Add metric has_event, update intel vendor events
Ian Rogers
reviewerv2 cover Add metric has_event, update intel vendor events
Add a new has_event function for metrics so that events that can be disabled by the kernel/firmware don't cause metrics to fail. Use this function for Intel transaction metrics fixing "perf all metrics test" on systems with TSX disabled. The update conversion script is posted in:
https://github.com/intel/perfmon/pull/90
Re-generate Intel vendor events using:
https://github.com/intel/perfmon/blob/main/scripts/create_perf_json.py
Adding rocketlake support, uncore and many core events for meteorlake, and smaller updates for cascakelakex, icelake, icelakex, sapphirerapids, skylake, skylakex and tigerlake.
v2. Handle failed memory allocated for evlist, John Garry.
Ian Rogers (12):
perf expr: Add has_event function
perf jevents: Support for has_event function
perf vendor metrics intel: Make transaction metrics conditional perf vendor events intel: Add rocketlake events/metrics perf vendor events intel: Update meteorlake to 1.03 perf vendor events intel: Update cascadelakex to 1.19 perf vendor events intel: Update icelake to 1.19
perf vendor events intel: Update icelakex to 1.21 perf vendor events intel: Update sapphirerapids to 1.14 perf vendor events intel: Update skylake to 57
perf vendor events intel: Update skylakex to 1.31 perf vendor events intel: Update tigerlake to 1.13
.../arch/x86/alderlake/adl-metrics.json | 8 +- .../arch/x86/cascadelakex/clx-metrics.json | 8 +- .../arch/x86/cascadelakex/frontend.json | 43 +- .../arch/x86/cascadelakex/pipeline.json | 17 +- .../x86/cascadelakex/uncore-interconnect.json | 2 +- .../arch/x86/cascadelakex/uncore-memory.json | 2 +- .../pmu-events/arch/x86/icelake/cache.json | 8 +- .../pmu-events/arch/x86/icelake/frontend.json | 32 +- .../arch/x86/icelake/icl-metrics.json | 8 +- .../pmu-events/arch/x86/icelake/pipeline.json | 6 +- .../arch/x86/icelakex/frontend.json | 32 +- .../arch/x86/icelakex/icx-metrics.json | 8 +- .../arch/x86/icelakex/pipeline.json | 4 +- .../x86/icelakex/uncore-interconnect.json | 2 +- tools/perf/pmu-events/arch/x86/mapfile.csv | 17 +- .../pmu-events/arch/x86/meteorlake/cache.json | 811 +++++++++ .../arch/x86/meteorlake/floating-point.json | 143 ++ .../arch/x86/meteorlake/frontend.json | 410 +++++ .../arch/x86/meteorlake/memory.json | 142 +- .../pmu-events/arch/x86/meteorlake/other.json | 57 +- .../arch/x86/meteorlake/pipeline.json | 1211 ++++++++++++- .../arch/x86/meteorlake/uncore-cache.json | 18 + .../x86/meteorlake/uncore-interconnect.json | 42 + .../arch/x86/meteorlake/uncore-memory.json | 126 ++ .../arch/x86/meteorlake/virtual-memory.json | 257 +++ .../pmu-events/arch/x86/rocketlake/cache.json | 894 ++++++++++ .../arch/x86/rocketlake/floating-point.json | 105 ++ .../arch/x86/rocketlake/frontend.json | 377 ++++ .../arch/x86/rocketlake/memory.json | 394 +++++ .../arch/x86/rocketlake/metricgroups.json | 113 ++ .../pmu-events/arch/x86/rocketlake/other.json | 242 +++ .../arch/x86/rocketlake/pipeline.json | 801 +++++++++ .../arch/x86/rocketlake/rkl-metrics.json | 1571 +++++++++++++++++ .../x86/rocketlake/uncore-interconnect.json | 74 + .../arch/x86/rocketlake/uncore-other.json | 9 + .../arch/x86/rocketlake/virtual-memory.json | 165 ++ .../arch/x86/sapphirerapids/pipeline.json | 2 +- .../arch/x86/sapphirerapids/spr-metrics.json | 8 +- .../arch/x86/sapphirerapids/uncore-cache.json | 308 ++++ .../sapphirerapids/uncore-interconnect.json | 2 +- .../pmu-events/arch/x86/skylake/frontend.json | 43 +- .../pmu-events/arch/x86/skylake/pipeline.json | 17 +- .../arch/x86/skylake/skl-metrics.json | 8 +- .../arch/x86/skylakex/frontend.json | 43 +- .../arch/x86/skylakex/pipeline.json | 17 +- .../arch/x86/skylakex/skx-metrics.json | 8 +- .../x86/skylakex/uncore-interconnect.json | 2 +- .../arch/x86/skylakex/uncore-memory.json | 2 +- .../arch/x86/tigerlake/frontend.json | 32 +- .../arch/x86/tigerlake/pipeline.json | 6 +- .../arch/x86/tigerlake/tgl-metrics.json | 8 +- tools/perf/pmu-events/metric.py | 8 +- tools/perf/tests/expr.c | 4 + tools/perf/util/expr.c | 21 + tools/perf/util/expr.h | 1 + tools/perf/util/expr.l | 1 + tools/perf/util/expr.y | 8 +- 57 files changed, 8506 insertions(+), 202 deletions(-) create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/floating-point.json create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/uncore-cache.json create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/uncore-interconnect.json create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/uncore-memory.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/cache.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/floating-point.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/frontend.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/memory.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/metricgroups.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/other.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/pipeline.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/rkl-metrics.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/uncore-interconnect.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/uncore-other.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/virtual-memory.json
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 1/12 perf expr: Add has_event function
Some events are dependent on firmware/kernel enablement. Allow such events to be detected when the metric is parsed so that the metric's event parsing doesn't fail.
