Skip to content

L
linux
Commit 59faeaf8 authored by Ilkka Koskinen's avatar Ilkka Koskinen Committed by Namhyung Kim
Browse files
Options

perf vendor events arm64: Fix for AmpereOne metrics 

This patch addresses review comments that were given for
705ed549 ("perf vendor events arm64: Add AmpereOne metrics")
but didn't make it to the original patch [1][2]

Changes include: A fix for backend_memory formula, use of standard metrics
when possible, using #slots, renaming metrics to avoid spaces in the names,
and cleanup.

[1] https://lore.kernel.org/linux-perf-users/e9bdacb-a231-36af-6a2e-6918ee7effa@os.amperecomputing.com/
[2] https://lore.kernel.org/linux-perf-users/20230826192352.3043220-1-ilkka@os.amperecomputing.com/

Fixes: 705ed549 ("perf vendor events arm64: Add AmpereOne metrics")
Signed-off-by: default avatarIlkka Koskinen <ilkka@os.amperecomputing.com>
Reviewed-by: default avatarIan Rogers <irogers@google.com>
Cc: James Clark <james.clark@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Leo Yan <leo.yan@linaro.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Dave Kleikamp <dave.kleikamp@oracle.com>
Cc: John Garry <john.g.garry@oracle.com>
Cc: D Scott Phillips <scott@os.amperecomputing.com>
Cc: linux-arm-kernel@lists.infradead.org
Link: https://lore.kernel.org/r/20230920061839.2437413-1-ilkka@os.amperecomputing.comSigned-off-by: default avatarNamhyung Kim <namhyung@kernel.org>
parent 29441ab3
Hide whitespace changes
Inline Side-by-side
Showing with 220 additions and 198 deletions
tools/perf/pmu-events/arch/arm64/ampere/ampereone/metrics.json
View file @ 59faeaf8
[ [
{ {
"MetricName": "branch_miss_pred_rate",
"MetricExpr": "BR_MIS_PRED / BR_PRED", "MetricExpr": "BR_MIS_PRED / BR_PRED",
"BriefDescription": "Branch predictor misprediction rate. May not count branches that are never resolved because they are in the misprediction shadow of an earlier branch", "BriefDescription": "Branch predictor misprediction rate. May not count branches that are never resolved because they are in the misprediction shadow of an earlier branch",
"MetricGroup": "Branch Prediction", "MetricGroup": "branch",
"MetricName": "Misprediction" "ScaleUnit": "100%"
}, },
{ {
"MetricExpr": "BR_MIS_PRED_RETIRED / BR_RETIRED", "MetricName": "bus_utilization",
"BriefDescription": "Branch predictor misprediction rate", "MetricExpr": "((BUS_ACCESS / (BUS_CYCLES * 1)) * 100)",
"MetricGroup": "Branch Prediction",
"MetricName": "Misprediction (retired)"
},
{
"MetricExpr": "BUS_ACCESS / ( BUS_CYCLES * 1)",
"BriefDescription": "Core-to-uncore bus utilization", "BriefDescription": "Core-to-uncore bus utilization",
"MetricGroup": "Bus", "MetricGroup": "Bus",
"MetricName": "Bus utilization" "ScaleUnit": "1percent of bus cycles"
}, },
{ {
"MetricExpr": "L1D_CACHE_REFILL / L1D_CACHE", "MetricName": "l1d_cache_miss_ratio",
"BriefDescription": "L1D cache miss rate", "MetricExpr": "(L1D_CACHE_REFILL / L1D_CACHE)",
"MetricGroup": "Cache", "BriefDescription": "This metric measures the ratio of level 1 data cache accesses missed to the total number of level 1 data cache accesses. This gives an indication of the effectiveness of the level 1 data cache.",
"MetricName": "L1D cache miss" "MetricGroup": "Miss_Ratio;L1D_Cache_Effectiveness",
"ScaleUnit": "1per cache access"
},
{
"MetricName": "l1i_cache_miss_ratio",
"MetricExpr": "(L1I_CACHE_REFILL / L1I_CACHE)",
"BriefDescription": "This metric measures the ratio of level 1 instruction cache accesses missed to the total number of level 1 instruction cache accesses. This gives an indication of the effectiveness of the level 1 instruction cache.",
"MetricGroup": "Miss_Ratio;L1I_Cache_Effectiveness",
"ScaleUnit": "1per cache access"
}, },
{ {
"MetricName": "Miss_Ratio;l1d_cache_read_miss",
