diff --git a/lib/libpmc/pmc.westmere.3 b/lib/libpmc/pmc.westmere.3 index 0b8934e2829c..b80aefaf3c23 100644 --- a/lib/libpmc/pmc.westmere.3 +++ b/lib/libpmc/pmc.westmere.3 @@ -92,15 +92,17 @@ Configure the Off-core Response bits. .It Li DMND_DATA_RD Counts the number of demand and DCU prefetch data reads of full and partial cachelines as well as demand data page table entry -cacheline reads. Does not count L2 data read prefetches or +cacheline reads. +Does not count L2 data read prefetches or instruction fetches. .It Li DMND_RFO Counts the number of demand and DCU prefetch reads for ownership -(RFO) requests generated by a write to data cacheline. Does not -count L2 RFO. +(RFO) requests generated by a write to data cacheline. +Does not count L2 RFO. .It Li DMND_IFETCH Counts the number of demand and DCU prefetch instruction cacheline -reads. Does not count L2 code read prefetches. +reads. +Does not count L2 code read prefetches. WB Counts the number of writeback (modified to exclusive) transactions. .It Li PF_DATA_RD @@ -181,7 +183,8 @@ All Store buffer stall cycles All store referenced with misaligned address .It Li STORE_BLOCKS.AT_RET .Pq Event 06H , Umask 04H -Counts number of loads delayed with at-Retirement block code. The following +Counts number of loads delayed with at-Retirement block code. +The following loads need to be executed at retirement and wait for all senior stores on the same thread to be drained: load splitting across 4K boundary (page split), load accessing uncacheable (UC or USWC) memory, load lock, and load @@ -225,9 +228,10 @@ ld_lat facility. In conjunction with ld_lat facility .It Li MEM_STORE_RETIRED.DTLB_MISS .Pq Event 0CH , Umask 01H -The event counts the number of retired stores that missed the DTLB. The DTLB -miss is not counted if the store operation causes a fault. Does not counter -prefetches. Counts both primary and secondary misses to the TLB +The event counts the number of retired stores that missed the DTLB. +The DTLB miss is not counted if the store operation causes a fault. +Does not counter prefetches. +Counts both primary and secondary misses to the TLB .It Li UOPS_ISSUED.ANY .Pq Event 0EH , Umask 01H Counts the number of Uops issued by the Register Allocation Table to the @@ -264,9 +268,11 @@ Load instructions retired remote DRAM and remote home-remote cache HITM Load instructions retired I/O (Precise Event) .It Li FP_COMP_OPS_EXE.X87 .Pq Event 10H , Umask 01H -Counts the number of FP Computational Uops Executed. The number of FADD, +Counts the number of FP Computational Uops Executed. +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 +DIVs, and IDIVs. +This event does not distinguish an FADD used in the middle of a transcendental flow from a separate FADD instruction. .It Li FP_COMP_OPS_EXE.MMX .Pq Event 10H , Umask 02H @@ -316,9 +322,9 @@ Counts number of loads dispatched from the Reservation Station that bypass the Memory Order Buffer. .It Li LOAD_DISPATCH.RS_DELAYED .Pq Event 13H , Umask 02H -Counts the number of delayed RS dispatches at the stage latch. If an RS -dispatch can not bypass to LB, it has another chance to dispatch from the -one-cycle delayed staging latch before it is written into the LB. +Counts the number of delayed RS dispatches at the stage latch. +If an RS dispatch can not bypass to LB, it has another chance to dispatch +from the one-cycle delayed staging latch before it is written into the LB. .It Li LOAD_DISPATCH.MOB .Pq Event 13H , Umask 04H Counts the number of loads dispatched from the Reservation Station to the @@ -329,13 +335,15 @@ Counts all loads dispatched from the Reservation Station. .It Li ARITH.CYCLES_DIV_BUSY .Pq Event 14H , Umask 01H Counts the number of cycles the divider is busy executing divide or square -root operations. The divide can be integer, X87 or Streaming SIMD Extensions -(SSE). The square root operation can be either X87 or SSE. +root operations. +The divide can be integer, X87 or Streaming SIMD Extensions (SSE). +The square root operation can be either X87 or SSE. Set 'edge =1, invert=1, cmask=1' to count the number of divides. Count may be incorrect When SMT is on .It Li ARITH.MUL .Pq Event 14H , Umask 02H -Counts the number of multiply operations executed. This includes integer as +Counts the number of multiply operations executed. +This includes integer as well as floating point multiply operations but excludes DPPS mul and MPSAD. Count may be incorrect When SMT is on .It Li INST_QUEUE_WRITES @@ -344,64 +352,72 @@ Counts the number of instructions written into the instruction queue every cycle. .It Li INST_DECODED.DEC0 .Pq Event 18H , Umask 01H -Counts number of instructions that