--- tools/perf/tests/expr.c | 4 ++++ tools/perf/util/expr.c | 21 +++++++++++++++++++++ tools/perf/util/expr.h | 1 + tools/perf/util/expr.l | 1 + tools/perf/util/expr.y | 8 +++++++- 5 files changed, 34 insertions(+), 1 deletion(-)
tools/perf/tests/expr.c +4 -0
@@ -254,6 +254,10 @@ static int test__expr(struct test_suite *t __maybe_unused, int subtest __maybe_uTEST_ASSERT_VAL ("source count" ,hashmap__size (ctx->ids) ==1 );TEST_ASSERT_VAL ("source count" ,hashmap__find (ctx->ids,"EVENT1" , &val_ptr)); +/* has_event returns 1 when an event exists. */ +expr__add_id_val (ctx,strdup ("cycles" ),2 ); + ret =test (ctx,"has_event(cycles)" ,1 ); +expr__ctx_free (ctx);return 0 ;
tools/perf/util/expr.c +21 -0
@@ -8,6 +8,7 @@#include "cpumap.h" #include "cputopo.h" #include "debug.h" +#include "evlist.h" #include "expr.h" #include "expr-bison.h" #include "expr-flex.h"
@@ -474,3 +475,23 @@ double expr__get_literal(const char *literal, const struct expr_scanner_ctx *ctxpr_debug2 ("literal: %s = %f\n" , literal, result);return result; } + +/* Does the event 'id' parse? Determine via ctx->ids if possible. */ +double expr__has_event (const struct expr_parse_ctx *ctx,bool compute_ids,const char *id) +{ +struct evlist *tmp; +double ret; + +if (hashmap__find (ctx->ids, id,/*value=*/ NULL )) +return 1.0 ; + +if (!compute_ids) +return 0.0 ; + + tmp =evlist__new (); +if (!tmp) +return NAN ; + ret =parse_event (tmp, id) ?0 :1 ; +evlist__delete (tmp); +return ret; +}
tools/perf/util/expr.h +1 -0
@@ -54,5 +54,6 @@ int expr__find_ids(const char *expr, const char *one,double expr_id_data__value (const struct expr_id_data *data);double expr_id_data__source_count (const struct expr_id_data *data);double expr__get_literal (const char *literal,const struct expr_scanner_ctx *ctx); +double expr__has_event (const struct expr_parse_ctx *ctx,bool compute_ids,const char *id);#endif
tools/perf/util/expr.l +1 -0
@@ -113,6 +113,7 @@ min { return MIN; } if { return IF; } else { return ELSE; } source_count { return SOURCE_COUNT; } +has_event { return HAS_EVENT; } {literal} { return literal(yyscanner, sctx); } {number} { return value(yyscanner); } {symbol} { return str(yyscanner, ID, sctx->runtime); }
tools/perf/util/expr.y +7 -1
@@ -37,7 +37,7 @@ } ids; } -%token ID NUMBER MIN MAX IF ELSE LITERAL D_RATIO SOURCE_COUNT EXPR_ERROR +%token ID NUMBER MIN MAX IF ELSE LITERAL D_RATIO SOURCE_COUNT HAS_EVENT EXPR_ERROR %left MIN MAX IF %left '|' %left '^'
@@ -199,6 +199,12 @@ expr: NUMBER } | ID { $$ = handle_id(ctx, $1, compute_ids, /*source_count=*/false); } | SOURCE_COUNT '(' ID ')' { $$ = handle_id(ctx, $3, compute_ids, /*source_count=*/true); } +| HAS_EVENT '(' ID ')' +{ + $$.val = expr__has_event(ctx, compute_ids, $3); + $$.ids = NULL; + free($3); +} | expr '|' expr { if (is_const($1.val) && is_const($3.val)) {
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 2/12 perf jevents: Support for has_event function
In reply to Ian RogersSupport for the new has_event function in metrics.
--- tools/perf/pmu-events/metric.py | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
tools/perf/pmu-events/metric.py +7 -1
@@ -408,6 +408,12 @@ def source_count(event: Event) -> Function: return Function('source_count', event) +def has_event(event: Event) -> Function: + # pylint: disable=redefined-builtin + # pylint: disable=invalid-name + return Function('has_event', event) + + class Metric: """An individual metric that will specifiable on the perf command line.""" groups: Set[str]
@@ -539,7 +545,7 @@ def ParsePerfJson(orig: str) -> Expression: r'Event(r"\1")', py) py = re.sub(r'#Event\(r"([^"]*)"\)', r'Literal("#\1")', py) py = re.sub(r'([0-9]+)Event\(r"(e[0-9]+)"\)', r'\1\2', py) - keywords = ['if', 'else', 'min', 'max', 'd_ratio', 'source_count'] + keywords = ['if', 'else', 'min', 'max', 'd_ratio', 'source_count', 'has_event'] for kw in keywords: py = re.sub(rf'Event\(r"{kw}"\)', kw, py)
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 3/12 perf vendor metrics intel: Make transaction metrics conditional
In reply to Ian RogersMake the transaction metrics conditional on the cycles-tx event being present. This event may not be present when TSX extensions have been disabled.
--- tools/perf/pmu-events/arch/x86/alderlake/adl-metrics.json | 8 ++++---- .../pmu-events/arch/x86/cascadelakex/clx-metrics.json | 8 ++++---- tools/perf/pmu-events/arch/x86/icelake/icl-metrics.json | 8 ++++---- tools/perf/pmu-events/arch/x86/icelakex/icx-metrics.json | 8 ++++---- .../pmu-events/arch/x86/sapphirerapids/spr-metrics.json | 8 ++++---- tools/perf/pmu-events/arch/x86/skylake/skl-metrics.json | 8 ++++---- tools/perf/pmu-events/arch/x86/skylakex/skx-metrics.json | 8 ++++---- tools/perf/pmu-events/arch/x86/tigerlake/tgl-metrics.json | 8 ++++---- 8 files changed, 32 insertions(+), 32 deletions(-)
tools/perf/pmu-events/arch/x86/alderlake/adl-metrics.json +4 -4
@@ -92,28 +92,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
tools/perf/pmu-events/arch/x86/cascadelakex/clx-metrics.json +4 -4
@@ -1830,28 +1830,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
tools/perf/pmu-events/arch/x86/icelake/icl-metrics.json +4 -4
@@ -1516,28 +1516,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
tools/perf/pmu-events/arch/x86/icelakex/icx-metrics.json +4 -4
@@ -1812,28 +1812,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
tools/perf/pmu-events/arch/x86/sapphirerapids/spr-metrics.json +4 -4
@@ -1938,28 +1938,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
tools/perf/pmu-events/arch/x86/skylake/skl-metrics.json +4 -4
@@ -1466,28 +1466,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
tools/perf/pmu-events/arch/x86/skylakex/skx-metrics.json +4 -4
@@ -1774,28 +1774,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
tools/perf/pmu-events/arch/x86/tigerlake/tgl-metrics.json +4 -4
@@ -1530,28 +1530,28 @@ }, { "BriefDescription": "Percentage of cycles in aborted transactions.", - "MetricExpr": "max(cpu@cycles\\-t@ - cpu@cycles\\-ct@, 0) / cycles", + "MetricExpr": "(max(cycles\\-t - cycles\\-ct, 0) / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_aborted_cycles", "ScaleUnit": "100%" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of elisions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@el\\-start@", + "MetricExpr": "(cycles\\-t / el\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_elision", "ScaleUnit": "1cycles / elision" }, { "BriefDescription": "Number of cycles within a transaction divided by the number of transactions.", - "MetricExpr": "cpu@cycles\\-t@ / cpu@tx\\-start@", + "MetricExpr": "(cycles\\-t / tx\\-start if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_cycles_per_transaction", "ScaleUnit": "1cycles / transaction" }, { "BriefDescription": "Percentage of cycles within a transaction region.", - "MetricExpr": "cpu@cycles\\-t@ / cycles", + "MetricExpr": "(cycles\\-t / cycles if has_event(cycles\\-t) else 0)", "MetricGroup": "transaction", "MetricName": "tsx_transactional_cycles", "ScaleUnit": "100%"
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 4/12 perf vendor events intel: Add rocketlake events/metrics
In reply to Ian RogersAdd RocketLake events added to Intel perfmon in:
https://github.com/intel/perfmon/commit/f5f47dc938d81d5cc491cf8ac1f90bee18e238cf
mapfile.csv is updated accordingly with the CPUID matching one previously associated with icelake.