"MetricExpr": "L1D_CACHE_LMISS_RD / L1D_CACHE_RD", "MetricExpr": "L1D_CACHE_LMISS_RD / L1D_CACHE_RD",
"BriefDescription": "L1D cache read miss rate", "BriefDescription": "L1D cache read miss rate",
"MetricGroup": "Cache", "MetricGroup": "Cache",
"MetricName": "L1D cache read miss" "ScaleUnit": "1per cache read access"
}, },
{ {
"MetricExpr": "L1I_CACHE_REFILL / L1I_CACHE", "MetricName": "l2_cache_miss_ratio",
"BriefDescription": "L1I cache miss rate", "MetricExpr": "(L2D_CACHE_REFILL / L2D_CACHE)",
"MetricGroup": "Cache", "BriefDescription": "This metric measures the ratio of level 2 cache accesses missed to the total number of level 2 cache accesses. This gives an indication of the effectiveness of the level 2 cache, which is a unified cache that stores both data and instruction. Note that cache accesses in this cache are either data memory access or instruction fetch as this is a unified cache.",
"MetricName": "L1I cache miss" "MetricGroup": "Miss_Ratio;L2_Cache_Effectiveness",
}, "ScaleUnit": "1per cache access"
{
"MetricExpr": "L2D_CACHE_REFILL / L2D_CACHE",
"BriefDescription": "L2 cache miss rate",
"MetricGroup": "Cache",
"MetricName": "L2 cache miss"
}, },
{ {
"MetricName": "l1i_cache_read_miss_rate",
"MetricExpr": "L1I_CACHE_LMISS / L1I_CACHE", "MetricExpr": "L1I_CACHE_LMISS / L1I_CACHE",
"BriefDescription": "L1I cache read miss rate", "BriefDescription": "L1I cache read miss rate",
"MetricGroup": "Cache", "MetricGroup": "Cache",
"MetricName": "L1I cache read miss" "ScaleUnit": "1per cache access"
}, },
{ {
"MetricName": "l2d_cache_read_miss_rate",
"MetricExpr": "L2D_CACHE_LMISS_RD / L2D_CACHE_RD", "MetricExpr": "L2D_CACHE_LMISS_RD / L2D_CACHE_RD",
"BriefDescription": "L2 cache read miss rate", "BriefDescription": "L2 cache read miss rate",
"MetricGroup": "Cache", "MetricGroup": "Cache",
"MetricName": "L2 cache read miss" "ScaleUnit": "1per cache read access"
}, },
{ {
"MetricExpr": "(L1D_CACHE_LMISS_RD * 1000) / INST_RETIRED", "MetricName": "l1d_cache_miss_mpki",
"MetricExpr": "(L1D_CACHE_LMISS_RD * 1e3) / INST_RETIRED",
"BriefDescription": "Misses per thousand instructions (data)", "BriefDescription": "Misses per thousand instructions (data)",
"MetricGroup": "Cache", "MetricGroup": "Cache",
"MetricName": "MPKI data" "ScaleUnit": "1MPKI"
}, },
{ {
"MetricExpr": "(L1I_CACHE_LMISS * 1000) / INST_RETIRED", "MetricName": "l1i_cache_miss_mpki",
"MetricExpr": "(L1I_CACHE_LMISS * 1e3) / INST_RETIRED",
"BriefDescription": "Misses per thousand instructions (instruction)", "BriefDescription": "Misses per thousand instructions (instruction)",
"MetricGroup": "Cache", "MetricGroup": "Cache",
"MetricName": "MPKI instruction" "ScaleUnit": "1MPKI"
}, },
{ {
"MetricExpr": "ASE_SPEC / OP_SPEC", "MetricName": "simd_percentage",
"BriefDescription": "Proportion of advanced SIMD data processing operations (excluding DP_SPEC/LD_SPEC) operations", "MetricExpr": "((ASE_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Instruction", "BriefDescription": "This metric measures advanced SIMD operations as a percentage of total operations speculatively executed.",
"MetricName": "ASE mix" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "CRYPTO_SPEC / OP_SPEC", "MetricName": "crypto_percentage",
"BriefDescription": "Proportion of crypto data processing operations", "MetricExpr": "((CRYPTO_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Instruction", "BriefDescription": "This metric measures crypto operations as a percentage of operations speculatively executed.",
"MetricName": "Crypto mix" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "VFP_SPEC / (duration_time *1000000000)", "MetricName": "gflops",
"MetricExpr": "VFP_SPEC / (duration_time * 1e9)",
"BriefDescription": "Giga-floating point operations per second", "BriefDescription": "Giga-floating point operations per second",