require decoder 0 to be decoded. Usually, -this means that the instruction maps to more than 1 uop +Counts number of instructions that require decoder 0 to be decoded. +Usually, this means that the instruction maps to more than 1 uop .It Li TWO_UOP_INSTS_DECODED .Pq Event 19H , Umask 01H An instruction that generates two uops was decoded .It Li INST_QUEUE_WRITE_CYCLES .Pq Event 1EH , Umask 01H This event counts the number of cycles during which instructions are written -to the instruction queue. Dividing this counter by the number of +to the instruction queue. +Dividing this counter by the number of instructions written to the instruction queue (INST_QUEUE_WRITES) yields the -average number of instructions decoded each cycle. If this number is less +average number of instructions decoded each cycle. +If this number is less than four and the pipe stalls, this indicates that the decoder is failing to decode enough instructions per cycle to sustain the 4-wide pipeline. If SSE* instructions that are 6 bytes or longer arrive one after another, -then front end throughput may limit execution speed. In such case, +then front end throughput may limit execution speed. +In such case, .It Li LSD_OVERFLOW .Pq Event 20H , Umask 01H Number of loops that can not stream from the instruction queue. .It Li L2_RQSTS.LD_HIT .Pq Event 24H , Umask 01H -Counts number of loads that hit the L2 cache. L2 loads include both L1D -demand misses as well as L1D prefetches. L2 loads can be rejected for -various reasons. Only non rejected loads are counted. +Counts number of loads that hit the L2 cache. +L2 loads include both L1D demand misses as well as L1D prefetches. +L2 loads can be rejected for various reasons. +Only non rejected loads are counted. .It Li L2_RQSTS.LD_MISS .Pq Event 24H , Umask 02H -Counts the number of loads that miss the L2 cache. L2 loads include both L1D -demand misses as well as L1D prefetches. +Counts the number of loads that miss the L2 cache. +L2 loads include both L1D demand misses as well as L1D prefetches. .It Li L2_RQSTS.LOADS .Pq Event 24H , Umask 03H -Counts all L2 load requests. L2 loads include both L1D demand misses as well -as L1D prefetches. +Counts all L2 load requests. +L2 loads include both L1D demand misses as well as L1D prefetches. .It Li L2_RQSTS.RFO_HIT .Pq Event 24H , Umask 04H -Counts the number of store RFO requests that hit the L2 cache. L2 RFO -requests include both L1D demand RFO misses as well as L1D RFO prefetches. +Counts the number of store RFO requests that hit the L2 cache. +L2 RFO requests include both L1D demand RFO misses as well as L1D RFO +prefetches. Count includes WC memory requests, where the data is not fetched but the permission to write the line is required. .It Li L2_RQSTS.RFO_MISS .Pq Event 24H , Umask 08H -Counts the number of store RFO requests that miss the L2 cache. L2 RFO -requests include both L1D demand RFO misses as well as L1D RFO prefetches. +Counts the number of store RFO requests that miss the L2 cache. +L2 RFO requests include both L1D demand RFO misses as well as L1D RFO +prefetches. .It Li L2_RQSTS.RFOS .Pq Event 24H , Umask 0CH -Counts all L2 store RFO requests. L2 RFO requests include both L1D demand +Counts all L2 store RFO requests. +L2 RFO requests include both L1D demand RFO misses as well as L1D RFO prefetches. .It Li L2_RQSTS.IFETCH_HIT .Pq Event 24H , Umask 10H -Counts number of instruction fetches that hit the L2 cache. L2 instruction -fetches include both L1I demand misses as well as L1I instruction -prefetches. +Counts number of instruction fetches that hit the L2 cache. +L2 instruction fetches include both L1I demand misses as well as L1I +instruction prefetches. .It Li L2_RQSTS.IFETCH_MISS .Pq Event 24H , Umask 20H -Counts number of instruction fetches that miss the L2 cache. L2 instruction -fetches include both L1I demand misses as well as L1I instruction -prefetches. +Counts number of instruction fetches that miss the L2 cache. +L2 instruction fetches include both L1I demand misses as well as L1I +instruction prefetches. .It Li L2_RQSTS.IFETCHES .Pq Event 24H , Umask 30H -Counts all instruction fetches. L2 instruction fetches include both L1I +Counts all instruction fetches. +L2 instruction fetches include both L1I demand misses as well as L1I instruction prefetches. .It Li L2_RQSTS.PREFETCH_HIT .Pq Event 24H , Umask 40H @@ -421,26 +437,30 @@ Counts all L2 requests for both code and data. .It Li L2_DATA_RQSTS.DEMAND.I_STATE .Pq Event 26H , Umask 01H Counts number of L2 data demand loads where the cache line to be loaded is -in the I (invalid) state, i.e. a cache miss. L2 demand loads are both L1D -demand misses and L1D