#��P�C�:3C"=�Y��P�C�:��
��Q�P�C�:���1n!�P�C�:"�����P�C�:��
�
�P�C�:KP�&1q�P�Co��:y�
� �q�P�Co��:�$'�$.� �P�Co��:D�����P�Co��:��M	�t�$��:	
$�]��t�$��Z�:4~-�
?�
t�$��Z�:�%�
���o��:-�9�
g� ��o��:
$q�����:�QgPM�
���:I~Q���
�qB����:7�!�����
�&���:�$a�#W��lq����:��
������:OC"V�� ���:�q�##�Q��
�:��� ���Q��
�:��B���O�Q��
�:?�LrP�qj�Q��
�:�q7r2�e���Q��
�:��R���:�:�Q��
�:XrN�$�r��L;�Q��
�:n�$��
�k�o��Q��
�:fn!#'��g�3� �Q��
�:r� D����R��n!u��+��m��
P�����
Ԉ��l�
bC"�v�#��#�� ��C"�r���P�� NPo!��$�dRqP��s#��D"�;[�#�˶+�
����.�F�F���D"�l)��4D���r����#`��;'�� S������;a�}�� �$~Qs#��D";o!m�� �;QKo!����� ��#����˶�;P�j�����<��+st��-�`R;<�����
���;'UDvD�#���Q��H�� �$+ST#s#��D"�����R��8m��$��
� ���Os����
�$W��R�L�ms���� ��E"����D���P��#|�ES��%\�ͺ��<M"�ͺ��0
$�"ͺ����$Gͺp_��vͺaM"H�
�"ͺ�+GqM"ͺT
$��PGͺi��]�ͺ7��]�ͺ`
$x
$�ͺ*����ͺ�P���ͺ�Z�vͺ:��"�0ͺ)w!���ͺX	� �ͺ 1_�ͺ�
$���ͺO��]�ͺ�$#}~�Pͺ(��
$�Pͺ?����Pͺ?� C��Pͺ@w!���Pͺ�M"�$#�PͺE��P�PͺR�#�Pͺ��~�Pͺ��

^�Pͺ+G�PͺS��~�PͺWw!&%�Pͺ1^x��Pͺf��~�Pͺ!���$
ͺ�M"��
ͺ;�41
ͺ��o�
ͺœ�Z
ͺ�_ ��vͺ�����vͺo	Q�vͺ�M"��J%ͺ����
J%ͺ��
�M"J%ͺg��
$�M"ͺ�_R��M"ͺ��N"�M"ͺܜ�
�M"ͺ|�V� �M"ͺ)�
C��$ͺ;#%��ͺ�����ͺ<�
l�
�ͺ��$v��ͺ(�D�S#ͺ�$#n#S#ͺ7Q/N"S#ͺ%#pGS#ͺ�Z��S#ͺ���S#ͺ��I�S#ͺm�i�S#ͺX1M^S#ͺ�TQS#ͺY�
�S#ͺ�
$SN"S#ͺ�Ga%S#ͺpw!��$S#ͺ!��%S#ͺwN" �${1ͺ�_��{1ͺ>���{1ͺ�#k{1ͺ�w!H�{1ͺ�G5%#{1ͺxQU�y�ͺw�_��ͺ�#�w!{1ͺ��q^{1ͺ�w!�{1ͺ�%�{1ͺ��^{1ͺz� ��{1ͺ����{1ͺ���Gh�ͺ²��h�ͺ�	z�h�ͺF���h�ͺ�� �Gh�ͺ
[H�� ͺ�#��� ͺղ�%�a�
� �%l�a�
� �#��a�
%[��
ga�
������a�
w@w~�
a�
_��N"��a�
�	���Qa�
 $�>Ha�
$�Q�a�
?���N"a�
Y%#��
X�a�
dw
&%$a�
 ;���a�
Q$���
a�
��
�r�a�
yw�w�	a�
7�J[�a�
o[�	��a�
@$�QD�$a�
�����a�
k�$��c$a�
W�r�Y�a�
	���$w�a�
�����[a�
!R� �^a�
��$����$a�
�w��]$a�
�^. �a�
2&�[*�a�
����a�
�^G���a�
Q�����Y��~$���Y#��$p%#�Y/�
�[_�Y�k�,_�Y"�
�)��Y�d�%#_��Y��$�w��$�Y<RH�ѝ�Y$�l��$�Yk��%#���Y`z�/�E�$�N".�A�E�$�H��S E�$_�#��E�$�w_ �HE�$F�
�$�[E�$�	\|�E�$� �$:�$E�$I&+\8�E�$��W�A� E�$��H�z�$E�$�$@_�$E�$�$Z��HE�$y�O"�E�$�$�%#k�E�$����d&E�$F�}�IE�$��.�IE�$a�
d_� E�$������%URN� �$%��$�$%p�
lR+�%��%Œ
%T���$Œ
%�%#N\I�%����I�%:$&#o�%Q$
o�%2O"I��$%CO"��
x%�
��x%��=Ia�%(%:&#a�%ExE%*
%?
^&#*
%i%r\�\%^�^x�\%�1��
%�%���
%�\#`� %��h$� %�xԳ�%�1���%�\aIQ�%�&�&#Q�%+�
�x�1%�%�$�1%�r� �_%�&C�
�_%�R�� �%���I�%�I�&�%%��
+��%%�%��&%��$���&%'�x�� %m��1�� %�R
�$%��(��$%�1��
��%�%��R%L���g�
%
�!g�
%ޭ v
g�
%
$w�
g�
%G`
�g�
%q���
g�
%��$�w!g�
%�� ��g�
%�_^O"��
%�%'�
%�
2�&#%��
?��&#%�x.
$�
%/2x!9x!J��x�R�O"J�y�\�$J��I4��$J�c�X��&#J�&���O"J�S����J�O�A'{�J�
]{�� J�����J�yn���J�%]���_J��_``*!J��`�`#&J���
%�ہJ�e'0��IJ�A!'#(�J��O"R
$1yJ�82L&G� J��O"��p&J�	J��_yJ�Hx!���J�u
$e�*'#J�T�'�$j�J�(J�
$j�J�>�'�$S�J��ex!M'#J���x!��J�E]��-SJ���'�$��J�h�P"J�J���#h'#��J��Y2�_J����y��J��@J��J��y�&�`J��&|���J���Γ
s�J�~2+���J��2�&O�J�ǖJ�$�
J���m�$�J�]!m�$�&J�ۖm�$�J�K]���'#�'y!�`�=�
� ͞���$�2Q��3P"�x!�
$���,���y�� �x!���b���$����a�3��������
�����x!�
�x!����y`��-�Į �w�cJ��y!e]ͺ�
$�$2�� ͺ��ҴHSE�ͺ2'��!�ͺZ�
p
�2��
ͺu(�_z{1ͺ��'y!��2ͺ+z��'#~]ͺ��� �
ͺU�g� 9�	^ͺ����
~'� ͺ����Cͺ����O���
ͺ�'�*�!\�ͺ����#]�
+ͺFy!�;"=�i{ͺe+U�$�c�
ͺ�y!kS�$��ͺs��SP"�ͺ�A`�JVaͺ$����ͺ ��$��ͺ�J�$�$Ŕ
ͺ���S�!a�ͺ��wP"3`�(ͺo`~^*��'#ͺf�_�\a�aͺ�
���+��$ͺ�^4D���ͺ�a�z!L)��ͺ"���
��ͺ��
Ο8�.�ͺ��#�*�ͺ�ai�FK �ͺ:{!�P"�bJ(#ͺ�m_t�
�ͺ�$>�
$�$7�ͺ��[3�)�Kͺ�{������ͺט�bi� ,$ͺÙ��5#��ͺ�b�+c��ͺ�K�$�cU|!ͺ(�Qcld|ͺq#��Q��P"ͺ)#��նs$ͺ�c`�$�|!�ͺ3�� �į ͺ�c���
>�ͺh�
��)�  �ͺ�}�_rdͺ�d��F4�ͺ;}��(��4ͺ��$�e�'�ͺ*#e� ��)�ͺ%��)��
ͺ�e�$w��_ͺ'e�S<~�*#ͺd��
�� /�
ͺ�Q"X��|!߰ ͺ�����,ͺ����$ͺ��	�����ͺ�
g�5�ͺ�,x~*g��ͺ�t�uL2+#ͺ����=R",�ͺ}�
"}!�L:�ͺ���$��d}!ͺ�����ͺyR"�R"��ͺ)&�`�$�ͺ?��-Hf��ͺ�� '�~��
ͺ�������4ͺ�+#y� �� �`ͺ��$;��m'ͺR�fg�'aͺ�a��*ʇͺ�$�L�g!�ͺ��%��$(ͺ$<*�(5ͺ�S�g��bͺ�R"�~��q�ͺ:)��X��TͺSUh8�R�
ͺ� Y���
ͺQ�����ͺ\$e����ͺ��Th����ͺ��A��� �ͺfV��
^W��ͺ��������ͺb*��5�5ͺt�Zc�h�cͺ��$��!~!ͺ�S"�����~!ͺ�h� [$
�ͺl6p�ė
��ͺ�,#�!�ͺ�
�-J*>dͺT!n�$��$�*ͺ�T"��U�E�ͺ�$����a�
�����d(�a�
����$،	�a�
:� �-#.�� �Y�i��
�$}��Y��P�07e�YJ�
�;�mj�Y<��).M��Y+e"�
+À�Y�U"�dE�I��Y/+L.W����Y^��ɘ
�-#E�$�$7��-#�fE�$�`��� \�E�$�$��=MͲ %K-�j1.L%�W"6���q�
%����$��%S7z���
%ƛw�X��jJ�������$q�J�`��� B+��J���M����J��7VM��kJ�������$J�ik�����X"J�t�K�4�E�J�b�2Y"�-#
$J�����f+J��f��+ѥJ�.#��
��d�J�"l$��oMJ�� �-�X$J�8���W��J�n�$���M�J�A8T�+��J��(J/j�J��g>�U�S�J���>�y�8�J��h��
�MJ���
NIY"�$J��W��J� NJ�9h��#[�!J�`.�l����!J��flY"�+�gJ����Y"{:,J��z.#M,�0J����g�$�:J���!q���Q�J��Y"��
��o,J��l�,�W�,J���T.�:�
J���0hNh$J� ���#��
��J���
�N���.#J��!�g#�l1J�-��m�:J��g��g.	�J��mp�M���J�'� P���� �Xށ�
B�
�����ܨZ������
��X���m3��.������hqh��h����
����!y!e]ͺ����
2�� ͺ��ҴHSE�ͺ2'��!�ͺZ�
p
�2��
ͺu(�_z{1ͺ��'y!��2ͺ+z��'#~]ͺ��� �
ͺU�g� 9�	^ͺ����
~'� ͺ����Cͺ����O���
ͺ�'�*�!\�ͺ����#]�
+ͺFy!�;"=�i{ͺe+U�$�c�
ͺ�y!kS�$��ͺs��SP"�ͺ�A`�JVaͺ$����ͺx���ͺ_� ��$l�!ͺ3��S�!�Yͺ�Z"~^*���ͺ��F$��!K�ͺK�Y��.#߃!ͺ��
_�\a�hͺ�4D�7-ͺ�����
m["ͺ@i0Z�
ͺ&;��1��ͺ�-Ο8��ͺ�
i�FK�.ͺ��m_t�c/ͺ�["�?h(�ͺ� >�
$�$�1ͺ��[3�)�ͺ{h����L�!ͺ���5#��ͺ�\"�<��ͺ���\ZM<ͺ�Z�+c�.ͺS�Qcl!jͺ�.#��ն��ͺ�o$/���ͺ��`�$�|!D2ͺ�20
�
?nͺ<����$�ͺ����CN��ͺiN�
��$ͺ����
҄!u�ͺ|�$�� �'0ͺ�$��)� �ͺ	/��F4�ͺ0 � ��xnͺ�!�p����ͺo��(�p[ͺ�ݻ�i&jͺ�<��)�=ͺ���$w�9�ͺ����$���ͺF�
�S<~�Nͺ�X��|!pͺ�����D� ͺo�	����
ͺ�!;#�]"ƶ ͺ�jg��kͺ�
t�uL�ͺ�
"}!�L��ͺe�$�$���[ͺS��pQ��ͺ����
ͺ���`�$W�ͺ$l '�~1ͺk�g3bmh�ͺ�3�4����
ͺ6<*�(]� ͺ��~����ͺ�1h8��Nͺ�!p�$�4��ͺ�0	�
_��1ͺ�>� Y��\ͺ85e��4�ͺu�A�����ͺ>��$�O��!ͺ����)�
ͺ���
^W�pͺ	2��5z�ͺ��!$��ͺ������ �� ͺp?��$ʑ^"ͺ~�����7qͺ�qĞ���^"ͺU���
���!ͺ�j��
l@�kͺ��`��22�!ͺ��p�ė
��ͺ���-J*�/#ͺ�2��U�=�ͺ��l��՜
ͺ�����a�
������(�a�
����$،��a�
:� �-#.�� �Y�i��
�$}��Y��P�07e�YJ�
�;�mj�Y<��).M��Y+e"�
+À�Y�U"�dE�I��Y/+L.W����Y^���$�-#E�$Rr7��-#�fE�$�`��� \�E�$�$��=MͲ %K-�j1.R�%�W"6���q�
%����$��%S7z���
%ƛw�X��jJ�������$q�J�`��� B+��J���M����J��7VM��kJ�������$J�ik�����X"J�t�K�4�E�J�b�2Y"�-#
$J�����f+J��f��+ѥJ�.#��
��d�J�"l$��oMJ�� �-��$J�8���W��J�n�$���M�J�A8T�+��J��(J/j�J��g>�U�S�J��
>�y�8�J��h��
�MJ���
NIY"�$J��W��J��J��0#h��#[�!J�`.���#łJ���l����!J��flY"�+�gJ����Y"{:,J��z.#M,�0J����g�$�:J���!q���Q�J��Y"��
��o,J��l�,�W�,J���T.�O�
J���0h�@$J� �%�B U$J��5��#d���J���
�N���.#J��!�g#�l1J�-��m�:J��g��g.�OJ��mp�M���J�'� ?��m��$��m�r�n�@���NPn�
��x$�$=�
�\��62#�3c����!�� d�!���1��;�����-s$30_"�X�e �H3�PP���� �XAn_"�n�%AUD�A�3�J�Q�� 4��BR�t�M$���!ށ�
B�
�����ܨZ������
��X���m�n%]��
��� �:)�ͺ����Kd��ͺ����Y�ͺ&2#�z:9"8ͺwdf9"fHͺܒ�7H]ͺ�0���ͺ�����6�ͺoD��z�Gͺ�G��a#�ͺ��0��~
ͺ�~
ۙ�d��ͺ���B�$H^ͺH^��$��$�~
ͺ���^k�ͺ�7��#&��lͺ�l�����ͺ��Zh!���9"ͺ��9��$��ͺ1��~
�oͺ
���#��ͺA�ȕ 0G�ͺ�
�J���ͺo^�z)�P�ͺ�$jh!n�w�ͺ�3�9"
�^ͺ�^A�U�/8ͺ����$�$��ͺ���� I��ͺ`�
I����ͺv y�����
ͺ��
?���ïͺï8
��#ͺ1_F�#�e�$ͺ�$V� �HͺHV�#�z�9"ͺ�9"�b_P�ͺP�P
D�'�ͺB�$p
*�$�ͺJ��_��
ͺx�\{�I�2#ͺ>1@HG3#ͺe�#�#q�ͺ����_"ͺO���
ͺ�ƳZ2Lsͺ����ͺg� �;"ͺ� ��
�Ktͺ�[��$ͺR�~h!}�#�3#ͺ�l�,�`_ͺד��#VG�ͺ�;"���
ͺU�$��@�ͺ�`�_ͺA`��P�ͺY��C$ͺ��KZ8ͺ��0�ͺ��4q�$z�ͺ)4��
��� ͺ��x�&4_`ͺ�)�~��Hͺ�<"	p�ͺ�� ���K�
ͺ�{����ͺ�#1h��ͺI4Z���ͺ?`"v!���ͺ�o8	���$ͺu4�4I�ͺ�t��<"�ͺR4��##ͺ##�3��
�ͺ�8�
��ͺ�|m��
�ͺ
��`}�4ͺ�4��:��##ͺ�##"pa��ͺ��4#aZFͺZF!��=��ͺ����
Lr�
ͺr�
�����ͺ��r���
r�ͺr��m��'�
ͺ'�
}Q��x�ͺx�_�Q��ͺ��2uͺu`���$�`"ͺ�`"e�!�?!Pͺ!P�
 ��8ͺ�8�BG($ͺ($�#	��!ͺ�RpX�a�
>��{G<�a�
� �di!��$a�
��$vJh���#a�
�r�j�U�a�
y�:#~��#a�
\up�
��
�i!a�
�u��
���a�
���|��a�
���G�
a�
Q�]h�<"�Y�O#��.��Y"�
-�
l��Y��"j!�
�Y$!"j!�
�Y�dE�Q�
�Y���L�G���Y��[�C5��$�Y��!f�h;x��Ye�ؠ�G��
�Y��������
�Y�!Fj!���YG�@jj!P��Y3�
!�
+�#�� �Y��P�#��d�Yw����� �Y?�6��="���Yg�
����="�Y@�$�!��Y�3�
`�#�Yݬ��MQ�$�Y�m���e/l�Y�M�:� �Y� f"8!�Y� >"��� �Y���j!�u�Y��=��B�Y�*� �Y��p#�!�Y��4�x
�C���{
�4#��\tot}t�4\�$�Ep���C�4���V_ot���4������u��d!�4Zj_�ot�e!�4�t	�#��<f!�4��|

7otF�4E���C�#�47w���$la�4-}
��p��4)�$�8�
k� �4�D�f!I7"�\�4��7d}
�\�4�LO����4E�c#�f!���4 700Z��4����|�#5� �4��E�
�7V�4�j}
ٶ�g!�4��#��$� ��4w��� � ���4�7t}
;D�#�4Y�)�Dla�4)#R�
U���40
��D�E�4/�@�W#�4%� 	�/0���4�*����� ]�4N�7�#�]�4���
���y�4�y��$��`G�4��
�]��?0�4FL�H��}
�4����$�#���4��$�H��$�4�4;^<�
u��$�48v3�J�
�7�4dF\�
���� �4�F;�#���4jc~
S����4A��Eٶ,�4l�
[�8���4�c��8n��4�
�cN�*�
�4�#q� �����4�#w�����
�4V�
�z.9"�4����S���4-~
�ES����4��
7�S��F�4R^�S�[~
�4^�0S���4g��GS�|~
�4���#S��� ؁
��E�$��@�VE�$=� ��"E�$}�
�сE�$��
�Jw5E�$E�A�#E�$�#^���`"E�$�H����E�$�L�$�E�$�q�*�E�$=���
g� E�$�M��CE�$�lc�9E�$�A���
E�$W�<f��E�$��h
�QE�$�6�# �E�$���MNE�$�����
��E�$،!������E�$��)D�!hPE�$hP�"�^E�$�^���
��
E�$��
��DR�`E�$�`E��REnE�$Envp��m$E�$m$���v�E�$v�V�!�kE�$k<$��!&�
E�$&�
�R�RB"E�$B"�� 7a"��E�$��G��9sa"E�$sa"��!��E�$���s�d�$E�$d�$w�(�E�$(��4#+S�E�$�gS.�YDE�$YD�a"�a"r$#E�$r$#��$zn%�E�$%��pR�
#�E�$#�$?"�#��
E�$�JfE�$QNi���E�$�#�J�#E�$�#�<$�E�$�
�<hfE�$hf�<!�$j�E�$8���pE�$����?�E�$����
E�$��D#�E�$�� -�#kNE�$h#��ނE�$Q�#,nz�E�$E�$o��9E�$�� "$t�E�$����E�$Pn,��SE�$�N����E�$��q$E�$�K���%�v#�$Q�%.��n��
%�+A�W�%�"S�
����$J�_��KG$J�5#Lm�
�6J���
�=�J�)�K�ݘ J�� �?"��J�3� R7m�#J�wX�
S�J�k��L��J���O��J����B�J�ǒ�#�fJ�[�$���?"J�o�|�
Y� J�w�\�� J���*O��f�
J�o�#��
J��M<wP:J�p:���
nNJ��>��
jtJ����`˓J����?"�7�J��:��$���KJ�q��
�#�J�5#�?"��J���4k!��J�?M� Xk!J��o�7��#J�+�Y���
J��$��}�J�!L�� ��J�����J�>�$Ś @L��J�E�"��$�\J��S�EJ�$cLJ�:�J��g�
�$J��$w�<�oJ��o��J�:�J�>�y�@"J��:�L�
�$J�"��
<�oJ��o`+�W�J��"��$J@"�OJ�"�/�#J�N�r� h%�OJ�P��DJ�m�$���l<J���L���#J��n~�
��Ӄ
J���i��g��J���g�L7 $J���K#�L6�J��
�=qH�J�@5#��
����$J��`��J�*J��h��#�3J�������
J��
��:�J�1b"V�n���J��!�A��m#J�m#{M��kJ�k��
��
3c"J�Va��
�{c"J�h�,�{�J�z $%�s�?�J��#E9}l!J�q9kpJ�Xp���AJ���
QPG�
b� J�U��J�o�y�
ƻJ���#����J����#����J��%��

J�TfP8���J�xMB�J�VT�ش�PJ�S�f��MJ���#�
4BJ��5���"�J�4�˓{�)�J��!@�X/�J����@"��J�IIN��J��M A"Y�$J����\�J���mNf�J��f�
�MJ�	C?&;kJ���k��J���
��� $J��<�� ԇ_5#J�O�
�L�J�ҵ��aqJ�
,��$J�d&��Π
J�ҥ�$�#J������q�J�����NʯJ�_���&�nJ�X"��TJ���@"��J���!�#YCcJ���$�#IN��J�١�&�� O�J�:� A"Y�$J�F6#��#��J�n�
��\�J�p�nk�k��J�m!mNf�J��6(��#�J�����$f�
�MJ��$��
�N/�J���N?&;kJ�T�
��k�J��,�k��J�S�$<NG��B"J��N�P�NJ��<c�5#�qJ�؝ ��J���$6���sJ���F�14OJ��5#��$�y�J�T�M�3!J�U�$I$m!J��B"&�#��J�׿�:��
m!J�q��
J�e9��$�P�v!���:6�����P�����:m��v!�\�PdO���:m�#�
[$�PwE���:	!$��=�P����:�Df�(!$�P�!$���:�G�
,�
�P��
���:�6ԏ!���PW���:��^8K"�PP����:�O�O�{�P]����:uoY�Y�P�����:���]��P�����:O�_���P�����:�c"��`��PN4���:"uz����Ps����:"� �YǏ
�P?u���:�$�Lo�P�6���:����P�!���:�����Pv!���:]�
�6#ڡ
�P�o���:�6#:7�$�P�����:��$G���P�����:��
NO��P/���:i����!$�P(�!���:���N�N�P}{���:�6#
�$1��Ul�Q�:�$��*^�U[� Q�:v�Y!��U�K"Q�:���	$���UQ�:B� ��	$�U"�
Q�:i����$�U��Q�:ҿ���/�U	�##�:i����!$�x
(�!�����N�N�x
}{����
NO��x
/���6#:7�$�x
������$G���x
����]�
�6#ڡ
�x
�o������x
�!�������x
v!���$�Lo�x
�6��uoY�Y�x
����"uz����x
s����G�
,�
�x
��
��m��v!�\�x
dO���O�O�{�x
]����c"��`��x
N4���Df�(!$�x
�!$��6�����x
�������]��x
����"� �YǏ
�x
?u���6ԏ!���x
W��m�#�
[$�x
wE��e9��$�x
�v!��	!$��=�x
�����^8K"�x
P���O�_���x
����v�Y!��x
�K"QB� ��	$�x
"�
Qi����$�x
��Q�6#
�$1��x
l�Q�$��*^�x
[� Qҿ0��/�x
	�##�����U�`w"$$�c"��c"u��+���m��
P�������
Ԉ�+d"�l�
 �$��4v�#��#P7��r������NPo!�4�"pqP��Zc��7�;[�w��D:"$(E�d"��d"+�
������.�F�F��}p�D"�l)���4D���X�!���#=���� S@��d"�;a���G~QZc���;o!m�� �Ko!�p��� q�#�����6#�;P�j��|����<�f��+st��<�-�`R^��[q����
�w�UDvDX��LEQ�����F5+ST#Zc�������["$��q�8m�5�0r�
� �E��5Os��"$���$���������[���$x�U��$���5��E������"$�C6�89�C6��cM����
�$W��q��L�ms��!��$� ��*� ���
v=ڐ!�q�����"$D��D���#�ES��:)�ͺ����Kd��ͺ����Y�ͺ&2#�z:9"8ͺwdf9"fHͺܒ�7H]ͺ�0���ͺ�����6�ͺoD��z�Gͺ�G��a#�ͺ��0��~
ͺ�~
ۙ�d��ͺ���B�$H^ͺH^��$��$�~
ͺ���^k�ͺ�7��#&��lͺ�l�����ͺ��Zh!���9"ͺ��9��$��ͺ1��~
�oͺ
���#��ͺA�ȕ 0G�ͺ�
�J���ͺo^�z)�P�ͺ�$jh!n�w�ͺ�3�9"
�^ͺ�^A�U�/8ͺ����$�$��ͺ���� I��ͺ`�
I����ͺv y�����
ͺ��
?���ïͺï8
��#ͺ1_F�#�e�$ͺ�$V� �HͺHV�#�z�9"ͺ�9"�b_P�ͺP�P
D�'�ͺB�$p
*�$�ͺJ��_��
ͺx�\{�I�2#ͺ>1@HG3#ͺe�#�#q�ͺ����_"ͺO���
ͺ��JQ�
�=ͺ��
 8*ͺbHZ�r�ͺƳZ2Lsͺ����ͺg� �;"ͺ� ��
�Ktͺ�[��$ͺR�~h!}�#�3#ͺ�l�,�`_ͺד��#VG�ͺ�;"���
ͺU�$��@�ͺ�`�_ͺA`��P�ͺY��C$ͺ��KZ8ͺ��0�ͺ��4q�$z�ͺ)4��
��� ͺ��x�&4_`ͺ�)�~��Hͺ�<"	p�ͺ�� ���K�
ͺ�{����ͺ�#1h��ͺI4Z���ͺ?`"v!���ͺ�o8	���$ͺu4�4I�ͺ�t��<"�ͺ�#��##ͺ##
�$��
�ͺ��##��ͺ�/0���Eͺ�E�Y}9�
ͺ9�
�O:��##ͺ�##1a��ͺ�+ZaZFͺZF6��=��ͺ��"VLr�
ͺr�
�E���ͺ������
r�ͺr�����'�
ͺ'�
�
��x�ͺx�@>Q��ͺ�(�$2uͺu���$�`"ͺ�`"r��?!Pͺ!P17# ��8ͺ�8�uG($ͺ($�#	��!ͺ�RpX�a�
>��{G<�a�
� �di!��$a�
��$vJh���#a�
�r�j�U�a�
y�:#~��#a�
\up�
��
�i!a�
�u��
���a�
���|��a�
���G�
a�
Q�]h�<"�Y�O#��.��Y"�
-�
l��Y��"j!�
�Y$!"j!�
�Y�dE�Q�
�Y���L�G���Y��[�C5��$�Y��!f�h;x��Ye�ؠ�G��
�Y��������
�Y�!Fj!���YG�@jj!P��Y3�
!�
+�#�� �Y��P�#��d�Yw����� �Y?�6��="���Yg�
����="�Y@�$�!��Y�3�
`�#�Yݬ��MQ�$�Y�m���e/l�Y�M�:� �Y� f"8!�Y� >"��� �Y���j!�u�Y��=��B�Y�*� �Y��p#�!�Y�`V�x
�C���{
`V#��\tot}t`V\�$�Ep���C`V���V_ot��`V������u��d!`VZj_�ot�e!`V�t	�#��<f!`V��|

7otF`VE���C�#`V7w���$la`V-}
��p�`V)�$�8�
k� `V�D�f!I7"�\`V��7d}
�\`V�LO���`VE�c#�f!��`V 700Z�`V����|�#5� `V��E�
�7V`V�j}
ٶ�g!`V��#��$� �`Vw��� � ��`V�7t}
;D�#`VY�)�Dla`V)#R�
U��`V0
��D�E`V/�@�W#`V%� 	�/0��`V�*����� ]`VN�7�#�]`V���
���y`V�y��$��`G`V��
�]��?0`VFL�H��}
`V����$�#��`V��$�H��$`V�4;^<�
u��$`V8v3�J�
�7`VdF\�
���� `V�F;�#��`Vjc~
S���`VA��Eٶ,`Vl�
[�8��`V�c��8n�`V���cN���`V���~z6`V�cn�8"X�`V	9"�c���
`VV�
�z.9"`V����S��`V-~
�ES���`V��
7�S��F`VR^�S�[~
`V^�0S��`Vg��GS�|~
`V���#S��� ؁
��E�$��@�VE�$=� ��"E�$}�
�сE�$��
�Jw5E�$E�A�#E�$�#^���`"E�$�H����E�$�L�$�E�$�q�*�E�$=���
g� E�$�M��CE�$�lc�9E�$�A���
E�$W�<f��E�$��h
�QE�$�6�# �E�$���MNE�$�����
��E�$،!��b��E�$���t�
hPE�$hP���^E�$�^��o� ��
E�$��
��
C�b0E�$b0�
9��E�$���t>_E�$_$#��E�$�`]��
��E�$��["�
�	$E�$�	$���� %�E�$%��0���E�$����}m$E�$m$<:vv�E�$v��
��kE�$k(�
�&�
E�$&�
n���B"E�$B"6���E�$��f7#'sa"E�$sa"B�$�$�E�$�ߚ�7#d�$E�$d�$��vvR�E�$R�AxVe"(�E�$(�D8lV�E�$���!r�YDE�$YD�v!bZr$#E�$r$#��I%�E�$%�� u�#�E�$#�$?"�#��
E�$�JfE�$QNi���E�$�#�J�#E�$�#�<$�E�$�
�<hfE�$hf�<!�$j�E�$8���pE�$����?�E�$����
E�$��D#�E�$�� -�#kNE�$h#��ނE�$Q�#,nz�E�$E�$o��9E�$�� "$t�E�$����E�$Pn,��SE�$�N����E�$��q$E�$�K���%�v#�$Q�%.��n��
%�+A�W�%�"S�
����$J�_��KG$J�5#Lm�
�6J���
�=�J�)�K�ݘ J�� �?"��J�3� R7m�#J�wX�
S�J�k��L��J���O��J����B�J�ǒ�#�fJ�[�$���?"J�o�|�
Y� J�w�\�� J���*O��f�
J�o�#��
J��M<wP:J�p:���
nNJ��>��
jtJ����`˓J����?"�7�J��:��$���KJ�q��
�#�J�5#�?"��J���4k!��J�?M� Xk!J��o�7��#J�+�Y���
J��$��}�J�!L�� ��J�����J�>�$Ś @L��J�E�"��$�\J��S�EJ�$cLJ�:�J��g�
�$J��$w�<�oJ��o��J�:�J�>�y�@"J��:�L�
�$J�"��
<�oJ��o`+�W�J��"��$J@"�OJ�"�/�#J�N�r� h%�OJ�P��DJ�m�$���l<J���L���#J��n~�
��Ӄ
J���i��g��J���g�L7 $J���K#�L6�J��
�=qH�J�@5#��
����$J��`��J�*J��h��#�3J�������
J��
��:�J�1b"V�n���J��!�A��m#J�m#{M��kJ�k��
��
3c"J�Va��
�{c"J�h�,�{�J�z $%�s�?�J��#E9}l!J�q9kpJ�Xp���AJ���
QPG�
b� J�U��J�o�y�
ƻJ���#����J����#����J��%��

J�TfP8���J�xMB�J�VT�ش�PJ�S�f��MJ���#�
4BJ��5���"�J�4�˓{�)�J��!@�X/�J����@"��J�IIN��J��M A"Y�$J����\�J���mNf�J��f�
�MJ�	C?&;kJ���k��J���
��� $J��<�� ԇ_5#J�O�
�L�J�ҵ��aqJ�
,��$J�d&��Π
J�ҥ�$�#J������q�J�����NʯJ�_���&�nJ�X"��TJ���@"��J���!�#YCcJ���$�#IN��J�١�&�� O�J�:� A"Y�$J�F6#��#��J�n�
��\�J�p�nk�k��J�m!mNf�J��6(��#�J�����$f�
�MJ��$��
�N/�J���N?&;kJ�T�
��k�J��,�k��J�S�$<NG��B"J��N�P�NJ��<c�5#�qJ�؝ ��J���$6���sJ���F�14OJ��5#��$�y�J�T�M�3!J�U�$I$m!J��B"&�#��J�׿�:��
m!J�q��
J��q���P���:y�H�Cm!�P��o��:w�e�z��P��o��:�k����P��o��:J�#����P��o��:$�����P��o��:��4��P�P��o��:�f��k�P��o��:�� Y�)Q�P��o��:�3��O�P��o��:��Sl9��P��o��:�C��$�m!�P��o��:�:%�C�P��o��:��
U���P���:��f����P��o��:����;��P��o��:�Q���
�P��o��:$��N��
�P��o��:�N���B"�P��o��:����
�P���:��
�:���P�C�:����#��P�C�:3C"=�Y��P�C�:��
��Q�P�C�:���1n!�P�C�:"�����P�C�:��
�
�P�C�:KP�&1q�P�Co��:y�
� �q�P�Co��:�$'�$.� �P�Co��:D�����P�Co��:��M	�t�$��:	
$�]��t�$��Z�:4~-�
?�
t�$��Z�:�%�
���o��:-�9�
g� ��o��:
$q�����:�QgPM�
���:I~Q���
�qB����:7�!�����
�&���:�$a�#W��lq����:��
������:OC"V�� ���:�q�##�Q��
�:��� ���Q��
�:��B���O�Q��
�:?�LrP�qj�Q��
�:�q7r2�e���Q��
�:��R���:�:�Q��
�:XrN�$�r��L;�Q��
�:n�$��
�k�o��Q��
�:fn!#'��g�3� �Q��
�:r� D�����U�`w"$$�c"��c"u��+���m��
P�������
Ԉ�+d"�l�
 �$��4v�#��#P7��r������NPo!�4�"pqP��Zc��7�;[�w��D:"$(E�d"��d"+�
������.�F�F��}p�D"�l)���4D���X�!���#=���� S@��d"�;a���G~QZc���;o!m�� �Ko!�p��� q�#�����6#�;P�j��|����<�f��+st��<�-�`R^��[q����
�w�UDvDX��LEQ�����F5+ST#Zc�������["$��q�8m�5�0r�
� �E��5Os��"$���$���������[���$x�U��$���5��E������"$�C6�89�C6��cM����
�$W��q��L�ms��!��$� ��*� ���
v=ڐ!�q�����"$D��D���#�ES��:)�ͺ���Kd��ͺ������Y�ͺ&2#�z:9"8ͺwdf9"fHͺfH�V
��>ͺc0�ͺ0�0�!vx^�
ͺ^�
#$��\�$ͺ����6�ͺ6�D��z�Gͺ�G��a#�ͺ��0��~
ͺ�~
ۙ^���ͺ���B��H^ͺH^�� &�~�ͺ~��cdr�ͺ�c�x��7#ͺ���$n��~
ͺ�����HͺO�Zh!I��9"ͺ��9����ͺ1��~
��
oͺ
���#��ͺA�ȕ ��~
ͺ�~
n�H�M�ͺ�$�>����ͺ�3������ͺ������$��ͺ`�
��8#��ͺ���r��l�ͺl�%FF�!��
ͺ��
I���	�ͺ�Vy�����
ͺ��
?���ïͺï8
��#ͺ�#F�#�e�$ͺ�$V� �HͺHV�#�z�9"ͺ�9"�b_P�ͺP�P
D�'�ͺB�$p
*�$>FͺJ��_��½ ͺx�+r�I��ͺ�x@H�rͺ��#�ͺsF���xͺi�!���ͺ#$ƳZ2%ͺ��>�ͺg� �;"��ͺ��
�K�ͺ�[���ͺ��}�#y8ͺ��r��cͺ��#VG�$ͺ�;"��ͺU�$�yͺ�`�ͺA`��?ͺY���rͺ��K�Fͺ��0�ͺ��4q�$z�ͺ)4��
��O�ͺ�e"x�&4_`ͺ�)�~��ͺ��<"	9dͺL?���K�ͺ��d1�� ͺ�� �{����ͺ���#1h]ͺ]I4Z��� ͺ�v!��ͺ��!8	����ͺ��u4�4�!ͺ�e"��%8#�6ͺ��dI�ͺ{����ͺ�6
e0sͺ
�$)�ɛͺ�##�$�Fͺ^�Rr�ͺ�8��a8#ͺ'� :��ͺf"[� O�ͺ�8#�� 4%ͺ��
��t�ͺ�p`G*�ͺls����ͺ��{#$�ͺ��$��8#ͺ�s7J�ͺ�Ee�aͺ����m�ͺ�V�Af"ͺ;#�?�ͺu�Fe�sͺ��t��ͺ	�$��9ͺj� Kt=�$ͺۣ
��$�Gͺ��$�8#�f"ͺ�#	��!ͺ�{G<�a�
� �di!��$a�
��$vJh���#a�
�r�j�U�a�
y�:#~��#a�
\u9#��1Ha�
1H~����Ha�
�H=7.�$��a�
'�@9#�5ya�
5y�t7��?a�
�p�
��
�i!a�
��
���a�
[�|�6�a�
�y��G�9a�
���H_7a�
s�	I&a�
Ų�
�qa�
��Q�]h�<"�Y�S�$���%�Y�O#��.��Y�t&�#$���Y(�"�
-�
l��Yl�$!"j!��Y��dE��y�Y/W�L�G���Y��[�C5��$�Y��!f�h;x��Ye�ؠ�G��
�Yk�$������
�Y��
�!Fj!���YG�@jj!�H�Y�t!�
+�#�� �Y��P�#��d�Y�dw����� �Y?�6��="���Y������="�Y�="@�$�!��Y��3�
`�#�Yݬ��MQ�$�Y�m���e/l�Y/l�M�:� �Y f"��YA�!>"��n"�Y��j!���Y��=���$�Y�*��Y��p#���Y��W�x
�C���{
�W#��\tot}t�W\�$�Ep���C�W���V_ot���W������u��d!�WZj_�ot�e!�W�t	�#��<f!�W��|

7otF�WE���C�#�W7w���$la�W-}
��p��W)�$�8�
k� �W�D�f!I7"�\�W��7d}
�\�W�LO��/�WǵP���D���WE�c#�f!���W 700X�#�W+&��k#X�#�W�r��u#Z��WXF��|�#5� �W��E�
�7V�W�j}
ٶ�g!�W��#��$� ��Ww��� � ���W�7t}
;D�#�WY�)�Dla�W)#R�
U���W0
��D���WV�-Ij�E�W/�@�W#�W%� 	�/0���W�*����� ]�WN�7�#�]�W���
���y�W�y��$��`G�W��
�]��?0�WFL�H��}
�W����$�#���Wq��H��$�W�4;^<�
u��$�W8v3�J�
�7�WdF\�
���� �W�F;�#��G�W�e:�JF��Wjc~
S����WA��Eٶ,�Wl�
[�8���W�c��8n��W�
�cN�*�
�W�#q� �����W�#w�����
�WV�
�z.9"�W����S���W-~
�ES����W��
7�S��F�WR^�S�[~
�W^�0S���Wg��GS�|~
�W���#S��L�$�E�$q�*�1JE�$1J=���
�fE�$�f�M�K�E�$K��l!KlJE�$lJ�A���E�$��W�<f��E�$����h
�#$E�$�#$�6�#��E�$���f''��E�$���MNE�$�����
��E�$��K'Ҿ '�E�$��s&�
E�$������E�$:����E�$����tE�$g"4��E�$-�v'
zE�$�K���%�v#f=Q�%Q�.��n��
%�+A�W�%�"�	�-SJ�-S�NS3�J�_��KG$J�5#Lm�
�6J��6��
�=�J��)�K�ݘ J�ݘ � �?"��J���3� R7m�#J�wX�
S�J�S�k��L��J�����O��J������B�J�B�ǒ�#�fJ��f[�$���?"J��?"o�|�
Y� J�Y� w�\�� J���*O��f�
J�o�#��
J�MMP:J�p:���nNJ��>��jtJ���i$˓J��?"�7�J���$���KJ�q��
�#�J�5#�?"��J����4k!��J���?M� Xk!J�Xk!�o�7��#J���#+�Y���
J���
�$��}�J�}�!L�� ��J�������J�>�$Ś @L��J�"��$�\J��S�EJ�cLx���!J��gU��$J��$w�>%9zJ���x�:�J�>�y�@"J��LU��$J�"��
>%9zJ�`+�W�J���$J@"�OJ��O"�/�#J�N�r� h%�OJ��Ob���J�P��DJ��-#����$J����5:J�A�L���#J��'J����J�~�
��Ӄ
J�aMJ��gIJ�i��g��J��3P�Lo�J��g�L�J�]:K#�L6�J�)�
�L
�J�8���aMJ��=qH�J�@5#��
����$J���J���J�h��#(�J�[�������J���V�n�
� J��A��m#J������!J�{M��Ȝ J�k��
��
�$J����
�{c"J�{c",�{�J�z $%�s�?�J��#E9}l!J�}l!q9kpJ�pXp���AJ�QPG�
b� J�o�y�
ƻJ����#����J�t��#����J�%��

J�fP���J�xMB�l!J��7�ش�PJ�S�f��MJ���#�
4BJ�p����"�J�4�˓{�)�J��$@�X/�J����@"$$J�����YC�	J��	IIN��J�Q�$�� 9J�9�M�lJ��J�W�J|�J����{�J�I$$��
ɝ�$J��$��?&�@J�����
�k�zJ��z�kEJ�O��fG�ä
J�ä
��
��+�
J���
�� ԇ1�J�1�O�
���J���ҵ��:�J�:�
,�J��d&����J���ҥ�$%CJ�%C�����J�����N��J�����&�pJ��pX"�8�$J��9�l!DB"J�t$$�N�P�NJ��<c�5#�qJ��q؝ w:J��z6���sJ���F�14OJ�ދ��$�y�J�y�T�M��JJ��$I$m!J��B"&�#��J��:��
m!J�q��
J�e9��$�P�v!���:6�����P�����:m��v!�\�PdO���:m�#�
[$�PwE���:	!$��=�P����:�Df�(!$�P�!$���:�G�
,�
�P��
���:�6ԏ!���PW���:��^8K"�PP����:�O�O�{�P]����:uoY�Y�P�����:���]��P�����:O�_���P�����:�c"��`��PN4���:"uz����Ps����:"� �YǏ
�P?u���:�$�Lo�P�6���:����P�!���:�����Pv!���:]�
�6#ڡ
�P�o���:�6#:7�$�P�����:��$G���P�����:��
NO��P/���:i����!$�P(�!���:���N�N�P}{���:�6#
�$���U)uQ�:�$��lu�U,�Q�:v�Y!+��U�7Q�:t���b9#�UE:#Q�:���	$���UG|Q�:B� ���u�U�WQ�:{��A�$�U2vQ�:i����$�U7�Q�:ҿ0��	�U	�Q�:5�
z�z�x
_v��^��!0
�x
Ռ�������
�x

����g���x
������:�8�x
�������
g�x
{g���
��:�x
P�
��$;��
��
�x
�8��&Xt/{�x
���;#%���x
B9��A�9��x
�$��Otft��x
���|t4A>�x
���K�v*�$�x
E���g�z�/��x
z�$����JA���x
�!��v�Y!+��x
�7QB� ���u�x
�WQi����$�x
7�Q{��A�$�x
2vQ�6#
�$���x
)uQ�$��lu�x
,�Q���	$���x
G|Qt���b9#�x
E:#Qҿ0��	�x
	�Q�9�� ��x
6g"��J;���$���e��9��g"u�u;+��;��l� �$��gqP^�!���:K�;[�w��D+�Q�����4Da���$~d�!Zc���;o!m���Ko!����� q�{�'`A�`A�#�������;T#j���-���^��^�+S�AZc��$$���D��D���#|�ES��:)�ͺ���Kd��ͺ������Y�ͺ&2#�z:9"8ͺwdf9"fHͺfH�V
��>ͺc0�ͺ0�0�!vx^�
ͺ^�
#$��\�$ͺ����6�ͺ6�D��z�Gͺ�G��a#�ͺ��0��~
ͺ�~
ۙ^���ͺ���B��H^ͺH^�� &�~�ͺ~��cdr�ͺ�c�x��7#ͺ���$n��~
ͺ�����HͺO�Zh!I��9"ͺ��9����ͺ1��~
��
oͺ
���#��ͺA�ȕ ��~
ͺ�~
n�H�M�ͺ�$�>����ͺ�3������ͺ������$��ͺ`�
��8#��ͺ���r��l�ͺl�%FF�!��
ͺ��
I���	�ͺ�Vy�����
ͺ��
?���ïͺï8
��#ͺ�#F�#�e�$ͺ�$V� �HͺHV�#�z�9"ͺ�9"�b_P�ͺP�P
D�'�ͺB�$p
*�$>FͺJ��_��½ ͺx�+r�I��ͺ�x@H�rͺ��#�ͺsF���xͺi�!�t�g"ͺ
� �A�
ͺ!h7���ͺ���=XͺƳZ2%ͺ��>�ͺg� �;"��ͺ��
�K�ͺ�[���ͺ��}�#y8ͺ��r��cͺ��#VG�$ͺ�;"��ͺU�$�yͺ�`�ͺA`��?ͺY���rͺ��K�Fͺ��0�ͺ��4q�$z�ͺ)4��
��O�ͺ�e"x�&4_`ͺ�)�~��ͺ��<"	9dͺL?���K�ͺ��d1�� ͺ�� �{����ͺ���#1h]ͺ]I4Z��� ͺ�v!��ͺ��!8	����ͺ��u4�4�!ͺ�e"
�$)�ɛͺ�##�$�Fͺ^�Rr�ͺ��C��'ͺKh�v:ͺFh"�A~h"ͺ)�$~��vͺ���
ͺJ�}�\wͺ�����ͺ/0�h;ͺ�Y�lKͺ�w�$�ͺ���t,;#ͺ��
��t�ͺls����ͺ�s7J�ͺ�Ee�aͺ����m�ͺ�V�Af"ͺ���w�ͺ��t��ͺ�t�*xͺOX
i"	Lͺ@>��ͺ(�$˳�;#ͺ�{8� �;#ͺ�.��ͺ��{��ͺ���$�ͺr� ���ͺ17#:�g�$ͺuuXt� ͺ��B�hͺ��$�$Fi"ͺcLL�v�ͺۣ
��$�Gͺ��$�8#�f"ͺ�#	��!ͺ�{G<�a�
� �di!��$a�
��$vJh���#a�
�r�j�U�a�
y�:#~��#a�
\u9#��1Ha�
1H~����Ha�
�H=7.�$��a�
'�@9#�5ya�
5y�t7��?a�
�p�
��
�i!a�
��
���a�
[�|�6�a�
�y��G�9a�
���H_7a�
s�	I&a�
Ų�
�qa�
��Q�]h�<"�Y�S�$���%�Y�O#��.��Y�t&�#$���Y(�"�
-�
l��Yl�$!"j!��Y��dE��y�Y/W�L�G���Y��[�C5��$�Y��!f�h;x��Ye�ؠ�G��
�Yk�$������
�Y��
�!Fj!���YG�@jj!�H�Y�t!�
+�#�� �Y��P�#��d�Y�dw����� �Y?�6��="���Y������="�Y�="@�$�!��Y��3�
`�#�Yݬ��MQ�$�Y�m���e/l�Y/l�M�:� �Y f"��YA�!>"��n"�Y��j!���Y��=���$�Y�*��Y��p#���Y��X�x
�C���{
�X#��\tot}t�X\�$�Ep���C�X���V_ot���X������u��d!�XZj_�ot�e!�X�t	�#��<f!�X��|

7otF�XE���C�#�X7w���$la�X-}
��p��X)�$�8�
k� �X�D�f!I7"�\�X��7d}
�\�X�LO��/�XǵP���D���XE�c#�f!���X 700X�#�X+&��k#X�#�X�r��u#Z��XXF��|�#5� �X��E�
�7V�X�j}
ٶ�g!�X��#��$� ��Xw��� � ���X�7t}
;D�#�XY�)�Dla�X)#R�
U���X0
��D���XV�-Ij�E�X/�@�W#�X%� 	�/0���X�*����� ]�XN�7�#�]�X���
���y�X�y��$��`G�X��
�]��?0�XFL�H��}
�X����$�#���Xq��H��$�X�4;^<�
u��$�X8v3�J�
�7�XdF\�
���� �X�F;�#��G�X�e:�JF��Xjc~
S����XA��Eٶ,�Xl�
[�8���X�c��8n��X���cN�X��X	9"�c���
�XV�
�z.9"�X����S���X-~
�ES����X��
7�S��F�XR^�S�[~
�X^�0S���Xg��GS�|~
�X���#S��L�$�E�$q�*�1JE�$1J=���
�fE�$�f�M�K�E�$K��l!KlJE�$lJ�A���E�$��W�<f��E�$����h
�#$E�$�#$�6�#��E�$���f''��E�$���MNE�$�����
��E�$����2�%$E�$?B5�LE�$:%$��!E�$���UE�$$#��
�E�$`]�TuE�$��F:�<#E�$�0p�uE�$(�
/v�:E�$n��Xj"E�$h(���E�$�(� �!E�$�X|��E�$����E�$A
n�E�$f7#YxE�$��hj"�j"E�$u%$'�$qBE�$�B��;E�$��\�!\�E�$Ax\��E�$�%$�:�(E�$D8���E�$��ɦ
:�E�$�xG�}
E�$U|�|8�E�$������E�$�v!�j"4YE�$�����E�$�K���%�v#f=Q�%Q�.��n��
%�+A�W�%�"�	�-SJ�-S�NS3�J�_��KG$J�5#Lm�
�6J��6��
�=�J��)�K�ݘ J�ݘ � �?"��J���3� R7m�#J�wX�
S�J�S�k��L��J�����O��J������B�J�B�ǒ�#�fJ��f[�$���?"J��?"o�|�
Y� J�Y� w�\�� J���*O��f�
J�o�#��
J�MMP:J�p:���nNJ��>��jtJ���i$˓J��?"�7�J���$���KJ�q��
�#�J�5#�?"��J����4k!��J���?M� Xk!J�Xk!�o�7��#J���#+�Y���
J���
�$��}�J�}�!L�� ��J�������J�>�$Ś @L��J�"��$�\J��S�EJ�cLx���!J��gU��$J��$w�>%9zJ���x�:�J�>�y�@"J��LU��$J�"��
>%9zJ�`+�W�J���$J@"�OJ��O"�/�#J�N�r� h%�OJ��Ob���J�P��DJ��-#����$J����5:J�A�L���#J��'J����J�~�
��Ӄ
J�aMJ��gIJ�i��g��J��3P�Lo�J��g�L�J�]:K#�L6�J�)�
�L
�J�8���aMJ��=qH�J�@5#��
����$J���J���J�h��#(�J�[�������J���V�n�
� J��A��m#J�����kvJ�{M��Ȝ J�k��
��
�$J����
�{c"J�{c",�{�J�z $%�s�?�J��#E9}l!J�}l!q9kpJ�pXp���AJ�QPG�
b� J�o�y�
ƻJ����#����J�t��#����J�%��

J�fP���J�xMB�l!J��7�ش�PJ�S�f��MJ���#�
4BJ�p����"�J�4�˓{�)�J��$@�X/�J����@"$$J�����YC�	J��	IIN��J�Q�$�� 9J�9�M�lJ��J�W�J|�J����{�J�I$$��
ɝ�$J��$��?&�@J�����
�k�zJ��z�kEJ�O��fG�ä
J�ä
��
��+�
J���
�� ԇ1�J�1�O�
���J���ҵ��:�J�:�
,�J��d&����J���ҥ�$%CJ�%C�����J�����N��J�����&�pJ��pX"�8�$J��9�l!DB"J�t$$�N�P�NJ��<c�5#�qJ��q؝ w:J��z6���sJ���F�14OJ�ދ��$�y�J�y�T�M��JJ��$I$m!J��B"&�#��J��:��
m!J�q��
J��q���P���:y�H�Cm!�P��o��:w�e�z��P��o��:�k�=#�P��o��:J�#��C�P��o��:�f�٧�P��o��:�� Y��:�P��o��:�3���P��o��:�CT;���P��o��:��SlZ� �P��o��:C	��$���P��o��:�%$c�$Z��P��o��:�:9����P��o��:����Q��P��o��:�Q��Y	�P��o��:G)�CB��P��o��:�NM����P��o��:$��N�
�P��o��:�Cz;�	�P��o��:��
���P��o��:��
����P��o��:��!��c�
�P��o��:'�iv��P���:���H��P���:��y�P���:��
�:��P�C�:��
��C�P�C�:KP�&1q�P�Co��:y�
� �q�P�Co��:�$'�$.� �P�Co��:D�����P�Co��:��M	��!t�$��:�����Yt�$��:������h)��$t�$��:�=#2k"M
�]=�t�$��Z�:Gk"-�
Ч
t�$��Z�:Zw�:�����:�%�
����o��:-�9�
����o��:
$q������:�QgPM�
���:I~Q���
�qB����:7�!�����
�&���:�$a�#W��l�Y���:�q�##�Q��
�:�
�Lr)�Q��
�:�C�$è�*F��Q��
�:_iZy�h)-D�Q��
�:M��r��*�Q��
�:1ck"e
�0��Q��
�:�C�$n��� �Q��
�:_iZy�%$6M_��Q��
�:M��eF�&$�Q��
�:1ck"��� ���Q��
�:��B���O�Q��
�:?�LrP�qj�Q��
�:�q7r2�e���Q��
�:��R���:�:�Q��
�:XrN�$�r��L;�Q��
�:n�$��
�k�o��Q��
�:fn!#'s��+3� �Q��
�:�
����wk"'� �Q��
�:r� D�$mw�;�Q��
�:z'$�Y�@<���Q��
�:��Z� ���;+�
�Q��
�:�|�Z<�'$��Q��
�:�k"�=#|��!iE�Q��
�:z'$�Y���s� �Q��
�:��Z� µ� 2� �Q��
�:�|�ny���k"�Q��
�:�k"�=#GM� �$�/�^�!���J;���$���e��9��g"u�u;+��;��l� �$��gqP^�!���:K�;[�w��D+�Q�����4Da���$~d�!Zc���;o!m���Ko!����� q�{�'`A�`A�#�������;T#j���-���^��^�+S�AZc��$$���D��D���#|�ES���	F�yͺ'�!�+=�!ͺZ��ۨͺ�:)�ͺ���|}",ͺ��Kd��ͺ����Y�ͺ�z:9"8ͺwdf9"fHͺ�V
��>ͺc0�ͺ0�!vx^�
ͺ�y�$<ͺ#$��\�$ͺ���6�ͺoD��z�Gͺ��a#�ͺ�0��~
ͺۙ^���ͺ�B��H^ͺ�� &�~�ͺ�cdr�ͺ�x��7#ͺ��$n��~
ͺ����HͺZh!I��9"ͺ9����ͺ�~
��
oͺ��#��ͺȕ ��~
ͺn�H�M�ͺ�>����ͺ������ͺ����$��ͺ��8#��ͺ�'$w<�iͺ�r��l�ͺ%FF�!��
ͺI���	�ͺy�����
ͺ?���ïͺ8
��#ͺF�#�e�$ͺV� �HͺV�#�z�9"ͺ�b_P�ͺP
D�'�ͺp
*�$>Fͺ�_��½ ͺ+r�IqMͺ��x@H��ͺ ���#-FͺsF���wͺi�!�)�$ͺ
� 
},ͺƳZ2%ͺ��>�ͺ��
�K�ͺ��}�#E<ͺ�y��r�V�ͺ��#VG�<ͺk�$�;"��ͺZl"U�$�yͺ�`��ͺ�wA`��W�ͺwxY���Fͺ��K�'$ͺ��0�ͺ4q�$z�ͺ��
��O�ͺx�&4_`ͺ)�~��ͺ�<"	9dͺ���K�ͺ�{����ͺ�#1h]ͺv!��ͺ�,����ͺu4�4�!ͺ{����ͺ
�$)�ɛͺ�##�$�Fͺ^�Rr�ͺ��C��'ͺKh�v:ͺFh"�A~h"ͺ)�$~��vͺ���
ͺJ�}�\wͺ�8��a8#ͺ�����ͺ/0�h;ͺ�Y�lKͺ�w�$�ͺ���t,;#ͺ'� :��ͺ��
��t�ͺ�p`G*�ͺls����ͺ��$��8#ͺ�s7J�ͺ�Ee�aͺ����m�ͺ�V�Af"ͺ���w�ͺ;#�?�ͺ�t�*xͺOX
i"	Lͺ@>��ͺ(�$˳�;#ͺ�{8� �;#ͺ�.��ͺ��{��ͺ���$�ͺr� ���ͺ17#:�g�$ͺuuXt� ͺ��B�hͺ��$�$Fi"ͺcLL�v�ͺۣ
��$�Gͺ��$�8#�f"ͺ�#	��!ͺ�{G<�a�
�di!��$a�
vJh���#a�
r�j�U�a�
:#~��#a�
9#��1Ha�
~����Ha�
=7.�$��a�
@9#�5ya�
�t7��?a�
[�|�6�a�
��G�9a�
���H_7a�
s�	I��a�
�
��a�
Q�]h�<"�YS�$���%�YO#��.��Y�I}�����Y&�#$���YR��l"t��Y"�
-�
l��Y�dE��y�Y��L�G���Y[�C5��$�Yf�h;x��Yؠ�G��
�Y������
�Y�!Fj!���Y@jj!�H�Y!�
+�#�� �YP�#��d�Yw����� �Y6��="���Y�i����="�Y@�$�!��Y�3�
`�#�Y��MQ�$�Y���e/l�Y�M�:� �Y f"���Y��>"��n"�Y��j!���Y���,��Y��=���$�Y�*��Y��p#���Y��� �x
�C���{
�� #��\tot}t�� \�$�Ep���C�� ���V_ot���� ������u��d!�� Zj_�ot�e!�� �t	�#��<f!�� ��|

7otF�� E���C�#�� 7w���$p��� )�$�8�
k� �� �D�f!I7"�\�� ��7d}
�\�� �LO��/�� ǵ
���D���� E�c#�f!����  700X�#�� +&��k#X�#�� �r��u#Z��� 6���|�#V�� �j}
ٶ�E�� /�@�W#�� %� 	�/0�7�� dF\�
���� �� �F;�#��G�� �e:�JF��� jc~
S����� A��Eٶ,�� l�
[�8���� �c��8n��� ���cN�X��� 	9"�c���
�� V�
�z.9"�� ����S���� -~
�ES����� ��
7�S��F�� R^�S�[~
�� ^�0S���� g��GS�|~
�� ���#S��L�$�E�$:kq�*�1JE�$=���
�fE�$�M�K�E�$�l!KlJE�$�A��E�$W�<f��E�$��h
�#$E�$�6�#��E�$�f''��$E�$���MNE�$����
��E�$Z�� 0�E�$����E�$� !-qlE�$(�
/v�:E�$n��Xj"E�$h(���E�$�(� �!E�$�X|��E�$����E�$A
n�E�$f7#YxE�$��hj"�j"E�$u%$'�$qBE�$�B��;E�$Ax\��E�$�%$�:�(E�$D8���E�$��ɦ
:�E�$�xG�}
E�$U|�|8�E�$������E�$�v!�j"4YE�$���v�E�$�K���%v#f=Q�%.��n��
%X���%_Zy]}%+A�W�%�����
J��	�<-SJ���NS3�J�_��KG$J�Lm�
�6J���
�=�J�)�K�ݘ J�� �?"��J�3� R7m�#J�wX�
S�J�k��L��J���O��J����B�J�ǒ�#�fJ�[�$���?"J�o�|�
Y� J�w�\�� J�*O��h�J�o�#��
J�MMP:J����nNJ���i$˓J��?"�7�J���$���KJ���
�#�J��?"��J�����FJ���4k!��J�?M� Xk!J��o�7��#J���mJ�+�Y���
J��$��}�J�!L�� ��J�����J�Ś @LS�J�S�"��$�\J�2z�}��J�����<J�U�S���!J�cLx���!J��gU��$J�w�>%�oJ���x�:�J�>�y�@"J�l��LU��$J��$�
>%�oJ�`+�W�J�&?��$J@"�OJ�"�/�#J�r� h%�OJ�P�($J����#J��<��=J�i��LݷJ��g�LոJ��=qH�J���
����$J���J���J�4�h��#(�J�������J�<�ɨ
K�J�V�n��=J��A��m#J������MJ��g���J�q��<�<J�{M��Ȝ J���
��
�$J�
���
���J�	k�LzJ�,�{�J�%�s�?�J�#E9}l!J�q9kpJ�Xp���AJ�QPG�
b� J�o�y�
ƻJ�����#����J�����FȹJ�Җ!Bm�J�,	�l"��J����!���J�X��}�}J��xMB�l!J�ө�($L�J��$�GJ�>y{$�$J�]GUy��J��ش�PJ�S�f��MJ���#�
4BJ�{�$��.J��TyZJ�����$J�zy8�e� J����"�J�˓{�)�J�@�]-/�J��y���!J�H��l!V	J����@"$$J���YC�	J�IIN��J�$�� 9J��M�lJ��W�J�l"J����{�J���
ɝ�$J���?&�@J���
�k�zJ��kEJ��fG�ä
J��� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&����J��N�P�NJ��N<c�5#�qJ���$�y�J�T�M�3!J�7�!�$��
m!J��B"&�#��J�x�:��
m!J�q��
J��q���P���:y�H�Cm!�P��o��:w�e�z��P��o��:�k�=#�P��o��:J�#��C�P��o��:�CT;���P��o��:C	��$���P��o��:����Q��P��o��:�Q��Y	�P��o��:G)�CB��P��o��:�NM����P��o��:$��N�
�P��o��:�Cz;�	�P��o��:��
���P��o��:��
����P��o��:��!��c�
�P��o��:)�_����P���:��S���H��P���:��
�:���P�C�:��
�
�Q�P�C�:��M	�t�$��:�����Yt�$��:������h)��$t�$��:�=#2k"M
�]=�t�$��Z�:Gk"-�
Ч
t�$��Z�:y�!gm9���:�%�
����:-���!g� ���:
$q�!����:�QgPM�
���:I~Q���
�qB����:7�!�����
�&���:�$a�#W��l�Y���:�y�
�Z���:�q�##�Q��
�:�
�L�-�Q��
�:�C�$è�*F��Q��
�:_iZy�h)vG�Q��
�:M��r��*�Q��
�:1ck"e
����Q��
�:�C�$n��_��Q��
�:_iZy�%$6M^=�Q��
�:M��eF�&$�Q��
�:1ck"��� ���Q��
�:��B���O�Q��
�:?�LrP�qj�Q��
�:�q7r2�e���Q��
�:��R���:�:�Q��
�:XrN�$�r��L;�Q��
�:n�$��
�k�o��Q��
�:fn!#'s��+3� �Q��
�:�
����q�'� �Q��
�:r� D�$mw�;�Q��
�:z'$�Y�@<���Q��
�:��Z� ���;+�
�Q��
�:�|�Z<�'$��Q��
�:�k"�=#|��!iE�Q��
�:z'$�Y���s� �Q��
�:��Z� µ� 2� �Q��
�:�|�ny���k"�Q��
�:�k"�=#�����U�u�]�+�����ll�b�qP��Zc��7�;[�w�+�Q���4Da��~d�!Zc�;o!m���Ko!������#�����;T#j��-���^��+S�AZc��� ���D����#|�ES�y!e]{�ͺҴ����!ͺ=�!�!�2-m"ͺ�.z�Zͺ�� �!��ͺ��#�
Jͺ�H�;"�!�ͺ��!p(�ͺ#[�y�=#ͺ���y�H?{ͺ/�~
ͺ��o
�ͺ���ͺ� q	(~ͺ�$�R�ͺB[*�!��ͺ%�>:>#ͺs[E��.ͺ�����ͺ1	f�!Rm"ͺ�>�	ͺV��>�/ͺ�~=ʪͺ5� a	��ͺ)$��/�
ͺx��H�ͺ���{N�ͺy�?#ͺ��>K�ͺe� 4
=%ͺqm���mͺ���
�ͺ�[&��yͺ��!���	ͺ‘H�0ͺ�
Xn
@#ͺF$��!y�ͺңO�ͺ+
<�
p
ͺ>��ͺ�
�0?ͺ]��n�ͺ��?�� ͺb�8�[ͺWO�%�nͺ�{|��ͺ+%�zE)$ͺ�0,�,?ͺ���tͺ�?�m"@ͺ_��z�~ͺ�
@d�ͺF$� �@#ͺ:o"�`� ͺ
1�^�ͺB@S|/�ͺ���|ioͺ��ҙ!��ͺ �Y%�Oͺ��71ͺ>�
$�$�ͺ{�`\ͺ��!zͺ%�T�Q�ͺ�%J�!��ͺ��
I:n"ͺ�� �|� ͺ“h�V
ͺ��'�n"ͺ�O�o%A#ͺ���!�ͺ�<x�
ͺ��\Zš!ͺ�����ͺ�
M�pͺ��S�FIͺq
J��ͺ=��Kͺ�%9�IJͺ�� ��pͺ��,]�1ͺ�h]o@ͺ�!}.�
ͺo$/��@ͺ`�$�|!pqͺ��
?�ͺs���޽ͺ��
��ͺ��Au�ͺ��ի
ͺ�)$��ͺ��%ͺ�
�)$m2ͺ?�
�qe%ͺ�����ͺN�
��ͺ�p��cͺ�A#C}�ͺ��3�o�ͺ!��%S�ͺ}�%ͺ�A#m�
�qͺ��
�A!o"ͺ0B�}�ͺ^��@�� ͺ$��A#cBͺ*$?0{ͺ�$w�:�!ͺ��$��}ͺ����2ͺ*��B%Aͺ"���!ͺ�o"k�Jͺ��ž�3ͺ̬&B#�%ͺܬ
L*$��ͺ�*$>%��ͺ	���!�ͺ;#�]"^4ͺ�o"��]ͺ�W�PͺbB#����!ͺ����W�ͺ���N�ͺ5��*$�ͺ�B#x�!�ͺ��\�8�ͺ�P�rr�ͺ�
C#p"ͺ�$����ͺ�pQ��p"ͺYC��z%ͺ���p"�AͺC~�C2q"ͺ�{��{ͺ{��ͺ��,Q�ͺ>�
L� ��ͺ�%��� ͺhQ�K�rͺ�'� v~ͺ-�ϔͺ�(�ͺ�d�_ͺ�^���^ͺ�4�*$d�ͺ	p�
e�ͺn�%3+$ͺ�����
ͺ�4_��Cͺ[CBr"ͺܕ}D=ͺ��>C��ͺ���ͺ�K5,$ͺ�4���ͺ��!ts�r"ͺe�
K�
ͺc� �Lͺ�k�
�ͺ�
�s"��ͺ����B#ͺt"��zCͺ_���ͺ/�΁B5ͺ�|�|
�ͺp�$�4K�ͺ�t��,$ͺa�)��ͺ@t"�'ͺ�����,$ͺK��ͺ�!?�
�ͺ�L�s�!ͺ�D{���ͺQ�{��� ͺ�e�Yͺ����Dͺ��!�5Y%ͺ������ͺ�s��ͺ9}���ͺ�՞!NC#ͺ��L�!ͺ���-$��ͺ��!9ͺ�-$�IEͺst"���ͺ�C#�2��ͺ���uͺ�N��t"ͺè��Eͺ���ͺt�� 3�ͺ&M��Qͺ)D#YD#�ͺY���:� ͺ"�˟!ͺJt�� ztͺ���Iͺ��$ʑ��ͺ���5'Rͺ�Ci}R�ͺ�t�Qͺuu",V%ͺh��ͺ���0ͺ�}�Dͺ�F��
�}ͺ�?�t�ͺ�u"�Fuͺ.��� �Dͺ���dԮ
a�
VM�u"%e~a�
Y����a�
$6�-#.��Y��
�$� �YP�07�!�Y�).�R�Y"�
+&��Y�dE���YL.W���Y�ɘ
�RE�$�u";8
E�$G�Q���E�$�6K�{�E�$��� E�$��.$�uE�$��!n�E�$C.$w��E�$d���,%E�$�%��ME�$d.$�
v"E�$ń>���E�$-�l�E�$-Nz�G_E�$%�
�I�
E�$�
�F%vE�$����!�.$E�$�%�!�E�$PN�!�E�$�����E�$.�3�zvE�$#�nN��E�$Q�m-�!E�$��7E�$�%R�4GE�$�_V�E�$�_���E�$_��D#
%E�$�� Ř�.$E�$��D#� E�$�v"�E�$R���vE�$s�=�E�$�_)�o�E�$Ԫ w�v"E�$|Gw"�E�$q�Zw"\wE�$y�
���E�$��w"��E�$�NP��EE�$U%��
��!E�$��T7E�$F�j�NE�$�����E�$
Q� �GE�$wȫ��E�$c��� E�$��j�ӯ
E�$�7�E�7E�$L/$��3E�$���_e�E�$��
E#E�$�2O7�
E�$H��E�$�w"���E�$�D�&FE�$~��!��E�$&�nO��E�$P���{%E�$��t/$��E�$+��:`E�$w�$�yFE�$O���2�E�$�FN�!8E�$ G��`E�$�8����E�$v�I��wE�$�w"S��
E�$����JGE�$}E#�E�$����E#E�$�%����E�$jx�xZE�$]aw�	%E�$:D�΅E�$����E�$� ��a	%E�$����E�$3� �/$��E�$�	%��>�E�$20�
E�$��V��	%E�$���G6F#E�$��O`�E�$�8��Z� E�$@�F#NSE�$�O�S9�E�$)
%P�8E�$P
%��	E�$�/$Κ\E�$��²��E�$l9HP��E�$t
%܁/HE�$���
�GE�$�6�dHE�$���ӱ
E�$�PbT�E�$�Hr��xE�$I�
Ƭ�E�$&��	E�$TB��� E�$w�X��J��� B+�9J�M��?�!J�VM��;
J�����o�
J������9J�K�4�ĂJ�2Y"�-#��J���;yJ���+!�J��y̲
�
%J��
,T��!J���
���J�$���PJ��-��J�����֢!J���W>� J����Mx"J�>�U�PJ�>�y��x"J�`+�
�J�7�!"�/�� J�h�Hu�!J�P:��J��J�Qh��#�J�T.NhI�J���#��
�HJ�jR~RyIJ��
%�x"PTJ��D��R�HJ�o�Q�ܥ!�yJ��{:bJ��%k�J�<�x"J�~���IJ��!���J��!�WNy"J�)<�$�J��TB���J��M,�RJ��a�I0�!�J����T�J�T�g.J��p�M��J������y"��T�!�:!
0$���T�!�:�F# b�I�9ST�!�:^�J5b�9ST�!�:��41z"�3��.8���
��h/%��x!T�
�� �����
b��-����w�U��Bz7<K� ͺ��Z��ͺ�z"}�
.Jͺ��
IS�ͺe����ͺ����ͺ�bw���ͺ��_�!mSͺV��ͺ��S��!ͺF��Sσͺ�
��ͺ[<RJ��ͺ9U���
ͺX�XzPUͺ���k�ͺ��p���!ͺk�z"�ͺ�%�
p<ͺ�~�Φ!ͺ����ͺ���S�bͺ���SȇͺT���ͺ���I�z"ͺ"0$�%�ͺ�|zśͺ'���
%ͺ6���ͺ���ͺ#�ͺ��40$5ͺH��zϳ
ͺ�%ͺ���!]� ͺ����
Uͺ���
�zͺ�b�AͺvJ �}� ͺ�W0$�ͺ*��� �ͺ!���ͺD�"�5Tͺ�z"mUͺ�!u1ͺۙ�� Bͺ�B��ͺ��<$ͺ�I�<�zͺ�� �U�UͺI�&�;�ͺ�
�� �ͺ�]�*�ͺ"�!Մe�ͺ�b3%��ͺ�9"��{0$ͺ�Jl{"ͺ3�!��7�ͺ=��<WTͺA�0�
R_ͺH�� �Jͺ��gԴͺ��T�
�ͺ�0${�0$ͺV��JI�!ͺ�4{t�ͺ���1{"ͺ�I��X�ͺU�V%w�
ͺ�T��bͺ�F#^�!�ͺy%�0$m�ͺ��I��ͺ-Cgͺ$�4� W{ͺjT�J�Jͺ���o{ͺ��0$E�ͺ�bK�ͺK��
n�ͺn��!��!ͺG#��
	=ͺp
q�R{"ͺ�_�_k�ͺG���cͺŧ!�%J�ͺm{"r���ͺ�0$�{�UͺJ)=̈́ͺ3�Bc�%ͺ��X� ͺ�7���ͺV�!ִ
ͺ�$1$ͺ5V��T=ͺP1$/Kw�ͺ��;)�ͺ G#�{"��ͺ���ͺ�gcVͺ:J�{�ͺ}1$��c�ͺ��a��%ͺ
[�:�ͺ�{��VͺKG#�e�ͺ�����ͺn��=ͺ��Ŝ�Vͺ���{2�ͺ��
7�
ͺ ����ͺ|y�mG#ͺ�=RJͺ�8|[Kͺ���1$�ͺ���{"�ͺ���n|ͺ��ͺ���
�|ͺ�K2
%�!ͺ�?[�ͺ>uJͺ�G#�|��ͺ�F�!uͺ0}���Kͺ�T��|�!ͺ9��1$�ͺ����!\}ͺ��%W#�ͺ���G#�G#ͺ�T2$��ͺ
��ZWͺį�ͺ2��W�ͺ�Tk�ˈͺ72$�&ͺT�� �ͺ�J
U�Jͺb�	��� ͺ�6ĵ
ͺOQ���
ͺ[�
�?�ͺ�� ���ͺ�:>6ͺl?�
�ͺ
1H#�ͺ�����ͺ����ͺ��G�Wͺ�}d2$��ͺ���{"o>ͺCU?�gH#ͺ��Kͺg
%U��Wͺu���3�ͺ)X���Kͺ�!�� 5Kͺ|A���ͺ� ۶@� ͺmKu�
 �ͺ�!\Lͺ�=L��ͺ�H#�NXͺ�}���ͺ2|"9�!�ͺ�}hU��ͺ`|"ى,
ͺ���� ͺ���J
ͺ�e�Uͺ�>�E�ͺ�
%n�!ͺeXs�ͺ*��2$�>ͺ��X��
ͺy���!�ͺP�1�� ͺ�� $��Xͺ�2$Y�|"ͺo
T��|"ͺO�~mcͺ�K�U�ͺ��sL�>ͺPYL�2$ͺ�����ͺ�H#�fͺ�
3� �ͺw��G�ͺ�E~i� ͺ��
��
ͺ���L��ͺ��
�ͺ����H#ͺ�K3��cͺV}"ͺ�L�c�ͺY�Y	�ͺE�?�ͺy�����ͺ�83$o�ͺz~�
%Jͺ�K�Y6�ͺm3$�� �ͺ�Y���ͺة!��L?ͺdLd&Lͺ�3$ͺ��8��ͺ�?��� ͺ�Z��ͺ�3$�3$8}"ͺ�d��/� ͺ��G�~� ͺmܷͺR �|�ͺnL�?	Mͺ~�?v}"ͺ�~(�
��ͺ�� &4$�ͺ1VH�LͺNZ�}"�!ͺ�d��ͺ;�!+�}ͺ6��|Zͺ�L'�҆ͺ�cVI#ͺ��?�ͺa��}"ͺ�R��ͺFM+�ͺ~"�V�ͺ^�
��ͺ.��Mͺ�Z��
O�ͺo�jI#�
%ͺ�V�� U�ͺŷ
W��ͺ��wM&ͺ6%\4$�I#ͺ�~&�$ͺ1���Yͺ5Wak�ͺ�M
Cͺ @�d\�ͺ\��Zf�!ͺH~"FMN@ͺ���eͺ�ͱ-� ͺ�X��ͺ�I#y� ��ͺl~"cW��ͺ���� ��
ͺ�^%i�ͺs���~"ͺ��M0�ͺcZ�i@ͺ�W�� 	�ͺzN��ͺn���ENͺ��J#��ͺ��G�ͺ|M2�
� ͺ����4$ͺ�%g�
ʈͺb��j� ͺ���ͺҊ�ͺ���ͺE9���
ͺTe�@=ͺ���ظ
ͺ�@�%�ͺ]��!�%ͺ��W5�ͺ�� 5J#�� ͺ�B�ͺ}e��ͺ>%���4$ͺp�e�Mͺ����Nͺ���o�ͺ��
�ͺ�W�Ѫ!ͺ2�jJ#�ͺ2��5$ͺAd�!ͺ�^�!7Aͺ=�O�� ͺ�~"�N��!ͺl%�N�~"ͺ��3�c�ͺ=� ��hAͺj� ��ͺ,O�$�ͺ�� �eN�
ͺ�X�J#ͺi��*[ͺQOW5$@"ͺO��ͺ�5$���J#ͺ8�Tͺ��
��˫!ͺ�5$�Aw[ͺ���ιͺ��[ͺf��ͺ�M���ͺ�!�
Nͺ���A�ͺ�%�	�ͺ�� �[]ͺB�s�ͺT����"ͺ���5$ͺ���Oͺ��%ͺ�#�%ͺ�6Bͺ�9f<�ͺ5�QX�Oa�
Ћ�O�Oa�
�#6$\a�
{�G�I�a�
���� h�a�
T��� a�
�i�Pa�
b\YN�a�
�'P� a�
-%�R%a�
�t�[R�a�
>���
��a�
��"�"a�
"��a�
E6$C��a�
3K#f�޹
a�
W6$�"z�a�
~N:�a�
LP6#a�
��MK#��a�
��	�
Ia�
O�Sc�a�
�N���a�
2,�
��a�
fPqK#��a�
�����a�
D�!sX�K#a�
`�YA�"a�
��	�s��Y������YbB�g�"�Y]�!-�|E�$p%�K#سE�$�N}PE�$��Ny�E�$[f���!E�$��"�P1E�$��
�F�E�$C�ǀ�E�$�K#L#u�E�$��"�fk�E�$v�XE�$�fΈ�\E�$��s6$
 E�$P�
���PE�$p���\E�$RL#v�
��
E�$ߺ���E�$�f�0� E�$ܺ
�OE�$;���E�$��\׀"E�$3���E�$W� M�E�$�X#���!E�$]���E�$G mE�$Y��X6�E�$@����E�$��VۊE�$�]�E�$� ��E�$���
c�E�$I]:O��E�$8 e ƁE�$wL#5�
�"E�$F�"�e�E�$�U�xBE�$r]}��!E�$�L#��
E�$��6$�!E�$��8�@E�$�X1E�$;��r�"E�$� 9�_E�$��gE�$O�r�� E�$pO��L#E�$�k�
�]E�$��
s��E�$�1gE�$���ǻ
E�$Q�L#��E�$a�%��E�$|��� ,YE�$�O�]��E�$��b�E�$��"0QdYE�$��R���E�$����E�$����E�$o���fQE�$	�Y�OE�$n��Y�E�$,	ĻE�$�%����E�$��ZJ�!E�$=��6$�QE�$� ���%��/�؉%�ҁ"�O%��B�%c�P��!%� ���B%҂�� �%J��"�%S�n�ʋ%Ϡ��&%�Q��7$%K� �%�Q6	�
%�����%� �O%9�6Z�%�B��
%���!��%ڌex!M'#J�aX	�J�4M#>�
��J����J�<P���"J��>M#!J��	*��%J�57$gg��!J�M�bM#\J��x�R�O"J�m�	�gJ�YZJP�QJ��	��BJ��!E7$J��g��	J��k�J�#!;!<�J��M#�j7$J�N�� ~J��mPa�J�w�7�^!J�o*�;�J���S�J���g��J�e�	^�7$J��%:%J�=�C+^J�w��P�7$J���g��J�[�g%9� J���!+^J�s�)�!�J����J�2z�M#�%J�>�M�kZJ����ZJ�!J�`M�S^J�"��
$�!J�����ZJ�n�y��J���	�^J���
�[
J��7$�$s�J��P�%�PJ�AR�J��P�g��J��%�(�J���M��ZJ�h��Z��J�
h�%J�%���7$J�3%���7$J��+J��^@R0J��cR�PJ�:CS�J��H�|�J�U�QJ�P%@�"��J���#�!��J�]����J��Rb�
�J���
@
��J�.��RhJ��!]� d
J�xM�M#|
J�i%��
��J�#[1Q��J�UQ��eJ���!�lJ�Z���:hJ�kQ�^PhJ��b�"�
J��� ��!G�J���"ݍw�J������^J������ J��ڮ!�M#J�vh��RJ��!���J�GC��RJ�a�N#J�شS��J��
��!��J�"�� /"J�eC{���"J�΂"��
ֽJ�� l��J���>SJ���8$J�@��1�!J����CVSJ��C8[Z�!J�	_��J�+s���J��c[:� J�ފ+8$D�J��NP8$�CJ�<cDV"J�\�{�J���y�!{�J���$��J�s8$"J�����
J���C�J��B"_� )�J��:��c�J�q�
�� J��C�"G�J��8$~%j���� � ��u�M�	D��l�����%�
�����!�� nS��%�Sq��#�"_,��O�{�D�-��"�[��8$�6��F_�����
	��Bz7<K� ͺ��Z��ͺ�z"}�
.Jͺ��
IS�ͺe����ͺ����ͺ�bw���ͺ��_�!mSͺV��ͺ��S��!ͺ�
��ͺF��Sσͺ[<RJ��ͺ9U���
ͺX�XzPUͺ���k�ͺ��p���!ͺk�z"�ͺ�%�
p<ͺ�~�Φ!ͺ����ͺ���S�bͺ���SȇͺT���ͺ���I�z"ͺ"0$�%�ͺ�|zśͺ6���ͺ'���
%ͺ���ͺ#�ͺ��40$5ͺH��zϳ
ͺ���!]� ͺ�%ͺ����
Uͺ���
�zͺ�b�Aͺ�W0$�ͺvJ �}� ͺ*��� �ͺ!���ͺD�"�5Tͺ�z"mUͺ�!u1ͺۙ�� Bͺ�B��ͺ��<$ͺ�I�<�zͺI�&�;�ͺ�
�� �ͺ�� �U�Uͺ�]�*�ͺ"�!Մe�ͺ�b3%��ͺ�9"��{0$ͺ3�!��7�ͺ=��<WTͺ�Jl{"ͺA�0�
R_ͺ��gԴͺ��T�
�ͺH�� �Jͺ�0${�0$ͺV��JI�!ͺ�4{t�ͺ���1{"ͺ�I��X�ͺU�V%w�
ͺ�T��bͺ�F#^�!�ͺy%�0$m�ͺ��I��ͺ-CgͺjT�J�Jͺ$�4� W{ͺ���o{ͺ��0$E�ͺ�bK�ͺn��!��!ͺK��
n�ͺG#��
	=ͺp
q�R{"ͺ�_�_k�ͺ�gcVͺ:J�{�ͺ}1$��c�ͺ��a��%ͺ
[�:�ͺ�{��VͺKG#�e�ͺ�����ͺ|y�mG#ͺ�=RJͺ�#�%ͺ�6Bͺ�9f<�ͺG���cͺŧ!�%J�ͺm{"r���ͺ�0$�{�UͺJ)=̈́ͺ3�Bc�%ͺ��X� ͺ�7���ͺV�!ִ
ͺ�$1$ͺ5V��T=ͺP1$/Kw�ͺ��;)�ͺ G#�{"��ͺ���ͺ�8|[Kͺ���1$�ͺ���{"�ͺ���n|ͺ��ͺ���
�|ͺ�K2
%�!ͺ�?[�ͺ>uJͺ�G#�|��ͺ�F�!uͺ0}���Kͺ�T��|�!ͺ9��1$�ͺ����!\}ͺ��%W#�ͺ���G#�G#ͺ�T2$��ͺ
��ZWͺį�ͺ2��W�ͺ�Tk�ˈͺ72$�&ͺT�� �ͺ�J
U�Jͺb�	��� ͺ�6ĵ
ͺOQ���
ͺ[�
�?�ͺ�� ���ͺ�:>6ͺl?�
�ͺ
1H#�ͺ�����ͺ����ͺ��G�Wͺ�}d2$��ͺ���{"o>ͺCU?�gH#ͺ��Kͺg
%U��Wͺu���3�ͺ)X���Kͺ�!�� 5Kͺ|A���ͺ� ۶@� ͺmKu�
 �ͺ�!\Lͺ�=L��ͺ�H#�NXͺ�}���ͺ2|"9�!�ͺ�}hU��ͺ`|"ى,
ͺ���� ͺ���J
ͺ�e�Uͺ�>�E�ͺ�
%n�!ͺeXs�ͺ*��2$�>ͺ��X��
ͺy���!�ͺP�1�� ͺ�� $��Xͺ�2$Y�|"ͺo
T��|"ͺO�~mcͺ�K�U�ͺ��sL�>ͺPYL�2$ͺ�����ͺ�H#�fͺ�
3� �ͺw��G�ͺ�E~i� ͺ��
��
ͺ���L��ͺ��
�ͺ����H#ͺ�K3��cͺV}"�ͺ�L�c�ͺY�Y	�ͺE�?�ͺy�����ͺ�83$o�ͺz~�
%Jͺ�K�Y6�ͺm3$�� �ͺ�Y���ͺة!��L?ͺdLd&Lͺ�3$ͺ��8��ͺ�?��� ͺ�Z��ͺ�3$�3$8}"ͺ�d��/� ͺ��G�~� ͺmܷͺR �|�ͺnL�?	Mͺ~�?v}"ͺ�~(�
��ͺ�� &4$�ͺ1VH�LͺNZ�}"�!ͺ�d��ͺ;�!+�}ͺ6��|Zͺ�L'�҆ͺ�cVI#ͺ��?�ͺa��}"ͺ�R��ͺFM+�ͺ~"�V�ͺ^�
��ͺ.��Mͺ�Z��
O�ͺo�jI#�
%ͺ�V�� U�ͺŷ
W��ͺ��wM&ͺ6%\4$�I#ͺ�~&�$ͺ1���Yͺ5Wak�ͺ�M
Cͺ @�d\�ͺ\��Zf�!ͺH~"FMN@ͺ���eͺ�ͱ-� ͺ�X��ͺ�I#y� ��ͺl~"cW��ͺ���� ��
ͺ�^%i�ͺs���~"ͺ��M0�ͺcZ�i@ͺ�W�� 	�ͺzN��ͺn���ENͺ��J#��ͺ��G�ͺ|M2�
� ͺ����4$ͺ�%g�
ʈͺb��j� ͺ���ͺҊ�ͺ���ͺE9���
ͺTe�@=ͺ���ظ
ͺ�@�%�ͺ]��!�%ͺ��W5�ͺ�� 5J#�� ͺ�B�ͺ}e��ͺ>%���4$ͺp�e�Mͺ����Nͺ���o�ͺ��
�ͺ�W�Ѫ!ͺ2�jJ#�ͺ2��5$ͺAd�!ͺ�^�!7Aͺ=�O�� ͺ�~"�N��!ͺl%�N�~"ͺ��3�c�ͺ=� ��hAͺj� ��ͺ,O�$�ͺ�� �eN�
ͺ�X�J#ͺi��*[ͺQOW5$@"ͺO��ͺ�5$���J#ͺ8�Tͺ��
��˫!ͺ�5$�Aw[ͺ���ιͺ��[ͺf��ͺ�M���ͺ�!�
Nͺ���A�ͺ�%�	�ͺ�� �[]ͺB�s�ͺT����"ͺ���5$ͺ���Oͺ��%ͺ5�QX�Oa�
Ћ�O�Oa�
�#6$\a�
{�G�I�a�
T��� a�
�i�Pa�
b\YN�a�
�'P� a�
-%�R%a�
���� h�a�
�t�[R�a�
>���
��a�
��"�"a�
"��a�
E6$C��a�
3K#f�޹
a�
W6$�"z�a�
~N:�a�
LP6#a�
��MK#��a�
��	�
Ia�
O�Sc�a�
�N���a�
2,�
��a�
fPqK#��a�
�����a�
D�!sX�K#a�
`�YA�"a�
��	�s��Y������YbB�g�"�Y]�!-�|E�$p%�K#سE�$�N}PE�$��Ny�E�$[f���!E�$��"�P1E�$��
�F�E�$C�ǀ�E�$�K#L#u�E�$��"�fk�E�$v�XE�$�fΈ�\E�$��s6$
 E�$P�
���PE�$p���\E�$RL#v�
��
E�$ߺ���E�$�f�0� E�$ܺ
�OE�$;���E�$��\׀"E�$3���E�$W� M�E�$�X#���!E�$]���E�$G mE�$Y��X6�E�$@����E�$��VۊE�$�]�E�$� ��E�$���
c�E�$I]:O��E�$8 e ƁE�$wL#5�
�"E�$F�"�e�E�$�U�xBE�$r]}��!E�$�L#��
E�$��6$�!E�$��8�@E�$�X1E�$;��r�"E�$� 9�_E�$��gE�$O�r�� E�$pO��L#E�$�k�
�]E�$��
s��E�$�1gE�$���ǻ
E�$Q�L#��E�$a�%��E�$|��� ,YE�$�O�]��E�$��b�E�$��"0QdYE�$��R���E�$����E�$����E�$o���fQE�$	�Y�OE�$n��Y�E�$,	ĻE�$�%����E�$��ZJ�!E�$=��6$�QE�$�	*��%%57$gg��!%M�bM#\%� ���%��/�؉%�ҁ"�O%��B�%c�P��!%� ���B%҂�� �%J��"�%S�n�ʋ%Ϡ��&%�Q��7$%K� �%]����%�Rb�
�%��
@
��%.��Rh%�!]� d
%�Q6	�
%�����%� �O%9�6Z�%�B��
%���!��%ڌex!M'#J�aX	�J�4M#>�
��J����J�<P����!J��>M#!J��x�R�O"J�m�	�gJ�YZJP�QJ��	��BJ��!E7$J��g��	J��k�J�#!;!<�J��M#�j7$J�N�� ~J��mPa�J�w�7�^!J�o*�;�J���S�J���g��J�e�	^�7$J��%:%J�=�C+^J�w��P�7$J���g��J�[�g%9� J���!+^J�s�)�!�J����J�2z�M#�%J�>�M�kZJ����ZJ�!J�`M�S^J�"��
$�!J�����ZJ�n�y��J���	�^J���
�[
J��7$�$s�J��P�%�PJ�AR�J��%�(�J��P�g��J���M��ZJ�h��Z��J�
h�%J�%���7$J��+J��^@R0J��cR�PJ�:CS�J��H�|�J�U�QJ�P%@�"��J���#�!��J�xM�M#|
J�i%��
��J�#[1Q��J�UQ��eJ���!�lJ�Z���:hJ�kQ�^PhJ��b�"�
J��� ��!G�J���"ݍw�J������^J������ J��ڮ!�M#J��C�"G�J�vh��RJ��!���J�GC��RJ�a�N#J�شS��J��
��!��J�"�� /"J�eC{���"J�΂"��
ֽJ�� l��J���>SJ���8$J�@��1�!J����CVSJ�	_��J��C8[Z�!J�+s���J��c[:� J�ފ+8$D�J��NP8$�CJ�<cDV"J�\�{�J���y�!{�J���$��J�s8$"J�����
J���C�J��B"_� )�J��:��c�J�q�
�� J�3%���7$J��8$~%j���� � ��u�M�	D��l�����%�
�����!�� nS��%�Sq��#�"_,���8$�6�O�{�D�-��"�[��F_�����
	���:)�ͺB���7د!ͺ������hͺ�S�z:9"8ͺ��wdf9"fHͺ��9�
6�ͺ;V��%���ͺ]�
�"�ͺ�h;��Qͺ�;�.N#ͺ��$��$�~
ͺ�%��^HͺF��#&���$ͺ��$���R�ͺR�Zh!����ͺ��9��$��ͺ�Q�~
�oͺ�� ��#��ͺ;�ȕ 0G�~
ͺǷ
�J��ͺj_�9"
seͺsen�m�ͺ�%�>�pͺ�hA�U�R_ͺR_��$�$��ͺ���� I��ͺ8iV�#�z�9"ͺ^�p
*�$:� ͺxi�_���jͺH��RVG�Rͺ�[�\%ͺ�_��ͺ�����"ͺ�_`Bͺq�
8���jͺ�_�V��ͺ��uN#�Kͺ���I�
ͺ�
�%@H��ͺ��f`�#]ͺ�8$��\�ͺ\��N#�7�ͺ(�@��[ͺ�%�8$�Ͱ!ͺ{\.��;"�ͺt�
[�K.�ͺ.��#[�6�ͺ6��}�#ݵͺ{�
G�r��ͺn���0�ͺ5�"t4p�$ͺ�"4q�$z�ͺ�`��
��O�ͺ�Vx�&4_`ͺ/)�~��Hͺ�8$�<"	9dͺ�����K�ͺ�� �d1ͺ�9$�{����ͺ{��#1h��ͺI�I4Z�d�ͺ�9$v!��ͺ*#8	����ͺu4�4]ͺ���<"�ͺ��pnͺď��"*�ͺHW*�ͺ�!;$*�ͺv���*�ͺQ��W*�ͺ�V*�ͺZ](�*�ͺ���*�ͺ�� �*�ͺ��*�ͺb��*�ͺ�N#*�ͺ\;$B*�ͺ�`„"*�ͺ}�R*�ͺ`��*�ͺ��� *�ͺ��Pk*�ͺ��
�*�ͺ�]��*�ͺ�WH*�ͺ�;$)S*�ͺ�k�*�ͺ-���*�ͺ�
�*�ͺW��k*�ͺ��"g�*�ͺ��*�ͺ���*�ͺ�`�N#*�ͺ�%ѥ*�ͺ6�" O#*�ͺ��*�ͺ/a�W*�ͺq�"H�*�ͺ����*�ͺ����
*�ͺ�,�*�ͺ�=�]*�ͺ��	
*�ͺ�u��*�ͺ9�!�*�ͺ��$eS�ͺ��6�uͺ�
�]�uͺ�T���uͺ��$��uͺ���uͺ���uͺ"� v��uͺ�%���uͺ�[O#�uͺ*Xo�!�uͺ.�
x
�uͺ:^Q� �uͺ��!���uͺ�� �k�uͺ)D0�uͺln^�uͺƆ�^�uͺ%r�uͺD�x��uͺ�����uͺ�$��uͺ�G��uͺdXC
�uͺ�^�S�uͺ֍ϋ�uͺ�\D�uͺ�$�O#�uͺ��"O�uͺg�څ"�uͺ�� ���uͺ��,�uͺ�
�D�uͺ
�4��uͺ�"���uͺ�;$n��uͺ�=���uͺ�(v�uͺv�g�uͺv�_�uͺ�����uͺ���O#�uͺ<$W�"�uͺ;#��ͺq��+ͺ���$+ͺ"+ͺGR��+ͺ�$Z_+ͺ���D+ͺ��9%+ͺ