--- tools/perf/pmu-events/arch/x86/mapfile.csv | 3 +- .../pmu-events/arch/x86/rocketlake/cache.json | 894 ++++++++++ .../arch/x86/rocketlake/floating-point.json | 105 ++ .../arch/x86/rocketlake/frontend.json | 377 ++++ .../arch/x86/rocketlake/memory.json | 394 +++++ .../arch/x86/rocketlake/metricgroups.json | 113 ++ .../pmu-events/arch/x86/rocketlake/other.json | 242 +++ .../arch/x86/rocketlake/pipeline.json | 801 +++++++++ .../arch/x86/rocketlake/rkl-metrics.json | 1571 +++++++++++++++++ .../x86/rocketlake/uncore-interconnect.json | 74 + .../arch/x86/rocketlake/uncore-other.json | 9 + .../arch/x86/rocketlake/virtual-memory.json | 165 ++ 12 files changed, 4747 insertions(+), 1 deletion(-) create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/cache.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/floating-point.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/frontend.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/memory.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/metricgroups.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/other.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/pipeline.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/rkl-metrics.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/uncore-interconnect.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/uncore-other.json create mode 100644 tools/perf/pmu-events/arch/x86/rocketlake/virtual-memory.json
tools/perf/pmu-events/arch/x86/mapfile.csv +2 -1
@@ -13,7 +13,7 @@ GenuineIntel-6-B6,v1.00,grandridge,core GenuineIntel-6-A[DE],v1.01,graniterapids,core GenuineIntel-6-(3C|45|46),v33,haswell,core GenuineIntel-6-3F,v27,haswellx,core -GenuineIntel-6-(7D|7E|A7),v1.18,icelake,core +GenuineIntel-6-7[DE],v1.18,icelake,core GenuineIntel-6-6[AC],v1.20,icelakex,core GenuineIntel-6-3A,v24,ivybridge,core GenuineIntel-6-3E,v23,ivytown,core
@@ -22,6 +22,7 @@ GenuineIntel-6-(57|85),v10,knightslanding,core GenuineIntel-6-A[AC],v1.01,meteorlake,core GenuineIntel-6-1[AEF],v3,nehalemep,core GenuineIntel-6-2E,v3,nehalemex,core +GenuineIntel-6-A7,v1.01,rocketlake,core GenuineIntel-6-2A,v19,sandybridge,core GenuineIntel-6-(8F|CF),v1.13,sapphirerapids,core GenuineIntel-6-AF,v1.00,sierraforest,core
tools/perf/pmu-events/arch/x86/rocketlake/cache.json new file +894 -0
894 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/floating-point.json new file +105 -0
105 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/frontend.json new file +377 -0
377 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/memory.json new file +394 -0
394 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/metricgroups.json new file +113 -0
113 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/other.json new file +242 -0
242 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/pipeline.json new file +801 -0
801 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/rkl-metrics.json new file +1571 -0
59 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/uncore-interconnect.json new file +74 -0
74 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/rocketlake/uncore-other.json new file +9 -0
Lines of this file are not shown. View the raw message.
tools/perf/pmu-events/arch/x86/rocketlake/virtual-memory.json new file +165 -0
165 lines not on this page · raw message
The rest of this diff is not shown. View the raw message.
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Part of the diff is not shown here. View the raw message.
Ian Rogers
reviewerPATCH v2 5/12 perf vendor events intel: Update meteorlake to 1.03
In reply to Ian Rogers1.03 events were released in:
https://github.com/intel/perfmon/commit/501a29e88b57e8b01d610168d0101d6181b15e28
It added a lot of events and all uncore events.
--- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- .../pmu-events/arch/x86/meteorlake/cache.json | 811 +++++++++++ .../arch/x86/meteorlake/floating-point.json | 143 ++ .../arch/x86/meteorlake/frontend.json | 410 ++++++ .../arch/x86/meteorlake/memory.json | 142 +- .../pmu-events/arch/x86/meteorlake/other.json | 57 +- .../arch/x86/meteorlake/pipeline.json | 1211 ++++++++++++++++- .../arch/x86/meteorlake/uncore-cache.json | 18 + .../x86/meteorlake/uncore-interconnect.json | 42 + .../arch/x86/meteorlake/uncore-memory.json | 126 ++ .../arch/x86/meteorlake/virtual-memory.json | 257 ++++ 11 files changed, 3116 insertions(+), 103 deletions(-) create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/floating-point.json create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/uncore-cache.json create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/uncore-interconnect.json create mode 100644 tools/perf/pmu-events/arch/x86/meteorlake/uncore-memory.json
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -19,7 +19,7 @@ GenuineIntel-6-3A,v24,ivybridge,core GenuineIntel-6-3E,v23,ivytown,core GenuineIntel-6-2D,v23,jaketown,core GenuineIntel-6-(57|85),v10,knightslanding,core -GenuineIntel-6-A[AC],v1.01,meteorlake,core +GenuineIntel-6-A[AC],v1.03,meteorlake,core GenuineIntel-6-1[AEF],v3,nehalemep,core GenuineIntel-6-2E,v3,nehalemex,core GenuineIntel-6-A7,v1.01,rocketlake,core
tools/perf/pmu-events/arch/x86/meteorlake/cache.json +811 -0
856 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/floating-point.json new file +143 -0
143 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/frontend.json +410 -0
425 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/memory.json +112 -30
159 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/other.json +33 -24
74 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/pipeline.json +1163 -48
1,296 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/uncore-cache.json new file +18 -0
@@ -0,0 +1,18 @@ +[ + { + "BriefDescription": "Number of all entries allocated. Includes also retries.", + "EventCode": "0x35", + "EventName": "UNC_HAC_CBO_TOR_ALLOCATION.ALL", + "PerPkg": "1", + "UMask": "0x8", + "Unit": "HAC_CBO" + }, + { + "BriefDescription": "Asserted on coherent DRD + DRdPref allocations into the queue. Cacheable only", + "EventCode": "0x35", + "EventName": "UNC_HAC_CBO_TOR_ALLOCATION.DRD", + "PerPkg": "1", + "UMask": "0x1", + "Unit": "HAC_CBO" + } +]
tools/perf/pmu-events/arch/x86/meteorlake/uncore-interconnect.json new file +42 -0
21 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/uncore-memory.json new file +126 -0
126 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/meteorlake/virtual-memory.json +257 -0
257 lines not on this page · raw message
The rest of this diff is not shown. View the raw message.
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Part of the diff is not shown here. View the raw message.
Ian Rogers
reviewerPATCH v2 6/12 perf vendor events intel: Update cascadelakex to 1.19
In reply to Ian RogersUpdates were released in:
https://github.com/intel/perfmon/commit/e4f83534f23a69e6da55c672c4d929919688c9b6
Adds the events IDQ.DSB_CYCLES_OK, IDQ.DSB_CYCLES_ANY, ICACHE_TAG.STALLS, DECODE.LCP, LSD.CYCLES_OK. Descriptions are also updated.