"MetricGroup": "Instruction", "MetricGroup": "InstructionMix"
"MetricName": "GFLOPS_ISSUED"
}, },
{ {
"MetricExpr": "DP_SPEC / OP_SPEC", "MetricName": "integer_dp_percentage",
"BriefDescription": "Proportion of integer data processing operations", "MetricExpr": "((DP_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Instruction", "BriefDescription": "This metric measures scalar integer operations as a percentage of operations speculatively executed.",
"MetricName": "Integer mix" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "INST_RETIRED / CPU_CYCLES", "MetricName": "ipc",
"BriefDescription": "Instructions per cycle", "MetricExpr": "(INST_RETIRED / CPU_CYCLES)",
"MetricGroup": "Instruction", "BriefDescription": "This metric measures the number of instructions retired per cycle.",
"MetricName": "IPC" "MetricGroup": "General",
"ScaleUnit": "1per cycle"
}, },
{ {
"MetricExpr": "LD_SPEC / OP_SPEC", "MetricName": "load_percentage",
"BriefDescription": "Proportion of load operations", "MetricExpr": "((LD_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Instruction", "BriefDescription": "This metric measures load operations as a percentage of operations speculatively executed.",
"MetricName": "Load mix" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "LDST_SPEC/ OP_SPEC", "MetricName": "load_store_spec_rate",
"BriefDescription": "Proportion of load & store operations", "MetricExpr": "((LDST_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Instruction", "BriefDescription": "The rate of load or store instructions speculatively executed to overall instructions speclatively executed",
"MetricName": "Load-store mix" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "INST_RETIRED / (duration_time * 1000000)", "MetricName": "retired_mips",
"MetricExpr": "INST_RETIRED / (duration_time * 1e6)",
"BriefDescription": "Millions of instructions per second", "BriefDescription": "Millions of instructions per second",
"MetricGroup": "Instruction", "MetricGroup": "InstructionMix"
"MetricName": "MIPS_RETIRED"
}, },
{ {
"MetricExpr": "INST_SPEC / (duration_time * 1000000)", "MetricName": "spec_utilization_mips",
"MetricExpr": "INST_SPEC / (duration_time * 1e6)",
"BriefDescription": "Millions of instructions per second", "BriefDescription": "Millions of instructions per second",
"MetricGroup": "Instruction", "MetricGroup": "PEutilization"
"MetricName": "MIPS_UTILIZATION"
},
{
"MetricExpr": "PC_WRITE_SPEC / OP_SPEC",
"BriefDescription": "Proportion of software change of PC operations",
"MetricGroup": "Instruction",
"MetricName": "PC write mix"
}, },
{ {
"MetricExpr": "ST_SPEC / OP_SPEC", "MetricName": "pc_write_spec_rate",
"BriefDescription": "Proportion of store operations", "MetricExpr": "((PC_WRITE_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Instruction", "BriefDescription": "The rate of software change of the PC speculatively executed to overall instructions speclatively executed",
"MetricName": "Store mix" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "VFP_SPEC / OP_SPEC", "MetricName": "store_percentage",
"BriefDescription": "Proportion of FP operations", "MetricExpr": "((ST_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Instruction", "BriefDescription": "This metric measures store operations as a percentage of operations speculatively executed.",
"MetricName": "VFP mix" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "1 - (OP_RETIRED/ (CPU_CYCLES * 4))", "MetricName": "scalar_fp_percentage",
"BriefDescription": "Proportion of slots lost", "MetricExpr": "((VFP_SPEC / INST_SPEC) * 100)",
"MetricGroup": "Speculation / TDA", "BriefDescription": "This metric measures scalar floating point operations as a percentage of operations speculatively executed.",
"MetricName": "CPU lost" "MetricGroup": "Operation_Mix",