�+ͺ�!��"+ͺ! �!+ͺ�F�+ͺ6�+ͺ7<$�+ͺpaO�!+ͺ�H+ͺY�� +ͺ��!��+ͺ���+ͺ�_s%+ͺ�X�+ͺ
P#p�+ͺ3l�+ͺ޿�+ͺ
E+ͺE�b�
+ͺW �+ͺ%���+ͺ� 7�+ͺB��%+ͺ��%+ͺ@��+ͺq�}+ͺx���+ͺ��
�X+ͺ�S͆"+ͺ%&Y�+ͺ��+ͺ�
�+ͺ
�+ͺu�ڿ
+ͺ��"+ͺ���+ͺj� KtK%ͺ�� �
7�ͺdv��7�ͺ��[�7�ͺ�$3� 7�ͺ�Im<$7�ͺ��
7�ͺB�"�a7�ͺ���7�ͺ� �a7�ͺ��J�7�ͺ� ��!7�ͺY&�&7�ͺ�&��7�ͺ���7�ͺ��&7�ͺǸ�7�ͺn� ��7�ͺ{%�
7�ͺ���7�ͺ�W�7�ͺ%�M�7�ͺ��!$7�ͺ�� T7�ͺ� �7�ͺY�[�7�ͺGT�� 7�ͺbl�_7�ͺrTb7�ͺ`�_7�ͺ����7�ͺ��T7�ͺ�l	Y7�ͺÐ��7�ͺ;`)
7�ͺ���7�ͺV7�ͺ$�
��7�ͺ�N
7�ͺJ$�!7�ͺ�
 7�ͺ�� P�
7�ͺ��DY7�ͺ�#	�/�ͺ����Tͺ��%���ͺ܌U�ͺ�D )�ͺ����/�a�
h ��a�
%U�{�c[a�
V��4�!�la�
�<$-���a�
��
��~Ya�
�>$�{G<�a�
N Q�Mb�<"�Y~ S�$���%�YӈO#��.��Yb�� I�EW�Y�k�
Zh�"�Y�Wpb����YOZ��
��"�Yu mm���Y� �":P#�Y`�_�!U���Y��@$��Y�
>�� :�"�Y@�P#�b�Y�� �Y<���s`@�Y�P# ��Y����Y��
�@$+'Bc�Y\�4���Y��aX��Y�%6 й�Yi�`'`_��Y�6�
"j!�c�Yw  �u�"� �Y���A$�Y!N�
� �Y�L�G���Yk[�C5��$�Y��f�h;x��Y��ؠ�G��
�Y��������
�Yd
�!Fj!���Y"
P�#��d�Y{'w����� �Y�`6��="���Y������="�Y�a��MQ�$�YK ���e/l�Y/l�M�:� �Y�% f"��Y�!>"��n"�YuA$��j!���Y��=���$�Y]��*��Y
��p#���Y��!y��ebE�$=d�L�ZE�$�� ؁
��E�$XX�
A�"E�$5�@�VE�$�X�J�E�$�6� �k E�$M
���E�$^���`"E�$d��H����E�$�m�L�$�E�$!q�*�M�E�$�P#=���
�E�$.�M�[dE�$���l!K�E�$��A�F�"E�$	W�<f͍E�$�� ��h
-!E�$n� �6�#�mE�$�A$�!�n� E�$� ε!���!E�$�b"I� E�$�Q#����E�$H"�B$*�E�$� �*�E�$�b��*�E�$�n!*�E�$�Xۓ*�E�$�!�*�E�$'[��*�E�$W�'�*�E�$�C�d*�E�$q�
�n*�E�$�[*�E�$C"q�*�E�$�A�!*�E�${ZL*�E�$5#U!*�E�$'�l
*�E�$L�(*�E�$�$�[*�E�$�c*�E�$R#k�*�E�$�"s"*�E�$�ZOL*�E�$\P��uE�$@R#-��uE�$ C$�uE�$}�!s��uE�$��P(�uE�$�
�uE�$@yBc�uE�$�$ӏ�uE�$p�!�"�uE�$�XxR#�uE�$ĩ �� �uE�$���%�uE�$$���uE�$DH"_�$�uE�$����uE�$�H"�R#�uE�$��$GE�uE�$�!#��uE�$����
�uE�$}c|��uE�$R�uE�$W�i�uE�$�H"�!+E�$�� [+E�$i�Y+E�$o3e+E�$>���+E�$��Z+E�$_�
< +E�$G\�
+E�$�H"MC$+E�$[�")[+E�$e�b�+E�$��"ݽ+E�$
��"+E�$;"I�+E�$u"w +E�$�P+E�$�+E�$���+E�$��"� +E�$�cQ�+E�$:�%+E�$��!e[+E�$>��"7�E�$�
u7�E�$�(Վ7�E�$��c7�E�$٪�(7�E�$��"l�7�E�$����7�E�$Q���
7�E�$C��(7�E�$'#��7�E�$��$�"7�E�$1)�"7�E�$!I"� 7�E�${��$7�E�$5�WY7�E�$k��!7�E�$J +d7�E�$Q�$97�E�$ $� 7�E�$So�Y7�E�$���7�E�$�!#v\7�E�$$?"�#��
E�$T�) �<hfE�$hf�EfE�$�R#�<��$E�$��$�[i��E�$ދ�J�#E�$�#�<!�$j�E�$��8���pE�$�[����?�E�$����
E�$�C$��D#�E�$}��� -�#kNE�$9�
h#��ނE�$ނƔ,nU#E�$E�$o��9E�$Ʋ�� "$t�E�$������E�$�Pn,��SE�$���N����E�$E���q$E�$*z���%.�@��#N�$%_�")�!�
J�w�\��J�ed*O��d�J�t�"o�#�YJ�'�D ��t�J�
MM%J�\����EJ��!��i$��J���?"�7X%J��o��$��t�J�
����J�;"Ś @L��J�� "��$�\J��2z�#�EJ����S��]J�cL_\�\J��gX)�J�w�	�"J���_\�\J�>�y�@"J�U��LX)�J��
	�"J��d$��J�>�`+�W�J�����$J@"�OJ�"�/�#J���r� h%�OJ�b���J�-#���J�J��%��J�J��gIJ�3P�L��J�)�
1�"
�J�8���aMJ��o��
��J�F�Zܔ:$J���S#^
�$J����"�
	#J���%S#�#��J��#X�\J����
����$J��o�#` oeJ�����J���J��h��#J�_�"�d��J������J���
3%�C$z)J�U�!%��m�J�V�$��J����A3Zm#J�{M��kJ�k��
��
2$J��C$��
�{c"J��$%�s�*J��#E9}l!J���
q9kpJ�\ZXp���AJ��dQPG�
b� J���#����J��
�ZC�AFJ�#�=�I�J�^]xMB�l!J��e��#�
4BJ�����!eJ��!_� A��J�˓{�)�J���@�]-/�J� E$���@"��J��FIIN��J��q�M A"Y�$J�`#���\�J��
��mNf�J��$�f�
�MJ�*	C?&;kJ�7�
��k��J����
��+�
J���� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&� ��J�ҥ�$%CJ�%C�����J�����N��J�����&�pJ��pX"�8�$J��e�9�l!DB"J��#�Sf�J����N�P�NJ��"F�14OJ���$�y�J�k�"�*�Z��J�h%��Z�]J�a��B"&�#P�J�� �:Yem!J���qz� �
J����f�x
�C���{
f#��\tot}tf\�$�Ep���Cf���V_ot��f��r��u��d!fZ���ot�e!f�t	�#��<f!f��|

7ot��f�e�%�7��f� 5'�rfi�LG��rf�g[�� � fJ �r|� fBS#��
]Glyf�*�%<�
lyfޏ��
f��#fwG�r�E$�#f![�B��f\H@���fxI)f�e#f�y��y�e#f��"��yFfE���C�#f7w���$laf-}
��p�f)�$�8�
k� f�D�f!I7"�\f��7d}
�\f�LO��/fǵP���D��fE�c#�f!��f 700X�#fdw��k#X�#f�� ��u#�#fg!�x5FJFR7"fw�Z�$"DZ�Z�fU���|�#�afV�\�g!�af;�]�a5� f��E�
�7Vf� j}
ٶ�g!f�I��$� �f���� � ��f�7t}
;D�#fY�)�Dlaf)#R�
U��f0
��D}}
f��#���fV��Djrfb��ݶf�7"��y�|�f�F�bwl��g!fW�
�8"�l��Ef/�@�Jf  �%��W#fJ�	�/02f�yt|?f#2f�%�%��L$fܸ!�%?f#��"fQ Kf#g�ɭf��ۚg�XJf�%��Sf#��f.�"��Sf#��f�Ȏ"���� ]fN�7�#�]f���
���yf�y��$��`Gf��
�]��Sff��s�?0f�JL�H�[\f�\�� H��}
fo
��$�#cf�טLF+zf�*H��f��1�
H��]f�0 H���f�f�H�+�f�%�H��$f�%;^<�
u��$f��3�J�
�7fdF\�
���� f�F;�#��Gf(�:�JF�fjc~
S���fA��Eٶ,fl�
[�8��f�c��8n�f�
�cN�*�
f�#q� ����f�#w���6fS'n�8"��
fV�
�z.9"f����S��f-~
�ES���f��
7�S��FfR^�S�[~
f^�0S��fg��GS�|~
f���#S����N�N�x
}{����
NO��x
/����$G���x
���������x
v!��"uz����x
s���m��v!�\�x
dO���O�O�{�x
]���6�����x
�������]��x
����"� �YǏ
�x
?u��m�#�
[$�x
wE��e9��$�x
�v!����^8K"�x
P���O�_���x
����v�Y!�'�x
sQB� ��	$�x
"�
Q���OdL"�x
�uQi����$�x
��Q�$��*^�x
[� Q���	$2�$�x
G|Qҿ0��/�x
	�##�����U���u���
�|��� �E$Q�`f#��f#�l�
���fv�#��#u��n��r�������NPo!+%����$���"�g�;[�w�1�+��lm&���"���]�&��E$4D������$���#{(�-g#�� S�(�!� �;a���
�-sP�
'���e m|����
Ko!�������#�����
�;+s֏"���u�����-�`R�(�M�����
�J��sUDvDX��;Q������� R�T#	^�f%���D���g#��#|�ES�H�y!e]��ͺ8%Ҵ��E�ͺή��!�ͺ�"p
�2��
ͺN
�_z@�"ͺL���#�
��ͺ���;"�!i2ͺa�"�!p�
ͺ�^#[�y%F$ͺ���y�H�ͺ�/�~�tͺ�'�� o
o�ͺW�
���ͺ�� q	
hͺ�tPh%ͺ5F$F)�Jͺ�� �t�'ͺ��zK��ͺx��� ͺ�$�
�!ͺ�S�!7�ͺwP"3`��ͺC�!*�9h#ͺ>�
t2ͺ����hͺ�b�2�)ͺ	�E� �ͺ��\ZQ�ͺ�$�
�ͺQcl�hͺ��Q�jF$ͺ�ն�F$ͺ�n(�� ͺ��� �ͺ�p��F�ͺ�e'�)ͺe� ����ͺO��)_ͺ�$w�/ͺ��$��\ͺ�
S%��ͺX��|!�� ͺ�����ͺ� ��
�ͺ	���s�!ͺx~e���ͺt�uL�h#ͺ��=R"��ͺ� �LS ͺ�$��ǟͺ��(��|ͺ%�-]q_ͺ�4��9�ͺ�g�|��ͺ�~����ͺ��X�)�ͺ��8�i]ͺvi�%V�ͺ�(� � ͺ��"�2d�ͺ���cͺe���]ͺTh���_ͺGG$���Kͺ���(N�ͺ��5ͺZc�h�ͺ����l!ͺ� ���'ͺ��%��ͺL:*<5uͺI^^���yG$ͺ������ͺs� �^�WM�ͺ{�i#�$~�ͺ�!H*B��'ͺ�^ULM,_ͺ:_���di#a�
�G$�-#.#
�Y(3��
�$� �Y��P�07G��YA
�;� �Y �).,_�Y*i#"�
+��!�Y��dE�_��Y�L.W�fL�Y� �ɘ
E�$M���Ͳ %K-�jHSL%�W"w�X��%J��i����$`�J��'�� B+�9J��"M��?�!J�_
VM��;
J�}i#����%J� ����S%J���K�4�ĂJ���2Y"�-#��J�l��� J�H$��+!�J�V`��
���J��	
$��#)J��b�-��dJ��!��W>� J��k#���M�J�M�!T�+�
�J�Z� (J/j�J��$>�U�S�J��_>�y��x"J�^Mh�HMuJ��%��J��J�X*h��#��J�?�T.NhS4J�
"%��#��
��J�m4jR8}�J�ő"�
%�x"Q� J�Q� D�l�\}J��Q�ܥ!sJ$J��{:��J����I0�!�-J����T�J��T�g.�J�/*p�M��QJ�����4Ǡ��3��.��"���
��hb�����x!T�
���- ����
P� �e�!���}��J$w�U,'%�}� ��	F�yͺ'�!�+=�!ͺZ��ۨͺ�:)�ͺ���|}",ͺ��Kd��ͺ����Y�ͺ�z:9"8ͺwdf9"fHͺ�V
��>ͺc0�ͺ0�!vx^�
ͺ�y�$<ͺ#$��\�$ͺ���6�ͺoD��z�Gͺ��a#�ͺ�0��~
ͺۙ^���ͺ�B��H^ͺ�� &�~�ͺ�cdr�ͺ�x��7#ͺ��$n��~
ͺ����HͺZh!I��9"ͺ9����ͺ�~
��
oͺ��#��ͺȕ ��~
ͺn�H�M�ͺ�>����ͺ������ͺ����$��ͺ��8#��ͺ�'$w<�iͺ�r��l�ͺ%FF�!��
ͺI���	�ͺy�����
ͺ?���ïͺ8
��#ͺF�#�e�$ͺV� �HͺV�#�z�9"ͺ�b_P�ͺP
D�'�ͺp
*�$>Fͺ�_��½ ͺ+r�Id'%ͺ5�x@H�K$ͺ��"��#k�
ͺsF��^nͺu)/l.ͺ�
ƳZ2eͺ��ͺ��
�K�Qͺ��}�#�`ͺR��#VGhaͺ���;"�tͺQ�U�$�R�ͺ�`ˤͺ�eA`��(�!ͺ,�Y����ͺ��K�.ͺ��0�ͺ4q�$z�ͺ��
��O�ͺx�&4_`ͺ)�~��ͺ�<"	9dͺ���K�ͺ�{����ͺ�#1h]ͺv!��ͺ�,����ͺu4�4�!ͺ
�"7/�6ͺ��dI�ͺ&B�"}fͺ{����ͺ�6
e0sͺ
�$)�ɛͺ�##�$�Fͺ^�Rr�ͺ��C�1,ͺ��)c7ͺR���
ͺ�a�}�Rͺ�� ~ͺ6���.SͺKh�v/�ͺFh"�A~h"ͺ)�$~��vͺ���
ͺJ�}��K$ͺF&U� ��!ͺ�'%�7�� ͺi���,ͺs/���m#ͺ����/ͺ�8��� ͺ�����ͺ/0�h��ͺ�Y��fͺ�w�Z�ͺ���t��ͺ'� :�� ͺf"[� O�ͺ1���)ͺ+Z0}�"ͺ�8#�� 4%ͺF0��ͺ��
�� ͺ6��
�ͺ�p`G*�ͺls����ͺ6�b�?�ͺ"V���ͺ��{#$�ͺ�n#��\�ͺ��$��aͺ�s7J�ͺ�Ee�aͺ����m�ͺ�V�Af"ͺ���w��ͺ;#�?"�ͺu�Fe�sͺ��t��ͺ�
b��ͺ	�$��9ͺ=b�~(%ͺ�n#�� �n#ͺj� Kt=�$ͺ�t�*xͺOX2o#	LͺB��#ͺ@>��ͺ���u��!ͺ��ĞPͺ��
	w&ͺ����Sͺ(�$˳lbͺ�{8� �;#ͺ�.��ͺ��{��ͺ���$�ͺ��ur0ͺ

�"`L$ͺd*C� �bͺNc�*;�ͺP

��ɸͺr� ���ͺTRT��ͺ#-[�!�ͺ�TA�"H�ͺ��(%no#ͺ17#:�a-ͺuuXt� ͺ��B�hͺ��$�$Fi"ͺcLL�v�ͺ�uv�*ͺ=',��0ͺE�#+�ͺ�g��:�ͺ�7�(%8ͺ��ͺh��~� ͺۣ
��$�Gͺ��$�8#�f"ͺ�#	��!ͺ�{G<�a�
�di!��$a�
vJh���#a�
r�j�U�a�
:#~��#a�
9#��1Ha�
~����Ha�
=7.�$��a�
@9#�5ya�
�t7��?a�
[�|�6�a�
��G�9a�
���H_7a�
s�	I��a�
�
��a�
Q�]h�<"�YS�$���%�YO#��.��Y�I}�����Y&�#$���YR��l"t��Y"�
-�
l��Y�dE��y�Y��L�G���Y[�C5��$�Yf�h;x��Yؠ�G��
�Y������
�Y�!Fj!���Y@jj!�H�Y!�
+�#�� �YP�#��d�Yw����� �Y6��="���Y�-����="�Y@�$�!��Y�3�
`�#�Y��MQ�$�Y���e/l�Y�M�:� �Y f"���Y��>"��n"�Y��j!���Y���,��Y��=���$�Y�*��Y��p#���Y�L�$�E�$:kq�*�1JE�$=���
�fE�$�M�K�E�$�l!KlJE�$�A��E�$W�<f��E�$��h
�#$E�$�6�#��E�$�f''��$E�$���MNE�$����
��E�$K'Ҿ '�E�$��s&�
E�$�^+TE�$�8�g+�E�$}�)%[1E�$������E�$$�j��E�$~

��!�E�$>)%:����E�$�n�+�L$q'E�$����tE�$g"4��E�$�9��1E�$����.E�$6��� �gE�$f�D<hE�$K��o#u�E�$����E�$��
�cpoE�$�/��E�$����!<E�$��)%ߓE�$��
�|hE�$�8�c�)%E�$-�v'�oE�$�K���%v#f=Q�%.��n��
%X���%_Zy]}%+A�W�%�����
J��	�<-SJ���NS3�J�_��KG$J�Lm�
�6J���
�=�J�)�K�ݘ J�� �?"��J�3� R7m�#J�wX�
S�J�k��L��J���O��J����B�J�ǒ�#�fJ�[�$���?"J�o�|�
Y� J�w�\�� J�*O��h�J�o�#ƙ J�MMP:J�����#J��>��LOJ���i$�J��?"�7�KJ���$���KJ���
�#�J��?"��J�����FJ���4k!��J�?M� Xk!J��o�7��#J���mJ�+�Y���
J��$��}�J�!L�� ��J�����J�Ś @LS�J�S�"��$=�J�2z�}1p#J�����!J�U�S�'pJ�cLx���!J��gU��$J�w�>%�oJ���x�:�J�>�y�@"J�l��LU��$J��$�
>%�oJ�`+�W�J�&?��$J@"�OJ�"�/�#J�r� h%�OJ�P�($J����#J��<��=J�i��LݷJ��g�LոJ��=qH�J���
����$J���J���J�4�h��#(�J�������J�<�ɨ
K�J�V�n��=J��A��m#J������MJ��g���J�q��<�<J�{M��Ȝ J���
��
�$J����
���J�	k�LzJ�,�{�J�%�s�?�J�#E9}l!J�q9kpJ�Xp���AJ�QPG�
b� J�o�y�
ƻJ�����#����J�����FȹJ�Җ!Bm�J�,	�l"��J�K���!���J�X��}�}J���xMB�l!J�ө�($L�J��$�GJ�>y{$�$J�]GUy��J��ش�PJ�S�f��MJ���#�
4BJ�{�$��.J��TyZJ�����$J�zy8�e� J����"�J�˓{�)�J�@�]-/�J��y���!J�H��l!V	J����@"$$J���YC�	J�IIN��J�$�� 9J��M�lJ��W�J�l"J����{�J���
ɝ�$J���?&�@J���
�k�zJ��kEJ��fG�ä
J��� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&����J��N�P�NJ��N<c�5#�qJ���$�y�J�T�M��J��p#�$��
m!J��B"&�#[�J��'�:��
m!J�q��
J��0��x
�C���{
0�#��\tot}t0�\�$�Ep���C0����V_ot��0�������u��d!0�Zj_�ot�e!0��t	�#��<f!0���|

7otF0�E���C�#0�7w���$p�0�)�$�8�
k� 0��D�f!I7"�\0���7d}
�\0��LO��/0�ǵ
���D��0�E�c#�f!��0� 700X�#0�+&��k#X�#0��r��u#Z�0�6���|�#V0��j}
ٶ�E0�/�@�W#0�%� 	�/0�70�dF\�
���� 0��F;�#��G0��e:�JF�0�jc~
S���0�A��Eٶ,0�l�
[�8��0��c��8n�0��
�cN�*�
0��#q� ����0��#w�����
0�V�
�z.9"0�����S��0�-~
�ES���0���
7�S��F0�R^�S�[~
0�^�0S��0�g��GS�|~
0����#S�e9��$�P�v!���:6�����P�����:m��v!�\�PdO���:m�#�
[$�PwE���:	!$��=�P����:�Df�(!$�P�!$���:�G�
,�
�P��
���:�6ԏ!���PW���:��^8K"�PP����:�O�O�{�P]����:uoY�Y�P�����:���]��P�����:O�_���P�����:�c"��`��PN4���:"uz����Ps����:"� �YǏ
�P?u���:�$�Lo�P�6���:����P�!���:�����Pv!���:]�
�6#ڡ
�P�o���:�6#:7�$�P�����:��$G���P�����:��
NO��P/���:i����!$�P(�!���:���N�N�P}{���:�6#
�$���U)uQ�:�$��lu�U,�Q�:v�Y!+��U�7Q�:t���b9#�UE:#Q�:���	$���UG|Q�:B� ���u�U�WQ�:{��A�$�U2vQ�:i����$�U7�Q�:ҿ0��	�U	�Q�:5�
z�z�x
_v��^��!0
�x
Ռ�������
�x

����g���x
������:�8�x
�������
g�x
{g���
��:�x
P�
��$;��
��
�x
�8��&Xt/{�x
���;#%���x
B9��A�9��x
�$��Otft��x
���|t4A>�x
���K�v*�$�x
E���g�z�/��x
z�$����JA���x
�!��v�Y!+��x
�7QB� ���u�x
�WQi����$�x
7�Q{��A�$�x
2vQ�6#
�$���x
)uQ�$��lu�x
,�Q���	$���x
G|Qt���b9#�x
E:#Qҿ0��	�x
	�Q�9�� ��x
6g"�������U�u�]�+�����ll�b�qP��Zc��7�;[�w�+�Q���4Da��~d�!Zc�;o!m���Ko!������#�����;T#j��-���^��+S�AZc��� ���D����#|�ES�Bz7<K� ͺ��Z��ͺ�z"}�
.Jͺ��
IS�ͺe����ͺ����ͺ�%�
p<ͺ���SȇͺT���ͺ���I�z"ͺ"0$�%�ͺ�|zśͺ'���
%ͺ6���ͺ���ͺ#�ͺ��40$5ͺH��zϳ
ͺ�%ͺ���!]� ͺ����
Uͺ���
�zͺ�b�AͺvJ �}� ͺ�W0$�ͺ*��� �ͺ!���ͺD�"�5Tͺ�z"mUͺ�!u1ͺۙ�� Bͺ�B��ͺ��<$ͺ�I�<�zͺ�� �U�UͺI�&�;�ͺ�
�� �ͺ�]�*�ͺ"�!Մe�ͺ�b3%��ͺ�9"��{0$ͺ�Jl{"ͺ3�!��7�ͺ=��<WTͺA�0�
R_ͺH�� �Jͺ��gԴͺ��T�
�ͺ�0${�0$ͺV��JI�!ͺ�4{t�ͺ���1{"ͺ�I��X�ͺU�V%w�
ͺ�T��bͺ�F#^�!�ͺy%�0$m�ͺ��I��ͺ-Cgͺ$�4� W{ͺjT�J�Jͺ���o{ͺ��0$E�ͺ�bK�ͺK��
n�ͺn��!��!ͺG#��
	=ͺp
q�R{"ͺ�_�_k�ͺG���cͺŧ!�%J�ͺm{"r���ͺ�0$�{�UͺJ)=̈́ͺ3�Bc�%ͺ��X� ͺ�7���ͺV�!ִ
ͺ�$1$ͺ5V��T=ͺP1$/Kw�ͺ��;)�ͺ G#�{"��ͺ���ͺ�gcVͺ:J�{�ͺ}1$��c�ͺ��a��%ͺ
[�:�ͺ�{��VͺKG#�e�ͺ�����ͺx�~p�� ͺ�f�<�ͺ��1�cͺ3/R��8ͺI��+j�ͺ�k���ͺ|y�mG#ͺ4 ����ͺ���h�

ͺ)��p#Mvͺ
d97�ͺ�W�{�ͺZ �svͺ�� ��)ͺ\�{�Edͺ�=RJͺ�p�
:M$ͺ��� ͺ��L� ͺ���{"�pͺ���edͺ��hͺ���
2ͺ�K2
%� ͺ�F�!��ͺ0}���

ͺ|�"�v�vͺ�T�� �dͺ9��T��ͺ����!u�ͺ��%W+�!ͺ���G#��ͺ�T2$�ͺ
����ͺ)��'ͺT�� qͺk
�M$ͺ�J���ͺb���J�
ͺ���M$ͺO؀|�
ͺ[�%��+ͺ��e�!Iqͺ�*��M$ͺ�d3,.iͺ����%�ͺ.�2)<ͺ�}��"wͺ��0N$� ͺCU?�� ͺ��Ke�ͺg
%U�_�ͺu���fN$ͺ)X���qͺ09u�
*%ͺ�!\Fwͺ��
��
�ͺ�R�Q2ͺ�H#�"�N$ͺ�}��k�ͺ2|"9�!��ͺΣ���2ͺ�}hUh)ͺ`|"ى
�ͺ�N$��ͺ�
%n�!��ͺsii,!ͺeX��U/ͺ*�8
��ͺ��X�/ͺy���!��ͺP�1:�ͺ�� $�8�ͺ�2$Yn
ͺ�� sL��ͺPYL��ͺ�yw��
ͺ���)u�ͺ�H#O$�)ͺ�
3� 8�
ͺw���2ͺ�E~�/ͺ��
��iͺ���LZ�ͺ\�}"��ͺ�L�c��ͺ'�j�2ͺY��� �ͺE��wv�ͺy����,ͺ�83$�wͺz~�
%q#ͺ�K�Y˕ͺm3$�� ^*%ͺ x��ͺ��8�ͺ�T�@Uͺ�?]x�Uͺ�M�*ͺ�3$�3$h�ͺ�d����ͺ��G���ͺm�xͺR �  ͺ)� &4$i� ͺ1VH�ͺT � :*ͺNZ93�/ͺ�d����ͺ;�!+�ۤͺ6����ͺ�L'�(0ͺ�cV��ͺ���xͺѡ�Vyͺ^�
�Lq#ͺ�q#2��ͺ.��*%ͺ�Z���ͺo�jI#Myͺ�V�� �qͺŷ
W�ͺ��wMf0ͺ6%\4$�*%ͺ6�
�q#ͺ @�d�Uͺo3;O$�� ͺ\�V�J�ͺH~"��3ͺ���
�ͺ�ͱͺ�X��ͺ�I#y� �,ͺl~"cWi9ͺh��M��!ͺcZ��9ͺ(�"�
� ͺ�W5��9ͺz�� -ͺn�r�eͺ�����9ͺ�qO$>�ͺ|MA���ͺ���
ͺ��?e�*ͺ}yށͺ��q�!ͺE���UͺTee�"e�
ͺ��� ͺ�@�%�3ͺ]��!�*%ͺ��W4ͺ�� 5J#�O$ͺ��e��ͺ����O$ͺv���
!�!ͺ��	r#�ͺ�*� �0ͺ�W��yͺ2�jJ#,:ͺ2����"ͺAdueͺ�^�!��ͺ�*3�#rͺ=� ����ͺP$�+ͺj� H+�yͺ,OL4-+%ͺ�� �ee�!ͺ�X��ͺi���eͺQOW5$�ͺO��4ͺ��A�iͺ���)�ͺS��0��
ͺ�5-v�ͺfFHP$ͺ�M��Q�ͺ�!�
ͺ���A+ͺ�%���!ͺ�� �[��ͺ?r#�җ"ͺ�4�ͺ���41ͺ�#�%ͺ�6Bͺ�9f<�ͺ5�QX�Oa�
Ћ�O�Oa�
�#6$\a�
{�G�I�a�
���� h�a�
T��� a�
�i�Pa�
b\YN�a�
�'P� a�
-%�R%a�
�t�[R�a�
>���
��a�
��"�"a�
"��a�
E6$C��a�
3K#f�޹
a�
W6$�"z�a�
~N:�a�
LP6#a�
��MK#��a�
��	�
Ia�
O�Sc�a�
�N���a�
2,�
��a�
fPqK#��a�
�����a�
D�!sX�K#a�
`�YA�"a�
��	�s��Y������YbB�g�"�Yh�~+��E�$k-B x E�$p%�e�"E�$^���E�$��N��E�$x�� ��!E�$��"����E�$`� 
�:
E�$C�ǀ
�!E�$�4'1dE�$��"�� zE�$b:wP$
5E�$�f@50jE�$H�"�� 6�E�$P�
�]+%E�$f��o
E�$RL#v�
y�"E�$]1�+%o�E�$�f��:�!E�$���1E�$;�ajE�$|,Vk�E�$3�h�!�E�$v5M>�E�$�X#�GfE�$���5��E�$G+�t�E�$�� �XcrE�$@����E�$�
#� BzE�$���xfE�$�r��P$E�$���
��E�$�-���5E�$8 �1�+E�$1�5�
G�
E�$F�"��+%E�$�1�-�E�$r]��
bVE�$.���E�$��6$�+E�$�6;.E�$yr#�E�$�j��VE�$� 9��P$E�$� ��
�
E�$O�!���E�$�Q$��E�$pO�P6E�$!��:b�E�$��
,��E�$�f�+%�r#E�$����!�fE�$w.��>,E�$a��+,%E�$�V�� ��!E�$�O�]�VE�$���
z,E�$��"�:ǾE�$	�!R�Y� E�$���E�$z�z�jE�$o�R��:E�$�� �Y��E�$n��YW�E�$�j�.{6E�$�%����"E�$P�Z�E�$=��6$âE�$� ���%��/�؉%�ҁ"�O%��B�%c�P��!%� ���B%҂�� �%>�\,%a�%J��"�%S�n�ʋ%Ϡ��&%�Q��7$%K� �%�Q6	�
%�����%�6�5�!%{�(W%�,IQ$KW%,2��$�z%eWlQ$I�!%,;����%~,%�.
%�Q$� �,%%��

M;%� �O%9�6Z�%�B��
%���!��%ڌex!M'#J�aX	�J�4M#>�
��J����J�<P���"J��>M#!J��	*��%J�57$gg��!J�M�bM#\J��x�R�O"J�m�	�gJ�YZJP�QJ��	��BJ��!E7$J��g��	J��k�J�#!;!<�J��M#�j7$J�N�� ~J��mPa�J�w�7�^!J�o*�;�J���S�J���g��J�e�	^�7$J��%:%J�=�C+^J�w��P�7$J���g��J�[�g%9� J���!+^J�s�)�!�J����J��  W�J�"��W�WJ�2z�M#�%J�>�M�kZJ����ZJ�!J�`M�S^J�"��
$�!J�����ZJ�n�y��J���	�^J���
�[
J��7$�$s�J��P�%�PJ�AR�J��P�g��J��%�(�J���M��ZJ�h��Z��J�
h�%J�%���7$J�3%���7$J��+J��^@R0J��cR�PJ�:CS�J��H�|�J�U�QJ�P%@�"��J���#�!��J�]����J��Rb�
�J���
@
��J�.��RhJ��!]� d
J�xM�M#|
J�i%��
��J�#[1Q��J�UQ��eJ���!�lJ�Z���:hJ�kQ�^PhJ��b�"�
J��� ��!G�J���"ݍw�J������^J������ J��ڮ!�M#J�vh��RJ��!���J�GC��RJ�a�N#J�شS��J��
��!��J�"�� /"J�eC{���"J�΂"��
ֽJ�� l��J���>SJ���8$J�@��1�!J����CVSJ��C8[Z�!J�	_��J�+s���J��c[:� J�ފ+8$D�J��NP8$�CJ�<cDV"J�\�{�J���y�!{�J���$��J�s8$"J�����
J���C�J��B"_� )�J��:��c�J�q�
�� J��C�"G�J��8$~%j���,%�6�,%��� � ��u�M�	D��l����{ܦ��%�
�����0
��k�!�r>  ���!�� nS��%�Sq��u;�
�� �;o!(kA�
�#�"_,��O�{�D�a�
�,7�-��"�[�6�r#���8$�6��F_�����
	��Bz7<K� ͺ��Z��ͺ�z"}�
.Jͺ��
IS�ͺe����ͺ����ͺ�%�
p<ͺ���SȇͺT���ͺ���I�z"ͺ"0$�%�ͺ�|zśͺ'���
%ͺ6���ͺ���ͺ#�ͺ��40$5ͺH��zϳ
ͺ�%ͺ���!]� ͺ����
Uͺ���
�zͺ�b�AͺvJ �}� ͺ�W0$�ͺ*��� �ͺ!���ͺD�"�5Tͺ�z"mUͺ�!u1ͺۙ�� Bͺ�B��ͺ��<$ͺ�I�<�zͺ�� �U�UͺI�&�;�ͺ�
�� �ͺ�]�*�ͺ"�!Մe�ͺ�b3%��ͺ�9"��{0$ͺ�Jl{"ͺ3�!��7�ͺ=��<WTͺA�0�
R_ͺH�� �Jͺ��gԴͺ��T�
�ͺ�0${�0$ͺV��JI�!ͺ�4{t�ͺ���1{"ͺ�I��X�ͺU�V%w�
ͺ�T��bͺ�F#^�!�ͺy%�0$m�ͺ��I��ͺ-Cgͺ$�4� W{ͺjT�J�Jͺ���o{ͺ��0$E�ͺ�bK�ͺK��
n�ͺn��!��!ͺG#��
	=ͺp
q�R{"ͺ�_�_k�ͺG���cͺŧ!�%J�ͺm{"r���ͺ�0$�{�UͺJ)=̈́ͺ3�Bc�%ͺ��X� ͺ�7���ͺV�!ִ
ͺ�$1$ͺ5V��T=ͺP1$/Kw�ͺ��;)�ͺ G#�{"��ͺ���ͺ�gcVͺ:J�{�ͺ}1$��c�ͺ��a��%ͺ
[�:�ͺ�{��VͺKG#�e�ͺ�����ͺ�
�=ͺ��Ŝ�Vͺ���{2�ͺ�n�7�
ͺ ����ͺx�~p�� ͺ�f�<�ͺ��1�cͺ3/R��8ͺI��+j�ͺ�k���ͺ|y�mG#ͺ
/M2T
ͺ4 ����ͺ���h�

ͺ)��p#Mvͺ
d97�ͺ�W�{�ͺZ �svͺ�� ��)ͺ\�{�Edͺ�=RJͺ�8|[Kͺ���1$�ͺ���{"�ͺ���n|ͺ��ͺ���
�|ͺ�K2
%�!ͺ���
[�ͺ>uJͺ�G#�,��ͺ�F�!uͺ0}���Kͺ�T��|�!ͺ9��1$�ͺ����!\}ͺ��%W#�ͺ���G#�G#ͺ�T2$��ͺ
��ZWͺį'{ͺ2��W�ͺ�Ts#ˈͺ72$�&ͺT�� �ͺ�J
U�Jͺb�	��� ͺ�6ĵ
ͺOQ���
ͺ[�
�?�ͺ�� ���ͺ�:>6ͺl���ͺ
1H#�ͺ���;��ͺ����ͺ��G�Wͺ�}d2$��ͺ���{"o>ͺCU?�gH#ͺ��Kͺg
%U��Wͺu���3�ͺ)X���Kͺ�!՘"5Kͺ|A���ͺ� 8�@� ͺmKu�
 �ͺ�!\Lͺ�=L��ͺ�H#�NXͺ�}���ͺ2|"9�!�ͺ�}hU��ͺ`|"ى,
ͺ���
�� ͺ���J
ͺ�-%�Uͺ�>�E�ͺ�
%n�!ͺeXs�ͺ*��2$�>ͺ��X��
ͺy���!�ͺP�1�� ͺ�� $��Xͺ�2$Y�|"ͺo
Mk�|"ͺO�~mcͺ�K�"�ͺ��sL�>ͺPYL�2$ͺ�����ͺ�H#�fͺ�
3� �6ͺw����ͺ�E~�Q$ͺ��
��ͺ���L��ͺ�2��ͺ����H#ͺ�Ku
�cͺV}"�ͺ�L�cq2ͺY�Y	�ͺE�?�ͺy�����ͺ�83$o�ͺz~�
%Jͺ�K�Y6�ͺm3$�� �ͺ�Y]{��ͺة!��L?ͺdn�&Lͺ�3$ͺ��8��ͺ�?��� ͺ�Z��ͺ�3$�3$8}"ͺ�d��/� ͺ��G�~� ͺmܷͺR �|�ͺnLR$	Mͺ~�?v}"ͺ�~����ͺ�� &4$�ͺ1VH�LͺNZ�}"�!ͺ�d��ͺ;�!+�}ͺ6��|Zͺ�L'�҆ͺ�cVI#ͺ��?�ͺa�Rͺ�R��ͺFMMR$�ͺ~"�V�ͺ^�
��ͺ.��Mͺ�Z��
O�ͺo�jI#�
%ͺ�V�� U�ͺŷ
W��ͺ��wM&ͺ6%\4$�I#ͺ�~��$ͺ1���Yͺ5W��k�ͺ�M
Cͺ @�d\�ͺ\��Zf�!ͺH~"FMN@ͺ���eͺ�ͱ-� ͺ�X��ͺ�I#y� ��ͺl~"cW��ͺ�����
ͺ�^%i�ͺs�R$�~"ͺ��M0�ͺcZ�i@ͺ�Wh  	�ͺz:7��ͺn�r�ENͺ������ͺ�qO$G�ͺ|MA�� ͺ���4$ͺ�%'-ʈͺb�K-%j� ͺ�]-ͺҊ?eͺ}y��ͺE9���
ͺTe�@=ͺ���ظ
ͺ�@�%�ͺ]��!�%ͺ��W5�ͺ�� 5J#�� ͺ�� B�ͺ}e��ͺ>%� �4$ͺp�e�Mͺ����Nͺ���o�ͺ��
�ͺ�W�Ѫ!ͺ2�jJ#�ͺ2��5$ͺAd�!ͺ�^�!7Aͺ=����� ͺ�~"�N��!ͺl%���~"ͺ��3�c�ͺ=� ��hAͺj� ��ͺ,O�$�ͺ�� �eN�
ͺ�X�J#ͺi��*[ͺQOW5$@"ͺO��ͺ�5$���J#ͺ8�Tͺ��
�-%˫!ͺ�5$�Aw[ͺ���ιͺ��[ͺf��ͺ�M���ͺ�!�
Nͺ���A�ͺ�%�	�ͺ�� �[]ͺB�s�ͺT����"ͺ���
�5$ͺ���Oͺ��%ͺ���41ͺ�#�%ͺ�6Bͺ�9f<�ͺ5�QX�Oa�
Ћ�O�Oa�
�#6$\a�
{�G�I�a�
���� h�a�
T��� a�
�i�Pa�
b\YN�a�
�'P� a�
-%�R%a�
�t�[R�a�
>���
��a�
��"�"a�
"��a�
E6$C��a�
3K#f�޹
a�
W6$�"z�a�
~N:�a�
LP6#a�
��MK#��a�
��	�
Ia�
O�Sc�a�
�N���a�
2,�
��a�
fPqK#��a�
�����a�
D�!sX�K#a�
`�YA�"a�
��	�s��Y������YbB�g�"�Y]�!-�|E�$p%�K#سE�$�N�N}PE�$�y�E�$[f���!E�$��"�P1E�$��ǀF�E�$C�
��E�$�K#L#u�E�$��"�fk�E�$vΈXE�$�f��\E�$��s6$
 E�$P�
���PE�$pv�
�\E�$RL#����
E�$ߺ���E�$�f�0� E�$ܺ
�OE�$;���E�$��\׀"E�$3���E�$W� #��E�$�XM��!E�$]���E�$G mE�$Y����WE�$@��XNE�$��VۊE�$�]�E�$� ��
�E�$��c�E�$I]:O��E�$8 e ƁE�$wL#��"E�$F�"5�
e�E�$�U�xBE�$r]}��!E�$�L#�6$
E�$����!E�$��8�@E�$�X1E�$;�9�r�"E�$� �_E�$��gE�$O�r�� E�$pOQ$�L#E�$�$�
�]E�$��
�{�E�$���!E�$���+%ǻ
E�$Q�L#��E�$a�%��E�$|��],YE�$�O�� ��E�$��b�E�$��"0QdYE�$�����E�$�R���E�$����E�$o���fQE�$	�Y�OE�$n��Y�E�$,	ĻE�$�%����E�$���6$J�!E�$=�Z�QE�$� ���%��/�؉%�ҁ"�O%��B�%c�P��!%� ���B%҂�� �%>�\,%a�%J��"�%S�n�ʋ%Ϡ��&%�Q��7$%K� �%�Q6	�
%�����%�6�5�!%{�(W%�,IQ$KW%,2��$�z%eWlQ$I�!%,;����%~,%�.
%�Q$� �,%%��

M;%� �O%9�6Z�%�B��
%���!��%ڌex!M'#J�aX	�J�4M#>�
��J����J�<P���"J��>M#!J��	*��%J�57$gg��!J�M�bM#\J��x�R�O"J�m�	�gJ�YZJP�QJ��	��BJ��!E7$J��g��	J��k�J�#!;!<�J��M#�j7$J�N�� ~J��mPa�J�w�7�^!J�o*�;�J���S�J���g��J�e�	^�7$J��%:%J�=�C+^J�w��P�7$J���g��J�[�g%9� J���!+^J�s�)�!�J����J��  W�J�"��W�WJ�2z�M#�%J�>�M�kZJ����ZJ�!J�`M�S^J�"��
$�!J�����ZJ�n�y��J���	�^J���
�[
J��7$�$s�J��P�%�PJ�AR�J��P�g��J��%�(�J���M��ZJ�h��Z��J�
h�%J�%���7$J�3%���7$J��+J��^@R0J��cR�PJ�:CS�J��H�|�J�U�QJ�P%@�"��J���#�!��J�]����J��Rb�
�J���
@
��J�.��RhJ��!]� d
J�xM�M#|
J�i%��
��J�#[1Q��J�UQ��eJ���!�lJ�Z���:hJ�kQ�^PhJ��b�"�
J��� ��!G�J���"ݍw�J������^J������ J��ڮ!�M#J�vh��RJ��!���J�GC��RJ�a�N#J�شS��J��
��!��J�"�� /"J�eC{���"J�΂"��
ֽJ�� l��J���>SJ���8$J�@��1�!J����CVSJ��C8[Z�!J�	_��J�+s���J��c[:� J�ފ+8$D�J��NP8$�CJ�<cDV"J�\�{�J���y�!{�J���$��J�s8$"J�����
J���C�J��B"_� )�J��:��c�J�q�
�� J��C�"G�J��8$~%j���� � ��u�M�	D��l����{ܦ��%�
�����0
����!�� nS��%�Sq��u;�
�� �;o!(kA�
�#�"_,��O�{�D�-��"�[�6�r#���8$�6��F_�����
	��Bz7<K� ͺ��Z��ͺ�z"}�
.Jͺ��
IS�ͺe����ͺ����ͺ�%�
p<ͺ���SȇͺT���ͺ���I�z"ͺ"0$�%�ͺ�|zśͺ'���
%ͺ6���ͺ���ͺ#�ͺ��40$5ͺH��zϳ
ͺ�%ͺ���!]� ͺ����
Uͺ���
�zͺ�b�AͺvJ �}� ͺ�W0$�ͺ*��� �ͺ!���ͺD�"�5Tͺ�z"mUͺ�!u1ͺۙ�� Bͺ�B��ͺ��<$ͺ�I�<�zͺ�� �U�UͺI�&�;�ͺ�
�� �ͺ�]�*�ͺ"�!Մe�ͺ�b3%��ͺ�9"��{0$ͺ�Jl{"ͺ3�!��7�ͺ=��<WTͺA�0�
R_ͺH�� �Jͺ��gԴͺ��T�
�ͺ�0${�0$ͺV��JI�!ͺ�4{t�ͺ���1{"ͺ�I��X�ͺU�V%w�
ͺ�T��bͺ�F#^�!�ͺy%�0$m�ͺ��I��ͺ-Cgͺ$�4� W{ͺjT�J�Jͺ���o{ͺ��0$E�ͺ�bK�ͺK��
n�ͺn��!��!ͺG#��
	=ͺp
q�R{"ͺ�_�_k�ͺG���cͺŧ!�%J�ͺm{"r���ͺ�0$�{�UͺJ)=̈́ͺ3�Bc�%ͺ��X� ͺ�7���ͺV�!ִ
ͺ�$1$ͺ5V��T=ͺP1$/Kw�ͺ��;)�ͺ G#�{"��ͺ���ͺ�gcVͺ:J�{�ͺ}1$��c�ͺ��a��%ͺ
[�:�ͺ�{��VͺKG#�e�ͺ�����ͺ:� �{�{ͺ�R$Ŝ�Vͺ�n�7�
ͺh��;(�ͺ ����ͺx�~p�� ͺ�f�<�ͺ��1�cͺ3/R��8ͺI��+j�ͺ�k���ͺ|y�mG#ͺ4 ����ͺ���h�

ͺ)��p#Mvͺ
d97�ͺ�W�{�ͺZ �svͺ�� ��)ͺ\�{�Edͺ�=RJͺ�8|[Kͺ���1$�ͺ���{"�ͺ���n|ͺ��ͺ���
�|ͺ�K2
%�!ͺ�?[�ͺ>uJͺ�G#�|��ͺ�F�!uͺ0}���Kͺ�T��|�!ͺ9��1$�ͺ����!\}ͺ��%W#�ͺ���G#�G#ͺ�T2$��ͺ
��ZWͺį�ͺ2��W�ͺ�Tk�ˈͺ72$�&ͺT�� �ͺ�J
U�Jͺb�	��� ͺ�6ĵ
ͺOQ���
ͺ[�
�?�ͺ�� ���ͺ�:>6ͺl?�
�ͺ
1H#�ͺ�����ͺ����ͺ��G�Wͺ�}d2$��ͺ���{"o>ͺCU?�gH#ͺ��Kͺg
%U��Wͺu���3�ͺ)X���Kͺ�!�� 5Kͺ|A���ͺ� ۶@� ͺmKu�
 �ͺ�!\Lͺ�=L��ͺ�H#�NXͺ�}���ͺ2|"9�!�ͺ�}hU��ͺ`|"ى,
ͺ���� ͺ���J
ͺ�e�Uͺ�>�E�ͺ�
%n�!ͺeXs�ͺ*��2$�>ͺ��X��
ͺy���!�ͺP�1�� ͺ�� $��Xͺ�2$Y�|"ͺo
T��|"ͺO�~mcͺ�K�U�ͺ��sL�>ͺPYL�2$ͺ�����ͺ�H#�fͺ�
3� �6ͺw����ͺ�E~�Q$ͺ��
��ͺ���L��ͺ��
�ͺ����H#ͺ�K3��cͺV}"�ͺ�L�cq2ͺY�Y	�ͺE�?�ͺy�����ͺ�83$o�ͺz~�
%Jͺ�K�Y6�ͺm3$�� �ͺ�Y���ͺة!��L?ͺdLd&Lͺ�3$ͺ��8��ͺ�?��� ͺ�Z��ͺ�3$�3$8}"ͺ�d��/� ͺ��G�~� ͺmܷͺR �|�ͺnL�?	Mͺ~�?v}"ͺ�~(�
��ͺ�� &4$�ͺ1VH�LͺNZ�}"�!ͺ�d��ͺ;�!+�}ͺ6��|Zͺ�L'�҆ͺ�cVI#ͺ��?�ͺa��}"ͺ�R��ͺFM+�ͺ~"�V�ͺ^�
��ͺ.��Mͺ�Z��
O�ͺo�jI#�
%ͺ�V�� U�ͺŷ
W��ͺ��wM&ͺ6%\4$�I#ͺ�~&�$ͺ1���Yͺ5Wak�ͺ�M
Cͺ @�d\�ͺ\��Zf�!ͺH~"FMN@ͺ���eͺ�ͱ-� ͺ�X��ͺ�I#y� ��ͺl~"cW��ͺ���� ��
ͺ�^%i�ͺs���~"ͺ��M0�ͺcZ�i@ͺ�W�� 	�ͺzN��ͺn���ENͺ��J#��ͺ��G�ͺ|M2�
� ͺ����4$ͺ�%g�
ʈͺb��j� ͺ���ͺҊ�ͺ���ͺE9���
ͺTe�@=ͺ���ظ
ͺ�@�%�ͺ]��!�%ͺ��W5�ͺ�� 5J#�� ͺ�B�ͺ}e��ͺ>%���4$ͺp�e�Mͺ����Nͺ���o�ͺ��
�ͺ�W�Ѫ!ͺ2�jJ#�ͺ2��5$ͺAd�!ͺ�^�!7Aͺ=�O�� ͺ�~"�N��!ͺl%�N�~"ͺ��3�c�ͺ=� ��hAͺj� ��ͺ,O�$�ͺ�� �eN�
ͺ�X�J#ͺi��*[ͺQOW5$@"ͺO��ͺ�5$���J#ͺ8�Tͺ��
��˫!ͺ�5$�Aw[ͺ���ιͺ��[ͺf��ͺ�M���ͺ�!�
Nͺ���A�ͺ�%�	�ͺ�� �[]ͺB�s�ͺT����"ͺ���5$ͺ���Oͺ��%ͺ���41ͺ�#�%ͺ�6Bͺ�9f<�ͺ5�QX�Oa�
Ћ�O�Oa�
�#6$\a�
{�G�I�a�
���� h�a�
T��� a�
�i�Pa�
b\YN�a�
�'P� a�
-%�R%a�
�t�[R�a�
>���
��a�
��"�"a�
"��a�
E6$C��a�
3K#f�޹
a�
W6$�"z�a�
~N:�a�
LP6#a�
��MK#��a�
��	�
Ia�
O�Sc�a�
�N���a�
2,�
��a�
fPqK#��a�
�����a�
D�!sX�K#a�
`�YA�"a�
��	�s��Y������YbB�g�"�Yh�~+��E�$]�!-�|E�$p%�K#سE�$�N}PE�$��Ny�E�$[f���!E�$��"�P1E�$��
�F�E�$C�ǀ�E�$�K#L#u�E�$��"�fk�E�$v�XE�$�fΈ�\E�$��s6$
 E�$P�
���PE�$p���\E�$RL#v�
��
E�$ߺ���E�$�f�0� E�$ܺ
�OE�$;���E�$��\׀"E�$3���E�$W� M�E�$�X#���!E�$]���E�$G mE�$Y��X�WE�$@���NE�$��VۊE�$�]�E�$� ��E�$���
c�E�$I]:O��E�$8 e ƁE�$wL#5�
�"E�$F�"�e�E�$�U�xBE�$r]}��!E�$�L#��
E�$��6$�!E�$��8�@E�$�X1E�$;��r�"E�$� 9�_E�$��gE�$O�r�� E�$pO��L#E�$�k�
�]E�$��
s��E�$�1gE�$���ǻ
E�$Q�L#��E�$a�%��E�$|��� ,YE�$�O�]��E�$��b�E�$��"0QdYE�$��R���E�$����E�$����E�$o���fQE�$	�Y�OE�$n��Y�E�$,	ĻE�$�%����E�$��ZJ�!E�$=��6$�QE�$� ���%��/�؉%�ҁ"�O%��B�%c�P��!%� ���B%҂�� �%>�\,%a�%J��"�%S�n�ʋ%Ϡ��&%�Q��7$%K� �%�Q6	�
%�����%�6�5�!%{�(W%�,IQ$KW%,2��$�z%eWlQ$I�!%,;����%~,%�.
%�Q$� �,%%��

M;%� �O%9�6Z�%�B��
%���!��%ڌex!M'#J�aX	�J�4M#>�
��J����J�<P���"J��>M#!J��	*��%J�57$gg��!J�M�bM#\J��x�R�O"J�m�	�gJ�YZJP�QJ��	��BJ��!E7$J��g��	J��k�J�#!;!<�J��M#�j7$J�N�� ~J��mPa�J�w�7�^!J�o*�;�J���S�J���g��J�e�	^�7$J��%:%J�=�C+^J�w��P�7$J���g��J�[�g%9� J���!+^J�s�)�!�J����J��  W�J�"��W�WJ�2z�M#�%J�>�M�kZJ����ZJ�!J�`M�S^J�"��
$�!J�����ZJ�n�y��J���	�^J���
�[
J��7$�$s�J��P�%�PJ�AR�J��P�g��J��%�(�J���M��ZJ�h��Z��J�
h�%J�%���7$J�3%���7$J��+J��^@R0J��cR�PJ�:CS�J��H�|�J�U�QJ�P%@�"��J���#�!��J�]����J��Rb�
�J���
@
��J�.��RhJ��!]� d
J�xM�M#|
J�i%��
��J�#[1Q��J�UQ��eJ���!�lJ�Z���:hJ�kQ�^PhJ��b�"�
J��� ��!G�J���"ݍw�J������^J������ J��ڮ!�M#J�vh��RJ��!���J�GC��RJ�a�N#J�شS��J��
��!��J�"��/"J�eCo���"J�΂"��
ֽJ�� l��J���>SJ���8$J�@��1�!J����CVSJ��C8[Z�!J�	_��J�+s���J��c[:� J�ފ+8$D�J��NP8$�CJ�<cDV"J�\�{�J���y�!{�J���$��J�s8$"J�����
J���C�J��B"_� )�J��:��c�J�q�
�� J��C�"G�J��8$~%j���,%�6�,%��� � ��u�M�	D��l����{ܦ��%�
�����0
��k�!�r�  ���!�� nS��%�Sq��u;�
�� �;o!(kA�
�#�"_,��O�{�D�a�
�,7�-��"�[�6�r#���8$�6��F_�����
	���:)�ͺB���7د!ͺ������hͺ�S�z:9"8ͺ��wdf9"fHͺ��9�
6�ͺ;V��%�gͺg]�
�";s#ͺ�h;��Qͺ�;�.N#ͺ�:��$��$�~
ͺ�%��^HͺF��#&���$ͺ��$���R�ͺR�Zh!����ͺ��9��$��ͺ�Q�~
�oͺ�� ��#��ͺ;�ȕ 0G�~
ͺǷ
�J��ͺj_�9"
seͺsen�m�ͺ�%�>�pͺ�hA�U�R_ͺR_��$�$��ͺ���� I��ͺ8iV�#�z�9"ͺ^�p
*�$:� ͺxi�_���jͺH��RVG�Rͺ�[�\%ͺ�_��ͺ�����"ͺ�_`Bͺq�
8���jͺ�_�V��ͺ��uN#�Kͺ���I�
ͺ�
�%@H��ͺ��f`�#]ͺ�8$��\�ͺ\���p7ͺ�2����ͺ�7932�ͺ2���
��ͺ���N#�7�ͺ(�@��[ͺ�%�8$�Ͱ!ͺ{\.��;"�ͺt�
[�K.�ͺ.��#[�6�ͺ6��}�#ݵͺ{�
G�r��ͺn���0�ͺ5�"t4p�$ͺ�"4q�$z�ͺ�`��
��O�ͺ�Vx�&4_`ͺ/)�~��Hͺ�8$�<"	9dͺ�����K�ͺ�� �d1ͺ�9$�{����ͺ{��#1h��ͺI�I4Z�d�ͺ�9$v!��ͺ*#8	����ͺu4�4]ͺ���<"�ͺ�{�����ͺ�� ���ͺ�3�+/ͺ8�
�s#+/ͺ5���\3ͺ��t#ͺV�6hHͺ���R$�rͺ8�
1XͺA�"41Xͺ"� G�ͺ�/�! ͺ��;�N8ͺ#�
8ͺ�L�~�"ͺm� S$~�"ͺQ4Y� |ͺ�|-��� ͺ'� �|�/ͺD�o!r�ͺ"pa��;ͺ�4#F �;ͺs�-��ͺIs4}�4ͺ�p��ͺď?�ͺ!��=Z�
ͺ��
h�Z�
ͺv<;01�ͺ@��
o}ͺ��$eSNS$ͺ���TN! ͺ���|sͺr�;�|sͺmh��!�Xͺ�<kY��!ͺ;#�[�ͺq�T�Yͺ�m���"ͺ}Q��
�"ͺ�-��z! ͺ��!�ͺi���$v�ͺz�"�
Bͺ�-%���Yͺ�k���Yͺ�����ͺ"
��"�ͺ	�

��ͺ�2ߧͺv0���-%ͺѣ�<�-%ͺ�4o�Rt#ͺ�
�^�
ͺ���H�ͺ�+��$+�!ͺ�4{�u#ͺ�k�}u#ͺA�#�� ͺ1" �u#�u#ͺ0V�Ӛ"ͺk$ �^�ͺ.��$(v#ͺf" "=(v#ͺ�8����ͺ��
7.�kͺ�� �}�0ͺ��
G0lͺZ�S��Zͺ,�"�.%�Zͺ~�8��ͺ%��d�"ͺ�#	�/�ͺ����Tͺ��%���ͺ܌U�ͺ�D )�ͺ����/�a�
h ��a�
%U�{�c[a�
V��4�!�la�
�<$D[�<�a�
��[
�0�� a�
��-���a�
��
��~Ya�
�>$�{G<�a�
N Q�Mb�<"�Y~ S�$���%�YӈO#��.��Yb�� I�EW�Y�k�
Zh�"�Y�Wpb����YOZ��
��"�Yu mm���Y� �":P#�Y`�_�!U���Y��@$��Y�
>�� :�"�Y@�P#�b�Y�� �Y<���s`@�Y�P# ��Y����Y��
�@$+'Bc�Y\�4���Y��aX��Y�%6 й�Yi�`'`_��Y�6�
"j!�c�Yw  �u�"� �Y���A$�Y!N�
� �Y�L�G���Yk[�C5��$�Y��f�h;x��Y��ؠ�G��
�Y��������
�Yd
�!Fj!���Y"
P�#��d�Y{'w����� �Y�`6��="���Y������="�Y�a��MQ�$�YK ���e/l�Y/l�M�:� �Y�% f"��Y�!>"��n"�YuA$��j!���Y��=���$�Y]��*��Y
��p#���Y��!y��ebE�$=d�L�ZE�$�� ؁
��E�$XX�
A�"E�$5�@�VE�$�X�J�E�$�6� �k E�$M
���E�$^���`"E�$d��H����E�$�m�L�$�E�$!q�*�M�E�$�P#=���
�E�$.�M�[dE�$���l!K�E�$��A�F�"E�$	W�<f͍E�$�� ��h
-!E�$n� �6�#�mE�$�A$�!�n� E�$� ε!���!E�$�b"I� E�$�Q#����E�$zl��tE�$�� ����E�$w�Z��v#E�$�l�l�E�$r.m[�E�$�h�h%iE�$� ]=��E�$I}�S$E�$U��
��E�$o�!$1�E�$�.%�=M~E�$D!Ȩ��E�$ަ9�iE�$��
s���E�$���=�E�$�/%�~E�$���" AtE�$T9�[�~E�$�9��!E�$N/%�" �E�$S��.��!E�$?��[_1E�$�� O�!# E�$m5�!E�$�X�
T�E�$Q��S$u�E�$����![E�$�1u# �tE�$��!�.�
E�$X/���"E�$��@� �
E�$�&T$�# E�$
�!>��E�$œbT$l�E�$\��!E�$�9-�
OE�$>�`m�E�$H�!vv�E�$j���J>E�$
:{/%x�!E�$�T$7�Z5E�$�|� R
E�$F:�i�iE�$�>����E�$r��T$�"E�$:�����!E�$��w:�
E�$���$�E�$ɧ2+�E�$��Pjf\E�$;
8���E�$��\Hw#E�$�/%�/%�jE�$���="E�$]t�!u
E�$>2��!80%E�$h�
s���
E�$$ m2�w#E�$�0%�5�E�$D��
��E�$�:�>E�$I$ H��jE�$?�
4x#E�$�"��r�E�$�:�m-/E�$y��"�/E�$��U�
��E�$�����mE�$���2|$ E�$���"��E�$�t��4uE�$�x#E�5;E�$$?"�#��
E�$T�) �<hfE�$hf�EfE�$�R#�<��$E�$��$�[i��E�$ދ�J�#E�$�#�<!�$j�E�$��8���pE�$�[����?�E�$����
E�$�C$��D#�E�$}��� -�#kNE�$9�
h#��ނE�$ނƔ,nU#E�$E�$o��9E�$Ʋ�� "$t�E�$������E�$�Pn,��SE�$���N����E�$E���q$E�$*��!�"#%�#>?Vn�%��U$�
Ɯ"%��0%0~k%��
z���%.�1�"@0�5%�5ݯ�k�� %�� @��#N�$%_�")�!�
J�w�\��J�ed*O��d�J�t�"o�#�YJ�'�D ��t�J�
MM%J�\����EJ��!��i$��J���?"�7X%J��o��$��t�J�
����J�;"Ś @L��J�� "��$�\J��2z�#�EJ����S��]J�cL_\�\J��gX)�J�w�	�"J���_\�\J�>�y�@"J�U��LX)�J��
	�"J��d$��J�>�`+�W�J�����$J@"�OJ�"�/�#J���r� h%�OJ�b���J�-#���J�J��%��J�J��gIJ�3P�L��J�)�
1�"
�J�8���aMJ��o��
��J�F�Zܔ:$J���S#^
�$J����"�
	#J���%S#�#��J��#X�\J����
����$J��o�#` oeJ�����J���J��h��#J�_�"�d��J������J���
3%�C$z)J�U�!%��m�J�V�$��J����A3Zm#J�{M��kJ�k��
��
2$J��C$��
�{c"J��$%�s�*J��#E9}l!J���
q9kpJ�\ZXp���AJ��dQPG�
b� J���#����J��
�ZC�AFJ�#�=�I�J�^]xMB�l!J��e��#�
4BJ�����!eJ��!_� A��J�˓{�)�J���@�]-/�J� E$���@"��J��FIIN��J��q�M A"Y�$J�`#���\�J��
��mNf�J��$�f�
�MJ�*	C?&;kJ�7�
��k��J����
��+�
J���� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&� ��J�ҥ�$%CJ�%C�����J�����N��J�����&�pJ��pX"�8�$J��e�9�l!DB"J��#�Sf�J����N�P�NJ��"F�14OJ���$�y�J�k�"�*�Z��J�h%��Z�]J�a��B"&�#P�J�� �:Yem!J���qz� �
J�������x
�C���{
��#��\tot}t��\�$�Ep���C�����V_ot������r��u��d!��Z���ot�e!���t	�#��<f!����|

7ot�����e�%�7����� 5'�r��i�LG��r���g[�� � �����r|� ������
]Gly���*�%<�
ly��ޏ��
f��#��wG�r�E$�#��![�B����\H@�����xI)f�e#���y��y�e#����"��yF��E���C�#��7w���$la��-}
��p���)�$�8�
k� ���D�f!I7"�\����7d}
�\���LO��/��ǵP���D����E�c#�f!���� 700X�#��G�"��k#X�#�����u#�#��g!
�!5FJFR7"��w��;"DZ�Z���U���|�#�a��V�\�g!�a��;�]�a5� ����E�
�7V��� j}
ٶ�g!���I��$� ������� � �����7t}
;D�#��Y�)�Dla��)#R�
U����0
��D}}
��������V��jr��b��ݶ���7"��y�|����F�bwl��g!��W�
�8"�l�c0���
�
l��E��/�@�J��  �%��W#��J�	�/02���yt|?f#2���%�%��L$��ܸ!�%?f#��"��Q Kf#g�ɭ����ۚg�XJ���%��Sf#����.�"��Sf#�����Ȏ"���� ]��N�7�#�]�����
���y���y��$��`G����
�]��Sf����s�?0���JL�H�[\���\�� H��}
��o
��$�#c���טLF+z���*H������1�
H��]���0 H������f�H�+����%�H��$���%;^<�
u��$����3�J�
:U$��x�I]9$�0��(
E!9$�7��dF\�
���� ���F;�#��G���:�JF���jc~
S�
1����a]>�S�y#��d��uY�X��y#��b�Z
v�S�@� ��9�$ `!X���!���k� ��!S�Ҟ"��2��2b�X�����A��Eٶ,��l�
[�8�����c��8n������cN�����zn��~z6����n�8"r=���l�
�ln1��ѱ�
#��l���]t3#�X������c���
��V�
�z.9"������S����-~
�ES�������
7�S��F��R^�S�[~
��^�0S����g��GS�|~
�����#S��%�
���o��:�%�
����o��:-�9�
g� ��o��:!9�
w6��o��:
$q�����:�QgPM�
���:I~Q�����
�&���:�$a�#W��lq����:��
������:M$=��U$�
�!��:>7����% ��:�1���1����:�4mR�_���:�����!��:�=��o�����:�y�
6v�A���:x���� ��
��:2�!���z#��:3q��
�U-2�:J�#���"�U-2�:�\?�U-2�:$��fv�U-2�:���
�U-2�:�������U-2�:��K��U-2�:��d^�3�U-2o��:d^1%�U-2o��:�Q�!