--- .../arch/x86/cascadelakex/frontend.json | 43 ++++++++++++++++--- .../arch/x86/cascadelakex/pipeline.json | 17 ++++++-- .../x86/cascadelakex/uncore-interconnect.json | 2 +- .../arch/x86/cascadelakex/uncore-memory.json | 2 +- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- 5 files changed, 54 insertions(+), 12 deletions(-)
tools/perf/pmu-events/arch/x86/cascadelakex/frontend.json +38 -5
@@ -7,6 +7,14 @@ "SampleAfterValue": "100003", "UMask": "0x1" }, + { + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to ILD_STALL.LCP]", + "EventCode": "0x87", + "EventName": "DECODE.LCP", + "PublicDescription": "Counts 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 chunk. [This event is alias to ILD_STALL.LCP]", + "SampleAfterValue": "2000003", + "UMask": "0x1" + }, { "BriefDescription": "Decode Stream Buffer (DSB)-to-MITE switches", "EventCode": "0xAB",
@@ -245,27 +253,34 @@ "UMask": "0x2" }, { - "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss.", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_TAG.STALLS]", "EventCode": "0x83", "EventName": "ICACHE_64B.IFTAG_STALL", "SampleAfterValue": "200003", "UMask": "0x4" }, { - "BriefDescription": "Cycles Decode Stream Buffer (DSB) is delivering 4 Uops", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_64B.IFTAG_STALL]", + "EventCode": "0x83", + "EventName": "ICACHE_TAG.STALLS", + "SampleAfterValue": "200003", + "UMask": "0x4" + }, + { + "BriefDescription": "Cycles Decode Stream Buffer (DSB) is delivering 4 Uops [This event is alias to IDQ.DSB_CYCLES_OK]", "CounterMask": "4", "EventCode": "0x79", "EventName": "IDQ.ALL_DSB_CYCLES_4_UOPS", - "PublicDescription": "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 IDQ.", + "PublicDescription": "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 IDQ. [This event is alias to IDQ.DSB_CYCLES_OK]", "SampleAfterValue": "2000003", "UMask": "0x18" }, { - "BriefDescription": "Cycles Decode Stream Buffer (DSB) is delivering any Uop", + "BriefDescription": "Cycles Decode Stream Buffer (DSB) is delivering any Uop [This event is alias to IDQ.DSB_CYCLES_ANY]", "CounterMask": "1", "EventCode": "0x79", "EventName": "IDQ.ALL_DSB_CYCLES_ANY_UOPS", - "PublicDescription": "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 IDQ.", + "PublicDescription": "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 IDQ. [This event is alias to IDQ.DSB_CYCLES_ANY]", "SampleAfterValue": "2000003", "UMask": "0x18" },
@@ -296,6 +311,24 @@ "SampleAfterValue": "2000003", "UMask": "0x8" }, + { + "BriefDescription": "Cycles Decode Stream Buffer (DSB) is delivering any Uop [This event is alias to IDQ.ALL_DSB_CYCLES_ANY_UOPS]", + "CounterMask": "1", + "EventCode": "0x79", + "EventName": "IDQ.DSB_CYCLES_ANY", + "PublicDescription": "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 IDQ. [This event is alias to IDQ.ALL_DSB_CYCLES_ANY_UOPS]", + "SampleAfterValue": "2000003", + "UMask": "0x18" + }, + { + "BriefDescription": "Cycles Decode Stream Buffer (DSB) is delivering 4 Uops [This event is alias to IDQ.ALL_DSB_CYCLES_4_UOPS]", + "CounterMask": "4", + "EventCode": "0x79", + "EventName": "IDQ.DSB_CYCLES_OK", + "PublicDescription": "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 IDQ. [This event is alias to IDQ.ALL_DSB_CYCLES_4_UOPS]", + "SampleAfterValue": "2000003", + "UMask": "0x18" + }, { "BriefDescription": "Uops delivered to Instruction Decode Queue (IDQ) from the Decode Stream Buffer (DSB) path", "EventCode": "0x79",
tools/perf/pmu-events/arch/x86/cascadelakex/pipeline.json +13 -4
@@ -361,10 +361,10 @@ "UMask": "0x1" }, { - "BriefDescription": "Stalls caused by changing prefix length of the instruction.", + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to DECODE.LCP]", "EventCode": "0x87", "EventName": "ILD_STALL.LCP", - "PublicDescription": "Counts 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 chunk.", + "PublicDescription": "Counts 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 chunk. [This event is alias to DECODE.LCP]", "SampleAfterValue": "2000003", "UMask": "0x1" },
@@ -488,11 +488,11 @@ "UMask": "0x1" }, { - "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder.", + "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder. [This event is alias to LSD.CYCLES_OK]", "CounterMask": "4", "EventCode": "0xA8", "EventName": "LSD.CYCLES_4_UOPS", - "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector).", + "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector). [This event is alias to LSD.CYCLES_OK]", "SampleAfterValue": "2000003", "UMask": "0x1" },
@@ -505,6 +505,15 @@ "SampleAfterValue": "2000003", "UMask": "0x1" }, + { + "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder. [This event is alias to LSD.CYCLES_4_UOPS]", + "CounterMask": "4", + "EventCode": "0xA8", + "EventName": "LSD.CYCLES_OK", + "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector). [This event is alias to LSD.CYCLES_4_UOPS]", + "SampleAfterValue": "2000003", + "UMask": "0x1" + }, { "BriefDescription": "Number of Uops delivered by the LSD.", "EventCode": "0xA8",
tools/perf/pmu-events/arch/x86/cascadelakex/uncore-interconnect.json +1 -1
@@ -6606,7 +6606,7 @@ "EventCode": "0x52", "EventName": "UNC_M3UPI_RxC_HELD.PARALLEL_SUCCESS", "PerPkg": "1", - "PublicDescription": "ad and bl messages were actually slotted into the same flit in paralle", + "PublicDescription": "ad and bl messages were actually slotted into the same flit in parallel", "UMask": "0x8", "Unit": "M3UPI" },
tools/perf/pmu-events/arch/x86/cascadelakex/uncore-memory.json +1 -1
@@ -2735,7 +2735,7 @@ "EventCode": "0x81", "EventName": "UNC_M_WPQ_OCCUPANCY", "PerPkg": "1", - "PublicDescription": "Counts 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 happening 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?", + "PublicDescription": "Counts 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 happening in the iMC for writes before they are actually issued to memory. High average occupancies will generally coincide with high write major mode counts.", "Unit": "iMC" }, {
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -5,7 +5,7 @@ GenuineIntel-6-(1C|26|27|35|36),v4,bonnell,core GenuineIntel-6-(3D|47),v28,broadwell,core GenuineIntel-6-56,v10,broadwellde,core GenuineIntel-6-4F,v21,broadwellx,core -GenuineIntel-6-55-[56789ABCDEF],v1.18,cascadelakex,core +GenuineIntel-6-55-[56789ABCDEF],v1.19,cascadelakex,core GenuineIntel-6-9[6C],v1.04,elkhartlake,core GenuineIntel-6-5[CF],v13,goldmont,core GenuineIntel-6-7A,v1.01,goldmontplus,core
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 7/12 perf vendor events intel: Update icelake to 1.19
In reply to Ian RogersUpdates were released in:
https://github.com/intel/perfmon/commit/f3d841189f8964bc240c86301f4c849845630b5b
A number of events are deprecated and event descriptions updated. Adds events ICACHE_DATA.STALLS, ICACHE_TAG.STALLS and DECODE.LCP.