"ScaleUnit": "1percent of operations"
}, },
{ {
"MetricExpr": "OP_RETIRED/ (CPU_CYCLES * 4)", "MetricName": "retired_rate",
"BriefDescription": "Proportion of slots retiring", "MetricExpr": "OP_RETIRED / OP_SPEC",
"MetricGroup": "Speculation / TDA", "BriefDescription": "Of all the micro-operations issued, what percentage are retired(committed)",
"MetricName": "CPU utilization" "MetricGroup": "General",
"ScaleUnit": "100%"
}, },
{ {
"MetricExpr": "OP_RETIRED - OP_SPEC", "MetricName": "wasted",
"BriefDescription": "Operations lost due to misspeculation", "MetricExpr": "1 - (OP_RETIRED / (CPU_CYCLES * #slots))",
"MetricGroup": "Speculation / TDA", "BriefDescription": "Of all the micro-operations issued, what proportion are lost",
"MetricName": "Operations lost" "MetricGroup": "General",
"ScaleUnit": "100%"
}, },
{ {
"MetricExpr": "1 - (OP_RETIRED / OP_SPEC)", "MetricName": "wasted_rate",
"BriefDescription": "Proportion of operations lost", "MetricExpr": "1 - OP_RETIRED / OP_SPEC",
"MetricGroup": "Speculation / TDA", "BriefDescription": "Of all the micro-operations issued, what percentage are not retired(committed)",
"MetricName": "Operations lost (ratio)" "MetricGroup": "General",
"ScaleUnit": "100%"
}, },
{ {
"MetricExpr": "OP_RETIRED / OP_SPEC", "MetricName": "stall_backend_cache_rate",
"BriefDescription": "Proportion of operations retired", "MetricExpr": "((STALL_BACKEND_CACHE / CPU_CYCLES) * 100)",
"MetricGroup": "Speculation / TDA",
"MetricName": "Operations retired"
},
{
"MetricExpr": "STALL_BACKEND_CACHE / CPU_CYCLES",
"BriefDescription": "Proportion of cycles stalled and no operations issued to backend and cache miss", "BriefDescription": "Proportion of cycles stalled and no operations issued to backend and cache miss",
"MetricGroup": "Stall", "MetricGroup": "Stall",
"MetricName": "Stall backend cache cycles" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "STALL_BACKEND_RESOURCE / CPU_CYCLES", "MetricName": "stall_backend_resource_rate",
"MetricExpr": "((STALL_BACKEND_RESOURCE / CPU_CYCLES) * 100)",
"BriefDescription": "Proportion of cycles stalled and no operations issued to backend and resource full", "BriefDescription": "Proportion of cycles stalled and no operations issued to backend and resource full",
"MetricGroup": "Stall", "MetricGroup": "Stall",
"MetricName": "Stall backend resource cycles" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "STALL_BACKEND_TLB / CPU_CYCLES", "MetricName": "stall_backend_tlb_rate",
"MetricExpr": "((STALL_BACKEND_TLB / CPU_CYCLES) * 100)",
"BriefDescription": "Proportion of cycles stalled and no operations issued to backend and TLB miss", "BriefDescription": "Proportion of cycles stalled and no operations issued to backend and TLB miss",
"MetricGroup": "Stall", "MetricGroup": "Stall",
"MetricName": "Stall backend tlb cycles" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "STALL_FRONTEND_CACHE / CPU_CYCLES", "MetricName": "stall_frontend_cache_rate",
"MetricExpr": "((STALL_FRONTEND_CACHE / CPU_CYCLES) * 100)",
"BriefDescription": "Proportion of cycles stalled and no ops delivered from frontend and cache miss", "BriefDescription": "Proportion of cycles stalled and no ops delivered from frontend and cache miss",
"MetricGroup": "Stall", "MetricGroup": "Stall",
"MetricName": "Stall frontend cache cycles" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "STALL_FRONTEND_TLB / CPU_CYCLES", "MetricName": "stall_frontend_tlb_rate",
"MetricExpr": "((STALL_FRONTEND_TLB / CPU_CYCLES) * 100)",
"BriefDescription": "Proportion of cycles stalled and no ops delivered from frontend and TLB miss", "BriefDescription": "Proportion of cycles stalled and no ops delivered from frontend and TLB miss",
"MetricGroup": "Stall", "MetricGroup": "Stall",