prefetches. +in the I (invalid) state, i.e. a cache miss. +L2 demand loads are both L1D demand misses and L1D prefetches. .It Li L2_DATA_RQSTS.DEMAND.S_STATE .Pq Event 26H , Umask 02H Counts number of L2 data demand loads where the cache line to be loaded is -in the S (shared) state. L2 demand loads are both L1D demand misses and L1D +in the S (shared) state. +L2 demand loads are both L1D demand misses and L1D prefetches. .It Li L2_DATA_RQSTS.DEMAND.E_STATE .Pq Event 26H , Umask 04H Counts number of L2 data demand loads where the cache line to be loaded is -in the E (exclusive) state. L2 demand loads are both L1D demand misses and +in the E (exclusive) state. +L2 demand loads are both L1D demand misses and L1D prefetches. .It Li L2_DATA_RQSTS.DEMAND.M_STATE .Pq Event 26H , Umask 08H Counts number of L2 data demand loads where the cache line to be loaded is -in the M (modified) state. L2 demand loads are both L1D demand misses and +in the M (modified) state. +L2 demand loads are both L1D demand misses and L1D prefetches. .It Li L2_DATA_RQSTS.DEMAND.MESI .Pq Event 26H , Umask 0FH -Counts all L2 data demand requests. L2 demand loads are both L1D demand +Counts all L2 data demand requests. +L2 demand loads are both L1D demand misses and L1D prefetches. .It Li L2_DATA_RQSTS.PREFETCH.I_STATE .Pq Event 26H , Umask 10H @@ -449,7 +469,8 @@ in the I (invalid) state, i.e. a cache miss. .It Li L2_DATA_RQSTS.PREFETCH.S_STATE .Pq Event 26H , Umask 20H Counts number of L2 prefetch data loads where the cache line to be loaded is -in the S (shared) state. A prefetch RFO will miss on an S state line, while +in the S (shared) state. +A prefetch RFO will miss on an S state line, while a prefetch read will hit on an S state line. .It Li L2_DATA_RQSTS.PREFETCH.E_STATE .Pq Event 26H , Umask 40H @@ -468,23 +489,27 @@ Counts all L2 data requests. .It Li L2_WRITE.RFO.I_STATE .Pq Event 27H , Umask 01H Counts number of L2 demand store RFO requests where the cache line to be -loaded is in the I (invalid) state, i.e, a cache miss. The L1D prefetcher +loaded is in the I (invalid) state, i.e, a cache miss. +The L1D prefetcher does not issue a RFO prefetch. This is a demand RFO request .It Li L2_WRITE.RFO.S_STATE .Pq Event 27H , Umask 02H Counts number of L2 store RFO requests where the cache line to be loaded is -in the S (shared) state. The L1D prefetcher does not issue a RFO prefetch. +in the S (shared) state. +The L1D prefetcher does not issue a RFO prefetch. This is a demand RFO request. .It Li L2_WRITE.RFO.M_STATE .Pq Event 27H , Umask 08H Counts number of L2 store RFO requests where the cache line to be loaded is -in the M (modified) state. The L1D prefetcher does not issue a RFO prefetch. +in the M (modified) state. +The L1D prefetcher does not issue a RFO prefetch. This is a demand RFO request. .It Li L2_WRITE.RFO.HIT .Pq Event 27H , Umask 0EH Counts number of L2 store RFO requests where the cache line to be loaded is -in either the S, E or M states. The L1D prefetcher does not issue a RFO +in either the S, E or M states. +The L1D prefetcher does not issue a RFO prefetch. This is a demand RFO request .It Li L2_WRITE.RFO.MESI @@ -536,21 +561,23 @@ is in the M (modified) state. Counts all L1 writebacks to the L2. .It Li L3_LAT_CACHE.REFERENCE .Pq Event 2EH , Umask 02H -Counts uncore Last Level Cache references. Because cache hierarchy, cache +Counts uncore Last Level Cache references. +Because cache hierarchy, cache sizes and other implementation-specific characteristics; value comparison to estimate performance differences is not recommended. See Table A-1. .It Li L3_LAT_CACHE.MISS .Pq Event 2EH , Umask 01H -Counts uncore Last Level Cache misses. Because cache hierarchy, cache sizes +Counts uncore Last Level Cache misses. +Because cache hierarchy, cache sizes and other implementation-specific characteristics; value comparison to estimate performance differences is not recommended. See Table A-1. .It Li CPU_CLK_UNHALTED.THREAD_P .Pq Event 3CH , Umask 00H 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 +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. see Table A-1 .It Li CPU_CLK_UNHALTED.REF_P @@ -569,7 +596,8 @@ Counts cycles of page walk due to misses in the STLB. .It Li DTLB_MISSES.STLB_HIT .Pq Event 49H , Umask 10H Counts the number of DTLB first level misses that hit in the second level -TLB. This event is only relevant if the core contains multiple DTLB levels. +TLB. +This event is only relevant if the core contains multiple DTLB levels. .It Li DTLB_MISSES.LARGE_WALK_COMPLETED .Pq Event 49H , Umask 80H Counts number of completed large page walks due to misses in the STLB. @@ -584,17 +612,22 @@ Counts number of hardware prefetch requests dispatched out of the prefetch FIFO. .It Li L1D_PREFETCH.MISS .Pq Event 4EH , Umask 02H -Counts number of hardware prefetch requests that miss the L1D. There are two -prefetchers in the L1D. A streamer, which predicts lines sequentially after +Counts number of hardware prefetch requests that miss the L1D. +There are two +prefetchers in the L1D. +A streamer, which predicts lines sequentially after this one should be fetched, and the IP prefetcher that remembers access -patterns for the current instruction. The streamer prefetcher stops on an +patterns for the current instruction. +The streamer prefetcher stops on an L1D hit, while the IP prefetcher does not. .It Li L1D_PREFETCH.TRIGGERS .Pq Event 4EH , Umask 04H Counts number of prefetch requests triggered by the Finite State Machine and -pushed into the prefetch FIFO. Some of the prefetch requests are dropped due +pushed into the prefetch FIFO. +Some of the prefetch requests are dropped due to overwrites or competition between the IP index prefetcher and streamer -prefetcher. The prefetch FIFO contains 4 entries. +prefetcher. +The prefetch FIFO contains 4 entries. .It Li EPT.WALK_CYCLES .Pq Event 4FH , Umask 10H Counts Extended Page walk cycles. @@ -626,31 +659,33 @@ Counts the number of cacheable load lock speculated or retired instructions accepted into the fill buffer. .It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND.READ_DATA .Pq Event 60H , Umask 01H -Counts weighted cycles of offcore demand data read requests. Does not -include L2 prefetch requests. +Counts weighted cycles of offcore demand data read requests. +Does not include L2 prefetch requests. Counter 0. .It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND.READ_CODE .Pq Event 60H , Umask 02H -Counts weighted cycles of offcore demand code read requests. Does not -include L2 prefetch requests. +Counts weighted cycles of offcore demand code read requests. +Does not include L2 prefetch requests. Counter 0. .It Li OFFCORE_REQUESTS_OUTSTANDING.DEMAND.RFO .Pq Event 60H , Umask 04H -Counts weighted cycles of offcore demand RFO requests. Does not include L2 -prefetch requests. +Counts weighted cycles of offcore demand RFO requests. +Does not include L2 prefetch requests. Counter 0. .It Li OFFCORE_REQUESTS_OUTSTANDING.ANY.READ .Pq Event 60H , Umask 08H -Counts weighted cycles of offcore read requests of any kind. Include L2 -prefetch requests. +Counts weighted cycles of offcore read requests of any kind. +Include L2 prefetch requests. Counter 0. .It Li CACHE_LOCK_CYCLES.L1D_L2 .Pq Event 63H , Umask 01H -Cycle count during which the L1D and L2 are locked. A lock is asserted when +Cycle count during which the L1D and L2 are locked. +A lock is asserted when there is a locked memory access, due to uncacheable memory, a locked operation that spans two cache lines, or a page walk from an uncacheable page table. -Counter 0, 1 only. L1D and L2 locks have a very high performance penalty and +Counter 0, 1 only. +L1D and L2 locks have a very high performance penalty and it is highly recommended to avoid such accesses. .It Li CACHE_LOCK_CYCLES.L1D .Pq Event 63H , Umask 02H @@ -665,9 +700,11 @@ Counts the number of completed I/O transactions. Counts all instruction fetches that hit the L1 instruction cache. .It Li L1I.MISSES .Pq Event 80H , Umask 02H -Counts all instruction fetches that miss the L1I cache. This includes +Counts all instruction fetches that miss the L1I cache. +This includes instruction cache misses, streaming buffer misses, victim cache misses and -uncacheable fetches. An instruction fetch miss is counted only once and not +uncacheable fetches. +An instruction fetch miss is counted only once and not once for every cycle it is outstanding. .It Li L1I.READS .Pq Event 80H , Umask 03H @@ -747,10 +784,10 @@ Counts all near call branches executed, but not necessarily retired. Counts taken near branches executed, but not necessarily retired. .It Li BR_INST_EXEC.ANY .Pq Event 88H , Umask 7FH -Counts all near executed branches (not necessarily retired). This includes -only instructions and not micro-op branches. Frequent branching is not -necessarily a major performance issue. However frequent branch -mispredictions may be a problem. +Counts all near executed branches (not necessarily retired). +This includes only instructions and not micro-op branches. +Frequent branching is not necessarily a major performance issue. +However frequent branch mispredictions may be a problem. .It Li BR_MISP_EXEC.COND .Pq Event 89H , Umask 01H Counts the number of mispredicted conditional near branch instructions @@ -791,9 +828,10 @@ Counts the number