--- .../pmu-events/arch/x86/icelake/cache.json | 8 ++--- .../pmu-events/arch/x86/icelake/frontend.json | 32 ++++++++++++++++--- .../pmu-events/arch/x86/icelake/pipeline.json | 6 ++-- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- 4 files changed, 36 insertions(+), 12 deletions(-)
tools/perf/pmu-events/arch/x86/icelake/cache.json +4 -4
@@ -155,18 +155,18 @@ "UMask": "0x21" }, { - "BriefDescription": "All requests that miss L2 cache. This event is not supported on ICL and ICX products, only supported on RKL products.", + "BriefDescription": "This event is deprecated.", + "Deprecated": "1", "EventCode": "0x24", "EventName": "L2_RQSTS.MISS", - "PublicDescription": "Counts all requests that miss L2 cache. This event is not supported on ICL and ICX products, only supported on RKL products.", "SampleAfterValue": "200003", "UMask": "0x3f" }, { - "BriefDescription": "All L2 requests. This event is not supported on ICL and ICX products, only supported on RKL products.", + "BriefDescription": "This event is deprecated.", + "Deprecated": "1", "EventCode": "0x24", "EventName": "L2_RQSTS.REFERENCES", - "PublicDescription": "Counts all L2 requests. This event is not supported on ICL and ICX products, only supported on RKL products.", "SampleAfterValue": "200003", "UMask": "0xff" },
tools/perf/pmu-events/arch/x86/icelake/frontend.json +28 -4
@@ -7,6 +7,14 @@ "SampleAfterValue": "100003", "UMask": "0x1" }, + { + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to ILD_STALL.LCP]", + "EventCode": "0x87", + "EventName": "DECODE.LCP", + "PublicDescription": "Counts 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 chunk. [This event is alias to ILD_STALL.LCP]", + "SampleAfterValue": "500009", + "UMask": "0x1" + }, { "BriefDescription": "Decode Stream Buffer (DSB)-to-MITE transitions count.", "CounterMask": "1",
@@ -213,10 +221,10 @@ "UMask": "0x1" }, { - "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss.", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss. [This event is alias to ICACHE_DATA.STALLS]", "EventCode": "0x80", "EventName": "ICACHE_16B.IFDATA_STALL", - "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity.", + "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity. [This event is alias to ICACHE_DATA.STALLS]", "SampleAfterValue": "500009", "UMask": "0x4" },
@@ -237,10 +245,26 @@ "UMask": "0x2" }, { - "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss.", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_TAG.STALLS]", "EventCode": "0x83", "EventName": "ICACHE_64B.IFTAG_STALL", - "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss.", + "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_TAG.STALLS]", + "SampleAfterValue": "200003", + "UMask": "0x4" + }, + { + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss. [This event is alias to ICACHE_16B.IFDATA_STALL]", + "EventCode": "0x80", + "EventName": "ICACHE_DATA.STALLS", + "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity. [This event is alias to ICACHE_16B.IFDATA_STALL]", + "SampleAfterValue": "500009", + "UMask": "0x4" + }, + { + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_64B.IFTAG_STALL]", + "EventCode": "0x83", + "EventName": "ICACHE_TAG.STALLS", + "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_64B.IFTAG_STALL]", "SampleAfterValue": "200003", "UMask": "0x4" },
tools/perf/pmu-events/arch/x86/icelake/pipeline.json +3 -3
@@ -318,10 +318,10 @@ "UMask": "0x40" }, { - "BriefDescription": "Stalls caused by changing prefix length of the instruction.", + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to DECODE.LCP]", "EventCode": "0x87", "EventName": "ILD_STALL.LCP", - "PublicDescription": "Counts 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 chunk.", + "PublicDescription": "Counts 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 chunk. [This event is alias to DECODE.LCP]", "SampleAfterValue": "500009", "UMask": "0x1" },
@@ -556,7 +556,7 @@ "BriefDescription": "TMA slots wasted due to incorrect speculation by branch mispredictions", "EventCode": "0xa4", "EventName": "TOPDOWN.BR_MISPREDICT_SLOTS", - "PublicDescription": "Number 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 misprediction.", + "PublicDescription": "Number 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 speculative path as well as the out-of-order engine recovery past a branch misprediction.", "SampleAfterValue": "10000003", "UMask": "0x8" },
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -13,7 +13,7 @@ GenuineIntel-6-B6,v1.00,grandridge,core GenuineIntel-6-A[DE],v1.01,graniterapids,core GenuineIntel-6-(3C|45|46),v33,haswell,core GenuineIntel-6-3F,v27,haswellx,core -GenuineIntel-6-7[DE],v1.18,icelake,core +GenuineIntel-6-7[DE],v1.19,icelake,core GenuineIntel-6-6[AC],v1.20,icelakex,core GenuineIntel-6-3A,v24,ivybridge,core GenuineIntel-6-3E,v23,ivytown,core
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Re: [PATCH v2 07/12] perf vendor events intel: Update icelake to 1.19
Quoted 4 lines · Ian Rogers
Updates were released in:
https://github.com/intel/perfmon/commit/f3d841189f8964bc240c86301f4c849845630b5b
A number of events are deprecated and event descriptions updated. Adds events ICACHE_DATA.STALLS, ICACHE_TAG.STALLS and DECODE.LCP.
why are the events marked as deprecated rather than just being removed?
Vince Weaver
[email protected]
Ian Rogers
reviewerRe: [PATCH v2 07/12] perf vendor events intel: Update icelake to 1.19
Quoted 9 lines · Vince Weaver
Quoted 4 lines · Ian Rogers
Updates were released in:
https://github.com/intel/perfmon/commit/f3d841189f8964bc240c86301f4c849845630b5b
A number of events are deprecated and event descriptions updated. Adds events ICACHE_DATA.STALLS, ICACHE_TAG.STALLS and DECODE.LCP.why are the events marked as deprecated rather than just being removed?
My guess would be so that people who used the deprecated event name don't suddenly get failures. The deprecated flag means that the events no longer show in "perf list" unless --deprecated is specified.
Thanks,
Ian
Quoted 2 lines
Vince Weaver
[email protected]
Ian Rogers
reviewerPATCH v2 8/12 perf vendor events intel: Update icelakex to 1.21
In reply to Ian RogersUpdates were released in:
https://github.com/intel/perfmon/commit/78d47cbbae48a0297a507ae4fea234ff37ff9960
Adds the events ICACHE_DATA.STALLS, ICACHE_TAG.STALLS and DECODE.LCP. Descriptions are also updated.