"MetricName": "Stall frontend tlb cycles" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "DTLB_WALK / L1D_TLB", "MetricName": "dtlb_walk_ratio",
"BriefDescription": "D-side walk per d-side translation request", "MetricExpr": "(DTLB_WALK / L1D_TLB)",
"MetricGroup": "TLB", "BriefDescription": "This metric measures the ratio of data TLB Walks to the total number of data TLB accesses. This gives an indication of the effectiveness of the data TLB accesses.",
"MetricName": "DTLB walks" "MetricGroup": "Miss_Ratio;DTLB_Effectiveness",
"ScaleUnit": "1per TLB access"
}, },
{ {
"MetricExpr": "ITLB_WALK / L1I_TLB", "MetricName": "itlb_walk_ratio",
"BriefDescription": "I-side walk per i-side translation request", "MetricExpr": "(ITLB_WALK / L1I_TLB)",
"MetricGroup": "TLB", "BriefDescription": "This metric measures the ratio of instruction TLB Walks to the total number of instruction TLB accesses. This gives an indication of the effectiveness of the instruction TLB accesses.",
"MetricName": "ITLB walks" "MetricGroup": "Miss_Ratio;ITLB_Effectiveness",
"ScaleUnit": "1per TLB access"
}, },
{ {
"MetricExpr": "STALL_SLOT_BACKEND / (CPU_CYCLES * 4)", "ArchStdEvent": "backend_bound"
"BriefDescription": "Fraction of slots backend bound",
"MetricGroup": "TopDownL1",
"MetricName": "backend"
}, },
{ {
"MetricExpr": "1 - (retiring + lost + backend)", "ArchStdEvent": "frontend_bound",
"BriefDescription": "Fraction of slots frontend bound", "MetricExpr": "100 - (retired_fraction + slots_lost_misspeculation_fraction + backend_bound)"
"MetricGroup": "TopDownL1",
"MetricName": "frontend"
}, },
{ {
"MetricExpr": "((OP_SPEC - OP_RETIRED) / (CPU_CYCLES * 4))", "MetricName": "slots_lost_misspeculation_fraction",
"MetricExpr": "100 * ((OP_SPEC - OP_RETIRED) / (CPU_CYCLES * #slots))",
"BriefDescription": "Fraction of slots lost due to misspeculation", "BriefDescription": "Fraction of slots lost due to misspeculation",
"MetricGroup": "TopDownL1", "MetricGroup": "Default;TopdownL1",
"MetricName": "lost" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "(OP_RETIRED / (CPU_CYCLES * 4))", "MetricName": "retired_fraction",
"MetricExpr": "100 * (OP_RETIRED / (CPU_CYCLES * #slots))",
"BriefDescription": "Fraction of slots retiring, useful work", "BriefDescription": "Fraction of slots retiring, useful work",
"MetricGroup": "TopDownL1", "MetricGroup": "Default;TopdownL1",
"MetricName": "retiring" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "backend - backend_memory", "MetricName": "backend_core",
"MetricExpr": "(backend_bound / 100) - backend_memory",
"BriefDescription": "Fraction of slots the CPU was stalled due to backend non-memory subsystem issues", "BriefDescription": "Fraction of slots the CPU was stalled due to backend non-memory subsystem issues",
"MetricGroup": "TopDownL2", "MetricGroup": "TopdownL2",
"MetricName": "backend_core" "ScaleUnit": "100%"
}, },
{ {
"MetricExpr": "(STALL_BACKEND_TLB + STALL_BACKEND_CACHE + STALL_BACKEND_MEM) / CPU_CYCLES ", "MetricName": "backend_memory",
"MetricExpr": "(STALL_BACKEND_TLB + STALL_BACKEND_CACHE) / CPU_CYCLES",
"BriefDescription": "Fraction of slots the CPU was stalled due to backend memory subsystem issues (cache/tlb miss)", "BriefDescription": "Fraction of slots the CPU was stalled due to backend memory subsystem issues (cache/tlb miss)",
"MetricGroup": "TopDownL2", "MetricGroup": "TopdownL2",
"MetricName": "backend_memory" "ScaleUnit": "100%"
}, },
{ {
"MetricExpr": " (BR_MIS_PRED_RETIRED / GPC_FLUSH) * lost", "MetricName": "branch_mispredict",
"MetricExpr": "(BR_MIS_PRED_RETIRED / GPC_FLUSH) * slots_lost_misspeculation_fraction",
"BriefDescription": "Fraction of slots lost due to branch misprediciton", "BriefDescription": "Fraction of slots lost due to branch misprediciton",
"MetricGroup": "TopDownL2", "MetricGroup": "TopdownL2",