of mispredicted near branch instructions that were executed, but not necessarily retired. .It Li RESOURCE_STALLS.ANY .Pq Event A2H , Umask 01H -Counts the number of Allocator resource related stalls. Includes register -renaming buffer entries, memory buffer entries. In addition to resource -related stalls, this event counts some other events. Includes stalls arising +Counts the number of Allocator resource related stalls. +Includes register renaming buffer entries, memory buffer entries. +In addition to resource related stalls, this event counts some other events. +Includes stalls arising during branch misprediction recovery, such as if retirement of the mispredicted branch is delayed and stalls arising while store buffer is draining from synchronizing operations. @@ -806,7 +844,8 @@ Counts the cycles of stall due to lack of load buffer for load operation. .Pq Event A2H , Umask 04H This event counts the number of cycles when the number of instructions in the pipeline waiting for execution reaches the limit the processor can -handle. A high count of this event indicates that there are long latency +handle. +A high count of this event indicates that there are long latency operations in the pipe (possibly load and store operations that miss the L2 cache, or instructions dependent upon instructions further down the pipeline that have yet to retire. @@ -816,7 +855,8 @@ start execution. .Pq Event A2H , Umask 08H This event counts the number of cycles that a resource related stall will occur due to the number of store instructions reaching the limit of the -pipeline, (i.e. all store buffers are used). The stall ends when a store +pipeline, (i.e. all store buffers are used). +The stall ends when a store instruction commits its data to the cache or memory. .It Li RESOURCE_STALLS.ROB_FULL .Pq Event A2H , Umask 10H @@ -828,7 +868,8 @@ floating-point unit (FPU) control word. .It Li RESOURCE_STALLS.MXCSR .Pq Event A2H , Umask 40H Stalls due to the MXCSR register rename occurring to close to a previous -MXCSR rename. The MXCSR provides control and status for the MMX registers. +MXCSR rename. +The MXCSR provides control and status for the MMX registers. .It Li RESOURCE_STALLS.OTHER .Pq Event A2H , Umask 80H Counts the number of cycles while execution was stalled due to other @@ -839,12 +880,14 @@ Counts the number of instructions decoded that are macro-fused but not necessarily executed or retired. .It Li BACLEAR_FORCE_IQ .Pq Event A7H , Umask 01H -Counts number of times a BACLEAR was forced by the Instruction Queue. The IQ -is also responsible for providing conditional branch prediction direction -based on a static scheme and dynamic data provided by the L2 Branch -Prediction Unit. If the conditional branch target is not found in the Target +Counts number of times a BACLEAR was forced by the Instruction Queue. +The IQ is also responsible for providing conditional branch prediction +direction based on a static scheme and dynamic data provided by the L2 +Branch Prediction Unit. +If the conditional branch target is not found in the Target Array and the IQ predicts that the branch is taken, then the IQ will force -the Branch Address Calculator to issue a BACLEAR. Each BACLEAR asserted by +the Branch Address Calculator to issue a BACLEAR. +Each BACLEAR asserted by the BAC generates approximately an 8 cycle bubble in the instruction fetch pipeline. .It Li LSD.UOPS @@ -856,22 +899,24 @@ Use cmask=1 and invert to count cycles Counts the number of ITLB flushes .It Li OFFCORE_REQUESTS.DEMAND.READ_DATA .Pq Event B0H , Umask 01H -Counts number of offcore demand data read requests. Does not count L2 -prefetch requests. +Counts number of offcore demand data read requests. +Does not count L2 prefetch requests. .It Li OFFCORE_REQUESTS.DEMAND.READ_CODE .Pq Event B0H , Umask 02H -Counts number of offcore demand code read requests. Does not count L2 -prefetch requests. +Counts number of offcore demand code read requests. +Does not count L2 prefetch requests. .It Li OFFCORE_REQUESTS.DEMAND.RFO .Pq Event B0H , Umask 04H -Counts number of offcore demand RFO requests. Does not count L2 prefetch -requests. +Counts number of offcore demand RFO requests. +Does not count L2 prefetch requests. .It Li OFFCORE_REQUESTS.ANY.READ .Pq Event B0H , Umask 08H -Counts number of offcore read requests. Includes L2 prefetch requests. +Counts number of offcore read requests. +Includes L2 prefetch requests. .It Li OFFCORE_REQUESTS.ANY.RFO .Pq Event 80H , Umask 10H -Counts number of offcore RFO requests. Includes L2 prefetch requests. +Counts number of offcore RFO requests. +Includes L2 prefetch requests. .It Li OFFCORE_REQUESTS.L1D_WRITEBACK .Pq Event B0H , Umask 40H Counts number of L1D writebacks to the uncore. @@ -880,38 +925,42 @@ Counts number of L1D writebacks to the uncore. Counts all offcore requests. .It Li UOPS_EXECUTED.PORT0 .Pq Event B1H , Umask 01H -Counts number of Uops executed that were issued on port 0. Port 0 handles -integer arithmetic, SIMD and FP add Uops. +Counts number of Uops executed that were issued on port 0. +Port 0 handles integer arithmetic, SIMD and FP add Uops. .It Li UOPS_EXECUTED.PORT1 .Pq Event B1H , Umask 02H -Counts number of Uops executed that were issued on port 1. Port 1 handles -integer arithmetic, SIMD, integer shift, FP multiply and FP divide Uops. +Counts number of Uops executed that were issued on port 1. +Port 1 handles integer arithmetic, SIMD, integer shift, FP multiply and +FP divide Uops. .It Li UOPS_EXECUTED.PORT2_CORE .Pq Event B1H , Umask 04H -Counts number of Uops executed that were issued on port 2. Port 2 handles -the load Uops. This is a core count only and can not be collected per +Counts number of Uops executed that were issued on port 2. +Port 2 handles the load Uops. +This is a core count only and can not be collected per thread. .It Li UOPS_EXECUTED.PORT3_CORE .Pq Event B1H , Umask 08H -Counts number of Uops executed that were issued on port 3. Port 3 handles -store Uops. This is a core count only and can not be collected per thread. +Counts number of Uops executed that were issued on port 3. +Port 3 handles store Uops. +This is a core count only and can not be collected per thread. .It Li UOPS_EXECUTED.PORT4_CORE .Pq Event B1H , Umask 10H -Counts number of Uops executed that where issued on port 4. Port 4 handles -the value to be stored for the store Uops issued on port 3. This is a core -count only and can not be collected per thread. +Counts number of Uops executed that where issued on port 4. +Port 4 handles the value to be stored for the store Uops issued on port 3. +This is a core count only and can not be collected per thread. .It Li UOPS_EXECUTED.CORE_ACTIVE_CYCLES_NO_PORT5 .Pq Event B1H , Umask 1FH Counts number of cycles there are one or more uops being executed and were -issued on ports 0-4. This is a core count only and can not be collected per -thread. +issued on ports 0-4. +This is a core count only and can not be collected per thread. .It Li UOPS_EXECUTED.PORT5 .Pq Event B1H , Umask 20H Counts number of Uops executed that where issued on port 5. .It Li UOPS_EXECUTED.CORE_ACTIVE_CYCLES .Pq Event B1H , Umask 3FH Counts number of cycles there are one or more uops being executed on any -ports. This is a core count only and can not be collected per thread. +ports. +This is a core count only and can not be collected per thread. .It Li UOPS_EXECUTED.PORT015 .Pq Event B1H , Umask 40H Counts number of Uops executed that where issued on port 0, 1, or 5. @@ -924,15 +973,18 @@ Counts number of Uops executed that where issued on port 2, 3, or 4. Counts number of cycles the SQ is full to handle off-core requests. .It Li SNOOPQ_REQUESTS_OUTSTANDING.DATA .Pq Event B3H , Umask 01H -Counts weighted cycles of snoopq requests for data. Counter 0 only +Counts weighted cycles of snoopq requests for data. +Counter 0 only Use cmask=1 to count cycles not empty. .It Li SNOOPQ_REQUESTS_OUTSTANDING.INVALIDATE .Pq Event B3H , Umask 02H -Counts weighted cycles of snoopq invalidate requests. Counter 0 only. +Counts weighted cycles of snoopq invalidate requests. +Counter 0 only. Use cmask=1 to count cycles not empty. .It Li SNOOPQ_REQUESTS_OUTSTANDING.CODE .Pq Event B3H , Umask 04H -Counts weighted cycles of snoopq requests for code. Counter 0 only. +Counts weighted cycles of snoopq requests for code. +Counter 0 only. Use cmask=1 to count cycles not empty. .It Li SNOOPQ_REQUESTS.CODE .Pq Event B4H , Umask 01H @@ -970,7 +1022,8 @@ Use MSR 01A7H. See Table A-1 Notes: INST_RETIRED.ANY is counted by a designated fixed counter. INST_RETIRED.ANY_P is counted by a programmable counter and is an -architectural performance event. Event is supported if CPUID.A.EBX[1] = 0. +architectural performance event. +Event is supported if CPUID.A.EBX[1] = 0. Counting: Faulting executions of GETSEC/VM entry/VM Exit/MWait will not count as retired instructions. .It Li INST_RETIRED.X87 @@ -985,8 +1038,9 @@ Counts the number of retired: MMX instructions. .It Li UOPS_RETIRED.ANY .Pq Event C2H , Umask 01H Counts the number of micro-ops retired, (macro-fused=1, micro- fused=2, -others=1; maximum count of 8 per cycle). Most instructions are composed of -one or two micro-ops. Some instructions are decoded into longer sequences +others=1; maximum count of 8 per cycle). +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. Use