--- .../arch/x86/icelakex/frontend.json | 32 ++++++++++++++++--- .../arch/x86/icelakex/pipeline.json | 4 +-- .../x86/icelakex/uncore-interconnect.json | 2 +- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- 4 files changed, 32 insertions(+), 8 deletions(-)
tools/perf/pmu-events/arch/x86/icelakex/frontend.json +28 -4
@@ -7,6 +7,14 @@ "SampleAfterValue": "100003", "UMask": "0x1" }, + { + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to ILD_STALL.LCP]", + "EventCode": "0x87", + "EventName": "DECODE.LCP", + "PublicDescription": "Counts 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 chunk. [This event is alias to ILD_STALL.LCP]", + "SampleAfterValue": "500009", + "UMask": "0x1" + }, { "BriefDescription": "Decode Stream Buffer (DSB)-to-MITE transitions count.", "CounterMask": "1",
@@ -213,10 +221,10 @@ "UMask": "0x1" }, { - "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss.", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss. [This event is alias to ICACHE_DATA.STALLS]", "EventCode": "0x80", "EventName": "ICACHE_16B.IFDATA_STALL", - "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity.", + "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity. [This event is alias to ICACHE_DATA.STALLS]", "SampleAfterValue": "500009", "UMask": "0x4" },
@@ -237,10 +245,26 @@ "UMask": "0x2" }, { - "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss.", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_TAG.STALLS]", "EventCode": "0x83", "EventName": "ICACHE_64B.IFTAG_STALL", - "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss.", + "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_TAG.STALLS]", + "SampleAfterValue": "200003", + "UMask": "0x4" + }, + { + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss. [This event is alias to ICACHE_16B.IFDATA_STALL]", + "EventCode": "0x80", + "EventName": "ICACHE_DATA.STALLS", + "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity. [This event is alias to ICACHE_16B.IFDATA_STALL]", + "SampleAfterValue": "500009", + "UMask": "0x4" + }, + { + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_64B.IFTAG_STALL]", + "EventCode": "0x83", + "EventName": "ICACHE_TAG.STALLS", + "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_64B.IFTAG_STALL]", "SampleAfterValue": "200003", "UMask": "0x4" },
tools/perf/pmu-events/arch/x86/icelakex/pipeline.json +2 -2
@@ -318,10 +318,10 @@ "UMask": "0x40" }, { - "BriefDescription": "Stalls caused by changing prefix length of the instruction.", + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to DECODE.LCP]", "EventCode": "0x87", "EventName": "ILD_STALL.LCP", - "PublicDescription": "Counts 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 chunk.", + "PublicDescription": "Counts 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 chunk. [This event is alias to DECODE.LCP]", "SampleAfterValue": "500009", "UMask": "0x1" },
tools/perf/pmu-events/arch/x86/icelakex/uncore-interconnect.json +1 -1
@@ -9311,7 +9311,7 @@ "EventCode": "0x50", "EventName": "UNC_M3UPI_RxC_HELD.PARALLEL_SUCCESS", "PerPkg": "1", - "PublicDescription": "Message Held : Parallel Success : ad and bl messages were actually slotted into the same flit in paralle", + "PublicDescription": "Message Held : Parallel Success : ad and bl messages were actually slotted into the same flit in parallel", "UMask": "0x8", "Unit": "M3UPI" },
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -14,7 +14,7 @@ GenuineIntel-6-A[DE],v1.01,graniterapids,core GenuineIntel-6-(3C|45|46),v33,haswell,core GenuineIntel-6-3F,v27,haswellx,core GenuineIntel-6-7[DE],v1.19,icelake,core -GenuineIntel-6-6[AC],v1.20,icelakex,core +GenuineIntel-6-6[AC],v1.21,icelakex,core GenuineIntel-6-3A,v24,ivybridge,core GenuineIntel-6-3E,v23,ivytown,core GenuineIntel-6-2D,v23,jaketown,core
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 9/12 perf vendor events intel: Update sapphirerapids to 1.14
In reply to Ian RogersUpdates were released in:
https://github.com/intel/perfmon/commit/a84850f1fec633002c35838ed34e51e1f0d6a2dd
Adds a large number of CXL events like UNC_CHA_TOR_OCCUPANCY.IA_MISS_RFO_CXL_ACC, UNC_CHA_TOR_OCCUPANCY.IA_MISS_DRD_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_MISS_DRD_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_MISS_RFO_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_MISS_LLCPREFRFO_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_MISS_RFO_PREF_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_MISS_DRD_PREF_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_MISS_LLCPREFDATA_CXL_ACC, UNC_CHA_TOR_OCCUPANCY.IA_MISS_DRD_PREF_CXL_ACC, UNC_CHA_TOR_OCCUPANCY.IA_MISS_LLCPREFDATA_CXL_ACC, UNC_CHA_TOR_OCCUPANCY.IA_MISS_LLCPREFRFO_CXL_ACC, UNC_CHA_TOR_OCCUPANCY.IA_MISS_RFO_PREF_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_MISS_CXL_ACC, UNC_CHA_TOR_INSERTS.IA_HIT_CXL_ACC, UNC_CHA_TOR_OCCUPANCY.IA_MISS_CXL_ACC, UNC_CHA_TOR_OCCUPANCY.IA_HIT_CXL_ACC.
--- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- .../arch/x86/sapphirerapids/pipeline.json | 2 +- .../arch/x86/sapphirerapids/uncore-cache.json | 308 ++++++++++++++++++ .../sapphirerapids/uncore-interconnect.json | 2 +- 4 files changed, 311 insertions(+), 3 deletions(-)
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -24,7 +24,7 @@ GenuineIntel-6-1[AEF],v3,nehalemep,core GenuineIntel-6-2E,v3,nehalemex,core GenuineIntel-6-A7,v1.01,rocketlake,core GenuineIntel-6-2A,v19,sandybridge,core -GenuineIntel-6-(8F|CF),v1.13,sapphirerapids,core +GenuineIntel-6-(8F|CF),v1.14,sapphirerapids,core GenuineIntel-6-AF,v1.00,sierraforest,core GenuineIntel-6-(37|4A|4C|4D|5A),v15,silvermont,core GenuineIntel-6-(4E|5E|8E|9E|A5|A6),v56,skylake,core
tools/perf/pmu-events/arch/x86/sapphirerapids/pipeline.json +1 -1
@@ -706,7 +706,7 @@ "BriefDescription": "TMA slots wasted due to incorrect speculation by branch mispredictions", "EventCode": "0xa4", "EventName": "TOPDOWN.BR_MISPREDICT_SLOTS", - "PublicDescription": "Number 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 misprediction.", + "PublicDescription": "Number of TMA slots that were wasted due to incorrect speculation by (any type of) branch mispredictions. This event estimates number of speculative operations that were issued but not retired as well as the out-of-order engine recovery past a branch misprediction.", "SampleAfterValue": "10000003", "UMask": "0x8" },
tools/perf/pmu-events/arch/x86/sapphirerapids/uncore-cache.json +308 -0
428 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/sapphirerapids/uncore-interconnect.json +1 -1
@@ -3326,7 +3326,7 @@ "EventCode": "0x50", "EventName": "UNC_M3UPI_RxC_HELD.PARALLEL_SUCCESS", "PerPkg": "1", - "PublicDescription": "Message Held : Parallel Success : ad and bl messages were actually slotted into the same flit in paralle", + "PublicDescription": "Message Held : Parallel Success : ad and bl messages were actually slotted into the same flit in parallel", "UMask": "0x8", "Unit": "M3UPI" },
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 10/12 perf vendor events intel: Update skylake to 57
In reply to Ian RogersUpdates were released in:
https://github.com/intel/perfmon/commit/1c3042c13bbfea05abe1ebb6910ae58b2172e9ef
Adds the events IDQ.DSB_CYCLES_OK, IDQ.DSB_CYCLES_ANY, ICACHE_TAG.STALLS, DECODE.LCP, LSD.CYCLES_OK. Descriptions are also updated.
--- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- .../pmu-events/arch/x86/skylake/frontend.json | 43 ++++++++++++++++--- .../pmu-events/arch/x86/skylake/pipeline.json | 17 ++++++-- 3 files changed, 52 insertions(+), 10 deletions(-)
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -27,7 +27,7 @@ GenuineIntel-6-2A,v19,sandybridge,core GenuineIntel-6-(8F|CF),v1.14,sapphirerapids,core GenuineIntel-6-AF,v1.00,sierraforest,core GenuineIntel-6-(37|4A|4C|4D|5A),v15,silvermont,core -GenuineIntel-6-(4E|5E|8E|9E|A5|A6),v56,skylake,core +GenuineIntel-6-(4E|5E|8E|9E|A5|A6),v57,skylake,core GenuineIntel-6-55-[01234],v1.30,skylakex,core GenuineIntel-6-86,v1.21,snowridgex,core GenuineIntel-6-8[CD],v1.12,tigerlake,core
tools/perf/pmu-events/arch/x86/skylake/frontend.json +38 -5
77 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/skylake/pipeline.json +13 -4
@@ -352,10 +352,10 @@ "UMask": "0x1" }, { - "BriefDescription": "Stalls caused by changing prefix length of the instruction.", + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to DECODE.LCP]", "EventCode": "0x87", "EventName": "ILD_STALL.LCP", - "PublicDescription": "Counts 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 chunk.", + "PublicDescription": "Counts 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 chunk. [This event is alias to DECODE.LCP]", "SampleAfterValue": "2000003", "UMask": "0x1" },
@@ -479,11 +479,11 @@ "UMask": "0x1" }, { - "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder.", + "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder. [This event is alias to LSD.CYCLES_OK]", "CounterMask": "4", "EventCode": "0xA8", "EventName": "LSD.CYCLES_4_UOPS", - "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector).", + "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector). [This event is alias to LSD.CYCLES_OK]", "SampleAfterValue": "2000003", "UMask": "0x1" },
@@ -496,6 +496,15 @@ "SampleAfterValue": "2000003", "UMask": "0x1" }, + { + "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder. [This event is alias to LSD.CYCLES_4_UOPS]", + "CounterMask": "4", + "EventCode": "0xA8", + "EventName": "LSD.CYCLES_OK", + "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector). [This event is alias to LSD.CYCLES_4_UOPS]", + "SampleAfterValue": "2000003", + "UMask": "0x1" + }, { "BriefDescription": "Number of Uops delivered by the LSD.", "EventCode": "0xA8",
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 11/12 perf vendor events intel: Update skylakex to 1.31
In reply to Ian RogersUpdates were released in:
https://github.com/intel/perfmon/commit/cdaa69afe7a48a217b1d89320a27efc6e650cec3
Adds the events IDQ.DSB_CYCLES_OK, IDQ.DSB_CYCLES_ANY, ICACHE_TAG.STALLS, DECODE.LCP, LSD.CYCLES_OK. Descriptions are also updated.
--- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- .../arch/x86/skylakex/frontend.json | 43 ++++++++++++++++--- .../arch/x86/skylakex/pipeline.json | 17 ++++++-- .../x86/skylakex/uncore-interconnect.json | 2 +- .../arch/x86/skylakex/uncore-memory.json | 2 +- 5 files changed, 54 insertions(+), 12 deletions(-)
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -28,7 +28,7 @@ GenuineIntel-6-(8F|CF),v1.14,sapphirerapids,core GenuineIntel-6-AF,v1.00,sierraforest,core GenuineIntel-6-(37|4A|4C|4D|5A),v15,silvermont,core GenuineIntel-6-(4E|5E|8E|9E|A5|A6),v57,skylake,core -GenuineIntel-6-55-[01234],v1.30,skylakex,core +GenuineIntel-6-55-[01234],v1.31,skylakex,core GenuineIntel-6-86,v1.21,snowridgex,core GenuineIntel-6-8[CD],v1.12,tigerlake,core GenuineIntel-6-2C,v4,westmereep-dp,core
tools/perf/pmu-events/arch/x86/skylakex/frontend.json +38 -5
77 lines not on this page · raw message
tools/perf/pmu-events/arch/x86/skylakex/pipeline.json +13 -4
@@ -361,10 +361,10 @@ "UMask": "0x1" }, { - "BriefDescription": "Stalls caused by changing prefix length of the instruction.", + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to DECODE.LCP]", "EventCode": "0x87", "EventName": "ILD_STALL.LCP", - "PublicDescription": "Counts 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 chunk.", + "PublicDescription": "Counts 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 chunk. [This event is alias to DECODE.LCP]", "SampleAfterValue": "2000003", "UMask": "0x1" },
@@ -488,11 +488,11 @@ "UMask": "0x1" }, { - "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder.", + "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder. [This event is alias to LSD.CYCLES_OK]", "CounterMask": "4", "EventCode": "0xA8", "EventName": "LSD.CYCLES_4_UOPS", - "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector).", + "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector). [This event is alias to LSD.CYCLES_OK]", "SampleAfterValue": "2000003", "UMask": "0x1" },
@@ -505,6 +505,15 @@ "SampleAfterValue": "2000003", "UMask": "0x1" }, + { + "BriefDescription": "Cycles 4 Uops delivered by the LSD, but didn't come from the decoder. [This event is alias to LSD.CYCLES_4_UOPS]", + "CounterMask": "4", + "EventCode": "0xA8", + "EventName": "LSD.CYCLES_OK", + "PublicDescription": "Counts the cycles when 4 uops are delivered by the LSD (Loop-stream detector). [This event is alias to LSD.CYCLES_4_UOPS]", + "SampleAfterValue": "2000003", + "UMask": "0x1" + }, { "BriefDescription": "Number of Uops delivered by the LSD.", "EventCode": "0xA8",
tools/perf/pmu-events/arch/x86/skylakex/uncore-interconnect.json +1 -1
@@ -6504,7 +6504,7 @@ "EventCode": "0x52", "EventName": "UNC_M3UPI_RxC_HELD.PARALLEL_SUCCESS", "PerPkg": "1", - "PublicDescription": "ad and bl messages were actually slotted into the same flit in paralle", + "PublicDescription": "ad and bl messages were actually slotted into the same flit in parallel", "UMask": "0x8", "Unit": "M3UPI" },
tools/perf/pmu-events/arch/x86/skylakex/uncore-memory.json +1 -1
@@ -1952,7 +1952,7 @@ "EventCode": "0x81", "EventName": "UNC_M_WPQ_OCCUPANCY", "PerPkg": "1", - "PublicDescription": "Counts 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 happening 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?", + "PublicDescription": "Counts 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 happening in the iMC for writes before they are actually issued to memory. High average occupancies will generally coincide with high write major mode counts.", "Unit": "iMC" }, {
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Ian Rogers
reviewerPATCH v2 12/12 perf vendor events intel: Update tigerlake to 1.13
In reply to Ian RogersUpdates were released in:
https://github.com/intel/perfmon/commit/9a3cd5ad68aee46078c663fe0cd9484e3956fd88
Adds the events ICACHE_DATA.STALLS, ICACHE_TAG.STALLS and DECODE.LCP. Descriptions are also updated.