"MetricName": "branch_mispredict" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "frontend - frontend_latency", "MetricName": "frontend_bandwidth",
"MetricExpr": "frontend_bound - frontend_latency",
"BriefDescription": "Fraction of slots the CPU did not dispatch at full bandwidth - able to dispatch partial slots only (1, 2, or 3 uops)", "BriefDescription": "Fraction of slots the CPU did not dispatch at full bandwidth - able to dispatch partial slots only (1, 2, or 3 uops)",
"MetricGroup": "TopDownL2", "MetricGroup": "TopdownL2",
"MetricName": "frontend_bandwidth" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "(STALL_FRONTEND - ((STALL_SLOT_FRONTEND - (frontend * CPU_CYCLES * 4)) / 4)) / CPU_CYCLES", "MetricName": "frontend_latency",
"MetricExpr": "((STALL_FRONTEND - ((STALL_SLOT_FRONTEND - ((frontend_bound / 100) * CPU_CYCLES * #slots)) / #slots)) / CPU_CYCLES) * 100",
"BriefDescription": "Fraction of slots the CPU was stalled due to frontend latency issues (cache/tlb miss); nothing to dispatch", "BriefDescription": "Fraction of slots the CPU was stalled due to frontend latency issues (cache/tlb miss); nothing to dispatch",
"MetricGroup": "TopDownL2", "MetricGroup": "TopdownL2",
"MetricName": "frontend_latency" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "lost - branch_mispredict", "MetricName": "other_miss_pred",
"MetricExpr": "slots_lost_misspeculation_fraction - branch_mispredict",
"BriefDescription": "Fraction of slots lost due to other/non-branch misprediction misspeculation", "BriefDescription": "Fraction of slots lost due to other/non-branch misprediction misspeculation",
"MetricGroup": "TopDownL2", "MetricGroup": "TopdownL2",
"MetricName": "other_clears" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "(IXU_NUM_UOPS_ISSUED + FSU_ISSUED) / (CPU_CYCLES * 6)", "MetricName": "pipe_utilization",
"MetricExpr": "100 * ((IXU_NUM_UOPS_ISSUED + FSU_ISSUED) / (CPU_CYCLES * 6))",
"BriefDescription": "Fraction of execute slots utilized", "BriefDescription": "Fraction of execute slots utilized",
"MetricGroup": "TopDownL2", "MetricGroup": "TopdownL2",
"MetricName": "pipe_utilization" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "STALL_BACKEND_MEM / CPU_CYCLES", "MetricName": "d_cache_l2_miss_rate",
"MetricExpr": "((STALL_BACKEND_MEM / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled due to data L2 cache miss", "BriefDescription": "Fraction of cycles the CPU was stalled due to data L2 cache miss",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "d_cache_l2_miss" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "STALL_BACKEND_CACHE / CPU_CYCLES", "MetricName": "d_cache_miss_rate",
"MetricExpr": "((STALL_BACKEND_CACHE / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled due to data cache miss", "BriefDescription": "Fraction of cycles the CPU was stalled due to data cache miss",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "d_cache_miss" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "STALL_BACKEND_TLB / CPU_CYCLES", "MetricName": "d_tlb_miss_rate",
"MetricExpr": "((STALL_BACKEND_TLB / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled due to data TLB miss", "BriefDescription": "Fraction of cycles the CPU was stalled due to data TLB miss",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "d_tlb_miss" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "FSU_ISSUED / (CPU_CYCLES * 2)", "MetricName": "fsu_pipe_utilization",
"MetricExpr": "((FSU_ISSUED / (CPU_CYCLES * 2)) * 100)",
"BriefDescription": "Fraction of FSU execute slots utilized", "BriefDescription": "Fraction of FSU execute slots utilized",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "fsu_pipe_utilization" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "STALL_FRONTEND_CACHE / CPU_CYCLES", "MetricName": "i_cache_miss_rate",