�U-2o��:�!V$�U-2o��:4�3j��U-2�:82ѴHm�U-2�:{#�
C��U-2�:{��^.��U-2�:�?Eo���U-2�:�"2�!�^�U�& �#�:[o2�!���U�& �#�:vo`$�
�U4m�:�!�$_�U4m�:�?��4�U4m�:F�!>U2�U4m�:���6���U4m�:m
AV$vo�6ږ�U4m�:m
AV$�#�v�U4m�:H_�2p_�U4m�:����o�U4m�:�$�v�m�U4m�:�C%�"�U4m�:b����#%�S4�U4m�:b���J� �& �o�U�-2�:�f� `��U�-2�:\�
�#
�U��-2�:����d��U|� -2�:!p�"K�U-�-2�:�_�_��U<>-2�:7pv�
���U��-2�:�"�4}��U�"-2�:��
w��U�#-2�:$�m4w�Um�!-2�:�!$�2�U!�-2�:h������U!�-2�:�"�?��U!�-2�:�67���U!�-2�:c?����U�?-2�:���0��U�-2�:g@�$`
�U�_-2�:��
��7$�U�& -2�:�$%>1%�U�1%-2�:�
�2�@�U�@-2�:�5%�m�UD�-2�:��8�"�#�U�1%-2�:��wZ��UA-2�:VV$3��U��-2�:��7����URp-2�:� 3%7�U�-2�:�
čdq�Ua�-2�:Q7@' ���UtV$-2�:,�#V' ��!�U��-2�:�gm��UM�4�:ȟ�o7�U'34�:�
BB��U��4�:�w����Uڍ4�:�
z?�U#n4�:��
�i3�UK%4�:l' �!q`�U�4�:�^�H�#�U�4�:�4��q��U>�!4�:3%�' ��!�U�4�:��?���U��
4�:�`?����U�4�:6X$��#r��U��4�:��4� {��Ù#4�:�)�
�4�U]X$4�:���Y�
�U?3%4�:�' �3�U��4�:
�e�#F��U54�:�`�%a�Un$4�:cn��I
�U��#4�:
��5��U�q4�:�����%�UJ�4�:5����!�U�64�:`a����Ud� 4�:�%&���U75%4�:o�&�'��U�?4�:P���7��UqB4�:$&L&y��U�A4�:N��6%�a�Uծ4�:N�"6s�Y$�U�n4�:FZ$v�"���Ufs4�:8�!V��3�UnZ$4�:���
�3�U�74�:��"$��3�U��4�:�w�o�3�U|&4�:�N&s� �U��4�:�'K8D�U�t4�:]�
���Uv�4�:�6\$#p�U{84�:.��f\$�U94�::fC�'�UsD4�:���\$~&�UF4�:��

���#�U�G4�:)��_(�Ua�!4�:��
"��UH44�:Ϡ"��� �UC:4�:�' ��"U��U( 4�:�#_v�a�U=�4�:��
<);�U_�4�:̣�p5<�U�v4�:&�"7%�p�U�(4�:��M� ]I�Uϰ4�:�;_b�w�U�*4�:��-q}� �UP�4�:���\$&]$�U�+4�:�b�ID��U�<4�:�>�5��!�U�4�:��<
� �U67%4�:��) ���U	4�:T�{��8%�U��4�:]q�&�b�U~�4�:G�޽(c�U�4�:�Qxϲ�Un�4�:+6Bx�I�U�4�:ɽ!9%;� �U_J4�:N�"��;�U�) 4�:S6h-�
�U��!4�:�����Uv�"�� �:*�#9��]$�Ul
�� �:'%'��
�UD�#�� �:�
�cO��Urx�� �:yA�L��U	��� �:e�
�c!��U�
�� �:k+ �x1�#�U՝��:�K����U&���:��,yk��U�>��:�������U�C��:�� 
�c�U�C��:�x���U���:�>
�U���:7y�+ �-�UWy��:оN�Q9%�U��
��:J{6�q�Uk��:�x�-�y�Ud��:'zHz���U�]$��:�K�x�9%�U���:�"L5��U
?��:�x�6��US�!��:<�c_$6��U����:�"� ��U&
��:s�� x_$�Uy�#��:}��qp��U�
��:+�"Y�
�U|���:�De�	� �U.��:U� ��l� �U<<��:n�ge�_$�U�� ��:��%L�
�U���:y�q2y�Ud���:	�:L�+ �U� ��:�+ gm���Uc���:!��Y�
�Uў��:��k�:'�U��#��:��
�����U~���:��
{y�9%�U���:��e�D
�U�q��:�?��
�?�U�?��:K�t
�#�U`$��:r�+ �� �UEL�L�:A�"gm��U�
�& �:h��+ ��
�U� �& �:7r]z�<�U6��& �::%�L���Uo�& �:�L�����U��& �:����)��U��& �:s��6)��UM�#�& �:�.:��
�U���& �:o'����U��& �:�y��!Sr�U���& �:Q���T�"�U���& �:E!��"�U��& �:S���}e�U�<�& �:*`$xE
�U`!�& �:�?V�r�U���& �:0:%E`$��!�U|��& �:�.Z`$G:%�U�:%�& �:�����U���u���
�|��� �E$Q�`f#��f#�l�
���fv�#��#u��n��r�������NPo!+%����$���"�g�;[�w�1�+��lm&���"���]�&��E$4D������$���#{(�-g#�� S�(�!� �;a���
�-sP�
'���e m|����
Ko!�������#�����
�;+s֏"���u�����-�`R�(�M�����
�J��sUDvDX��;Q������� R�T#	^�f%���D���g#��#|�ES�H��:)�ͺB���7د!ͺ������hͺ�S�z:9"8ͺ��wdf9"fHͺ��9�
6�ͺ;V��%�gͺg]�
�";s#ͺ�h;��Qͺ�;�.N#ͺ�:��$��$�~
ͺ�%��^HͺF��#&���$ͺ��$���R�ͺR�Zh!����ͺ��9��$��ͺ�Q�~
�oͺ�� ��#��ͺ;�ȕ 0G�~
ͺǷ
�J��ͺj_�9"
seͺsen�m�ͺ�%�>�pͺ�hA�U�R_ͺR_��$�$��ͺ���� I��ͺ8iV�#�z�9"ͺ^�p
*�$:� ͺxi�_���jͺH��RVG�Rͺ�[�\%ͺ�_��ͺ�����"ͺ�_`Bͺq�
8���jͺ�_�V��ͺ��uN#�Kͺ���I�
ͺ�
�%@H��ͺ��f`�#]ͺ�8$��\�ͺ\���p7ͺ�2����ͺ�793҈#ͺ҈#��
��ͺ���e��O�ͺ��N#�7�ͺ(�@��[ͺ�%�8$�Ͱ!ͺ{\.��;"�ͺt�
[�K.�ͺ.��#[�6�ͺ6��}�#ݵͺ{�
G�r��ͺn��y��.ͺ0������ͺM�i�ͺ�e��}
ͺ��s#	�"ͺ:���߳ͺl�%�T�ͺ<M�'��ͺ5���"fͺEf�/rzͺ2��Da�ͺ\�
��!`�ͺD��!��ͺhf�D�6ͺ��
�z !ͺ@D!A�ͺ��
ڒh�ͺ��
e��/ͺ�r�r��ͺ~��
��ͺ�ya>�ͺ�y�~!ͺ�ʣ"�fͺ�>
7�'ͺ�fI7�!ͺ, i�N�ͺ@�#.�!�ͺo`$�>zͺ0;%��
o�#ͺ0�7, 8@ͺ���`$�!ͺ\@���fͺ=�T�(ͺq, [z^Mͺ3�
E�zͺ<!�`$�zͺ��#�a$ͺ6
�R$��ͺi�!�
[�ͺi���"ͺ�z4��ͺ�� T;%ͺ��e(�7ͺW�
{1gͺ&���`!ͺLE�M�7ͺK�|�Z
ͺ�;%0�ͺ�${sͺ����"<%ͺ�7�
H{ͺ'�*�\�
ͺrE�!O0ͺޓZ<%�@ͺs��
3a$ͺ�@{g�(ͺ&85���!ͺ�>�a$��#ͺd��E�Mͺ�0�<%~sͺ&�, R8ͺ��
���, ͺ
��8;- ͺ�a$)
��ͺ�d
�a$ͺ���_{ͺ<��w�ͺ�
�!Rͺ�8 Nb�ͺp��- �ͺ�E!?�gͺ�0�sN�ͺ��{�ͺ/��
�0ͺ�
L���ͺ9-��EͺƤ"S$"!ͺI1Y�hͺΠ)yFͺo1@��<%ͺ���
tͺ\?މ#=%ͺ b$�{Hb$ͺ���"��ͺ��!Ht�!ͺ+ 
S 
8�ͺ��!U�!ͺ�@)��ͺ �;�o{ͺ���#Aͺ��
���
ͺ[N�- ��ͺ. Y=%�b$ͺ�F�1O�ͺV�#�1aAͺ�{>. r�ͺ�h�{�hͺ|I|4�
ͺ��
B�!ͺ>9}�"�ͺu�~?�tͺZ�w�}�ͺ��
-�"�b$ͺ��
�
�Fͺy�#�N?�
ͺ����f9ͺx�!��
�Nͺ�A&��ͺ2G/c$ͺ(2�?��ͺ��
��i�"ͺPGG2{�ͺ�{� 
��!ͺ��h�=%ͺRc$��#��ͺ�2z. ��ͺ�. xG�tͺ�. �
�ͺs�!�!��ͺ5iM�Oͺ�A@��|ͺ�2ߊ#��ͺ�|jO�Aͺ>%]�!�#ͺ��!a���
ͺإ"`!��ͺ<�#{���ͺ�?��!ͺ�G��4�ͺ�zc$�ͺ� 
���9ͺ|�f�#ͺ��9kͺ����!��ͺ(�}.�!ͺ� 
�t"@ͺ��z)<}ͺ�A�O��ͺ!:�B�ͺ!u�O��"ͺ%>%B��ͺ�. SBq@ͺN�"�B��#ͺ��#�B��ͺ!
��!R�ͺ_i�o!�ͺ�)a�u�"ͺ��4}�@ͺ�!F �ͺt��-��ͺ�P/ͺ��!�#�)ͺ�2VP�Gͺ��CB:ͺ5�Hu@�ͺ���2>Aͺ@C�:HͺkC#/ U�#ͺ|i���ͺ���A�ͺ�:�!^/ ͺ^}�ަ"ͺ�A���ͺP�U�ͺ"3�]�ͺ�C�A��ͺ�iO>%��ͺ���#�Pͺ
��:�Cͺ܌#�/ "�ͺ�P�}�c$ͺ,B)�"�Cͺ�#QBFDͺ2��/ fDͺlH�H0 ͺ�is��Dͺ �ҩ{�ͺ�H��R�ͺ�H<|H�ͺ*d$4!
ͺ�i�Pd�"ͺj��ͺ�>%�=��ͺ4Q�
Gd$ͺG3h��}ͺ%j;0T!
ͺ�
�*�ͺ��9�
;ͺU����ͺ;~G�4*ͺ���ͺI��ͺ�D�aQͺ�d$B�uͺE��*�ͺY;*�ͺ�BEt�
ͺl����Qͺ����ͺa~'���
ͺ�!
w3��ͺ!(!*�ͺ��!�;*�ͺ,���Bͺ�B�x*ͺQ0 �0 �Qͺ�u�� c!ͺR���?Eͺ��Pk*�ͺ�
�*�ͺ4Z�RIͺ�!���ͺ�!
ͷ�~ͺ�*���Eͺw|g���"ͺ�;
�*�ͺZ�k*�ͺ�!�T�ͺHR��d$ͺe�kY�*ͺrR;�Χ"ͺ�E��!g+ͺآF�uͺZ��R�+ͺ�>%v�ͺ�;~�H�ͺ�R[O#<ͺ>F/�
z�ͺ�3o�!�!ͺ$�#��M<ͺ�>%x
�uͺWe$Q� �uͺOj��:vͺ�~e$b�ͺzv"CG�"ͺj�
 "
ͺ�3�~ujͺ�v���uͺ�jZ�uͺ�|T�!��ͺ�>%n^F�ͺ�<�e$�ͺ���^!ͺ����Qͺ��r�uͺ�x��uͺ�j,�e$ͺ�C
["
ͺ�]C�CͺO!�SkͺmF�O#ͺ�ϋ�uͺ�v\D�uͺ�T��"ͺP�#����
ͺ���Jͺ��!��
4�ͺ����5f$ͺ�|��+ͺ��!�f$��ͺ���"�0 ͺ޸��1 ͺ���"Tkͺ
�
N,��ͺ�!�!8�
ͺ�!�+?%ͺ��!F�+ͺ�6�+ͺ�"
;�,ͺ(��|�!ͺ��F�Iͺ��kD�ͺ��"��<ͺv�#�f$+ͺ=P�+ͺ(�y?%�kͺ�"
�� wͺ�v�,I1 ͺ4��J�ͺ-�]�ͺ%l�+ͺ��s%+ͺ���"
ͺ���>4ͺ3=.}�Cͺ��#E��ͺ4-����ͺ�!b�
+ͺ�1 �+ͺ���R��ͺ�1 u	Sͺ�4g$\-ͺ3#
�I��ͺ"J���ͺ�-.�7�ͺ�1 6أͺ����ͺ�FeS�ͺ�4�a"wͺHl�4�ͺ��
J�5ͺ2D��H�ͺ/�"��!7�ͺN�!�&7�ͺi��M�"ͺ`w��O�ͺ���Z#
ͺ�FUD�Sͺ�w�-Ǎ#ͺ�#
��"7�ͺnl��7�ͺ.Wg$��ͺ�l��wͺ���R5ͺ��"���g$ͺI�W�"�ͺ�5�
7�ͺ�1 �7�ͺ���"�#ͺ�lTA�ͺ�D�!�lͺ�5 �qͺCJ�_�wͺ`�[�7�ͺG�� 7�ͺ8G"�

�ͺi}��#ͺ����"2 ͺ�5���g$ͺ�#
���}ͺ�?%[G%�"ͺ	6�:.ͺ5!��ͺN�Y!t�"ͺ�!�S!Tͺ�h$m(ͺ�"=�#�.ͺ�2 fJ-6ͺ�D�?%%�"ͺ@%Nm%�"ͺ�6���ͺ�l�!�TͺU���!ͺ<��h$�ͺ�m~�ͺB@%m�%�"ͺ����%�"ͺ�D��ͺ7����ͺb���"xͺ���
��ͺ>!8�ͺ�"%%�"ͺx�#E%�"ͺ�J./@�"ͺ�J3~l�ͺo~��
�~ͺ��"S�;Eͺ+�
��j/ͺ�x�J%�"ͺ�x��!%�"ͺ�G2ͺh@%���"ͺs�
�+�ͺ���<��!ͺ��o���ͺ�G,K�vͺ�5�Fͺ�x���6ͺ���#
^=ͺ�yV�
ͺ���2 ��ͺ$
�E�/ͺ%3 p��ͺ�@%�@%�vͺ����vͺ$7��@%ͺ�s�!�mͺ"n�G3�ͺJ0Ly��ͺ�� HgKͺ��!��!�vͺ΀��
�vͺKA%��#EUͺ[H�yr�ͺ�I7�=ͺ8��Pnͺ�!;�!�h$ͺ�K`�vͺ���vͺѬ���Uͺ��
�
��ͺO$
���ͺv�!�Ei�ͺLi$"���ͺ>�n�vͺ�y!�vͺV�
��$pA%ͺ��{��ͺ�H�u#�yͺu�
�}�nͺwo#�{0ͺ��}$
�ͺE��X�ͺ��
�A%]�ͺ@���UͺDz��2FͺU��Fڎ#ͺk�"�!?>ͺwi$��oͺ��7�ͺЙ��ͺ�Hkz�i$ͺ�!A�e�ͺ�6�#��!ͺ}�q�#�7ͺ��
8��ͺ��P3 �ͺ��"�$
�ͺ�o�z �ͺ��#��8ͺ��JV�ͺ��>��
ͺF�
�i$	!ͺ�!n	!�ͺӏ#m�
�ͺs�,j$�0ͺ$�����ͺ��B%ͺhj$�F�$
ͺa�Y8��ͺ�p�ͺЀ�j$�ͺ�j$ �uB%ͺJ���$��"ͺI17.ͺ�"=��
ͺ'���%�!ͺ��!���,ͺG�3 �ͺP�Q%
�ͺ��#�K�ͺ�HN�%Gͺ�8���Gͺ��V�Hͺ���B%�#ͺ��#�V�,ͺ3C%$��,ͺP���ͺ�K�uͺ���>�Gͺ�	!��	!ͺYC%�3 �%
ͺ#Qp�,ͺ^��z�,ͺs�!�ͺ{��?ͺ4 ����ͺ/4 ��(LͺS�u1I�"ͺ����,ͺt�X��,ͺI
!�iIͺ�V)ͺ7W`4 �8ͺ~��‚ͺB�
�j$}?ͺ����"�,ͺ1����,ͺ�LGW�ͺfHS�W�ͺ����Lͺn
!�.%�pͺ�C%�8��ͺ�0�;�ͺk��m�
ͺ.9��
��ͺqҁ�4 ͺD�H{ͺ����C%ͺD�I��ͺ�%
q��ͺ��1;�ͺ�
!�1;�ͺ����P�ͺ�{%���ͺ�4 :q�?ͺ�&
<�ͺuqeW�{ͺN9D%;�ͺH@-|;�ͺ���kͺ�"<��ͺ����#,ͺ�W��
�
!ͺ�I�qͺ�W�";�ͺʛ5 ;�ͺ !O�"I�ͺ��-k$p9ͺ��X5 %2ͺ���2,ͺ�؃��ͺa�J;�ͺx��;�ͺ��0�ͺ5�"t4p�$ͺ�"4q�$z�ͺ�`��
��O�ͺ�Vx�&4_`ͺ/)�~��Hͺ�8$�<"	9dͺ�����K�ͺ�� �d1ͺ�9$�{����ͺ{��#1h��ͺI�I4Z�d�ͺ�9$v!��ͺ*#8	����ͺu4�4]ͺ���<"�ͺ�{����ͺ�:�� ���ͺ�:�3���ͺ�:�q��V�ͺ�:8�
�s#�5 ͺ�:5����Wͺ�:%��Hͺ�:��'�ͺ�:6�
��"��ͺ�:V&
�/�ͺ�:E��D�2ͺ�:��!ik$ͺ�:���!��ͺ�:�#�DBJͺ�:��!�z�ͺ�:��D!�ͺ�:��"ڒ�ͺ�:�He�J�ͺ�:(�r�9ͺ�:9�
�
FD%ͺ�:=Ma>2rͺ�:�&
��k$ͺ�:A!ʣ"Kͺ�:l$
7sMͺ�:II7<�ͺ�:(:i��&
ͺ�:'
.�!��ͺ�:���>2�ͺ�:�5 ��
�ͺ�:�D%7, L'
ͺ�:a�`$��ͺ�:�M��j@ͺ�:6 T���ͺ�:�#[z_�ͺ�:CXE��ͺ�:���`$��"ͺ�:W6 �2	ͺ�:�'
�M��ͺ�:˯�V�#ͺ�:;�{���ͺ�:��6 �!ͺ�:a:�rf�!ͺ�:�D%O��ͺ�:u�
���ͺ�:��#�'
yXͺ�:�E%�@ͺ�:��Cl$3ͺ�:V�6hNIͺ�:���R$;�ͺ�:8�
��ͺ�:(
���ͺ�:A�"4�ͺ�:"� �:ͺ�:�/e(*ͺ�:E
{��ͺ�:w��5Nͺ�:�r�M�Iͺ�:��
|�K(
ͺ�:�N0�ͺ�:{�${e�ͺ�:ܾ��"�Jͺ�:���
r!ͺ�:�r*�ӂͺ�:��!�Xͺ�:�Z<%>�ͺ�:��
�
n3ͺ�:}l${g��ͺ�:A5�L�ͺ�:�!�a$�ͺ�:��E�Iͺ�:+s�<%�(
ͺ�:�N�, g|ͺ�:�"��'Yͺ�:�#�8�ͺ�:+�
)
��ͺ�:�!d
�6 ͺ�:�:����ͺ�:��';ͺ�:z;�!��ͺ�:, N!�ͺ�:LE%�- ��ͺ�:7 !?�ͺ�:i��sd�
ͺ�:�l$�{�Yͺ�:y���
��ͺ�:n��|��!ͺ�:��*�#��ͺ�:4$��
�	ͺ�:�;�|�	ͺ�:�OA��ͺ�:�;�l$.!ͺ�:R�6}hͺ�:T}�Q7 ͺ�:&�
0�
'�ͺ�:��;��E%ͺ�:#�
�ͺ�:�L��ͺ�:q}-���ͺ�:m� S$X�ͺ�:Q4Y��ͺ�:�|)�!ͺ�:$�@�^sͺ�:&�
�ͺ�:k�މ#�E%ͺ�:�!�{t�!ͺ�:��"8�ͺ�:�NHt9
ͺ�:�JS 
��ͺ�:�A�!�!ͺ�:j�
@)��!ͺ�:?�;���ͺ�:���#��ͺ�:�7 ���ͺ�:n��- H�"ͺ�:�(
Y=%��ͺ�:��
�1,)
ͺ�:�A�1�
ͺ�:��>. σͺ�:BJ�{d�ͺ�:I|��ͺ�:O�B�+m$ͺ�:��}��sͺ�:K~?V�ͺ�:tw�5<ͺ�:��-�"7Oͺ�:e�#�
��"ͺ�:&K�N
!ͺ�:�7 ��-�ͺ�:��"��
-�"ͺ�:�Y&��ͺ�:�G��ͺ�:n��O�ͺ�:^K��Kͺ�:���31tͺ�:.#Q��ͺ�:A�����ͺ�:�}�A%Bͺ�:����ͺ�:^�wJ�}ͺ�:F"o
!�
!ͺ�:��J;�ͺ�:�8�K|)
ͺ�:�<�?ͺ�:|m���
ͺ�:�G2�m$ͺ�:�`� 
��ͺ�:�<�h�!ͺ�:��#��#S�ͺ�:�tz. �Oͺ�:'�xG{ͺ�:{����ͺ�:�
!!�!�Bͺ�:��M���ͺ�:Z@��Bͺ�:4ߊ#Pͺ�:�jO8 ͺ�:�]�!0Cͺ�:AZa��<ͺ�:��`!ϒ#ͺ�:�m${�94ͺ�:C��!��ͺ�:��"��#�#ͺ�:w�#zc$"�ͺ�:k�!��q�ͺ�:6��)
ͺ�:1=�9Džͺ�:��
��!5�ͺ�:3!}*
ͺ�:l*
�tcPͺ�:�Pz)�Pͺ�:d�O��"ͺ�:���̱ͺ�:��O�tͺ�:��#B�
ͺ�:m!SBoZͺ�:n$�B>~ͺ�:#F%�B��ͺ�:5Q��!�ͺ�:�CT�j�ͺ�:�~��H�ͺ�:'L�"�*
ͺ�:uDu�ͺ�:=�v���ͺ�:��l8 ��ͺ�:���X�
ͺ�:��0��*
ͺ�:�~x��~ͺ�:U�k���ͺ�:'� �|RF%ͺ�:D�o!��!ͺ�:"pa��4ͺ�:W+
4}�Cͺ�:�4#F ��ͺ�:s�-e=ͺ�:IsPCDͺ�:3��#;�ͺ�:&�VP�ͺ�:eLC�Zͺ�:�Hu�F%ͺ�:���2��ͺ�:�u�:+ͺ�:�#/ �4ͺ�:�=��[ͺ�:�=�A�8 ͺ�:�u�!�!ͺ�:25��uͺ�:!���ͺ�:��U�+
ͺ�:�J�V�ͺ�:��Ac�ͺ�:ݫO>%Rͺ�:��!��#�+
ͺ�:�D�:��ͺ�:�F%�/ dQͺ�:�8 �}��
ͺ�:>)�"bͺ�:6vQBA>ͺ�:�
�/ �ͺ�:1�H�Dͺ�:d5s�Kͺ�:��ҩ�ͺ�:pK���5ͺ�:O�<|��ͺ�:�>d$ͺ�:�L�Pb[ͺ�:EG%[G%ͺ�:@�"�K�ͺ�:�LH�
�ͺ�:���F"<�
ͺ�:��s�*�ͺ�:����
5�ͺ�:���G%�!ͺ�:����[ͺ�:�]��ͺ�:?�$��k�ͺ�:�pɆͺ�:ď.,
ͺ�:!��=Fn$ͺ�:�!�
�ͺ�:��
h���ͺ�:v<;0��ͺ�:@�
��Kͺ�:v�9�
��ͺ�:Q���ͺ�:��G�
�ͺ�:;��ݓ#ͺ�:��I�ͺ�:v�"�4�ͺ�:,9 B\ͺ�:b��IEͺ�:�N#��ͺ�:NE��
ͺ�:����<Lͺ�:�f�ͺ�:a�'�p�ͺ�:;�w3�ͺ�:�(!�Lͺ�:"��;V�ͺ�:}�+�ͺ�:_\�,
ͺ�:-!�0 �n$ͺ�:`��� �Lͺ�:e9 ��i!ͺ�:��Pk�G%ͺ�:��
��ͺ�:�;$Z�ȇͺ�:t���>ͺ�:�
ͷ�,
ͺ�:����4-
ͺ�:6�#g���ͺ�:�
��ͺ�:W��k¬ͺ�:�!��(
ͺ�:��-
ͺ�:cv�W��ͺ�:��� ��"ͺ�:�v��
�Lͺ�:�vNMt�ͺ�:�M,��Mͺ�:����-
ͺ�:�=�]��ͺ�:��	
��ͺ�:��$eS�\ͺ�:���T�
ͺ�:�����ͺ�:%wkY�ͺ�:r�;�S�ͺ�:mh��!�n$ͺ�:�<F1Nͺ�:��$�R��ͺ�:�v��ͺ�:�%~���ͺ�:N.
[O#��ͺ�:[�/�
~�ͺ�:'�o�!f�ͺ�:�.
��o�#ͺ�:.�
x
Ŵͺ�::^Q� R�ͺ�:R���_wͺ�:Ȕ#~e$�5ͺ�:�E"CW�
ͺ�:� "
Po$ͺ�:���~?H%ͺ�:RM��
�"ͺ�:�Z�
ͺ�:)DT�!A�ͺ�:�!n^��ͺ�:���e$)ͺ�:��
�^�Qͺ�:a�"����ͺ�:%r?ͺ�:D�x�.�ͺ�:�,h?ͺ�:RC
�wͺ�:�9 ]C�9 ͺ�:χ�S��"ͺ�:6: �O#�ͺ�:֍ϋڀͺ�:�\D�Mͺ�:o: 6�Eͺ�:.�4���ͺ�:�#��7!ͺ�:�=�$��ͺ�:��(v��ͺ�:�k�&�ͺ�:��
gG6ͺ�:�o$�O#xͺ�:v�_R�ͺ�:�����o$ͺ�:;#�rͺ�:q�T�?ͺ�:�m��5Fͺ�:��"���ͺ�:}Q��
�!ͺ�:�-����ͺ�:��!��@ͺ�:GR�f$�ͺ�:�$��"/�"ͺ�:
��}�"ͺ�:�\��"��ͺ�:�.
N,�ͺ�:�6�!]ͺ�:ix�8�ͺ�:�F��ͺ�:6�M�ͺ�:QR;�Nͺ�:Fp$�|{p$ͺ�:��F��ͺ�:^@�k7�ͺ�:�x"��Rͺ�:�R�f$)ͺ�:�NP��6ͺ�:�y?%��ͺ�:��� D�#ͺ�:�: �,67ͺ�:'����!ͺ�:���u]ͺ�:�����ͺ�:�_s%�ͺ�:3l��ͺ�:�x���ͺ�:b�.}\�ͺ�:6�
E�p$ͺ�:y���H%ͺ�:E�b�
�Mͺ�:W ���ͺ�:1��: F�ͺ�:=y��Sͺ�:; �Xvyͺ�:%&���H%ͺ�:^; ��]ͺ�:
On��ͺ�:����7ͺ�:$^�"yͺ�:
�^Sͺ�:u�ڿ
��#ͺ�:j� KtGOͺ�:�� �R?�ͺ�:�u^^ͺ�:�g$�#ͺ�:���I��ͺ�:m�!���Fͺ�:��.�q�
ͺ�:�$6�ͺ�:�I��.
ͺ�:�eS0�ͺ�:��a?�#ͺ�:��#�4AI%ͺ�:�SJ��7ͺ�:���D8ͺ�:� ��!i�ͺ�:Y&�&;�ͺ�:s��*Nͺ�:���–#ͺ�:���N!ͺ�:��UD�ͺ�:���-�yͺ�:9��"�8ͺ�:+q$���Oͺ�:�Wg$%zͺ�:�!���
ͺ�:����!ͺ�:0	��ozͺ�:@�W���ͺ�:{%�
�zͺ�:���A/
ͺ�:��!�"��!ͺ�:!�T̳"ͺ�:Q��!�Sͺ�:�/
 ���ͺ�:���_�; ͺ�:Y�[��Fͺ�:GT�� {ͺ�:dz���ͺ�:�@����ͺ�:��)
��!ͺ�:Viq�ͺ�:!G���"ͺ�:.�f	�^ͺ�:_N
��ͺ�:<TP�
VGͺ�:J$�!ͺ�:�
 ڶͺ�:i���$�I%ͺ�:z�"�
*�ͺ�:�-%���
ͺ�:z���"�	ͺ�:�k���8ͺ�:�����ͺ�:"
[GqTͺ�:Vq$�tNͺ�:q�"�7�ͺ�:9�Y!;9ͺ�:�9�Sf�ͺ�:	�m��ͺ�:[{=�#�ͺ�:SfJh�
ͺ�:���?%��"ͺ�:��Nm�; ͺ�:	���G�ͺ�:��"l�!�ͺ�:�	��w�ͺ�:t��h$��ͺ�:҃�m.�ͺ�:��m��ͺ�:�T���Tͺ�:����ͺ�:-�"����ͺ�:?_��>�ͺ�:p���
�Nͺ�:
�>!�/
ͺ�:�O%��ͺ�:?�!E�@ͺ�:�./B�ͺ�:PU3~�#ͺ�:�{��
	Pͺ�:�GS�$!ͺ�:#O��[Oͺ�:�q$�J>�ͺ�:�!��!;< ͺ�:5
����
ͺ�:����z_ͺ�:�q$����ͺ�:�q$� Aͺ�:-�wAD�ͺ�:e�"��dPͺ�:j��Pb�ͺ�:���P��ͺ�:p
�O�Oͺ�:j�#�U�ͺ�:	�

�;�ͺ�:�23r$ͺ�:v0���_ͺ�:0

���"ͺ�:ѣ�<FPͺ�:�4o��Aͺ�:�,K�
ͺ�:�+5��r$ͺ�:����!ͺ�:�I%�#
k�ͺ�:�r$y�Gͺ�:��2 D�ͺ�:,`�E��ͺ�:AHp���ͺ�:�P�@%�Pͺ�:���s$ͺ�:��t�
ͺ�:�s�!(Qͺ�:r�!�G6ͺ�:�< Ly0�ͺ�:� H��#ͺ�:��"��!�Qͺ�:_s$��
�ͺ�:.���#��ͺ�:��y�9ͺ�:�UI7ͺ�:���yHͺ�:�{;�!�< ͺ�:j``4�ͺ�:��wͺ�:|����ͺ�:�s$�
�ͺ�:��`�ͺ�:�s$�E��#ͺ�:J%"��Hͺ�:�`�n�`ͺ�:��!y!�Qͺ�:��X�#'�ͺ�:S0
H�"ͺ�:��:��ͺ�:��$����ͺ�:9��ڢͺ�::Q��
&Rͺ�:���� ͺ�:!�!g ͺ�:s�t$�ͺ�:C��
�0
ͺ�:����
ͺ�:�+��$#aͺ�:�4{�>|ͺ�:<��u#v�ͺ�:�k�}$= ͺ�:A�#�Ct$ͺ�:1" }$
��#ͺ�:�����ͺ�:1��A%�ͺ�:����ͺ�:��!���Uͺ�:���F��ͺ�:E��!oaͺ�:�R���ͺ�:?��71�"ͺ�:�a�/Iͺ�:�!kzzQͺ�:�RA��|ͺ�:�6�#�ͺ�:�|q�#�Qͺ�:�R8q�ͺ�:~P3 B�!ͺ�:A!�$
�0
ͺ�:�t$�z��!ͺ�:p!��}ͺ�:��JV��"ͺ�:�t$�>I�
ͺ�:JJ%�i$˱ͺ�:
�n	!�#ͺ�:	Bm�
�aͺ�:^�,j$9�ͺ�:�
��n�ͺ�:QV��!ͺ�:��FV�ͺ�:"SY8Gbͺ�:�bpŹͺ�:�!�j$��ͺ�:��b3�ͺ�:c����ͺ�:׶"��>cͺ�:J?}���ͺ�:61
^���ͺ�:Y�8Bͺ�:�7�8�#ͺ�:o1
�cX:ͺ�:����ͺ�:5�
�"ޅͺ�:0V��Vͺ�:k$ ���#ͺ�:.��$�cͺ�:~J%7.�:ͺ�:f" "=L�ͺ�:�8��*!ͺ�:��
��-dͺ�:a�3 �!ͺ�:��
Q%
�ͺ�:���K2�ͺ�:�dN���ͺ�:�d��5�ͺ�:O�"�V�1
ͺ�:<��B%Qͺ�:|= �V��ͺ�:� $�r�ͺ�:r}��t$ͺ�:~IK�ţͺ�:2
�>��"ͺ�:�B��}ͺ�:���3 >�ͺ�:�dQp�Vͺ�:�= �z�ͺ�:VSs��!ͺ�:A2
����ͺ�:�����ͺ�:����#ͺ�:'Wu1߆ͺ�:�B��ͺ�:�= X�)�ͺ�:����ͺ�:C�Vy2
ͺ�:Ou$`4 ��ͺ�:�J%��2
ͺ�:<��j$1�ͺ�:�S��"�ͺ�:�I����ͺ�:
;_WE�#ͺ�:���t�ͺ�:��|�g�ͺ�:)!�?��
ͺ�:��#�q!~ͺ�:t�,eͺ�:_~��!�~ͺ�:/3
��
> ͺ�:1@~7�ͺ�:`!"R3�ͺ�:�� �}5�
ͺ�:��
G�ͺ�:Z�S�܋ͺ�:��r> ͺ�:,�"�.%�Sͺ�:~�8�J%ͺ�:%�0��ͺ�:N��� ͺ�:���
D!ͺ�:.�ҁͺ�:-��GK%ͺ�:���lͺ�:ɺIE;ͺ�:ݚ#q��ͺ�:
�1Tͺ�:�1�> ͺ�:^����"ͺ�:�K%%�bTͺ�:�u$:q? ͺ�:�#&
�Tͺ�:<�eW�ͺ�:\RD%\�ͺ�:�;-|��ͺ�:U���ͺ�:�<�K�#ͺ�:=C��#6�"ͺ�:#!��
�Wͺ�:�K%�I[�ͺ�:��#�"tCͺ�:��"5 ��"ͺ�:p�O�"�!ͺ�:a�-k$]!ͺ�:�IX5 geͺ�:�u$�2��ͺ�:�R؃h�!ͺ�:p�J��!ͺ�:����u$ͺ�: J�C�"ͺ�:�F
}�
ͺ�:>U����ͺ�:)�
��l3
ͺ�:�!�,��ͺ�:���WJͺ�::��!4ͺ�:;���b�"ͺ�:k� �;Jv$ͺ�:N�����"ͺ�:�#	�/�ͺ����Tͺ��%���ͺ܌U�ͺ�D )�ͺ��C�x
�C���{
�C#��\tot}t�C\�$�Ep���C�C���V_ot���C��r��u��d!�CZ���ot�e!�C�t	�#��<f!�C��|

7ot���C�e�%�7���C� 5'�r�C��LG��r�C�R[�� � �C|U�r|� �CD��
]Gly�C�*�%<�
ly�Cޏ��
f��#�CwG�r�E$�#�C![�B���C\H@����CxI)f�e#�C�y��y�e#�C��"��yF�CE���C�#�C7w���$la�C-}
��p��C)�$�8�
k� �C�D�f!I7"�\�C��7d}
�\�C�LO��/�CǵP���D���CE�c#�f!���C 700X�#�CG�"��k#X�#�C���u#�#�Cg!
�!5FJFR7"�Cw��;"DZ�Z��CU���|�#�a�CV�\�g!�a�C;�]�a5� �C��E�
�7V�C� j}
ٶ�g!�C�I��$� ��C���� � ���C�7t}
;D�#�CY�)�Dla�C)#R�
U���C0
��D}}
�C�����CV��jr�Cb��ݶ�C�7"��y�|��C�F�bwl��g!�CW�
�8"�l�c0�C�
�
l��E�C/�@�J�C  �%���3
�C�����W#�C�	�/02�C�yt|?f#2�C�%�%��L$�Cܸ!�%?f#��"�CQ Kf#g�ɭ�C��ۚg�XJ�C�%��Sf#���C.�"��Sf#���C�Ȏ"���� ]�CN�7�#�]�C���
���y�C�y��$��`G�C��
�]��Sf�C��s�?0�C�JL�H�[\�C�\�� H��}
�Co
��$�#c�C�טLF+z�C�*H���C��1�
H��]�C�0 H����C�f�H�+��C�%�H��$�C�%;^<�
u��$�C��3�J�
:U$�Cx�I]9$�0�C(
E!9$�7�CdF\�
���� �C�F;�#��G�C�:�JF��Cjc~
S�
1�C��a]>�S�y#�Cd��uY�X��y#�Cb�Z
v�S�@� �C9�$ `!X���!�C�k� ��!S�Ҟ"�C2��2b�X����CA��Eٶ,�Cl�
[�8���C�c��8n��C���cN����Czn��~z6�C��n�8"n��Ch��q�lr=�C�l�
�lhq�C@�����#d�C�����n1�Cѱ�
#��l�C�]t3#�X��C���c���
�CV�
�z.9"�C����S���C-~
�ES����C��
7�S��F�CR^�S�[~
�C^�0S���Cg��GS�|~
�C���#S�����/�a�
h ��a�
%U�{�c[a�
V��4�!�la�
�<$D[��ea�
�q[
�0�!a�
5-���a�
��
��~Ya�
�>$�s�v$I�a�
4
�JL%%!a�
�
�s�!
ta�
n!�{G<�a�
N Q�Mb�<"�Y~ S�$���%�YӈO#��.��Yb�� I�EW�Y�k�
Zh�"�Y�Wpb����YOZ��
��"�Yu mm���Y� �":P#�Y`�_�!U���Y��@$��Y�
>�� :�"�Y@�P#�b�Y�� �Y<���s`@�Y�P# ��Y����Y��
�@$+'Bc�Y\�4���Y��aX��Y�%6 й�Yi�`'`_��Y�6�
"j!�c�Yw  �u�"� �Y���A$�Y!N�
� �Y�L�G���Yk[�C5��$�Y��f�h;x��Y��ؠ�G��
�Y��������
�Yd
�!Fj!���Y"
P�#��d�Y{'w����� �Y�`6��="���Y������="�Y�a��MQ�$�YK ���e/l�Y/l�M�:� �Y�% f"��Y�!>"��n"�YuA$��j!���Y��=���$�Y]��*��Y
��p#���Y��!y��ebE�$=d�L�ZE�$�� ؁
��E�$XX�
A�"E�$5�@�VE�$�X�J�E�$�6� �k E�$M
���E�$^���`"E�$d��H����E�$�m�L�$�E�$!q�*�M�E�$�P#=���
�E�$.�M�[dE�$���l!K�E�$��A�F�"E�$	W�<f͍E�$�� ��h
-!E�$n� �6�#�mE�$�A$�����"E�$y�	<��E�$��Vt��
E�$t�'��E�$(L%��E�$r����E�$��Z��E�${�0�!1�E�$�Jb��v$E�$Λ#��#��E�$W#V�
w�E�$K7�#ހE�$	w$��K�E�$�#�W��
E�$6KPL%�tE�$�m[�qE�$��>��E�$r�#�h�E�$ύ\5
��#E�$��0X��E�$�#����
E�$erE<u�"E�$����#E�$6�[dBw$E�$m? j�dKE�$��]=1�E�$�#k�!�dE�$��$$4�E�$OlX*�E�$Y�5
�XE�$x�}��E�$��
�"E�$�r[���!E�$������!E�$�t���L%E�$��_$��E�$��ۻ(YE�$�? �!r�
E�$�L%"M%�!E�$�����E�$�r�=?�!E�$��!��
eE�$�5
Ȩ�YE�$]M%�g�gE�$ƂN�2�E�$��%�
�E�$�$��=�E�$��!��E�$����E�$�$��#6
E�$�? 9�E�$��#�K̈E�$�<��ʼE�$�K�M%�E�$�? �"�!E�$%@ s�sE�$Z�=?�E�$��Ks4�E�$=�%��E�$�s����!E�$H�s�tE�$f!P�"E�$z��vw$E�$�s��=ZE�$+�����E�$���" HeE�$A��!X�E�$,h�[FE�$TuN%6�E�$�M%��"�eE�$���N%E�$'fɳ�uE�$_���ZE�$��!L`�
E�$�sn��E�$����
[E�$�t��E�$�vs[E�$�!H@ dhE�$���w$F6
E�$§�� ��E�$4StE�$�h�[#�!E�$e�!	i*�E�$Ci%�x�E�$ni�i>vE�$��	��<E�$l���
!E�$܉�%
�E�$��!"x$��E�$�%�vҽE�$o��"��E�$a�Ɏ��!E�$v!<L>�E�$�!�6
�#E�$��1�"U�#E�$�6
�t�E�$� �!=E�$efʞ#��E�$]x$��E�$17
E�$D��#�iE�$մ�=�7
E�$�x$p�
�
E�$�t��x$E�$���" ��E�$���.[�!E�$%�F�!�E�$�fA�#uE�${�#٨ �E�$)\~Vh�E�$���v}N%E�$���#xLE�$��3�4�E�$G�8%y$E�$D�O�!��
E�$Oy$ߟ#��E�$�%5�@ E�$wq��fE�$_�yy$GjE�$Ϗl�"�#E�$8
�BuE�$n��LME�$ڊ��S�E�$;8
����E�$���
��"E�$�N%�=X\E�$g�y$�@ E�$k8
�!�y$E�$fA Cg��E�$��S$�ME�$��
��!sE�$���
��E�$�\oz$
�E�$~g��uE�$�gLM�E�$�x��E�$�8
U!�gE�$��c>hE�$��l
*�E�$��.��E�$��#(*�E�$8/�0&E�$��vY�
E�$
��
O%E�$����#�=E�$,���
n�E�$���M�&E�$K��\*�E�$��,]*�E�$�&@� ��!E�$��#9���E�$CvV>E�$��!��
E�$Fwf]c�
E�$!�#&T$�hE�$�8
>�"E�$�j�]��E�$NH�#qwE�$��##�
�]E�$��_�$U�"E�$ �G�!7�E�$gO%��wE�$�A ��
��"E�$!�R#�uE�$�����"E�$Y�GE�uE�$'-�
DŽE�$u���
�O%E�$��j��E�$I�ŵ��E�$���)�E�$��D'�E�$����uE�$
�t�uE�$�
`mCE�$��S���#E�$%�#O�#'E�$���&^E�$X�9
$�E�$J!vv��E�$B���I�"E�$��#�A ��!E�$�z$��#�E�$A9
~.�>E�$�I�B�E�$�#~v�#E�$l�w ��E�$�'� ��E�$B P+E�$��"7���E�$�+E�$l�|� �^E�$���/�
E�$����E�$&�?*(E�$x��!Ͼ"E�$Dx��X?E�$'/B +E�$��D�+E�$|x_��E�$��>_�E�$3�"�v:E�$�jB �!E�$8��ZE�$�B z_��E�$Z��B )�E�$m!�v�#E�$����vE�$�h��.NE�$�?�$�E�$�I�H�E�$��WY�z$E�$+w$� �xE�$a��!7�E�$��#h��?E�$�!+d7�E�$�ߣ#��
E�$����!E�$X�"�!�hE�$(�͐GiE�$�i�B �E�$�6!XwE�$�O%?7�E�$��
7�E�$^��i�_E�$�xܶ
yE�$z��~�E�$k�'�E�$/@�w{$E�$�i��Y@E�$���T$�(E�$ц�{$�yE�$�wM`�E�$�)��E�$!P%��xE�$m9
���xE�$!z�i��E�$�`n�!E�$|��{$%�"E�$;�w:7�E�$���#%�"E�$�`���E�$�9C �`E�$�9
uC \zE�$pN����"E�$����
GE�$&��C �@E�$�a�%�"E�$2j�%�"E�$�9
2+�E�$S�a�:
E�$�a��DkE�$�"}��NE�$2|$�k��E�$u�Pj�:
E�$l8��E�$D�hj�E�$q!%�I�"E�$�����:
E�$M��\YAE�$T�#C)8�E�$��AO9E�$��!�C 'D E�$(�!�!�vE�$��"�/%��E�$��Z��vE�$y��E�$��!_�E�$�z��؇E�$ʍ�j.�E�$�a�"G;
E�$�
Ny&bE�$�y���vE�$��D �vE�$�t�!�E�$��Ԫ�E�$�J�
��
E�$VP%\|$�E�$�y�
|OE�$����!��E�$"{s�X�!E�$���;
i�E�$��P%�|$E�$��#MaE�$1�e�b�E�$���D C�
E�$9����E�$=��A�E�$ϒ9l�E�$�;
�5~)E�$�"E �E�$W��

!E�$ȫ����#E�$���yp�E�$�b���E�$G�tl��E�$��-<
��!E�$h<
�|$�E�$�A��E�$���:K{E�$���#�E�$�P%!z�{E�$�#�n�E�$��)��E�$�jH��E�$Ŏ�
0�E�$/Bu!�<
E�$gt���
E�$��M\�E�$�
*]O�E�$�?E TBE�$@�
$}$�!E�$���b��E�$��g|�,E�${�
�m_}$E�$�l�|�,E�$����"�E�$\zk�E�$��5�V�#E�$��!����
E�$��!q� *E�$h�@k�E�$x�\��,E�$���*�,E�$�P%U�
��E�$B�"�*�}$E�$�!��zE�$Y~$�OzE E�$�4P�|E�$!��ϥ#E�$���2(�#E�$�l�!��E�$�<
h�"��E�$^���	F E�$<�mE�$+{O�E�$���:�E�$��#oP1�
E�$\���
;�E�$��#���~$E�$�B{k;�E�$��E�#�E�$��"I+��E�$�U{	Q%E�$
�P�"E�$A�����E�$�BC�E�$���;�E�$�
�!;�E�$�!�n� E�$� ε!���!E�$�b"I� E�$�Q#����E�$zl��lF E�$�:-C	<xQ%E�$�:bCVt~�E�$�:_}'�E�"E�$�:֏���E�$�:�� ���E�$�:w�Z���E�$�:\�
0�!�E�$�:�r!b���E�$�:���#��E�$�:��V�
=
E�$�:~7�#��E�$�:�k��:�E�$�:,��W��E�$�:�r!PL%��E�$�:r.m[�E�$�:s�>��
E�$�:�h�h�#E�$�:�\5
�+E�$�:M�!0Xp�E�$�:�F ��֐E�$�:�PE<U�#E�$�:?��!�E�$�:�C[d��"E�$�:?�j�G E�$�:� ]=b�
E�$�:�{k�!�E�$�:4�$$��E�$�:C�lXۊE�$�:9��5
�+E�$�:I}Y�E�$�:U��
=,E�$�:T�
[�}=
E�$�:�Y����#E�$�:�-��eG E�$�:�C_$
cE�$�:�ۻ�G E�$�:��
�!mE�$�:3�"M%o�E�$�:_x����E�$�:�.%�=1QE�$�:���
�"E�$�:D!Ȩ��E�$�:���g�Q%E�$�:�=
N��,E�$�:C�%�
��!E�$�:d�����E�$�:�����!E�$�:GŽ�{E�$�:v���#DE�$�:ަ90R%E�$�:im�K(H E�$�:X����kE�$�:�D�M%��E�$�:~R%��E�$�:��
s��QE�$�:���=l�E�$�:�"KsmcE�$�:��
%�QE�$�:aK���mE�$�:h�!s��E�$�:��f!_�"E�$�:�!�RlE�$�:�m����
E�$�:
Z���}E�$�:���" ��!E�$�:���!��E�$�:T9�[5nE�$�:�N%�~$E�$�:��"-E�$�:>
�f-E�$�:�Dɳ��E�$�:Q$��E�$�:F�L�}E�$�:�nn�T~E�$�:�c��
8RE�$�:��t:�E�$�:��v�E�$�:a>
H@ N�E�$�:��w$��E�$�:�>
�� �cE�$�:��4�#E�$�:p��[�~E�$�:��	i��"E�$�:�R%��nE�$�:j�#�i��#E�$�:2o	���E�$�:�R%��
q�E�$�:\�
�%�"E�$�:E"x$G�E�$�:so�v0!E�$�:m�"��".�E�$�:�Ɏ�RE�$�:'�<L��E�$�:���6
�>
E�$�:D1�"-|E�$�:��t�E�$�:�! �!�!E�$�:@Eʞ#EE�$�:�O�!E�$�:�9�R%E�$�:?�#Q�E�$�:���=��E�$�:�p���E�$�:A�M�E�$�:N/%�" �E�$�:S��.�EE�$�:�BF�EE�$�:�ZA�#2FE�$�:3$٨ ��E�$�:��
~VĭE�$�:p�ve�E�$�:��!��#�oE�$�:�$3��!E�$�:�s!8rE�$�:�� O�!�E�$�:� ߟ#�E�$�:m5�E�$�:��"q�pE�$�:��yy$AS%E�$�:��"l�"��E�$�: ���E�$�:8d�L~FE�$�:�$���#E�$�:tp���E�$�:�X�
S!E�$�:���=!SE�$�:���y$�|E�$�:��!�E�$�:�!Cg+�E�$�:Q��S$-?
E�$�:����!�pE�$�:W��
f�E�$�:�Coz$��E�$�:��c�#E�$�:7�L(�E�$�:�x��E�$�:|H U!�H E�$�:'�"c+�E�$�:'�l
��E�$�:��!�.ՋE�$�:L�(�|E�$�:X/��-E�$�:Qv�lE�$�:Ē�M}E�$�:q��#��#E�$�:���
zdE�$�:yS�M�
E�$�:F��\8E�$�:�},]
�E�$�:��@� �E�$�:��9��!E�$�:=�Cvj�"E�$�:m�!5!E�$�:�!f]��!E�$�:�&T$��E�$�:
�!>�FE�$�:�$�]H�!E�$�:p�H�#�E�$�:$#�
h�E�$�:z�_�$��!E�$�:�?
G�!�dE�$�:��!�,�#E�$�:����
��E�$�:�H"�R#�$E�$�:\� �E�$�:��$GE.I E�$�:�9-�
��#E�$�:D���
�E�$�:;��jq�E�$�:�ŵ��#E�$�:�}��7.E�$�:;GD'm�$E�$�:j�#��k�E�$�:��t��#E�$�:>�`m=�E�$�:XmS��.E�$�:�O�#�GE�$�:?�
���E�$�:�9
��E�$�:H�!vv�E�$�:j���ր$E�$�:G\�A G�!E�$�:�H"��#-�$E�$�:
�~.��"E�$�:�SI�|�$E�$�:�#~v��E�$�:T�#w �!E�$�:T� �"E�$�:�Pہ$E�$�:�T$7�0 !E�$�:��S%E�$�:�|� }�"E�$�:@T��E�$�:��!��E�$�:�?�E�$�:T��!��!E�$�:�S%��5�$E�$�:Oe/B ��E�$�:��D�y�
E�$�:٪_��E�$�:7�>_��E�$�:f��v� !E�$�:�.jB ��E�$�:���//E�$�:x�z_�GE�$�:�m�B F�!E�$�:Q����!E�$�:C���&HE�$�:!I"��E�$�:Ȗ�$�?
E�$�:��!I���E�$�:BT%WYpHE�$�:�e$� �TE�$�:k��!�eE�$�:��$h��E�$�:J +d�E�$�:Xqߣ#��E�$�:�
��I E�$�:�T%�!�!E�$�:$U͐�I E�$�:�
�B IJ E�$�:�J 6!~/E�$�:�?=�E�$�:Y�

kE�$�:F:�i��E�$�:gUܶ�E�$�:����E�$�:��Z�E�$�:I~�wU�E�$�:�>��ړE�$�:r��T$��#E�$�:a��{$�E�$�:��M`��E�$�:�),�E�$�:J���-�E�$�:'�����E�$�:$ �i�E�$�:�~�`�UE�$�:��{$�!E�$�:��w:l�E�$�:��#��E�$�:���v�E�$�:f9C ��E�$�:+�uC �T%E�$�:� !��E�$�:l ��
��"E�$�:F��C � E�$�:��!���E�$�:���[�E�$�:ɧ2ق$E�$�:�a�E�E�$�:�����mE�$�:/�}�"!E�$�:*�$�kP�"E�$�:��PjnE�$�:;
8�*U%E�$�:�hj��
E�$�: %���E�$�:-z���HE�$�:s��\,IE�$�:��C)�E�$�:��AO�E�$�:��"�C ��"E�$�:��!knE�$�:�/%�/%l�E�$�:>�
Z��
E�$�:���`fE�$�:a�$�!�#E�$�:����;�E�$�:���j��E�$�:��"ԔE�$�:�J NyU�#E�$�:�U%��!@
E�$�:y��D C�
E�$�:]t�!A�E�$�:1!!Ԫ�U%E�$�:��J�
�qE�$�:��\|$��
E�$�:7K �
��E�$�:>2��!tE�$�:h�
s�jK E�$�:L��;
K�"E�$�:�$�P%+�E�$�:�I"M��$E�$�:�ne�y�E�$�:!��D ρE�$�:0����E�$�:�K �A��E�$�:z�9lP�E�$�:�0%�5�E�$�:�E �fE�$�:D��
!gE�$�:��"���K E�$�:�!�y�E�$�:c�!��E�$�:!tlM�E�$�:�g-<
�IE�$�:S��|$�E�$�:A���@
E�$�:�:b�E�$�:�g�#��E�$�:��!!z��E�$�:��!#���"E�$�:���)��E�$�:I$ H�טE�$�:?�
��E�$�:��u!��E�$�:�t��/E�$�:KJ"M\d!!E�$�:o*]00E�$�:��
?E ��E�$�:\L $}$��$E�$�:���b�E�$�:
�g|,�E�$�:�:�mhE�$�:[u!�|��E�$�:y��"ghE�$�:kkR�E�$�:�~5��@
E�$�:!�!����E�$�:Hoq�4�"E�$�:��@k�L E�$�:V\�D�
E�$�:p��*!E�$�:��U�
'E�$�:��*�E�$�:��!��oE�$�:��"�O԰E�$�:z�4P�!E�$�:�����qE�$�:���2�IE�$�:-$�!�E�$�:��h�"7�E�$�:E{����E�$�:����E�$�:��{�0E�$�:c�$��-�E�$�:��oP:�E�$�:���
�oE�$�:�t��HJE�$�:��{kEpE�$�:�x#E�CVE�$�:KA
I+V%E�$�:�hU{l�!E�$�:���P��E�$�:*���t�E�$�:iC��E�$�:Wi��رE�$�:xV%�!��#E�$�:$?"�#��
E�$T�) �<hfE�$hf�EfE�$�R#�<��$E�$��$�[i��E�$ދ�J�#E�$�#�<!�$j�E�$��8���pE�$�[����?�E�$����
E�$�C$��D#�E�$}��� -�#kNE�$9�
h#��ނE�$ނƔ,nU#E�$E�$o��9E�$Ʋ�� "$t�E�$������E�$�Pn,��SE�$���N����E�$E���q$E�$*��!�"#%�#>?Vn�%��U$�
Ɯ"%��0%0~k%��
z���%.�1�"@0�5%�5ݯ�k�� %�� ����7�%��$�ML�%��"�I�%B�{�`�%a��6 2�%���r�J%�"O���%!��k�%�!!�'
�%
E%��
%tCl$�A
%��6h��%�V��
�%t��|�p%�V%*�#ނ%�A
��
U�%ď�|�i%�!!OA""!%5"�l$�"!%#�6}�V%%�"!}��V%��!0�
�#%K;���%M �OPW%%&#!���%���3�%WQ��%)W���i%��A�%�p�T�%�
wJԲ%��o
!R#!%Tj�J��%/��K�%J�T�VW%����3%�W%�"1q%�0Du�W%ӄ$v�u�%z#!l8 �%$K�=r%9�"0���!%O�x�Y%�k�E�%L��|7M %��$G%'�$%B��KgM %e�"H�J�%���F"�r%�M s��W%%|j��
l"%�#!�G%J�%����v�%VK]��%�r��s%���%���5s%���%{K�W%�%�q� H�#%k���
�W%�KNMe�%�q,�w�$%�#!��t�%,X%�]*�%��"	
*�%�eS�%Ε6�%��4��M %FX���s%,N �$Ã%%�(v
$!%��#k�V�%�Kg�%���O#WN %XX%_�u%x����u%L�.�%��: E�%|X��U$!%��X%%l����j%�����
%��n�s%�j���%�N �")k%Л�+%��ڿ
+%�qKtzI%��z���%����:�%��)
��%{�iq&t%?����X%%�$!f	��%��N
��"%|kP�
�$!%˄$�!7�%(�
 7�%ː��$�%P�
��մ%!�����%~�
���X%%�X��"%��wAR�%̅$��-�%*��P�X%%!�P�%��$�O%�"%��U%�"%|
�nU%�N X�#XY%U�!H��%��:*�$%A#���%��$��k%=���
\�%1��Ů#%D%!!l%�N t$�v%Ι�
�v%e�����%�A
�bK�%O1����%	�"��ҵ%g�}�/�#%�Y^��Y%B
��
%†$7�6B
%���ct%!%z1��%4��"�%���
 #%�_W��#%ll�F�%r|��$%���?�%CZ�q!��%p�,eKY%%��!��!:�"%�%!��
ct%�l~�,%�Y%"R�,%���}Q�%�l�C�%`�F
��!%%���7%M���[�%��,sZ%V���!%��!&!%��"���l%���;;�%f�$��;�%_�")�!�
J�w�\��J�ed*O��d�J�t�"o�#�YJ�'�D ��t�J�
MM%J�\����EJ��!��i$��J���?"�7X%J��o��$��t�J�
����J�;"Ś @L��J�� "��$�\J��2z�#�EJ����S��]J�cL_\�\J��gX)�J�w�	�"J���_\�\J�>�y�@"J�U��LX)�J��
	�"J��d$��J�>�`+�W�J�����$J@"�OJ�"�/�#J���r� h%�OJ�b���J�-#���J�J��%��J�J��gIJ�3P�L��J�)�
1�"
�J�8���aMJ��o��
��J�F�Zܔ:$J���S#^
�$J����"�
	#J���%S#�#��J��#X�\J����
����$J��o�#` oeJ�����J���J��h��#J�_�"�d��J������J���
3%�C$z)J�U�!%��m�J�V�$��J����A3Zm#J�{M��kJ�k��
��
2$J��C$��
�{c"J��$%�s�*J��#E9}l!J���
q9kpJ�\ZXp���AJ��dQPG�
b� J���#����J��
�ZC�AFJ�#�=�I�J�^]xMB�l!J��e��#�
4BJ�����!eJ��!_� A��J�˓{�)�J���@�]-/�J� E$���@"��J��FIIN��J��q�M A"Y�$J�`#���\�J��
��mNf�J��$�f�
�MJ�*	C?&;kJ�7�
��k��J����
��+�
J���� ԇ1�J�O�
���J�ҵ��:�J�
,�J�d&� ��J�ҥ�$%CJ�%C�����J�����N��J�����&�pJ��pX"�8�$J��e�9�l!DB"J��#�Sf�J����N�P�NJ��"F�14OJ���$�y�J�k�"�*�Z��J�h%��Z�]J�a��B"&�#P�J�� �:Yem!J���qz� �
J����%�
���o��:�%�
����o��:-�9�
g� ��o��:!9�
w6��o��:
$q�����:�QgP����:I~Q�����
�&���:�$a�#W��lq����:��
������:��
�����:s��������:�B
�
i����B
�:��
Ar���B
�:g��������B
�:s��������B
�:�1�
8L���B
�:P�����
����B
�:P���k#~��r���:O�~��r���#�:"��Y%k#~�l&!���#�:"��Y%M$=��U$�
�!��:>7����% ��:�1���1����:�4mR�_���:�����!��:�Z�#�
�Ϸ��:f�h��L�����:�B
�i�����:��#(O ��Z��:�1��!�������:�=��o�����:�y�
6v�A���:���&!�}���:��������:����L����:x���� ��
��:2�!���z#��:3q��
�U-2�:J�#���"�U-2�:�\?�U-2�:$��fv�U-2�:���
�U-2�:�������U-2�:��K��U-2�:��d^�3�U-2o��:d^1%�U-2o��:�Q�!

�U-2o��:�!V$�U-2o��:4�3j��U-2�:82ѴHm�U-2�:{#�
C��U-2�:{��^.��U-2�:�?Eo���U-2�:�"2�!�^�U�& �#�:[o2�!���U�& �#�:�#�v�U4m�:H_�2p_�U4m�:����o�U4m�:�$�v�m�U4m�:�C%�"�U4m�:�Z���#%�S4�U4m�:�Z��vo`$�
�U4m�:�!�$_�U4m�:�?��4�U4m�:F�!>U2�U4m�:���6���U4m�:*�AV$vo�6ږ�U4m�:*�AV$J� �& �o�U�-2�:�f� `��U�-2�:\�
�#
�U��-2�:����d��U|� -2�:!p�"K�U-�-2�:�_�_��U<>-2�:7pv�
���U��-2�:�"�4}��U�"-2�:��
w��U�#-2�:$�m4w�Um�!-2�:�!$�2�U!�-2�:h������U!�-2�:�"�?��U!�-2�:�67���U!�-2�:c?����U�?-2�:���0��U�-2�:g@�$`
�U�_-2�:��
��7$�U�& -2�:�$%>1%�U�1%-2�:�
�2�@�U�@-2�:�5%�m�UD�-2�:��8�"�#�U�1%-2�:��wZ��UA-2�:VV$3��U��-2�:��7����URp-2�:� 3%7�U�-2�:�
čdq�Ua�-2�:Q7@' ���UpO -2�:,�#V' ��!�U��-2�:����"�U'�"-2�:�&!�[���U�#-2�:�gm��UM�4�:ȟ�o7�U'34�:�
BB��U��4�:�w����Uڍ4�:�
z?�U#n4�:��
�i3�UK%4�:l' �!q`�U�4�:�^�H�#�U�4�:�4��q��U>�!4�:3%�' ��!�U�4�:��?���U��
4�:�`?����U�4�:6X$��#r��U��4�:��4� {��Ù#4�:�)�
�4�U]X$4�:���Y�
�U?3%4�:�' �3�U��4�:
�e�#F��U54�:�`�%a�Un$4�:cn��I
�U��#4�:
��5��U�q4�:�����%�UJ�4�:5����!�U�64�:`a����Ud� 4�:�%&���U75%4�:o�&�'��U�?4�:P���7��UqB4�:$&L&y��U�A4�:N��6%�a�Uծ4�:N�"6s�Y$�U�n4�:FZ$v�"���Ufs4�:8�!V��3�UnZ$4�:���
�3�U�74�:��"$��3�U��4�:�w�o�3�U|&4�:�N&s� �U��4�:�'K8D�U�t4�:]�
���Uv�4�:�6\$#p�U{84�:.��f\$�U94�::fC�'�UsD4�:���\$~&�UF4�:��

���#�U�G4�:)��_(�Ua�!4�:��
"��UH44�:Ϡ"��� �UC:4�:�' ��"U��U( 4�:�#_v�a�U=�4�:��
<);�U_�4�:̣�p5<�U�v4�:&�"7%�p�U�(4�:��M� ]I�Uϰ4�:�;_b�w�U�*4�:��-q}� �UP�4�:���\$&]$�U�+4�:�b�ID��U�<4�:�>�5��!�U�4�:��<
� �U67%4�:��) ���U	4�:T�{��8%�U��4�:]q�&�b�U~�4�:G�޽(c�U�4�:�Qxϲ�Un�4�:+6Bx�I�U�4�:ɽ!9%;� �U_J4�:N�"��;�U�) 4�:S6h-�
�U��!4�:�����Uv�"�� �:*�#9��]$�Ul
�� �:'%'��
�UD�#�� �:�
�cO��Urx�� �:yA�L��U	��� �:e�
�c!��U�
�� �:k+ �x1�#�U՝��:�K����U&���:��,yk��U�>��:�������U�C��:�� 
�c�U�C��:�x���U���:�>
�U���:7y�+ �-�UWy��:оN�Q9%�U��
��:J{6�q�Uk��:�x�-�y�Ud��:'zHz���U�]$��:�K�x�9%�U���:�"L5��U
?��:�x�6��US�!��:<�c_$6��U����:�"� ��U&
��:s�� x_$�Uy�#��:}��qp��U�
��:+�"Y�
�U|���:�De�	� �U.��:e�
����U2���:U� ��l� �U<<��:n�ge�_$�U�� ��:��%L�
�U���:�1'!1'!�U�� ��:y�q2y�Ud���:	�:L�+ �U� ��:�+ gm���Uc���:!��Y�
�Uў��:��k�:'�U��#��:��
�����U~���:�#	���U�B
��:�[g��1�U�Y%��:/���$�U���:��
{y�9%�U���:��e�D
�U�q��:�?��
�?�U�?��:K�t
�#�U`$��:r�+ �� �UEL�L�:A�"gm��U�
�& �:h��+ ��
�U� �& �:7r]z�<�U6��& �::%�L���Uo�& �:�L�����U��& �:����)��U��& �:s��6)��UM�#�& �:�.:��
�U���& �:o'����U��& �:�y��!Sr�U���& �:Q���T�"�U���& �:E!��"�U��& �:S���}e�U�<�& �:*`$xE
�U`!�& �:�?V�r�U���& �:0:%E`$��!�U|��& �:�.Z`$G:%�U�:%�& �:�����U���u���
�|��� �E$Q�`f#��f#�l�
���fv�#��#u��n��r�������NPo!+%����$���"�g�;[�w�1�+��lm&���"���]�&��E$4D������$���#{(�-g#�� S�(�!� �;a���
�-sP�
'���e m|����
Ko!�������#�����
�;+s֏"���u�����-�`R�(�M�����
�J��sUDvDX��;Q������� R�T#	^�f%���D���g#��#|�ES�H��H�}�ͺB����ͺ�̑Tmx�ͺo�~�
/Q M2ͺ�����SQ ͺ��zRM8ͺ����9�
6�ͺ;V��%�Y
ͺ��]�
�"}Q ͺ��h;�R�ͺ�t��$��$�~
ͺ�%��^HͺF��#&���$ͺ�2���R�ͺk�9��$��ͺ�Q�~
�oͺ�� ��#��ͺ;�ȕ 0G�~
ͺ�
�J��ͺj_��$�$��ͺ���� I��ͺ8i��~M��
ͺ�$�$n�4�ͺE�V�#�z�9"ͺ^�p
*�$��ͺ���RC
�Rͺ���[��\%ͺ����M�ͺ�����"ͺ\��_`Bͺq�
8���jͺ�_�V��ͺ��uN#�Kͺ���Is�ͺE��%@Hjsͺ�Mf`�#[�ͺ��8$����ͺ��(�@��2ͺ�%�8$MNͰ!ͺ{\.�9��ͺt�
[~'.�ͺ���#�[6�ͺÁ��ݵͺ{�
G�o�$�ͺn����"1C
ͺgu�='�#ͺ�C
h���ͺ��9�
��ͺ�
;0�'!ͺ�u�s��ͺ2��Q X�"ͺP���
ͺ&�;�8�ͺZ��RR ͺ%t��!u�ͺ\j�Ӆͺ��
1��C
ͺ����Q�ͺm���
��ͺ��!�f$,(!ͺ�#��)\ͺqN�m�
ͺ����pͺ����(!ͺy�
����ͺ�)!��"��ͺ�����!ͺ
��<��
ͺ��5��ͺ��o��ͺ�#�\��"ͺPtkt�mͺ�R {�8�#ͺ�$�}�ͺ�D
��ͺ�N#��$ͺ	�"���$ͺ1Z%3�ߘͺ��I2�ͺ�1 6��$ͺ�u���D
ͺG��$^�ͺ9�CE
։$ͺ��0�ͺ5�"t4p��ͺ����
��O�ͺ�Vx�&4_`ͺ/�<"HZ%�uͺ����%<�ͺ���#%]�#ͺU�v!o�Qͺ�u4�N^�"ͺ�3�
�N�
ͺ��"�v%�TͺO���f��ͺ��܌ن�ͺ�%��)!)�ͺc���FlZ%a�
T�!����4a�
+�$�}�Oa�
���{�
�Z%a�
�)!�4�� �"a�
�R D[2<�a�
��[
�%�� a�
��-���>a�
��Q]'va�
�%Q�Mb�<"�Y~ S�$9�Y�wO#t]9��Yn� I�EW�Y�k�
Zh�"�Y�Wpb����YOZ��
��"�Yy6u mm���Y&x� �]���YE
_�!��"���Y�x�(���Y�
>��]�'�Y/�#@�6�b�Y���� �Y<���+!��Y��
�@$G�$Bc�Y?y\�<F
��Y,!��qF
��Yl��%��й�Yi�`'`_��Y�6�
���c�Y` �u�"� �Yٞ���A$�Y~�!N�
� �Y�y)�����$�Y3(`��+��Y,��!Fj!���Y[�Ȃ���
�Yr�F��T Y��Y,�w����� �Y�T ]��="V�Y���o�i�Y/l�M����Y�% f��O\%�Y��!>"��(�Y�,!�
)�t�Y��#<)	u�r�E$�<)�u@��F<)E���p�<))�$�8�
k� <)�D�f!I7"�\<)�\%7d}
J�!<)	7�ٶ�/<)ǵP���D��<)�\%c#�f!X�#<)�U ��k#�#<)g!�#5FJFZ�<)U���|�#5� <)��E�
�7V<)^j}
ٶ�g!<)�I��$� �<)���� � ��<)�7t}
;D�#<)Y�)�Dla<))#R�
U��<)0
��D}}
<)����<)V��jr<)b��ݶ<)�7"��y�|�<)�F�bwl��g!<)W�
�8"�l�c0<)�
�
l��E<)/�@�J<)  �%���3
<)�����W#<)�	�/0L$<)ܸ!�%?f#��"<)��Kf#g�ɭ<)��ۚg�XJ<)�%��Sf#��<)���Sf#��!<)7nA7Sf#��<)�Ȏ"���� ]<)N�7�#�]<)���
���y<)�y��$��`G<)��
�]��Sf<)��s�?0<)�JL�H�[\<)�\�� H��}
<)o
��$�#c<)�טLF+z<)�*H��]<)�0 H���<)�f�H�+�<)�%�H��$<)�;^<�
u��7<)dF\�
���� <)�F;�#��G<)�:�JF�<)jc~
S�
1<)��h�#>�S��y#<)b�"wv�S���!<)�k����!S���<)�Eٶ,<)l�
[�8��<)�c��8n�<)�
�cN���
<)V�
�z.�<)-~
�ES���<)��
7�S��F<)R^�S�[~
<)^�0S��<)g��GS�|~
<)���#S��F
<)H)�F
S���<).�$S�<�<)>W ÜS�y��ebE�$=d.�"�ZE�$�� 1�LzE�$���
p�"E�$5�^VE�$�\%M
[���E�$��#^���`"E�$u)�HJ���E�$��L�$P7E�$rq�*�M�E�$�P#=���
�E�$.�M�[dE�$���l!K�E�$��A�F�"E�$	W�<f͍E�$�� ��h
-!E�$n� �6�#�mE�$�A$܂��!��!E�$�z7a"lW E�$�z��W E�$��Z�"��E�$8+S�E�$��".�՜E�$T�M]%4�#E�$f�!�!�)E�$�!�nTE�$ε!�o�E�$��#$?"��_wE�$��) ��hfE�$�-!�E^�E�$*�R#���$E�$��"ދ�]%�#E�$��<�j�E�$�o8��]%�pE�$��#��!.!�oE�$G������7E�$�D.!5;E�$2�Q�#���!E�$�E�$m��9E�$_.!�� �t�E�$�.!���7�E�$��Pn��SE�$���N�"��E�$��#��{�q$E�$�z��!�"#%�F
>?Vn�%YU$�
Ɯ"%�G
�k�%��{��%/!{��%�eS�%�Y{��%V�#Y{��%L�A^%|�$��^%!/!��^%��$��%�]%��&�%�{��&�%��
�%�"%����8%���8%Z���)%�{��"�{%�P��"�{%�qKtf�%��ŝ%��ŝ%;/!�8#/�%9*��$�P%Y/!��$�P%_�"?H
J���ˌ$���J�O*w�\;�J�*8)���
J�L�D �� |J�i��"�#��"J�G�MM��J�A�V|o|��J������nNJ�t�#��i$˓J�ʇ�?"�7�J�vp��p�wJ�˔�"���\J�$�!�>X (8J�aX U�Qw8J����޽�J�p�!�׌$wJ���k��]J�/��]�:QJ�÷#��_\�\J�kH
�*�+J��]%>�y�@"J��8�LX)�J�x�`+�W�J�f�"�/�#J���b�V_%�J�-#M��J�J�����J�J��QIJ�3P����J�)�
�w
�J�8�>aMJ��o��
��J�F�Zܔ:$J���S#^
�$J����"�
	#J���%S#݅4�#J��X ��
����$J��o�#` oeJ�����J�+�"J��"h��#u/!J��"�d�H>J���J��J���$�Q����#J����n�Y_J�a%����"J��V���D�J���
�?�J�i{M��kJ�`�"��
��
2$J��C$��
�}�J��H
�s�2J��q9kpJ�\Z��p�!g+J��Xp���AJ�5I
QPG�
b� J���!��#����J��
�_��!eJ��>��`�J�����#?�4BJ���_b��J�˓�_%)�J�AY @�����J���#���_%y�J��Q�+��J���	��R�J���؈�I
J����|����J�Z ��@"��J��F�!IN�/!J��q�+ A"�J�o�#�mN�RJ��p�>f�
�wJ�*x?&�J�7�
,xk��J��I
��
��+�
J��_%�� ԇ1�J��!O�
���J�%�ҵ��:�J�,
,�J���nJ
�$%CJ�%C�_%���J���`%N��J����|�&�pJ��pX"�8�$J��e�9�l!DB"J��Sf�J����N�PExJ��"��$�>�J�k�"�*��J�h%`x&�#P�J�.��:CZ m!J���qG �
J�?�!u���|��� �E$Q�%����l�
���f�r��sx���NPl��x����$]���-�+��lm&��/!���]:0!��E$4D������$�� SZ��!� �;a�?��
�-sQuZ �Ş�;+s֏"�2��u��_�'�-�`R�(�M�UDvDX��;Q������� R�T#���J
���D���g#��#|�ES�H��H�}�ͺB����ͺ�̑Tmx�ͺo�~�
/Q M2ͺ�����SQ ͺ��zRM8ͺ����9�
6�ͺ;V��%�Y
ͺ��]�
�"}Q ͺ���$��$�~
ͺ�%9��$��ͺ�Q�~
�oͺ�� ��#��ͺ;�ȕ 0G�~
ͺ��9"
seͺz ��$�$��ͺ���� I��ͺ8i��~M��
ͺ�$�$n�4�ͺE�V�#�z�9"ͺ^�P
D�'�ͺ�9p
*�$��ͺ���RC
�Rͺ���[��\%ͺ����M�ͺ�����"ͺ\��_`Bͺq�
8���jͺ�_�V��ͺ��uN#�Kͺ��/�@H��"ͺf`�#[�ͺ��8$����ͺ�����I��"ͺ(�@��2ͺ�%�8$MNͰ!ͺ{\.�9��ͺt�
[~'.�ͺ���#�[6�ͺÁ��ݵͺ{�
G�o�$�ͺn�P���
ͺ�آkYW`ͺ�����Q�ͺ���0�ͺ5�"t4p�$ͺ�"��
��O�ͺ�Vx�&4_`ͺ/�<"HZ%9dͺ����%�ͺ�� �{����ͺ{��#%]��ͺI�v!o�7�ͺ�Z 8	����ͺu4�N^�"ͺ�3�
�N|�ͺ�+�v%�TͺO���f��ͺ��܌ن�ͺ�%��)!)�ͺc���FlZ%a�
T�!����4a�
+�$�}�Oa�
���{�
�Z%a�
�)!�4�� �"a�
�R D[2<�a�
��[
�%�� a�
��-���>a�
��Q]'va�
�%Q�Mb�<"�Y~ S�$9�Y�wO#t]9��Yn� I�EW�Y�k�
Zh�"�Y�Wpb����YOZ��
��"�Yy6u mm���Y&x� �]���YE
_�!��"���Y�x�(���Y�
>��]�'�Y/�#@�6�b�Y���� �Y<���+!��Y��
�@$G�$Bc�Y?y\�<F
��Y,!��qF
��Yl��%��й�Yi�`'`_��Y�6�
���c�Y` �u�"� �Yٞ���A$�Y~�!N�
� �Y�y)�����$�Y3(`��+��Y,��!Fj!���Y[�Ȃ���
�Yr�F��T Y��Y,�w����� �Y�T ]��="V�Y���o�i�Y/l�M����Y�% f��O\%�Y��!>"��(�Y�,!�
)�t�Y�=d.�"�ZE�$�L�$P7E�$rq�*�M�E�$�P#=���
�E�$.�M�[dE�$���l!K�E�$��A�F�"E�$	W�<f͍E�$�� ��h
-!E�$n� �6�#�mE�$�A$�!�n� E�$� $?"��_wE�$��) ��hfE�$�-!�E^�E�$*�R#���$E�$��"ދ�]%�#E�$��<�j�E�$�o8��]%�pE�$��#��!.!�oE�$G������7E�$�D.!5;E�$2�Q�#���!E�$�E�$m��9E�$_.!��!�"#%�F
>?Vn�%YU$�
Ɯ"%�G
;/!�8#/�%�_�"?H
J���ˌ$���J�O*w�\;�J�*8)���
J�L�D �� |J�i��"�#��"J�G�MM��J�A�V|o|��J������nNJ�t�#��i$˓J�ʇ�?"�7�J�vp��p�wJ�˔�"���\J�$�!�>X (8J�aX U�Qw8J����޽�J�p�!�׌$wJ���k��]J�/��]�:QJ�÷#��_\�\J�kH
�*�+J��]%>�y�@"J��8�LX)�J�x�`+�W�J�f�"�/�#J���b�V_%�J�-#M��J�J�����J�J��QIJ�3P����J�)�
�w
�J�8�>aMJ��o��
��J�F�Zܔ:$J���S#^
�$J����"�
	#J����{2`%J�}�%S#݅4�#J��X �#��\J����
����$J��o�#` oeJ�����J�+�"J���!h��#u/!J���!�d�H>J���J��J���$��n�Y_J�a%����"J��V���D�J���
�?�J�i{M��kJ�`�"��
��
2$J��C$��
�{c"J��x�s�2J��q9kpJ�\Z��p�!g+J��Xp���AJ�5I
QPG�
b� J���!��#����J��
�_��!eJ��!>��`�J�����#?�4BJ���_b��J�˓�_%)�J�AY @�����J���#���_%y�J��Q�+��J���	��R�J���؈�I
J����|����J�Z ��@"��J��F�!IN�/!J��q�+ A"�J�o�#�mN�RJ��p�>f�
�wJ�*x?&�J�7�
,xk��J��I
��
��+�
J��_%�� ԇ1�J��!O�
���J�%�ҵ��:�J�,
,�J���nJ
�$%CJ�%C�_%���J���`%N��J����|�&�pJ��pX"�8�$J��e�9�l!DB"J��Sf�J����N�PExJ��"��$�>�J�k�"�*��J�h%`x&�#P�J�.��:CZ m!J���qG �
J�?�!F|0!E���$la|0!)#R�
�p�|0!K
�I7"k� |0!�D�f!I7"�||0!�\%c#"�"X�#|0!�U ��� Z�|0!����>5� |0!��E�
K
V|0!^j}
ٶ�g!|0!�I��$� �|0!���� � ��|0!�7t}
�R�#|0!Y�)(K
��|0!0
�{�}}
|0!����#�E|0!/�@��3
|0!���"�$W#|0!�	�����|0!Ȏ"��Nq ]|0!N�7�#!��$|0!m�;^��u��]|0!���
��y|0!�y��$�Sf|0!��s�`%?0|0!�JL���}
|0!o
��$���|0!�f���7|0!dF\�
���� |0!�F;�#��G|0!�:��>�|0!jc~
S���|0!9��Eٶ,|0!l�
[�8��
|0!V�
�z.�|0!-~
�ES���|0!��
7�S��|0!g��GS�|~
|0!���#S�u���|��� �E$Q�%����l�
���f�r��sx���NPl��x����$]���-�+��lm&��/!���]:0!��E$4D������$�� SZ��j��;a�?����-sQuZ �Ş�;+s֏"�2��u��_�'�-�`R�(�M�UDvDX��;Q������� R�T#���J
���D���g#��#|�ES�H��H�}�ͺB����ͺ�̑Tmx�ͺo�~�
/Q M2ͺ�����SQ ͺ��zRM8ͺ����9�
6�ͺ;V��%���ͺ1K
]�
�"}Q ͺ���$��$�~
ͺ�%��^HͺF��#&���$ͺ�29��$��ͺ�Q�~
�oͺ�� ��#��ͺ;�ȕ 0G�~
ͺ�
�J��ͺj_��$�$��ͺ���� I��ͺ8i��~M��
ͺ�$�$n�4�ͺE�V�#�z�9"ͺ^�p
*�$��ͺ���_���ͺ�RC
�Rͺ���[��\%ͺ����M�ͺ�����"ͺ\��_`Bͺq�
8���jͺ�_�V��ͺ��uN#�Kͺ��/�@Hjsͺ�M�I�ͺ�%@H<ͺf`�#[�ͺ��8$����ͺ�����Is�ͺE���p7ͺ�2����ͺ�7932�ͺ2���
��ͺ���e��"�ͺ�`%�N#��a%ͺߡ(�@��2ͺ�%�8$MNͰ!ͺ{\.�9��ͺt�
[~'.�ͺ���#�[6�ͺÁ��ݵͺ{�
G�o�$�ͺn��y�ßͺ�[ �Ri\ ͺgu�=��ͺW�����ͺS��ͺ�`�T�a%ͺP���ͺ&�;�6�ͺآkYdb%ͺ�|�u�ͺr�՝��ͺ�>�`,�"ͺ ~�y�ͺ��T��"ͺ����� ͺ&�$dq@�ͺ��] ��ͺD�$"�
,ͺ�K
�!O,ͺ_yP��ͺ������ͺ��"��"P�ͺ��a�$s!ͺF@��#ͺ��[}ͺ�,??Sͺ����ͺ5"�#4Tͺ���V"ͺrZ?�ͺ��0�ͺ5�"t4p��ͺ��4q�$�`ͺ����
��O�ͺ�Vx�&4_`ͺ/�<"HZ%�uͺ����%<�ͺ���d{��ͺ��#%]�#ͺU�I4�?;] ͺt#v!o�Qͺ��
�N�
ͺ��"�v%�TͺO���f��ͺ��܌ن�ͺ�%��)!)�ͺc���FlZ%a�
T�!����4a�
+�$�}�Oa�
���{�
�Z%a�
�)!�4�� �"a�
�R D[2<�a�
��[
�%�� a�
��-���>a�
��Q]'va�
�%Q�Mb�<"�Y~ S�$9�Y�wO#t]9��Yn� I�EW�Y�k�
Zh�"�Y�Wpb����YOZ��
��"�Yy6u mm���Y&x� �]���YE
_�!��"���Y�x�(���Y�
>��]�'�Y/�#@�6�b�Y���� �Y<���+!��Y��
�@$G�$Bc�Y?y\�<F
��Y,!��qF
��Yl��%��й�Yi�`'`_��Y�6�
���c�Y` �u�"� �Yٞ���A$�Y~�!N�
� �Y�y)�����$�Y3(`��+��Y,��!Fj!���Y[�Ȃ���
�Yr�F��T Y��Y,�w����� �Y�T ��o�i�Y/l�M����Y�% f��O\%�Y��!>"��(�Y�,!�
)�t�Y��#�	u�r�E$���u@��F�E���C�#�GU���$la��?��p��)�$�8�
k� ��D�f!I7"�\��\%7d}
J�!�	7�ٶ�/�ǵP���D����\%c#�f!X�#��U ��k#�#�g!�#5FJFZ��U���|�#5� ���E�
�7V�^j}
ٶ�g!��I��$� ������ � ����7t}
;D�#�Y�)�Dla�)#R�
U���0
��D}}
������V��jr�b��ݶ��7"��y�|���F�bwl��g!�W�
�8"�l�c0��
�
l��E�/�@�J�  �%���3
������W#��	�/0L$�ܸ!�%?f#��"���Kf#g�ɭ���ۚg�XJ��%��Sf#������Sf#��!�7nA7Sf#����Ȏ"���� ]�N�7�#�]����
���y��y��$��`G���
�]��Sf���s�?0��JL�H�[\��\�� H��}
�o
��$�#c��טLF+z��*H��]��0 H�����f�H�+���%�H��$��;^<�
u�:U$�x�I]9$�0�(
E!9$�7�dF\�
���� ��F;�#��G��:�JF��jc~
S�
1���h�#>�S��y#�b�"wv�S���!��k����!S�����Eٶ,�l�
[�8����c��8n�����cN�������~z6��] n�8"n��&�q�lr=�uy�
�lhq�@�����#d������n1����
#��l��t3#�X�����c���
�V�
�z.��	��qS���-~
�ES��F�R^�S���g��GS�=d.�"�ZE�$�L�$P7E�$rq�*�M�E�$�P#=���
�E�$.�M�[dE�$���l!K�E�$��A�F�"E�$	W�<f͍E�$�� ��h
-!E�$n� �6�#�mE�$�A$܂��E�$��$�y'	E�$�z6��E�$=�J�?E�$�z��	E�$�}�
E�$����aE�$��a#zE�$)�#����E�$z�f@��E�$%�mUE�$T���~E�$�@��ȤE�$r���$�,E�$�@��$�aE�$�$��-E�$U�!��\�E�$f�!5�$�
E�$)�;�x�E�$�!�nTE�$ε!�o�E�$��#�b"�"E�$�@�Q#��|�!E�$�b$?"��_wE�$��) ��hfE�$�-!�E^�E�$*�R#���$E�$��"ދ�]%�#E�$��<�j�E�$�o8��]%�pE�$��#��!.!�oE�$G������7E�$�D.!5;E�$2�Q�#���!E�$�E�$m��9E�$_.!��!�"#%�F
>?Vn�%YU$�
Ɯ"%�G
L
��c%7��~��"�~%�] �UK�V%ʥo�$�qH�%�b%�<�u�%�A��$=�#c�%�0!���!��%�!�k�%��a�C%/!c�$%
�c%,_ %�eS�%�a�#�%V�#��
��%�����c%%�1!J��%&r�!^�%��E�z%����%���1!�%L�~A^%|�$���%!/!� ��$%({�Xҡ%�c�_ ��%#OV��!%|C�q��%!���o�%�]%(` ��%�{�`��"%=a�%V��C%�V��#�%'D�A{%�qKtf�%Ő$4�t%8�����%o�c` �C%�1!�D�%��p��%%���m.%7�$e%��ǣ�` %�d%���%r2!�V%�0�Q�%;/!�8#/�%_�"?H
J���ˌ$���J�O*w�\;�J�*8)���
J�L�D �� |J�i��"�#��"J�G�MM��J�A�V|o|��J������nNJ�t�#��i$˓J�ʇ�?"�7�J�vp��p�wJ�˔�"���\J�$�!�>X (8J�aX U�Qw8J����޽�J�p�!�׌$wJ�,r�k��]J�/��]�:QJ�÷#��_\�\J�kH
�*�+J��]%>�y�@"J��8�LX)�J�x�`+�W�J�f�"�/�#J���b�V_%�J�-#M��J�J�����J�J��QIJ�3P����J�)�
�w
�J�8�>aMJ��o��
��J�F�Zܔ:$J���S#^
�$J����"�
	#J���%S#݅4�#J��X ��
����$J��o�#` oeJ�����J�+�"J��"h��#u/!J��"�d�H>J���J��J���$��n�Y_J�a%����"J��V���D�J���
�?�J�i{M��kJ�`�"��
��
2$J��C$��
�}�J��H
�s�2J��q9kpJ�\Z��p�!g+J��Xp���AJ�5I
QPG�
b� J���!��#����J��
�_��!eJ��>��`�J�����#?�4BJ���_b��J�˓�_%)�J�AY @�����J���#���_%y�J��Q	��R�J���؈�I
J����|����J�Z ��@"��J��F�!IN�/!J��q�+ A"�J�o�#�mN�RJ��p�>f�
�wJ�*x?&�J�7�
,xk��J��I
�� ԇ1�J��!O�
���J�%�ҵ��:�J�,
,�J���nJ
�$%CJ�%C�_%���J���`%N��J����|�&�pJ��pX"�8�$J��e�9�l!DB"J��Sf�J����N�PExJ��"��$�>�J�k�"�*��J�h%`x&�#P�J�.��:CZ m!J���qG �
J�?�!��c*E���:�;� �$���:���"�����:O"f"����:�d�����:��
��^�$���:OC"|"O����:���$ּ#���:!k�v����:>���r���:�&��Da ���:��"�#Da ���:%L
9�Da ���:
'�o���:�W��o���:O�	� ���:8�#0�� ���:��fEכ���:
$q��r���:Q��������:P�
6v���:�W�!6v���:<T2!�����:�r��"�����:f�����"���:!�z�5����:k#�d%
s���:��
F����:s����"���:�Z�$'���:f�J'As���:�c�����:�B
=L
_2!���:�1U�"���:�����$�U-2�:g@�$�/�U-2�:82Ѵwa �U-2�:{#�
���U-2�:�?Eo8�"�U-2�:SL
0؉�U-2�:3q�"d�U-2�:��
w��U-2�:$�m	�U-2�:�{�2!e	�U-2�:4�3��U-2�:{��W`��U-2�:�5%�m�U-2�:��8�"�#�U-2�:��wZ��U-2�:y���20�U-2�:�{p��	�U-2�:{E�i�U-2�:��*����U-2�:���E�E�U-2�:X�$�2!���U-2�:w�$F��U-2�:��W�U-2�:�+���"�U-2�:n0ǟ�2!�U-2�:�0���	�U-2�:�d%	|�U-2�:3!��"<��U-2�:�d% ��W�U-2�:�a ��"��U-2�:�osԜ�U-2�:R���#��U-2�:b��"kd�U-2�:�&!+���U-2�:�d�a "��U-2�:<�q����U-2�:Ћu����U-2�:
b O�#��U-2�:��013!�U-2�:�s�0�s�U-2�:p��	��U-2�:k0�U��U-2�:�X��"�U-2�::
��
e%�U-2�:��rL
�U-2�:�C����U-2�:�dF<X�U-2�:_
�L
"�U-2�:�0*����U-2�:���"k�#�U-2�:��L����U-2�:��$�s{
�U-2�:
�"��t��U-2�:���G"�U-2�:��#%�"��U-2�:����L
�U-2�:G�s�0�U-2�:[1�1���U-2�:��do�U-2�:�sG|���U-2�:��#�G��U-2�:��_e%G��U-2�:e2��#�U-2�:�
��$
��U-2�:N��{e%�U-2�:�3!M
���U-2�:I�#c|,��U-2�:>e|h��U-2�:h�#��ٝ�U-2�:(��"#t�U-2�:��7M
�|�U-2�:�"}A�"�U-2�:j�ϒ$��U-2�:O
����"�U-2�:��#0}�t�U-2�:>����U-2�:�$�3!�X�U-2�:��$�L}�U-2�:z
�1"F�U-2�:@��e%�
�U-2�:�+b �"�U-2�:����F��U-2�:�e%�B��U-2�:������U-2�:^e����#�U-2�:Π2�`'�U-2�:SM
�vM
�U-2�:h���N��U-2�:���Gb �U-2�:�gm��U4�:�#�#�U4�:H_�2�#�U4�:����#�U4�:�$�v�#�U4�:vo`$�'�U4�:�!�$�'�U4�:�?���'�U4�:F�!>�'�U4�:$f%�
�b �U4�:_�ь�b �U4�:�1���b �U4�:�M
�e�b �U4�:����$�U4�:�' �3�$�U4�:
�e�#�$�U4�:�`�%�$�U4�:�`?�(��U4�:6X$��#(��U4�:��4� (��U4�:�)�
(��U4�:�b�I�}�U4�:�>�5�}�U4�:��<
�}�U4�:��) �}�U4�:��M� ��U4�:�;_b��U4�:��-q��U4�:���\$��U4�:g�A	�M
�U4�:GN
o��M
�U4�:Gf%��M
�U4�:%��M
�U4�:.���b �U4�::fC�b �U4�:���\$�b �U4�:��

��b �U4�:�N&q	�U4�:�'K8q	�U4�:]�
q	�U4�:�6\$q	�U4�:�1�$�U4�:����$�U4�:vF�X�$�U4�:�F���$�U4�:b�$N�(��U4�:����(��U4�:X�#�F(��U4�:�	iN
(��U4�:G�t�#�U4�:~�Kc �#�U4�:�N
4�#�U4�:�3!�}�#�U4�:��$���'�U4�:�"�	�'�U4�:{c o�'�U4�:�"u�'�U4�:�F���b �U4�:y��e�b �U4�:�3!+��b �U4�:#~���b �U4�:��$��Q��U4�:ڊ[��b �U4�:4u�
�b �U4�:4!���b �U4�:[uif%�b �U4�:�N
F~q	�U4�:�"q	�U4�:����q	�U4�:@"Yq	�U4�:JY�}�U4�:�f%���}�U4�:o�"��}�U4�:�����}�U4�:f"���U4�:2���U4�:�e����U4�:��qY��U4�:;4!֡�M
�U4�:�"�'�M
�U4�:(K��M
�U4�:�Y�#�M
�U4�:Ϟq�*
�U4�:�
G��U4�:�w@(��U4�:��z?�U4�:��
D�i3�U4�:�̩f�U4�:'�-2d�U4�:t��Y��U4�:�����U4�:�^����U4�:ғ$[f���U4�:�f%]2]4!�U4�:*��K��U4�:�f%4���"�U4�:��?
��U4�:3%�' ��U4�:�4��?��U4�:ȟ�"���U4�:l' ����$�U4�:W��uc��U�� �:*�#9��]$�U�� �:�u�4!�N
�U�� �:��P�~f�U�� �:IG	��U�� �:e�
�u!��U�� �:p(gmt
�U�� �:��V�e��U�� �:�
�cO��U�� �:yA�L��U�� �:gG��#@�U�� �:�u��"q�U�� �:о�21��U��:J��l��U��:�x��*��U��:'z�(�c �U��:K��c @��U��:�#	�~G�U��:�q����U��:v~�0��U��:e�
��y��U��:�1��
�U��:��gm��Ut��:k��0���Ut��:�Ggm���U�L�:�����f�U��:o'��P�$�U�& �:Q������U�& �:[o�"��$�U�& �:�.Z`$��U�& �:A�"gm-g�U�& �::%�L��#�U�& �:E!�G�U�& �:*`$x��$�U�& �:�y��!{��U�& �:d �
�Ua�-2�:�G���Ua�-2�:��$�2ڿ#�U�~-2�:���2���U�~-2�:�x��2d �U�d ��:�q����Q��
�:u���|��� �E$Q�%����l�
���fv�#�e �n��r��sx���NPl��x����$]���-�+��lm&��/!���]:0!��E$4D������$���#
�-g#�� SZ��!� �;a�?��
�-sQuZ �Ş�;+s֏"�2��u��_�'�-�`R�(�M�UDvDX��;Q������� R�T#���J
���D���g#��#|�ES�H�Ԣe]�Yͺ�2�$2�iZͺ:�We �!i�ͺ#����'ͺ��Hphe ͺ�$���qͺ�n���+�ͺGp
�2сͺ��_z�ͺ.�R�Z�"ͺ1�Qg��ͺ���R�#Hͺ�4!u�n�ͺO
��g%/g%ͺR���q�ͺ�"���ͺ��.�#ugͺ�����(ͺ'y!��(ͺ��'#��ͺ� \5!ͺg� 9��g%ͺ��
~'��ͺ��q�ͺ~h!̍�uͺ�*�!��$ͺ��#]�rͺ���;"=�\�ͺh"U�$�jͺkS�$�ͺ�SP"��ͺA`�J��ͺ�~�$ͺ���� h%ͺu���"ͺ���e ͺQ�#
�ZͺV���ͺ3vm��Hͺ�R�g��ͺgu�=��ͺ�C
h��ͺ��
Gvͺ��9�
][ͺ�
;0\�ͺ�5!Р'�ͺ�[��e ͺ4�";�s�#ͺ?"kY�f ͺ���RX�ͺ����!�h%ͺ�`Рu�ͺP����ͺ&�;��ͺآkY��$ͺZ��R��ͺ%t��!u"ͺ�ei%!�ͺ����P
ͺm���
 \ͺ�|��ͺ��!�f$L)ͺ�#��)ͺ����"ͺ�$��J�ͺB�$�7"ͺ�"!G"�O
ͺh�"�)$�ͺ��
�g ͺڣ�P
ŁͺH�#�'hͺ�v`�ǩͺ���ͺ
��<�"ͺԐ
�:*ͺ��5��P
ͺ��o�N�ͺ��wͺ�R {�wͺ�$�}Z
ͺqh �u#YQ
ͺ�D
��i%ͺ�N#�~ͺ��>�ͺ��]�$6�ͺt�#����ͺ��X5W�ͺ��"��"�hͺ��a�$�iͺF@�6j%ͺ{���ͺ���}�ͺr���ͺ��~�ͺ/x��Kxͺ��$�5!�\ͺ�5��#�ͺ���i]ͺ���x�ͺ�M�k%ͺ�*16!|"ͺ����a�
����ya�
����$،��a�
Q�.�#�Y���
�$��$�Y����ޫ�YP�07A��Y�+���Y�;�fy�Y�).
R
�Yuj"�
+�h �Y��$�dE�\�YV6!��$�yE�$7��-#"�E�$`��� ��E�$ת�"��E�$�H�"��E�$t��7!E�$�h �7!E�$܂\�#��#E�$��$\�#��#E�$�I��$E�$�h I��$E�$�zI.�E�$=�IUk%E�$�z�k%/+E�$�}�k%/+E�$l&�-"E�$��&�-"E�$v�����E�$�"����E�$8��#E�$}���#E�$��"&l�E�$HI&l�E�$*8!�i E�$bl�i E�$�d�4�E�$�+d�4�E�$T�jR
��E�$�@jR
��E�$��"�E�$a�"�E�$���E�$�@���E�$�$���E�$f�!%�$��E�$)�%�$��E�$
�x�#�zE�$�k%x�#�zE�$y���"E�$l%��"E�$N(���%��"{l�%R�{l@�%��[H�%��"[��%x��(�l%%�]K����%�R
���+%�5.�#=8!%6���7%����$��%a�$z�h^%�^I��%fI���"%����"%3�$0��8!%���l�l%�l%����%�k�%��{C%/!{C%r�eS~%�Y{��%�Y{��%�K��^%�eS��%�Y{
�%V�#Y{��#%'�K�7"%L���%|�$��%!/!��%F�
�#L�#%P�� %��$W�%��5%X�O�$+�%��O�$+�%U�"%��
��%������%�����%Z����%�{��"��%�P��"��%2m�l%{I%��"c�s�%��a��%�qKt,%�z�R
&�#%�_{}%Ő$�t%8�����%o�����%��X�%;/!�8#/�%9!U���#%q,�,�"%��$�,�"%�l%��Nm%�.�%w�X�:6J�C�� B+�9J�M��7JJ�VM��;
J�������J�����S%J�K�4�ĂJ�2Y"�-#��J���(J���+�"J�S��
���J�$��PJ��-��dJ���W>� J����M+9!J��q�>�J�<T�+���J��"(J/�� J���>�U�b�J���>�y��R
J����,U�b�J������
�J�NIY"�,J���J��JJ��h��#B-J�z.#�/�J��6OS
m%J�q�si �9!J���
;��J���$�O�m%J�0h�@l�J�%�B KJ���#d���J���Ҥxn%J��	"��l"��J���l"��J���w��6J�-:!l"��J�ٔ�-C{J�L:!q�"�J�����o%J��-��mJ�>&���J���_XJ���$�o%��#J���"��J�k��nK�J�^�'��J�w��p:!ݖJ��_�0���J���"B�K�J�7����J�f���#�J�l��"�J��`�i �{sS
J��S
��#ѝ$J����o%"�J��N��>�$J���T�m�o%J���g.��J�p�M��J����&�#}�J����$���o��:���$����o��:-�k�g� ��o��:!k�w6��o��:
$q�����:��
�������:OC"|"�� ���:f�����"��K��:>���r��"��:%L
9�Da �,p%��:�&��Da �,p%��:��"�#Da �,p%��:Q������T
��:!�z�5������:��fEכ��i ��:
'�o�����:�W��o�����:P�
6v�O��:�W�!6v�O��:O�	� �T���:8�#0�� �T���:<T2!����:!��:�r��"����:!��:3q��
�U-2�:J�#���"�U-2�:�\?�U-2�:$�tp%fv�U-2�:tp%��
�U-2�:��57���U-2�:57K��U-2�:����3�U-2o��:�1%�U-2o��:�Qg�

�U-2o��:g�V$�U-2o��:4�3j��U-2�:{��W.��U-2�:���6���U4m�:m
AV$vo�6ږ�U4m�:m
AV$�C%�"�U4m�:b����#%�S4�U4m�:b���H_�2p_�U4m�:����o�U4m�:�$�v�m�U4m�:�!�$_�U4m�:�?��4�U4m�:F�!>U2�U4m�:�s�0�s�U�-2�:��
w��U�#-2�:$�m	�U
�-2�:���Gb �U�{-2�:�{�2!e	�U��-2�:�5%�m�U,a-2�:��8�"�#�U8�-2�:��wZ��U7M-2�:y���20�U�?!-2�:I�#c|,��U`-2�:
�"��t��Ud7-2�:ԇ~��@!�U�n-2�:��ϯ���UEj -2�:�{p��	�U�-2�:{E�i�UF-2�:p��	��U��-2�:�E�A!�U�--2�:�p%8a��U�-2�:X�$�2!���U��-2�:�X��"�U��-2�:��W�UT-2�::
��
e%�Uɗ-2�:��{N/T
�U��-2�:�oE8g��U��-2�:��#0}�t�U�-2�:>����U�-2�:�0���	�UI-2�:�C����U8�-2�:T
����U -2�:��#!�.�U/-2�:�������U��-2�:>a8���U��-2�:���"k�#�U��-2�:�0*����U��-2�:�d%	|�U��-2�:3!��"<��U�-2�:�dF<X�U/k -2�:(��"#t�U��-2�:^e����#�U��-2�:_
�L
"�U�~-2�:h�#��ٝ�Ua�-2�:�d% ��W�U�/-2�:��013!�U-2�:��7M
�|�U��-2�:z
�1"F�U-2�:Π2�`'�Uݦ-2�:�a ��"��U��-2�:����L
�U�-2�:�}ّ���U|�$-2�:�N2���U-2�:�osԜ�U}8-2�:�N��sb�U�T
-2�:�UA!t�$�U��-2�:R���#��U�-2�:b��"kd�UN�-2�:��p%V��U	-2�:�d�a "��U�0-2�:���G"�U��-2�:<�q����U��"-2�:Ћu����U$l -2�:
b O�#��Uʨ-2�:@��e%�
�U��-2�:h���N��U��-2�:�gm��UqA!4�:Ϟq�*
�U!4�:ȟ�"���U�b4�:�
G��U��4�:�w@(��Uɟ$4�:��z?�U�4�:��
D�i3�U�"4�:�̩f�U�14�:'�-2d�U��4�:t��Y��U�A!4�:�����U��4�:l' ����$�Ub�4�:�^����U�p%4�:�4��?��U��"4�:3%�' ��UXB!4�:��?
��U��"4�:�f%4���"�Uw�$4�:*��K��U�
"4�:�f%]2]4!�U^4�:ғ$[f���U�94�:�`?�(��Uש4�:6X$��#(��Uŧ4�:��4� (��U�}4�:�)�
(��U�N4�:b�$N�(��U�4�:����(��U�4�:X�#�F(��U�24�:�	iN
(��U(�4�:����$�U~q%4�:�' �3�$�U�4�:
�e�#�$�U�~4�:�`�%�$�U2�4�:�1�$�U�r%4�:����$�U��4�:vF�X�$�Uc4�:�F���$�U�O4�:$f%�
�b �Ux!4�:_�ь�b �U��4�:�1���b �U6�4�:�M
�e�b �U��4�:�F���b �U�4�:y��e�b �U�4�:�3!+��b �U�4�:#~���b �U�34�:��$���'�U�4�:�"�	�'�U��4�:{c o�'�UØ4�:�"u�'�U�$4�:G�t�#�U�$4�:~�Kc �#�U�B!4�:�N
4�#�U�44�:�3!�}�#�U�$4�:��$��Q��U]�4�:�N&q	�UP�4�:�'K8q	�U/�4�:]�
q	�U.m 4�:�6\$q	�U�$4�:�N
F~q	�U:4�:�"q	�U4�:����q	�U'n 4�:@"Yq	�U�4�:.���b �U�"4�::fC�b �U��#4�:���\$�b �U�4�:��

��b �U6�"4�:ڊ[��b �U��4�:4u�
�b �U�U
4�:4!���b �U��4�:[uif%�b �U��4�:g�A	�M
�Uq�4�:GN
o��M
�U�4�:Gf%��M
�U
�"4�:%��M
�Ud4�:;4!֡�M
�Uβ4�:�"�'�M
�U�"4�:(K��M
�U��4�:�Y�#�M
�U�C!4�:��M� ��U�4�:�;_b��U"4�:��-q��Uw4�:���\$��U#�4�:f"���U��4�:2���UҪ4�:�e����U�4�:��qY��U�4�:�b�I�}�U��4�:�>�5�}�U�"4�:��<
�}�U�"4�:��) �}�U�"4�:JY�}�Uę4�:�f%���}�Um4�:o�"��}�U4�:�����}�Uɀ4�:��P�~f�U�D!�� �:p(gmt
�U��#�� �:*�#9��]$�Ul
�� �:�u�4!�N
�Uq��� �:��V�e��U���� �:�
�cO��Urx�� �:yA�L��U	��� �:gG��#@�U�o�� �:IG	��U���� �:W��uc��U�"�� �:�u��"q�U�5�� �:e�
�u!��U�
�� �:�q����UŤ$��:v~�0��U���:��gm��U>t��:k��0���U�t��:�����f�U6��:�q����Q����
�:"$�#b�u������l=;�#���v�#�m�����r�n	���#NPn�
��#��d�$=�
���:"$L�EE!�z�+��e o �4D����
�^�����s%�� d�!�o ��E!�;�����-s$3f���uF!5�����$g	,����~V
�h��*�#)p Ɖ��$��e �������P����Q6��V���#��;8p P��-��V
������������UD�A��{�Q�� �t%�"R�t�׫�p�p���ށ�
�����ܨq����
��(���$2�iZͺ:�p
�2сͺ��_z�ͺ.�1�Qg��ͺ���R�#Hͺu	u�n�ͺ�pR���q�ͺ�"���ͺF���.�#ugͺ[�����(ͺ'y!���ͺ� \5!ͺg� 9��g%ͺ��
~'��ͺ��q�ͺ~h!̍�uͺ��#]�rͺ���;"=�\�ͺh"�*16!|"ͺF�P���p��PK
�k� �Pj,�f!5� �P��E�
���P�7t}
$�P/�`G�P�F!�]�7�PdF\�
�� �P�7;�#��P��~
,�P_;[�����a�
��$،��a�
Q�.a�#�Y���
�$��$�Y����ޫ�YP�07A��Y�+���Y�;�fy�Y�).
R
�Yuj�dE�\�YV6!��$�qE�$Y������$E�$��h ����$E�$��z��.�E�$�=���Uk%E�$��z�F!/+E�$��}�F!/+E�$�N(���%��"{l�%R�{l@�%��[H�%��"[��%x��(�l%%�]K����%�R
���+%�5.�#=8!%6���7%����$��%a�$z�h^%r�eS~%��eS��%�L���%�w�X�:6J�C�� B+�9J�M��7JJ�VM��;
J�������J�����S%J�K�4�ĂJ�2Y"�-#��J���(J���+�"J�S��
���J�$��PJ���W>� J����M+9!J��q�>�J�<T�+���J��"(J/�� J���>�U�b�J���>�y��R
J����,U�b�J������
�J�NIY"�,J���J��JJ��h��#B-J���$�O�PJ���Ҥxn%J����l"��J���l"��J���w��6J�-:!l"��J�ٔ�-C{J�L:!q�"�J�����o%J��-��mJ�>&���J���_XJ���$�o%��#J���"��J�k��nK�J�^�'��J�w��p:!ݖJ��_�0���J���"B�K�J�l��"�J��`�i �{sS
J��S
��#ѝ$J����o%"�J�p�M��J��"$�#b�u������l=;�#����r�n	���#NPn�
��#��d�$=�
���:"$L�EE!�z��� d�!�o ��E!�;�����-s$3f���uF!5�����$g	,����~V
�h��*�#)p Ɖ��$���P����Q6��V���#��;8p P��UD�A��{�Q�� �t%�"R�t�׫�pށ�
�����ܨq����
��(���"jQ�	u�r�E$dfjQ�E���f"jQ�)�$�8�
+jQ��D�f!I7"�jQ��\%7d}
�7jQ�	7�ٶ��jQǵ�P���D�"jQ��\%c#�f!�t%jQ�q;��k#H8jQg!�ܱ5FJF��"jQ�U���|�#T�jQ���E�
�70�jQ�^j}
ٶA$jQ��I��$� ��jQ����� � �<jQ��7t}
;D
jQ�Y�)�D]p jQ�)#R�
U��#jQ�0
��DZ�jQ���#��jQV���Dj`
jQb�����$jQ�7"���y�q�jQ�F��bwl��jQW�
��8"�l���jQ�/�@���jQ��8�%���jQ��������#jQ��	�/0��jQ�ܸ!�%?f#rjQ���Kf#g��jQ���ۚg��jQ��%��Sf#�V
jQ����Sf#@�jQ�7nA7Sf#�jQ��Ȏ"�����"jQ�N�7�#�jQ����
��K�jQ��y��$��W
jQ���
�]���jQ���s�VjQ��JL�H���jQ��\�� H��jQ�o
��$�#`�jQ��טLF�"jQ��*H�C�jQ���1�
H��
jQ��0 H��jQ��f�H�ڊjQ��%�H���jQ�vQ;^<�
u� �jQ�dF\�
���jQ��F;�#�,�jQ�(�:�JF�QjQ�jc~
S�5�jQ��EٶjQ�l�
[�8�"jQ��c��8�F!jQ��
�cN� �jQ��#w���V�jQ�V�
�z.��jQT�I����<jQ�T�a�"p���jQu�$T����u��jQ�T��J��$jQT��p ��XW
jQT�:��f!XW
jQT����8XW
jQT��$7v�jQT�6��'�jQT�Ĭ�?G!jQT�_r��ERjQT��f�r�jQT��rP��p jQT�hW
O�B�jQT��7t}
��"jQT���)��jQT���#��$��jQT�w��� 	q jQT���E�
%jQT��r�z��jQT�s-%�#jQT�_�Re�jQT�eG!<s�G!jQT���[�PsjQT�\�W
�jQT��W
Zq ��jQT���"~
�8jQT���[�γjQT�dF\�
k�jQT�g�F��jQT���|��jQT�:����=jQT��W
̺]�jQT�_���;"jQT�����=jQT�!X
�R�"jQT����jQ	��qS��jQ-~
�ES���jQ��
7�S��FjQR^�S�[~
jQ^�0S��jQg��GS�|~
jQ���#S��F
jQH)�F
S���jQ.�$S�<�jQ>W ÜS�R�Z�ͺT�1�Qg3u%ͺT���R��G!ͺT�u�n��ͺT���g%fX
ͺT�R���w�ͺT��"��[�ͺT���.�#˭ͺT�'y!�NH!ͺT���
~'��ͺT�~h!̍_�#ͺT�R���$]ͺT�,9��+ͺT��R���ͺT�
>��(>ͺT���#]��H!ͺT��;"=���ͺT���_��X
ͺT�����>ͺT����hq ͺT���#�I�ͺT��Y
G9ͺT���su9ͺT�L�y:ͺT���$Bj�ͺT��SP"Ԩ$ͺT����D�ͺT������ͺT��*16!Z
ͺT��H���#ͺ������ͺ�̑TmK�ͺ�~�
/Q ͺ��/Q pͺ����7�$ͺ��zRM�fͺ���ݚSͺ��U�lr ͺ���
��ͺ���$��$l�ͺ���^��ͺ���#&�/�#ͺ�9��$pͺ��~
��ͺ���#�
�ͺ�ȕ 0GE%ͺ�
�J��ͺ��9"
(�ͺ�A�U���ͺ���$�$�ͺ��� I�sͺ���~M@Sͺ��$n�xSͺ�p
*�$%�"ͺ��RC
�"ͺ��[��2�ͺ���M�Z
ͺ����sͺ��_`��ͺ�8��:�"ͺ��V�pͺ�uN#�Kr�$ͺ��,I!QI!ͺ�/�@H/�ͺ��I�r ͺ��%@H/�ͺ�f`�#^s ͺ��8$���ͺ����I�r ͺ����$ͺ��N#��ͺ�(�@���ͺ��8$MN3�ͺ�.�9���ͺ�
[~'ͺ��#�[c�#ͺ����ͺ�G�o�$q�ͺ��Sg%�"ͺ�:g��͡ͺ�P��X�ͺ�آLNgͺ�����#I�ͺ�t��"%�"ͺ�0����ͺ���
[
�#ͺ��<"��|;ͺ���I�%�"ͺ��#t 7�ͺ�����%�"ͺ���g�ͺ�����S�#ͺ�܌d�$#[
ͺ�����ͺ�������a�
T���$،�a�
T�҄��%�"a�
����F�;a�
��%�?%�"a�
����"%�"a�
����#%�"a�
�u���$%�"a�
�����a�
��}���a�
��{�
4t a�
��4���Sa�
�-����Ua�
�Q]�ga�
�Q�Mb���YT��).��$�YT��dE��"�YT��I!�����Y�O#�u%0��Y�� I�Zt�Y�k�
Z���Y�pb�YW�Y���
��Y�u mm��Y�� �W��Y�_�!��#���Y�����#�Y�>�C�"�t�Y�@��h��Y��� �u%�Y��0�
�Y��@$�$�
�Y�\�����Y����<�Y��%��v%�Y�`'`�W�Y�v ����Y�e���՛�Y�N�$N�
(v �Y�)���yw%�Y�`��I!x�"�Y��!Fj!�I!�Y�Ȃ�"�Y�F�^i!�Y�w���_[
�Y�]��c�Y���u���Y��M����Y� f�����Y�>"��i�Y��
)��"�Y��[
ݎ{�E�$T�`�$�[
�E�$T���x�$�<E�$T��i
j,�"E�$T��w%�%
�E�$T���$�w%E�$T��zJ��"E�$T��z�v �E�$T�=d.�"H"E�$��L�$��#E�$��x%��E�$��?�v ��$E�$��v �x%%�"E�$�,�+�%�"E�$�q�*�1jE�$�=���
�%E�$��M���E�$��l!K�x%E�$��A�E�$�W�<fh&E�$���h
BuE�$��6�#�E�$���:�ǣE�$��z��� E�$��ze��$E�$��eS^�%T�L��"%T�;/!�8#HX%T�ˌ$��#�[
%�����?%��X�X%�"%�>\
�$%�"%�8�%�"%�'vx%�"%��W��=%�L�#
w %�;/!��j�%�K��j%�"%�w�X�T�#J�T��� B+-"J�T�M����$J�T��B+�uJ�T���+�XJ�T�T�+��w J�T�(J/�J�T�>�y�Z�J�T�`+��w J�T�"�/�J�T���J�x�J�T�z.#���J�T�����J�T�q�si S\
J�T�0h�@��J�T�!�7���J�T�%�B LvJ�T����� J�T���#d��J�T�.�$���"J�T���8J!�<J�T���l"YAJ�T�����w J�T���w���J�T�,'��#�"J�T�ٔ�-��J�T�L:!q�"9wJ�T������J�T��-�"�J�T�>&��x J�T���_�J�T���$�o%�J�T�k��n(�J�T�^�'�M�J�T��p:!kJ�T��0��!J�T���"B�\�J�T�7���=J�T�f���#BJ�T�l��"I'J�T��i �{�J�T��S
��#Y"J�T����o%�J�T�T�m��"J�T���g.�y J�T�p�M��J�T���&�#�y J�T�_�"��#9�J������#AYJ��]"��RJ!J���Y]�%�"J��w�\�#J��8)��"J��D ����J����"�#��J��MM�y%J��V|o|K�$J������J����i$�BJ���?"�7ѼJ������J���"�y%��J����z%J��U��$i�J��!��#҅J���S��J����$�$םJ���]��>J����_\lz%J�� #�Y%�"J���'�J!%�"J���*��>J��>�y�!J��`+��J��"�/5z J��b�V_%�J��-#M�ðJ��J����J��J��Q��$J��3P���z%J��8�>��J���o��
2 J��Zܔ��"J��S#^
R�#J���"�
L�J���{��#J��%S#݅
�J���#� �kJ����J�RJ��h��#l?J��fX�%�"J���d��J!J����J�i�J��b��\
%�"J��%��]
J��V��w�J��K!�]
%�"J��r���$J���C%�"J��.K!u"%�"J��}��ܭ$J���"���J�����%�"J����uz %�"J���H�"%�"J��ݑ+�%�"J������J��{MY�6�J����
�"�#J���s�ϐJ��q9km�#J�����wP�J��Xp��J��QPG�
�YJ����#��J�����%�"J���_����J����#?��'J���_b�EK!J�����_%H�J����� 3CJ���+����J��F�$%�"J��	��RE�$J���؈�J���k�%�"J���~�${%J���!m�"ZJ��ű��kJ���Q��]
J����$��#I�J��xx� J��,x�"sCJ���� ԇˮ$J��O�
��z J��ҵ��aJ��
,�J��nJ
�$J���_%�J��`%NJ�J���|�&z�J��M{%"�7#J��X"��z J���9�l!��J���SEJ���N��}�J����K!lJ��	^
M�%�"J��p�3l%�"J��`x&�#dlJ��ۖCZ ��J���:CZ νJ��
$gm�����:�K!�ɞ���:�#^
F{ ���:���:
!���:b{%k�D(���:�Ce�BZ���: �A��"���:�CZ8x���:�l׸��#���:�l'�#r(���:%�x{%����:�?�?�#���:|Z��v����:���L�?���:t{ e����:�,y�C���:��"�7����:|����(���:8@
���:bx�#���:�{%
$�$���:��q��C���:�Zq�2�#���:�q�����:��#q�����:ҿ���{%�U#"�:B� ���u�UQ�:�$��lu�UQ�:v�Y!v��UQ�:�l=;�{%�T�4D�(���T�ppD���T�u���#�����E$��"�����lx�$���v�#�K!=����rʦ�D��NP��"7!���$��$�l��+�*�������D���4Ds�%����N@E���� ��$)���;N����-sm�����;^E<������Z���-�V�����UD0">��Q�$���R�q@d����H�}�ͺB����ͺ�̑Tmx�ͺo�~�
/Q B|%ͺ�/Q M2ͺ�����SQ ͺ��zRM8ͺ����ݚ6�ͺ;V��J�̺ͺ]�
9m}Q ͺ��U��ͺ|x��
j)ͺ���$��$�~
ͺ�%��^HͺF��#&���$ͺ�2���R�ͺk�9��$��ͺ�Q�~
�oͺ�� ��#��ͺ;�ȕ 0G�~
ͺ�
�J��ͺj_�9"
;�ͺy�A�U���ͺ����$�$��ͺ���� I��ͺ8i��~M��
ͺ�$�$n�4�ͺE�p
*�$}6"ͺ�RC
�Rͺ���[��\%ͺ����M�ͺ�����"ͺ\��_`Bͺq�
8���jͺ�_�V��ͺ��uN#�Kͺ���,I!�ͺ�/�@Hjsͺ�Mf`�#[�ͺ��8$����ͺ�����Is�ͺE���q�ͺ�)�œ�Eͺ�793҈#ͺ҈#����ͺ�e���xͺ�$�N#�7�ͺߡ(�@��2ͺ�%�8$MNͰ!ͺ{\.�9��ͺt�
[~'.�ͺ���#�[6�ͺÁ��ݵͺ{�
G�o�$�ͺn��y&^
��ͺ�@�Sg�ͺ:g��Dyͺ��$�R-i\ ͺgu�|%��ͺW�gy��ͺ���ͺ�`3$�a%ͺP���ͺ&��E6�ͺآLdb%ͺ�|�yu�ͺr�\���ͺ�>�,�"ͺ �y�ͺ��|%��"ͺ����#� ͺ&�$��#@�ͺ������ͺ�K
�#O,ͺ���!ͺ��"��ͺ��\m�ͺFK^
Ұ$ͺ�����$ͺ�,$�)�ͺ�(���ͺ5"�Z�|%ͺ��~%��ͺri�ͺt��"<�ͺ0��0�ͺ5�"��
[
O�ͺ�V�<"��*�ͺ��I�Y~%ͺ�#t �ͺ����S"ͺ��g�TͺO�������ͺ��܌d�$�ͺ�%��)�ͺc҄��o$a�
���FlZ%a�
T�!�%�?m�a�
���"�$a�
���#��a�
u���$k�a�
����4a�
+�$�}�Oa�
���{�
�Z%a�
�)!�4�� �"a�
�R D[2<�a�
��[
�%�� a�
��-���>a�
��Q]'va�
�%���$�$a�
Z�-*�a�
��`�X�#a�
!�L�o�a�
�y[ҵa�
#c�P*�a�
ײ$A��"a�
��I!����$�Y�oO#�u%9��Yn� I�EW�Y�k�
Zh�"�Y�Wpb����YOZ��
��"�Yy6u mm���Y&x� �W���YE
_�!��#���Y�x����Y�
>�C�"�'�Y/�#@���b�Y���� �Y<��0���Y��
�@$�$Bc�Y?y\�����Y,!�����Yl��%�й�Yi�x"e�y�Y`'`_��Y�v ���#�Ye����c�YR�"N�$N�
� �Y�y)�����$�Y3(`��I!+��Y,��!Fj!���Y[�Ȃ���
�Yr�F�^iY��Y,�w����� �Y�T ]��V�Y���ui�Y/l�M����YF f��O\%�Y��!>"��(�Y�,!�
)�t�Y�=d.�"�ZE�$�L�$P7E�$r�x%��E�$�?�v IE�$�v �x%��$E�$,�+�2�E�$q�*�M�E�$�P#=���
�E�$.�M�[dE�$���l!K�E�$��A�F�"E�$	W�<f͍E�$�� ��h
-!E�$n� �6�#�mE�$�A$��:���E�$܂�m�E�$�z���E�$�z'z�	E�$)�#���E�$z�����E�$�@�!��E�$U�!�4AE�$f�!Q��
E�$)�s*x�E�$$?"��_wE�$��) ��hfE�$�-!�E^�E�$*�R#���$E�$��"ދ�]%�#E�$��<�j�E�$�o8��]%�pE�$��#D.!5;E�$2�Q�#���!E�$�E�$m��9E�$_.!�X�X�$%cz5��"%���k�%���*C%/!/��$%
��,_ %�W��%����%V�#����%&�^�%���$�z%��j�%LY�A^%|�$���%!/!��$%|Cl���%�"L!%V��B%�V�^
�%Ő$4��!%8�%KF%o��ֿ%����`%���p�%7����%���{ ��%�"��"�*%�d%����%r�pz%;/!��/�%����^
%6�
)�% �K��j)"%L<[g{J�XL!aƶJ�_�"��#�FJ�����#J���]"���mJ���'�$�"J����Y]�۶J�ˌ$��#��J�O*w�\;�J�*8)���
J�L�D �� |J�i��"�#��"J�G�MM��J�A�V|o|��J������nNJ�t�#��i$˓J�ʇ�?"�7�J�vp���wJ�˔�"�y%�\J�$�!��(8J�aX U��$w8J�����{ �J�8"!��#`[J���S��]J�/���$�$t�J����+���J�	G�C.J��~% G|{��#J��{ �]�:QJ�÷#��_\�\J�kH
 #�Y�"J��'�J!��J��*�+J��]%>�y�@"J��"`+�W�J�f�"�/�#J���b�V_%�J�-#M��J�J�����J�J��QIJ�3P����J�8�>aMJ��o��
��J�F�Zܔ:$J���S#^
�$J����"�
	#J����{
�J�%S#݅4�#J��X �#�|�J�9G�#� oeJ�����J�+�"J���!h��#u/!J���!fX�mL!J��d�nJ���J��L!J��Gb��\
w�J�%����"J���#DC[_
J�V��wD�J���K!�]
��J�r�?�J�i�C%J�.K!u"ΊJ�}���J�K�$�"���J�-�"����J���uz �#J��H�"4�#J�ݑ+���"J����{c"J��x{MY�kJ�`�"��
�"2$J��C$�s�2J��q9kpJ�\Z���wg+J��Xp���AJ�5I
QPG�
b� J���!��#����J��
����"J��_��!eJ��!��#?�4BJ���_b��J����_%y�J�3���� �+J���+��J���"F�$�L!J�	��R�J���؈�I
J����k��J��~�>SJ�ɷ�!m�"VSJ�D%ű�iCJ�G��Q��{J����$��#o_
J���xx��$J��L!,x�"(�J��_
��
��+�
J��_%�� ԇ1�J��!O�
���J�%�ҵ��:�J�,
,�J���nJ
�$%CJ�%C�_%���J���`%N��J����|�&�pJ��pM{%"�8�$J��eX"��CJ��9�l!DB"J��Sf�J����N��ExJ��"��K!��J��_
	^
M��_
J�p�3lԼJ�`x&�#P�J�.�ۖCZ m!J����:CZ �"J�
$`
M!���:Q�gm��$���:�1�+U���:����"���:�ޕּ#���:!�Cv����:
'�'o���:�W%o���:O�| � ���:8�#,� ���:P�
6v���:�W�!6v���:<T2!�����:�r��"�����:����MM!���:$`
q%MM!���:��
E`
���:�1w`
�"���:s���M�#���:��F��"���:~�8��"���:f��C��#���:��
�%"���:OC"DV����:����"�����:!�*�����:>���r���:�������:�H�`
�����:�% ,�����:��=,
s���:[,�"
s���: �"��^�$���:d�\�O����:,(5����:�|�\����:G�Է5����:��M!5����:�{�`
5����:IG�5����:%Dd�5����:5%�:n���:�;D�����:�n.�XD���:�%�M!
s���:�{8�"
s���:A�"gm��U�& �:[o�| ��$�U�& �:�����f�U��:p(gmy��U�� �:��gm�[�Ut��:��[|��U4�:�#��#���U4�:H_+I�D�U4�:����#�U4�:�$�_��U4�:8�!;��3�U4�:�����3�U4�:��"��3�U4�:�w��3�U4�:G�D��U4�:~��| ��U4�:�N
���U4�:�3!�%��U4�:��eIu��U4�:�`
E�u��U4�:��!Eu��U4�:%(9u��U4�:+6���U4�:ɽ�$���U4�:N�"и���U4�:S6�E���U4�:
������U4�:s,�$���U4�:G�#�M!���U4�:|����U4�:�gm�,�U��:�Ggm�,�U�L�:82Ġ���U-2�:{#O%T��U-2�:�?o�#�[�U-2�:SL
f�$K��U-2�:\����$�U-2�:3gm�,�U-2�:��E#�%�U-2�:�-�N(�U-2�:�-�| �U-2�:_
�nK��U-2�:��2��U-2�:�s���n�U-2�:��#�E�I�U-2�:�������U-2�:�&!)|G|�U-2�:G��M!�U-2�:[1��'\�U-2�:�"5�7��U-2�:��}\�U-2�:���|��U-2�:�$�����U-2�:ΠG"�`
�U-2�:h�#V�"��U-2�:�m%#�U�& �:��#�3��U�& �:������U�& �:�?'a
KN!�U�& �:������U�& �:��#ߴ$	��U�& �:`�,���U�& �:"-t�"��U�& �:�.����U�& �:0:%o��#�U�& �:*`$H����U�& �:6o�����U�& �:�\\�vN!�U�& �:��"(�I�U�& �:y�|����U�& �:<-Է�(�U�& �:�|�M!(J�U�& �:���`
F�U�& �:S�G�I�U�& �:o'PP�$�U�& �:Q��\���U�& �:Da
]e"�U�& �:�yT-Sr�U�& �:<F�0]�U�& �:5} No�U�& �:��_�%�U�& �:�(�N!;��U�& �:e��]m�U�& �:�oR} ӗ�U�& �:��#:�0��U�& �:�%��|�U�& �:h��]���U�& �::%�L��#�U�& �:�"��[a
�U�& �:�#����U�& �:q-�W��U�& �:��!��"�U�& �:s�?���U�& �:�.�"]��U�& �:@������U�& �:��"��6��U�& �:��$�n��U�& �:�N!"��"�U�& �:S������U�& �:��ċ��U�& �:gG��#@�U�� �:�
�c�a
�U�� �:yA�9��U�� �:IG| ��U�� �:e�
�!��U�� �:k��0���Ut��:vokJ��"�U4�:�!�(��U4�:�?7�-�U4�:F�!��a
�U4�:��$�N!��U4�:�";�"��U4�:{c "��U4�:�"�%��U4�:�b")D��U4�:�>���!�U4�:��u�"� �U4�:�m����U4�:��\F]I�U4�:�;�$�w�U4�:���]}� �U4�:����&]$�U4�:JY��}�U4�:�f%R�$�}�U4�:o�"q�}�U4�:����}�U4�:f"o} ��U4�:28���U4�:�eϋ��U4�:���%��U4�:���Y�
�U4�:�' ���U4�:
���F��U4�:�`��a�U4�:�1̪��U4�:��}��U4�:vFϘ��U4�:�FO!��U4�:.�F"f\$�U4�::Q��'�U4�:����"~&�U4�:��
���#�U4�:��s� �U4�:�'�D�U4�:]!�#���U4�:���"#p�U4�:ڊS�S}�U4�:4u�} S}�U4�:4!\�S}�U4�:[u�-S}�U4�:�N
?
.�U4�:���
.�U4�:��	�
.�U4�:@"�]
.�U4�:$f%	��b �U4�:_� �b �U4�:�1���b �U4�:�M
U��b �U4�:�Fb
�b �U4�:y��o�b �U4�:�3![�#�b �U4�:#~Z��b �U4�:�
�J��U4�:�w�U4�:���z?�U4�:��
��$i3�U4�:�#�"f�U4�:'�r�d�U4�:t�����U4�:�����U4�:l' �F"&�U4�:�+ m%\)�U��:!�^
Y�
�U��:���"k��U��:�K�} ���U��:������U��:�>�F�U��:�x0����U��:�� ĵ$�c�U��:7yLO!�-�U��:о	Q9%�U��:Jy�q�U��:�x0��y�U��:'z'����U��:�K/^�9%�U��:�"�}5��U��:�x���U��:<�C�6��U��:�"ڵ$��U��:s��x_$�U��:}���p��U��:bO!
p�"�U��:N j [�"�U��:(pEb
�}�U��:=���Fp�U��:�� wp�U��:C��wp�U��:[��Fwp�U��:[��wp�U��:��#F���%�U��:_&�}�p�U��:���&wp�U��:h�@��&�U��:�%cb
�p�U��:r�+ q�U�L�:�b
�$ú�U�L�:�b
�} ��U�L�:~O!�0^��U�L�:���0��U-2�:45q�F�U-2�:{������U-2�:\�
�&
�U-2�:��(Gd��U-2�:!p�&K�U-2�:�_����U-2�:G^"@�U-2�:�0��%�U-2�:�s�0�s�U-2�:~ G�%Gb �U-2�:�b
�����U-2�:n.a^���U-2�:~ �����U-2�:y������U-2�:�{'�O!�U-2�:{E^���U-2�:���)���U-2�:�����J�U-2�:X�$��#7~ �U-2�:��K�U-2�:n0�O!��#�U-2�:�0�)���U-2�:ī!'�U-2�:3!�O!�"�U-2�:�d%����U-2�:�&���U-2�:��"$���$�U-2�:~��"P!�U-2�:Sq� �)�U-2�:��"e�%��"�U-2�:n~ �.S'�U-2�:� �~ mq�U-2�:� �3��U-2�:���*�U-2�:�#4~w�U-2�:����%�q�U-2�:FGx'aG�U-2�:��'י�U-2�:����d��U-2�:R~D�c��U-2�:�G�q�^�U-2�:���q<*�U-2�:N��qPK�U-2�:�^�q���U-2�:��qh��U-2�:0P!�q��U-2�:�"�q��U-2�:�G�q���U-2�:�G�q�'�U-2�:OP!�q"�U-2�:p�1�?��U-2�:khP!n~�U-2�:�O���U-2�:�p%�~ ~K�U-2�:���6�$�U-2�::
�q�'�U-2�:����A"�U-2�:�o�^B��U-2�:��^m��U-2�:�d{���U-2�:�G�~ �b
�U-2�:����"�U-2�:�06����U-2�:�b���U-2�:���~ �~�U-2�:ɖ�.��U-2�:��$���U-2�:ԇ�����U-2�:��T��~ �U-2�:~"�.���U-2�:
�"9 S��U-2�:e!"!�U-2�:�
�~���U-2�:N�r�$�b
�U-2�:�3!��$�b
�U-2�:I�#_���U-2�:
;_�.�U-2�:�"$(H�U-2�:>e>c
B(�U-2�:�qY���U-2�:��w�k��U-2�:�a i*C�"�U-2�:�Ӣ�K�U-2�:��$϶$R��U-2�:�*r��U-2�:x(-�#"/�U-2�:�"9r�*�U-2�:R���R��U-2�:bZ!��U-2�:I�W w�"�U-2�:�d\c
u �U-2�:<�Wr��"�U-2�:Ћ� b/�U-2�:
b �"���U-2�:r�չ�*�U-2�:x!K�#�U-2�:����%��%�U-2�:���,�U-2�:3�/��U-2�:��ur�!�U-2�:��r�Y_�U-2�:��(��$�U-2�:��i�#�/�U-2�:�r���_�U-2�:�/�P!�K�U-2�:M�����U-2�:9H���_�U-2�:����_�U-2�:WH���"�U-2�:���(܍�U-2�:(L�(&`�U-2�:���(zc
�U-2�:CL�(�(�U-2�:���(�$�U-2�:j�(�c
�U-2�:�/�(�r�U-2�:0�(�"�U-2�:�%�(	s�U-2�:�(�(���U-2�:F�U�$��U-2�:3��! �"�U-2�:�NU`=�U-2�:����(�U-2�:	�����U-2�:`"��Y��U-2�:s�$s`�`�U-2�:���#ٻ�U-2�:��� ^�"�U-2�:��=s��U-2�:�%��$`�U-2�:���\��U-2�:z�P!��#�U-2�:� /�+�U-2�:��l�*�%�U-2�:��^L��"�U-2�:̚Q+���U-2�:��#���U-2�:�}����U-2�:�N�[s�U-2�:���s��U-2�:�M���U-2�:o+�7)�U-2�:��"|L�`�U-2�:�"^�%�`�U-2�:$��!u �U-2�:k���u �U-2�:�+���U-2�:��c
��U-2�:����U-2�:�"<0�`�U-2�:�6Z0 ��U-2�:(�:�"#t�U-2�:��j��|�U-2�:�"�����U-2�:p)O����U-2�:���P!-��U-2�:��$x0��U-2�:^em��)�U-2�:z
a���U-2�:uH׷$��U-2�:@�/��c
�U-2�:h�F�Hd
�U-2�:9a�/��U-2�:d��"L��U-2�:SM
�+�"�U-2�:n��)�+�U-2�:���s`a�U-2�:	���d
�U-2�:���ė�U-2�:|�%�L��%�U-2�:�)*����U-2�:D	�a݂%�U-2�:���!���U-2�:ێ�����U-2�:j�j���U-2�:O
����"�U-2�:���d
F��U-2�:�e%ҺB��U-2�:����U-2�:��d
�L�U-2�:�0��n	�U-2�:�=���U-2�:�"H��U-2�: �%���U-2�:�a�0��U-2�:i�>�%�H�U-2�:X�"\�%[��U-2�:
������U-2�::",M��#�U-2�:6�@��U-2�:��
�)��U-2�:*@*,�U-2�:T
��"��U-2�:>^���U-2�:��"+�g,�U-2�:���0���U-2�:|_"��U-2�:d�I����U-2�:�H�s��U-2�:�	��g��U-2�:1�a
� �U-2�:M���"�U-2�:���,}"�U-2�:��#��"�.�U-2�:��$�b�U-2�:�$��p"�U-2�:p��P!p��U-2�:�P!z�e
�U-2�:���"2��U-2�:z�%��^*�U-2�:JMT���U-2�:��"/��U-2�:���"�#�U-2�:	
��%_b�U-2�:,1��r��U-2�:ce
�"��U-2�:��O��e
�U-2�:hM�,w� �U-2�:��"ă%�b�U-2�:Hc�st�U-2�:ȏ��%�U-2�:vt���"�U-2�:-������U-2�:(��m��U-2�:o�#���U-2�:�t� /#�U-2�:X�%�*���U-2�:'�X1U��U-2�:%Iy�%�U-2�:D���%K��U-2�:����Q��U-2�:���#���U-2�:�t�����U-2�:ԝۣ�U-2�:��f�8�U-2�:#Q!܄%1��U-2�:v1��#�e
�U-2�:�#� ]��U-2�:��6
��%�U-2�:2f
�����U-2�:�"T
�,�U-2�:)"��#��#�U-2�:E�$�����U-2�:EQ!Q"��U-2�:�scD�U-2�:%��c`f
�U-2�:)� '���U-2�:*�)�#�1�U-2�:|�W���U-2�:���*u��U-2�:o"M-s�$�U-2�:����#�U-2�:E��tP� �U-2�:���$e��U-2�:r
tQ!��U-2�:uY�%�M�U-2�:��(y��U-2�:Ծ�qN�U-2�:���(CI�U-2�:��q��"�U-2�:�gm��Q��
�:u���#�|��� �E$��"%����lx�$���fv�#�K!e �n��rʦsx���NP��"�x����$��$��-�+�*�m&��/!���D:0!��E$4Ds�����$�N@
�-g#�� ��$Z��!� �;N�?��
�-smuZ �Ş�;^E֏"�2���Z�_�'�-�V��(�M�UD0"X��;Q�$����� R�q@���J
Bus�ٻk-w�%+�
k-G�#�� �f
k-�2ڑ�k-�"r-�
k-g��
�ͺW�#�!�ͺ�� T2!d�#ͺ�Q!��"~�ͺ�Q!āE$ͺ�f
gP��ͺ+�q �ͺ@N�
�� ͺ ����f
ͺ}-��$A�ͺ�c���ͺ��%� ��ͺ�Q!2� ͒ͺqI�$�ͺ��"�c�Q!ͺTuR!,ͺ����ͺٞ�u�ͺ��#�I��ͺR!�Iҙͺ�c�"�ͺx�puN�$ͺH��cUg
ͺρ6�=� ͺ�g
*�%�ͺd/d2R!ͺ2{R!U�ͺr��#�ͺ��"�$�g
ͺ�g
5�h
ͺ���IL�ͺ�I��#�ͺ������#ͺ!+��"ͺu��K�ͺ�ʮͺo��umͺ/2��"�uͺ\N���ͺ�"ɂzNͺ+�$�".�ͺK�$��S2ͺ5�"]���ͺ�R!> "��ͺÚ�$�ͺ�
�N�$ͺ�R!�0�ͺ\%���ͺ�-k�$s�ͺT�"_�#e�"ͺ��OT "ͺ�-����$�ͺ�-
v�"�-ͺ�-����0ͺ�-<+4v�-ͺ�-b.!Oͺ�-�C�0�������Oh
��_��� �Jv��#ֽ�I�\h
O�v�.�H�EdP+�����$7�.[+�����u�#F��.����%�����ȹ$���%���߂@���#�K�p����"]	����i��R!Yd�� �S�"�2~ "��
�	�$i
���� i%��� y+�2��I���"�ß�v�+��R!��+���օ%`�X��&O�::i
	�`�X��&O�:���`�X�ǯ&O�:\i
od`�X�܂ &O�: ���`�X�~i
&O�:{�"`�X�~i
&O�:S!�`�X�~i
&O�:���`�X�~i
&O�:�$�`�X�~i
&O�:/� 	�`�X�~i
&O�:�d�`�X�~i
&O�:�"r�`�X�~i
&O�:D3#S!v%a�
l�J��#\�a�
�]3�a�
��B�
��a�
�/����"�da�
� "�
�11�a�
\]!��w3a�
��$]� ���a�
x����v��#a�
-O�d0��a�
3�x�gm��#a�
�x!a�
�n�����a�
6S!�"�1�i
a�
��""Jc�a�
�
"J��
wa�
ۦ�9J)�a�
e�:,NJy�a�
m��!�%�$a�
�1�S!%�":�"a�
���%UN� a�
�Qt���a�
�i
��
0$�%a�
	�$�����a�
t�#j
���a�
H��""����a�
�6��d^�"a�
7�#��w�� a�
����#cJ��a�
���""R��S!a�

���3�a�
����}���a�
�2
�%e�""E�$X#"8Q�E�$���3�E�$��"T!3E�$���qeE�$��"��"E�$�?��E�$!T!¢Z,E�$���%U�E�$NQ_�wE�$��5�%s#"E�$"��#"K�%E�$�"v�G3E�$�%�#"�#E�$
&�5T!E�$YT!9�|,E�$m�%"�#g�E�$�!��xJE�$��ȼ$�3E�$3�#�j
cQE�$K�#/e��"E�$u����JE�$��I��"E�$ǧآ�j
E�$���,�E�$��޼$E�$"�$9�"��E�$D�2��E�$9�F�$VE�$�g�EeE�$�<w�QE�$�J�Q%���#"��%}��� ��%���3�,%\�$0&%�#"U�k�%�e�j
,�%F&_�T!%j3�3�3%���Gw�T!�e�!-�Q�3��3��� ,-��v&���� U!7� �O�"��Q
$"��H��faf��)$"]�-U!���%ɂn4��3��4��%��x� �F$"�����J\��3�c�#;�Lw��J��n�"������!��-dw����-Q���tw������Gw��-4��І%���"�&	R߆%uf��$�u�_�A4��4�� ٜ�ߧ�"���#K�� ��^U!�w�
R�������#����:ʲ��.�B-����:/K��.�Bus��%k-w�%+Rk-G�#�� �f
k-�2ڑ�k-�"��
k-����� k-]��&��k-�-v��wk-�Q!2� ͒k-qI�$�ͺ��"�c�Q!ͺTuR!,ͺ����ͺٞ�u�ͺ��#�I��ͺR!�Iҙͺ�c�"�ͺx�puN�$ͺH��cUg
ͺρ6�=� ͺ�g
*�%�ͺd/d2R!ͺ2{R!U�ͺr��#�ͺ��"�$�g
ͺ�g
5�h
ͺ���IL�ͺ�I��#�ͺ������#ͺ!+��"ͺu��K�ͺ�ʮͺo��umͺ/2��"�uͺ\N���ͺ�"ɂzNͺ+�$�".�ͺK�$��S2ͺ5�"]���ͺ�R!> "��ͺÚ�$�ͺ�
�N�$ͺ�R!�0�ͺ\%���ͺ�-k�$s�ͺT�"_�#e�"ͺg��
�ͺW�#�!�ͺ�� T2!d�#ͺ�Q!��"~�ͺ�Q!āE$ͺ(x�-��ͺ̖��#�4ͺ�-�-P4ͺ���fX�ͺ+�q �ͺ@N�
�� ͺ ����f
ͺ}-��$A�ͺ�c���ͺ��%� ��ͺ�J�Qͺ���#"��ͺ��OT "ͺ�-����$�ͺ�-
v�"�-ͺ�-����0ͺ�-<+4v�-ͺ�-b.!Oͺ�-�C�0�������Oh
���_��� �Jv��#ֽ�I�\h
O�v�.�H�EdP+�����$7�.[+�����u�#F��.����%�����ȹ$���%���߂@���#�K�p����"]	����i��R!Yd�� �S�"�2~ "��[R�4.�
�	�$i
���� i%��� y+�2��I���"�ß�v�+��R!������օ%`�X��&O�::i
	�`�X��&O�:���`�X�ǯ&O�:\i
od`�X�܂ &O�: ���`�X�~i
&O�:{�"`�X�~i
&O�:S!�`�X�~i
&O�:���`�X�~i
&O�:�$�`�X�~i
&O�:/� 	�`�X�~i
&O�:�d�`�X�~i
&O�:�"r�`�X�~i
&O�:�
�1�a�
Cx]!��!za�
��$]� ��a�
x����v��a�
-O�d0��a�
3�"J��/�%a�
؇%U��a�
��t��a�
Fz
0$�&a�
r��sU!a�
j
���%a�
�""��95a�
6��dA.a�
��w�.a�
��cJ��%a�
��R���a�
�""�3�a�
��#}���%a�
dz���j
a�
y�%���a�
��#lR� a�
���4��a�
"ɗ�E�$³׆w$"��E�$�,k
�#E�$��$"��"E�$
�%e��%E�$�R
$6E�$�h�E�$X#"8Q5�%E�$���3�RE�$��"T!�RE�$ߗ�} �E�$��#<���E�$[�#��d�#E�$���qeE�$�,"�"E�$��"�/E�$��"�?�nE�$!T!¢��E�$���%��#E�$�z��$j'E�$G�7"�zE�$K6����E�$��#p�4�E�$�{�/E�$NQ_��E�$��5�%&�E�$"��#"e�E�$Ê%v��'E�$�%�#"�E�$
&��E�$YT!9���E�$��"�#��#E�$�!��$�E�$��ȼ$�3E�$3�#�j
cQE�$K�#/e��"E�$u����JE�$��I��"E�$ǧآ�j
E�$���,�E�$&�#M�0E�$��:�E�$%V!9�"�'E�$5�2�Ƚ$E�$9�F�$VE�$�g�EeE�$9�#�4��E�$Bk
�%R0E�$s6�0Q�E�$����$"E�$;�w�4E�$��hk
4{E�$5L��^�E�$[L�6�LE�$�	�%�6E�$��Ķb�E�$�<w�QE�$4�"���E�$��05
%��{CV!%�%ڶ��%A� �0bS%^�#��|V!%�0�f@�%���7%��#��#�{%�B�c�%~k
M"��%R�$��B�%A�"1�$"%���� �%}��3��%��0&�,%\�$U�%�#"�j
z�%F&_�T!%j3�3�3%���Gw�T!�e�!-�Q�3��3��� ,-��v&���� U!7� �O�"��Q
$"��H��faf��)$"]�-U!���%ɂn4��3��4��%��x� �F$"�����J\��3�c�#;�Lw��J��n�"������!��-dw����-Q���tw������Gw�-4��І%�m� �&	R߆%uf��$�u�_�A4��4�� ٜ�ߧ�"���#K�� ��^U!�w�
R�������#����:ʲ��.�B-����:/K��.�Bus��%k-w�%+Rk-G�#�� �f
k-��#2ڑ�k
k-�S�"��
k-���0�#k-]��&/W!k-�-v��%"k-�Q!2� ͒k-qI�$�ͺ��"�c�Q!ͺTuR!,ͺ����ͺٞ�u�ͺ��#�I��ͺR!�Iҙͺ�c�"�ͺx�puN�$ͺH��cUg
ͺρ6�=� ͺ�g
*�%�ͺd/d2R!ͺ2{R!U�ͺr��#�ͺ��"�$�g
ͺ�g
5�h
ͺ���IL�ͺ�I��#�ͺ������#ͺ!+��"ͺu��K�ͺ�ʮͺo��u�W!ͺ/2��"��ͺ\N���ͺ�"ɂݝͺ+�$�".�ͺK�$��-(ͺ5�"]�X�ͺ�R!> "��ͺÚ�$�ͺ�
�N�$ͺ�R!�0�ͺ\%���ͺ�-k�$s�ͺT�"_�#�W!ͺg��
�ͺW�#�!�ͺ�� T2!�ͺ�Q!��"ۤͺ�k
ā{�ͺ(x�-��ͺ�]���ͺ̖��#�� ͺ�-�-K�%ͺ���f�ͺ+�q �ͺ@N�
�� ͺ ����f
ͺ}-��$A�ͺ�c���ͺ��%� ��ͺ��71ͺ�S����ͺ���7ͺ�J�Qͺ���#"��ͺ��L|�"ͺ(&"B� @�ͺ��s��ͺ��OT "ͺ�-����H&"ͺ�-
v�"�-ͺ�-����0ͺ�-<+;5�-ͺ�-J�>��&"ͺ�-�C�0���Y� ����$���_��� �Jv��#ֽ�I�6O�v�.�H�EdP+�����$7�.[+�����u�#F��.����%�����ȹ$���%���߂@���#���#K�p����%]	����i��R!Yd�� �S�"Y1������2~ "��
�	�$i
���� i%�c"T��"���� y+�2��I���"�ß�v�+�sX!�G���օ%`�X��&O�::i
	�`�X��&O�:���`�X�ǯ&O�:\i
od`�X�܂ &O�: ���`�X�~i
&O�:{�"`�X�~i
&O�:S!�`�X�~i
&O�:���`�X�~i
&O�:�$�`�X�~i
&O�:/� 	�`�X�~i
&O�:�d�`�X�~i
&O�:�"r�`�X�~i
&O�:�
�1�a�
Cx]!��!za�
��$]� ��a�
x����v��a�
-O�d0��a�
3�"J��/�%a�
؇%UZ�a�
��t�e�#a�
Fz
0$xa�
r���a�
j
���%a�
�""��95a�
6��dA.a�
��w�.a�
��cJ�fa�
��R�{�a�
�""�3��a�
��#}���%a�
dz���j
a�
y�%���a�
��#lR� a�
���4��a�
"ɗ�E�$³׆w$"��E�$�,k
�#E�$��$"��"E�$
�%e�E�$�R
gE�$���"E�$X#"8QT�E�$���3|E�$��"T!�E�$ߗ�} #'"E�$��#<�)�E�$[�#��d�#E�$���qeE�$��?�"E�$��"�/E�$��"l�o1`�E�$�����E�$�)E�$�?�nE�$!T!¢��E�$���%��#E�$���$d� E�$�"7"iE�$̙���X!E�$�����%E�$�p��E�$>�{�� E�$ҥ�F'"E�$�X!y���E�$O�md�%E�$"��k
�1E�$�(�SOgE�$�gT#8E�$NQ_�H�E�$��5�%�E�$"��#"r�E�$Ê%v�ȉE�$�%�#"�LE�$
&��E�$YT!9���E�$��"�#<�E�$�!��$�E�$��ȼ$�3E�$3�#�j
cQE�$K�#/e��"E�$u����JE�$��I��"E�$ǧآ�j
E�$�|���1E�$�X!V�%TE�$&�#M�0E�$��:�E�$%V!9�"�gE�$5�2�\�E�$9�F�$VE�$�g�EeE�$$M�4�(E�$b8�%hE�$���0��E�$�"�|
�E�$]�����E�$|�w��E�$��%hk
cTE�$�"����E�$�T�6DME�$Ѝ%���8E�$�8	�%��E�$F5Ķ��E�$��#��<�E�$_5�<w�QE�$2�$)E�$��05
%��i��)%oh	�"�%����8%�0�f�%���ئ%�T��#�%�B�c�%�MM"�%#��)2%T�1�%�k
�3�%�"0&�,%UU�l�%�� �j
��%��_�N�"%���3��%���Gw�T!�e�!L��Q�3��3��� ,-��v&���� ��$7� �O�"��Q
$"��H��faf��)$"]�-U!���%ɂn4�(���4��%��x� �F$"�����J\��3�c�#)Lw��J������'"�����-dw���-Q����������? ����%�Gw:�-4��&���&l
��"�����6�|	��|*�d��ƷJU���І%��'"�&	R߆%uf��$�u�_�A4��4�� ٜ�ߧ�"����MGl
`U�^U!�w��)����#���#����:ʲ��.�B-����:/K��.�w�%+FNk-G�#�� -� k-2ڑ�)k-_���%k-��#ֽ��#k-O�vy�k-EdP+hUk-7�.**k-����%�6k-ȹ$�53k-߂@��k-Y1���3k-�2~ "F�k-c"T�K�k-���#��k-�����k-�n�ͺ�����hͺl�o1��ͺ��{&�ͺip���%ͺ+�#����ͺ����ͺ�67"��"ͺ0i��$�ͺR��}ͺȾ$�N@�ͺ^�"T|�ͺ�N�S��ͺH#�|}�"ͺϊ�k
2}ͺ�mͺ޷y�7ͺ�N*��ͺP����ͺSi����ͺh#��.�ͺ������ͺ����#ͺN�JUw�ͺ5�2�B�ͺ%V!9�"r�ͺ����ͺ����ڎ%ͺ�g��� ͺ9�F�$�ͺ� �N�$ͺ�#w*�ͺ�����'"ͺ\l
�|�%ͺui�0Mͺ��%��ͺ���4�3ͺ�h�d7ͺ�9�*՝ͺM�#Ķ�� ͺI�	�%�iͺr�%���iͺ�*�6��ͺ7("���}ͺ�hk
�#ͺ��
+ͺ�N���"ͺ��#���$ͺ�c�"�7ͺR!�I��%ͺ��#�I��ͺٞ��%ͺ����}ͺTuR!��ͺ��"�c�9ͺqI�$S("ͺ6���$,�ͺx�puLͺ�U��Nͺ5�"]�<$ͺK�$��5�ͺ+�$�"Oͺx�\��$ͺ�R!> "Y!ͺ�"ɂ3�ͺÚ�$�$ͺ\N��ͺ/2��"h�#ͺo��u�"ͺ�ʮ��"ͺ:�%)�"ͺ�
�N)ͺ�a�w�ͺSjQ%5#ͺyl
�Ug%ͺ��~��#ͺK�B�:	ͺ
8Y�%I4ͺ�����ͺ�("��6� ͺX��C+ͺ8�$��?�ͺe����Uͺ 8ij"~ͺ�%�	#ͺ�+�O?8ͺ����l
ͺ��#c�O�$ͺV&N�ͺ�~y��4ͺ�O:v�ͺ�� k�$��ͺ��#�#��ͺA���Oͺ_����	ͺ&&o��$ͺ�5�#�#ͺֈ o�%�~ͺè��%�ͺcY!� �
ͺ��_�#��ͺ�� T2!5�ͺ�Q!��"��ͺ��~Y!ͺ�S��žͺ��79m
ͺ��s��8ͺ(&"B� � ͺ��L� ͺ�w0ͺ�-�j��$ͺ�-�����ͺ�-t�.��ͺ�-s#am
wm
ͺ�-o�j�ͺ�-ި&���ͺ�-�m
)�#��ͺ�-D���{�ͺ�-��,�ͺ�-9<>Pͺ�-R�۞ͺ�-�$l�b�ͺ�-��$z�jͺ�-<���/Vͺ�-'��m
n
ͺ�-�Y!�5�ͺ�-
�%e?�#��R
N���0:�ߗ�} ����#<��Y!��<w,���C�0k�Y� ��6� �
�	�(�$���� q�#��5�P4k���#��%��nn
�#������#k��<�/,M�$��n
�#d����5�P�6K:)9�[��P��#�ß�vW:��I�i:�sX!�J�����("`�X�Τ&O�:D��`�X�H� &O�:��%`�X���&O�:�V�k`�X��:&O�:�/�`�X��:&O�:�("��`�X�h�&O�:���#`�X��� &O�:I9�#`�X�Ȑ%&O�:�VG�`�X��&O�:�("��`�X�E,&O�:@&6`�X�_�&O�:���#`�X��&O�:�V܉ `�X��Y!&O�:J�o�`�X���#&O�:)"e&`�X�l�&O�:�n
A`�X�&O�:�%0�#`�X�;)"&O�:�V�)"`�X���&O�:� ;`�X��,&O�:7�%!W`�X���&O�:��4�`�X�L�&O�:���`�X�\�%&O�:� �n
`�X���&O�:�)"m9`�X���&O�:��$��%`�X��)"&O�:�,�9`�X���&O�:H�#`�X�|&&O�:G�*"`�X�e�&O�:X�,`�X��&&O�:��A�`�X�:&O�:S4;`�X���&O�:l���`�X��n
&O�:�P�k`�X�Q&O�:L;&*"`�X�H�#&O�:Y�KQ`�X���&O�:�#��$`�X�-&O�:�k�`�X�W�&O�:�l`�X���#&O�:}��`�X���%&O�:����`�X�m�&O�:x�9`�X��&O�:��#`�X��&O�:q;��`�X�>� &O�:��ɋ`�X��&O�:��#6�`�X��;&O�:�� ��`�X�9W&O�:>*"-6`�X��&O�:o
�`�X�'l&O�:!��`�X��&O�:tl
�%`�X��9&O�:G��l`�X�e*"&O�:�W�`�X�:&O�:m��W`�X��&O�:�l�`�X���&O�:�lc-`�X�ɟ&O�:{-Z!`�X��#&O�:Ao
�`�X��W&O�:�*"��`�X���&O�:*#�`�X�N�&O�:�5`�X�ho
&O�:���o
`�X��$&O�:f��`�X�'&O�:�; �$`�X��&O�:�l��`�X��&O�:"�%
`�X��-&O�:,Z!8�$`�X�V:&O�:V���`�X��&O�:��`�X��-&O�:Q# 
`�X�M&O�:�� E6`�X�ڊ &O�:����`�X��:&O�:��ʪ`�X��&O�:6� cQ`�X�VZ!&O�:e'm`�X���&O�:P�$z�$`�X���&O�:�W�#`�X��&O�:a�o�#`�X�4�&O�:`� �`�X�{Q&O�:X�*"`�X�#�&O�:8
�*"`�X��&O�:��{#`�X�
+"&O�:<]6`�X��Z!&O�:b
l+"`�X�;&O�:.ܫ`�X���&O�:m��`�X�b�&O�:��#$�`�X��'&O�:�� ی`�X�Z.&O�:[!�`�X���&O�:�QDm`�X��#&O�:3��.`�X�Y�&O�:�o
��`�X��� &O�:<��`�X�^�&O�:����`�X�u6&O�:m;�`�X��&O�:L�%S�`�X�5�&O�:��$#�`�X��;&O�:6<��`�X���&O�:���o
`�X��+"&O�:�.��`�X���$&O�:?Xw�%`�X�k�&O�:�'��`�X���#&O�:p
�$`�X�2�$&O�:���`�X��+"&O�:jX�X`�X�Ȭ&O�:\m�`�X�<[!&O�:��%�`�X���%&O�:�#�[!`�X�@p
&O�:;�a<`�X��&O�:(��`�X���$&O�:� �[!`�X�>,"&O�:Y�`�X��&O�:�
B�`�X�>�&O�:��?� `�X��&O�:�X�6`�X�=�&O�:g�`�`�X�Z�&O�:x��m`�X��&O�:���.`�X��&O�:�
�6`�X�֦&O�:�[!�%`�X�y�&O�:D(�,"`�X��#&O�:���;`�X��&O�:7o�#`�X��[!&O�:�Xf�`�X���&O�:~�	`�X�#&O�:��R`�X���&O�:/R`�X��&O�:L��,"`�X���&O�:��$��`�X�d�&O�:��07`�X�6�&O�:��<`�X���&O�:غx�`�X�'<&O�:
��`�X���&O�:@/2R`�X�H7&O�:Y�`�X�W� &O�:JR]�`�X��7&O�:p(�,"`�X���#&O�:!J�(a�
���1/�a�
����?a�
��x!�� a�
r����a�
Fz
0$U\!a�
��t�Ыa�
؇%UW�a�
�%�"qRa�
"J��}#a�
��#}�z<a�
�""�3��a�
��R��%a�
��cJR�a�
�\!�ûa�
�g/#�a�
ĭ,Y��a�
�|/�\!a�
��78a�
�,"�ϧa�
-8��֌ a�
�
�$��a�
��\!��a�
y<�#1�a�
�mW���#a�
�$��a�
u��/A�a�
���(K�a�
9��(O�a�
�<�m�<a�
���(�/a�
)y�*)a�
[)A\�a�
� �"� a�
[8j���a�
����Ra�
�p
y���a�
-�$V�a�
�p
�\!})a�
���-"a�
�;-"�a�
=6�%AYa�
P-"�#��a�
,�l�a�
;��m2�a�
�\!���/a�
��A���a�
�#��7=a�
�p
K�%�ma�
u���a�
����&�a�
`�%iYV�a�
���a�
|�%_��)a�
ȎO�$c�a�
�p
&Sa�
!�#k�a�
��%�#
]!a�
d�$ SO=a�
��[�q
a�
��t�;a�
�t=��a�
�#nB�a�
��v�9�a�
k�v-"R�a�
?q
.�:]!a�
����a�
!n�<:�a�
p�}8��a�
�)�8da�
��$��$Cna�
�)#H� a�
b]!��v�a�
�=��$��$a�
�]!�<��a�
�����a�
_q
~YA�#a�
`�����a�
��%hn�<a�
��$p� �$a�
�-"���]!a�
%#���a�
v��=�� a�
G�$g�$t�#a�
�� Ӯ�-"a�
}n����a�
C�C#��a�
��n|�$a�
��%=a�
}q
X#��$a�
����#c�a�
�nE~�a�
��$�Y�a�
���a�
ٍ �ԓ%a�
5S��Ăa�
�/���na�
�=0�$a�
�Y�20a�
:=��#m#a�
��SSd0a�
Z*�q
a�
�#���Ya�
o$*9*a�
]��]!hSa�
����a�
����a�
K�+�a�
`�{�5a�
���g."a�
1�����#a�
�=
��8a�
�#Z=�q
a�
�0�%�a�
�A��� a�
%`�Sa�
�."�r
�0a�
���r
�a�
.���0a�
141 �a�
�r
��$��a�
 o"� ra�
��#e*R�a�
z*�I1a�
�8�0�a�
�8�L�a�
����Ya�
�r
t�>oa�
��$�*�#a�
p1~���a�
���49a�
��#lR�*a�
y�%�Ea�
dz��ra�
�S^U!�#a�
<9��a�
�q��=�*>u�:�%k��=�*L�u�:��$k��=�*�u�:�k��=�*�u�:�$���=�*7� u�:m����=�*�r
u�:s
���=�*��u�:��e��=�*�u�:*�e��=�*�1u�:poe��=�*O9u�:�M	�=�*}�u�:v9M	�=�*o=u�:�SM	�=�*�]!u�:"ɗ#E�$��"T!oE�$���3��E�$X#"8Q�E�$���q$s
E�$��?ZE�$�!��1>E�$��"�#^!E�$YT!9�_>E�$
&�V�E�$�%�#"�E�$Ê%v�(�E�$"��#"�*E�$��5�%��E�$W�#�� �E�$NQ_���E�$�X!V�Fs
E�$�|���E�$�=��E�$�k
ā{�E�$���f�."E�$�-�-��E�$̖��#��E�$�]�o�E�$9�"�E�$�-v��SE�$]��&�E�$���/"E�$�oQ��oE�$��M/"%X�05��%t�#�0�$%
�ڶ�%��1�'�%�1~#B�%�	�"G�%=+i���%T�1�%#��G�%�MM"ֽ%�T��#5	%�����#%�0�fy	%����/"%���3ͩ%��_�<�%�� �j
tZ%UU��	%�k
�3�Z%�ĐF�$%�>�1�%�Zېg�%a�J���8��J�Gw�>.j��J�Gw?u�j��#J�Gw??pj��/"J�Gw���^!j�@�J�Gw�^!�^!j��1J�Gww+{���#'�J�Gw�����#��J�Gw������9J�Gw���T���9J�Gwt�ps
���^!J�Gw��D���?J�Gw9_!Y��s
��J�Gw�����s
ل�Gw�T!�e�!L��Q-��T��=,-���+��ٰ��˭�6�::��"[�'p
$"��3��+�%af��0"�5�"����p_!��"�?��^:�a2��%��k�������%��x��?�����:�%�:��@�e��J����%����'"���#��d��}���G�����͑�Gw? ����Gw:�-4��Si&�6�>�h#�"6�@��H#��6��_!��N��$6�
�N�G�6��T�����6�=�iV[6�0@�+�#Y� 6����B>6��2��6q6���0iT�6�m���:�&	R�s
�9�߆%	R���NQu�	R��W�#�4	R��$��=ߧ	R�:��1Gl
`U#��S�N0"�Y� �)`0"���#�:��
�#.�A����#�:�:",.�n>���#�:�<��.��� ���#�:�0"_�.�F,���#�:��C�.���$���#�:)#�� .��>���#�:~�c2".�����#�:X��>.�M�����:���.�Ύ ���#�:��%�2".�x@����: ���.��s
���#�:!���.�[���#�:���.�'��Gw�:6��[�����:t
<t
��Mt
�Gw�:���[���2�Gw�:t
�[��`�%��_!�:3g�������_!�:��$
���>��_!�:(�<
��Mt
�Gw�:���[�����:t
<t
���_!��:�[����;3��:�@����Bus�ٻk-w�%+�
k-��%"jpͺ�%'�
%�1��
��%�%��R%N
�U��x
(��#��
NO��x
/��~�e����x
{���-�#�T���x
?����%�+,�x
���2" \�T���x
�p��g��<��x
� �F�G�$�D�&��~E�,'I�~E�,'�Y��l�'&_�6�`e(7"�6�`e(z�6�`e(���6�`e(��6�`e(j�����(N�����(J�$����(�����(�(ώ���(�(Jt���(�(�
~E��j)�68_���)7"�d!�`E*d�"�
��*`�X��t!+��X��t!+��r��Hz+�r��Hz+v�r��Hz+#r��L�+�d!���$,�E���$,�d!����-&�
����-�Y����-[�[��X;-��r��h�-s�r���1.�Yr���.�Y���@/�x
��4�/Ю��Hd2���Hd2����2ϵ��Hd2�/���3����3�x
����3��x#4<_���B4����B4m��� �4�d!r��D�5������5&7"���5�d!��X#6�r��D�5N_��Ъ6ͮ���} ?�x
	URE<�#�>�x
D?_�x�6'�3�=���
���ohV@�
���6�T
I���op	d%&&
(&�o���o����oH�6����YY@�% ��$4�\�#j,��)��D��#ɖ֖q��686"ɖ�s<�6V&f&v&�&�&�&�&�&�&�&�&&&&&6&F&V&f&v&�&�&�&�&�&�&�&�&&&&&6&F&V&f&v&�&�&�&�&�&�&�&�&&&&&6&F&FreeBSD clang version 18.1.5 (https://github.com/llvm/llvm-project.git llvmorg-18.1.5-0-g617a15a9eac9)Linker: LLD 18.1.5 (FreeBSD llvmorg-18.1.5-0-g617a15a9eac9-1400006)libpmc.so.5.debug�x+.shstrtab.note.tag.dynsym.gnu.version.gnu.version_r.gnu.hash.hash.dynstr.rel.dyn.rel.plt.rodata.gcc_except_table.eh_frame_hdr.eh_frame.text.init.fini.plt.ctors.dtors.jcr.init_array.data.rel.ro.dynamic.got.relro_padding.data.got.plt.bss.comment.gnu_debuglinktt������o�+���o���:���op	p	�DPJTTIR	B����
[	B@�
@�
�d2��
��
<�l��%��%t~p�%p�%����%��% � �% �%�U�&&�(&(&�@&@&�P%&P&�X%&X&�`%&`&�d%&d&�h%&h&Ԍ�<�6<�6���6�6,�H�6H�6��H�6H�6|���6ē6���6��6��
0��6�<�6T�6"

Anon7 - 2021