cmask=1 and invert to count active cycles or stalled cycles @@ -1006,7 +1060,8 @@ Counts the number of machine clears due to memory order conflicts. .Pq Event C3H , Umask 04H Counts the number of times that a program writes to a code section. Self-modifying code causes a sever penalty in all Intel 64 and IA-32 -processors. The modified cache line is written back to the L2 and L3caches. +processors. +The modified cache line is written back to the L2 and L3caches. .It Li BR_INST_RETIRED.ANY_P .Pq Event C4H , Umask 00H See Table A-1. @@ -1063,23 +1118,25 @@ cache. .It Li MEM_LOAD_RETIRED.OTHER_CORE_L2_HIT_HITM .Pq Event CBH , Umask 08H Counts number of retired loads that hit in a sibling core's L2 (on die -core). Since the L3 is inclusive of all cores on the package, this is an L3 -hit. This counts both clean or modified hits. +core). +Since the L3 is inclusive of all cores on the package, this is an L3 hit. +This counts both clean or modified hits. .It Li MEM_LOAD_RETIRED.L3_MISS .Pq Event CBH , Umask 10H -Counts number of retired loads that miss the L3 cache. The load was -satisfied by a remote socket, local memory or an IOH. +Counts number of retired loads that miss the L3 cache. +The load was satisfied by a remote socket, local memory or an IOH. .It Li MEM_LOAD_RETIRED.HIT_LFB .Pq Event CBH , Umask 40H Counts number of retired loads that miss the L1D and the address is located -in an allocated line fill buffer and will soon be committed to cache. This -is counting secondary L1D misses. +in an allocated line fill buffer and will soon be committed to cache. +This is counting secondary L1D misses. .It Li MEM_LOAD_RETIRED.DTLB_MISS .Pq Event CBH , Umask 80H -Counts the number of retired loads that missed the DTLB. The DTLB miss is -not counted if the load operation causes a fault. This event counts loads -from cacheable memory only. The event does not count loads by software -prefetches. Counts both primary and secondary misses to the TLB. +Counts the number of retired loads that missed the DTLB. +The DTLB miss is not counted if the load operation causes a fault. +This event counts loads from cacheable memory only. +The event does not count loads by software prefetches. +Counts both primary and secondary misses to the TLB. .It Li FP_MMX_TRANS.TO_FP .Pq Event CCH , Umask 01H Counts the first floating-point instruction following any MMX instruction. @@ -1087,15 +1144,15 @@ You can use this event to estimate the penalties for the transitions between floating-point and MMX technology states. .It Li FP_MMX_TRANS.TO_MMX .Pq Event CCH , Umask 02H -Counts the first MMX instruction following a floating-point instruction. You -can use this event to estimate the penalties for the transitions between +Counts the first MMX instruction following a floating-point instruction. +You can use this event to estimate the penalties for the transitions between floating-point and MMX technology states. .It Li FP_MMX_TRANS.ANY .Pq Event CCH , Umask 03H Counts all transitions from floating point to MMX instructions and from MMX -instructions to floating point instructions. You can use this event to -estimate the penalties for the transitions between floating-point and MMX -technology states. +instructions to floating point instructions. +You can use this event to estimate the penalties for the transitions between +floating-point and MMX technology states. .It Li MACRO_INSTS.DECODED .Pq Event D0H , Umask 01H Counts the number of instructions decoded, (but not necessarily executed or @@ -1105,14 +1162,15 @@ retired). Counts the cycles of decoder stalls. .It Li UOPS_DECODED.MS .Pq Event D1H , Umask 02H -Counts the number of Uops decoded by the Microcode Sequencer, MS. The MS -delivers uops when the instruction is more than 4 uops long or a microcode -assist is occurring. +Counts the number of Uops decoded by the Microcode Sequencer, MS. +The MS delivers uops when the instruction is more than 4 uops long or a +microcode assist is occurring. .It Li UOPS_DECODED.ESP_FOLDING .Pq Event D1H , Umask 04H Counts number of stack pointer (ESP) instructions decoded: push , pop , call , ret, etc. ESP instructions do not generate a Uop to increment or decrement -ESP. Instead, they update an ESP_Offset register that keeps track of the +ESP. +Instead, they update an ESP_Offset register that keeps track of the delta to the current value of the ESP register. .It Li UOPS_DECODED.ESP_SYNC .Pq Event D1H , Umask 08H @@ -1122,7 +1180,8 @@ value of the ESP register. .It Li RAT_STALLS.FLAGS .Pq Event D2H , Umask 01H Counts the number of cycles during which execution stalled due to several -reasons, one of which is a partial