"MetricExpr": "((STALL_FRONTEND_CACHE / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled due to instruction cache miss", "BriefDescription": "Fraction of cycles the CPU was stalled due to instruction cache miss",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "i_cache_miss" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": " STALL_FRONTEND_TLB / CPU_CYCLES ", "MetricName": "i_tlb_miss_rate",
"MetricExpr": "((STALL_FRONTEND_TLB / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled due to instruction TLB miss", "BriefDescription": "Fraction of cycles the CPU was stalled due to instruction TLB miss",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "i_tlb_miss" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "IXU_NUM_UOPS_ISSUED / (CPU_CYCLES / 4)", "MetricName": "ixu_pipe_utilization",
"MetricExpr": "((IXU_NUM_UOPS_ISSUED / (CPU_CYCLES * #slots)) * 100)",
"BriefDescription": "Fraction of IXU execute slots utilized", "BriefDescription": "Fraction of IXU execute slots utilized",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "ixu_pipe_utilization" "ScaleUnit": "1percent of slots"
}, },
{ {
"MetricExpr": "IDR_STALL_FLUSH / CPU_CYCLES", "MetricName": "stall_recovery_rate",
"MetricExpr": "((IDR_STALL_FLUSH / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled due to flush recovery", "BriefDescription": "Fraction of cycles the CPU was stalled due to flush recovery",
"MetricGroup": "TopDownL3", "MetricGroup": "TopdownL3",
"MetricName": "recovery" "ScaleUnit": "1percent of slots"
},
{
"MetricExpr": "STALL_BACKEND_RESOURCE / CPU_CYCLES",
"BriefDescription": "Fraction of cycles the CPU was stalled due to core resource shortage",
"MetricGroup": "TopDownL3",
"MetricName": "resource"
}, },
{ {
"MetricExpr": "IDR_STALL_FSU_SCHED / CPU_CYCLES ", "MetricName": "stall_fsu_sched_rate",
"MetricExpr": "((IDR_STALL_FSU_SCHED / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled and FSU was full", "BriefDescription": "Fraction of cycles the CPU was stalled and FSU was full",
"MetricGroup": "TopDownL4", "MetricGroup": "TopdownL4",
"MetricName": "stall_fsu_sched" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "IDR_STALL_IXU_SCHED / CPU_CYCLES ", "MetricName": "stall_ixu_sched_rate",
"MetricExpr": "((IDR_STALL_IXU_SCHED / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled and IXU was full", "BriefDescription": "Fraction of cycles the CPU was stalled and IXU was full",
"MetricGroup": "TopDownL4", "MetricGroup": "TopdownL4",
"MetricName": "stall_ixu_sched" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "IDR_STALL_LOB_ID / CPU_CYCLES ", "MetricName": "stall_lob_id_rate",
"MetricExpr": "((IDR_STALL_LOB_ID / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled and LOB was full", "BriefDescription": "Fraction of cycles the CPU was stalled and LOB was full",
"MetricGroup": "TopDownL4", "MetricGroup": "TopdownL4",
"MetricName": "stall_lob_id" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "IDR_STALL_ROB_ID / CPU_CYCLES", "MetricName": "stall_rob_id_rate",
"MetricExpr": "((IDR_STALL_ROB_ID / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled and ROB was full", "BriefDescription": "Fraction of cycles the CPU was stalled and ROB was full",
"MetricGroup": "TopDownL4", "MetricGroup": "TopdownL4",
"MetricName": "stall_rob_id" "ScaleUnit": "1percent of cycles"
}, },
{ {
"MetricExpr": "IDR_STALL_SOB_ID / CPU_CYCLES ", "MetricName": "stall_sob_id_rate",
"MetricExpr": "((IDR_STALL_SOB_ID / CPU_CYCLES) * 100)",
"BriefDescription": "Fraction of cycles the CPU was stalled and SOB was full", "BriefDescription": "Fraction of cycles the CPU was stalled and SOB was full",
"MetricGroup": "TopDownL4", "MetricGroup": "TopdownL4",
"MetricName": "stall_sob_id" "ScaleUnit": "1percent of cycles"
} }
] ]
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7