flag register stall. A partial register +reasons, one of which is a partial flag register stall. +A partial register stall may occur when two conditions are met: 1) an instruction modifies some, but not all, of the flags in the flag register and 2) the next instruction, which depends on flags, depends on flags that were not modified @@ -1135,28 +1194,34 @@ was partially written by previous instruction. .It Li RAT_STALLS.ROB_READ_PORT .Pq Event D2H , Umask 04H Counts the number of cycles when ROB read port stalls occurred, which did -not allow new micro-ops to enter the out-of-order pipeline. Note that, at +not allow new micro-ops to enter the out-of-order pipeline. +Note that, at this stage in the pipeline, additional stalls may occur at the same cycle -and prevent the stalled micro-ops from entering the pipe. In such a case, +and prevent the stalled micro-ops from entering the pipe. +In such a case, micro-ops retry entering the execution pipe in the next cycle and the ROB-read port stall is counted again. .It Li RAT_STALLS.SCOREBOARD .Pq Event D2H , Umask 08H Counts the cycles where we stall due to microarchitecturally required -serialization. Microcode scoreboarding stalls. +serialization. +Microcode scoreboarding stalls. .It Li RAT_STALLS.ANY .Pq Event D2H , Umask 0FH Counts all Register Allocation Table stall cycles due to: Cycles when ROB read port stalls occurred, which did not allow new micro-ops to enter the -execution pipe. Cycles when partial register stalls occurred Cycles when +execution pipe. +Cycles when partial register stalls occurred Cycles when flag stalls occurred Cycles floating-point unit (FPU) status word stalls -occurred. To count each of these conditions separately use the events: +occurred. +To count each of these conditions separately use the events: RAT_STALLS.ROB_READ_PORT, RAT_STALLS.PARTIAL, RAT_STALLS.FLAGS, and RAT_STALLS.FPSW. .It Li SEG_RENAME_STALLS .Pq Event D4H , Umask 01H Counts the number of stall cycles due to the lack of renaming resources for -the ES, DS, FS, and GS segment registers. If a segment is renamed but not +the ES, DS, FS, and GS segment registers. +If a segment is renamed but not retired and a second update to the same segment occurs, a stall occurs in the front- end of the pipeline until the renamed segment retires. .It Li ES_REG_RENAMES @@ -1176,16 +1241,18 @@ or return branch. .Pq Event E6H , Umask 01H 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. This can occur +corrected by the Branch Address Calculator at the front end. +This can occur if the code has many branches such that they cannot be consumed by the BPU. Each BACLEAR asserted by the BAC generates approximately an 8 cycle bubble -in the instruction fetch pipeline. The effect on total execution time -depends on the surrounding code. +in the instruction fetch pipeline. +The effect on total execution time depends on the surrounding code. .It Li BACLEAR.BAD_TARGET .Pq Event E6H , Umask 02H Counts number of Branch Address Calculator clears (BACLEAR) asserted due to conditional branch instructions in which there was a target hit but the -direction was wrong. Each BACLEAR asserted by the BAC generates +direction was wrong. +Each BACLEAR asserted by the BAC generates approximately an 8 cycle bubble in the instruction fetch pipeline. .It Li BPU_CLEARS.EARLY .Pq Event E8H , Umask 01H @@ -1195,7 +1262,8 @@ The BPU clear leads to 2 cycle bubble in the Front End. .It Li BPU_CLEARS.LATE .Pq Event E8H , Umask 02H Counts late Branch Prediction Unit clears due to Most Recently Used -conflicts. The PBU clear leads to a 3 cycle bubble in the Front End. +conflicts. +The PBU clear leads to a 3 cycle bubble in the Front End. .It Li THREAD_ACTIVE .Pq Event ECH , Umask 01H Counts cycles threads are active. @@ -1258,12 +1326,13 @@ Counts number of Super Queue LRU hints sent to L3. Counts the number of SQ lock splits across a cache line. .It Li SQ_FULL_STALL_CYCLES .Pq Event F6H , Umask 01H -Counts cycles the Super Queue is full. Neither of the threads on this core -will be able to access the uncore. +Counts cycles the Super Queue is full. +Neither of the threads on this core will be able to access the uncore. .It Li FP_ASSIST.ALL .Pq Event F7H , Umask 01H Counts the number of floating point operations executed that required -micro-code assist intervention. Assists are required in the following cases: +micro-code assist intervention. +Assists are required in the following cases: SSE instructions, (Denormal input when the DAZ flag is off or Underflow result when the FTZ flag is off): x87 instructions, (NaN or denormal are loaded to a register or used as input from memory, Division by 0 or