--- tools/perf/pmu-events/arch/x86/mapfile.csv | 2 +- .../arch/x86/tigerlake/frontend.json | 32 ++++++++++++++++--- .../arch/x86/tigerlake/pipeline.json | 6 ++-- 3 files changed, 32 insertions(+), 8 deletions(-)
tools/perf/pmu-events/arch/x86/mapfile.csv +1 -1
@@ -30,7 +30,7 @@ GenuineIntel-6-(37|4A|4C|4D|5A),v15,silvermont,core GenuineIntel-6-(4E|5E|8E|9E|A5|A6),v57,skylake,core GenuineIntel-6-55-[01234],v1.31,skylakex,core GenuineIntel-6-86,v1.21,snowridgex,core -GenuineIntel-6-8[CD],v1.12,tigerlake,core +GenuineIntel-6-8[CD],v1.13,tigerlake,core GenuineIntel-6-2C,v4,westmereep-dp,core GenuineIntel-6-25,v3,westmereep-sp,core GenuineIntel-6-2F,v3,westmereex,core
tools/perf/pmu-events/arch/x86/tigerlake/frontend.json +28 -4
@@ -7,6 +7,14 @@ "SampleAfterValue": "100003", "UMask": "0x1" }, + { + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to ILD_STALL.LCP]", + "EventCode": "0x87", + "EventName": "DECODE.LCP", + "PublicDescription": "Counts 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 chunk. [This event is alias to ILD_STALL.LCP]", + "SampleAfterValue": "500009", + "UMask": "0x1" + }, { "BriefDescription": "Decode Stream Buffer (DSB)-to-MITE transitions count.", "CounterMask": "1",
@@ -213,10 +221,10 @@ "UMask": "0x1" }, { - "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss.", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss. [This event is alias to ICACHE_DATA.STALLS]", "EventCode": "0x80", "EventName": "ICACHE_16B.IFDATA_STALL", - "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity.", + "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity. [This event is alias to ICACHE_DATA.STALLS]", "SampleAfterValue": "500009", "UMask": "0x4" },
@@ -237,10 +245,26 @@ "UMask": "0x2" }, { - "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss.", + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_TAG.STALLS]", "EventCode": "0x83", "EventName": "ICACHE_64B.IFTAG_STALL", - "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss.", + "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_TAG.STALLS]", + "SampleAfterValue": "200003", + "UMask": "0x4" + }, + { + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache miss. [This event is alias to ICACHE_16B.IFDATA_STALL]", + "EventCode": "0x80", + "EventName": "ICACHE_DATA.STALLS", + "PublicDescription": "Counts cycles where a code line fetch is stalled due to an L1 instruction cache miss. The legacy decode pipeline works at a 16 Byte granularity. [This event is alias to ICACHE_16B.IFDATA_STALL]", + "SampleAfterValue": "500009", + "UMask": "0x4" + }, + { + "BriefDescription": "Cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_64B.IFTAG_STALL]", + "EventCode": "0x83", + "EventName": "ICACHE_TAG.STALLS", + "PublicDescription": "Counts cycles where a code fetch is stalled due to L1 instruction cache tag miss. [This event is alias to ICACHE_64B.IFTAG_STALL]", "SampleAfterValue": "200003", "UMask": "0x4" },
tools/perf/pmu-events/arch/x86/tigerlake/pipeline.json +3 -3
@@ -335,10 +335,10 @@ "UMask": "0x80" }, { - "BriefDescription": "Stalls caused by changing prefix length of the instruction.", + "BriefDescription": "Stalls caused by changing prefix length of the instruction. [This event is alias to DECODE.LCP]", "EventCode": "0x87", "EventName": "ILD_STALL.LCP", - "PublicDescription": "Counts 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 chunk.", + "PublicDescription": "Counts 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 chunk. [This event is alias to DECODE.LCP]", "SampleAfterValue": "500009", "UMask": "0x1" },
@@ -564,7 +564,7 @@ "BriefDescription": "TMA slots wasted due to incorrect speculation by branch mispredictions", "EventCode": "0xa4", "EventName": "TOPDOWN.BR_MISPREDICT_SLOTS", - "PublicDescription": "Number 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 misprediction.", + "PublicDescription": "Number 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 speculative path as well as the out-of-order engine recovery past a branch misprediction.", "SampleAfterValue": "10000003", "UMask": "0x8" },
Signature · 1 line
2.41.0.162.gfafddb0af9-goog
Hi Ian,
Quoted 15 lines · Ian Rogers
Add a new
has_eventfunction for metrics so that events that can be disabled by the kernel/firmware don't cause metrics to fail. Use this function for Intel transaction metrics fixing "perf all metrics test" on systems with TSX disabled. The update conversion script is posted in:
https://github.com/intel/perfmon/pull/90Re-generate Intel vendor events using:
https://github.com/intel/perfmon/blob/main/scripts/create_perf_json.py
Adding rocketlake support, uncore and many core events for meteorlake, and smaller updates for cascakelakex, icelake, icelakex, sapphirerapids, skylake, skylakex and tigerlake.v2. Handle failed memory allocated for evlist, John Garry.
Quoted 13 lines
Ian Rogers (12):
perf expr: Addhas_eventfunction
perf jevents: Support forhas_eventfunction
perf vendor metrics intel: Make transaction metrics conditional perf vendor events intel: Add rocketlake events/metrics perf vendor events intel: Update meteorlake to 1.03 perf vendor events intel: Update cascadelakex to 1.19 perf vendor events intel: Update icelake to 1.19
perf vendor events intel: Update icelakex to 1.21 perf vendor events intel: Update sapphirerapids to 1.14 perf vendor events intel: Update skylake to 57
perf vendor events intel: Update skylakex to 1.31 perf vendor events intel: Update tigerlake to 1.13
My tigerlake laptop now passes the all metrics test with this. It used to fail like below:
event syntax error:
'{cpu/cycles-t,metric-id=cpu!3cycles!1t!3/,cpu/tx-start,m..'
\___ unknown term 'cycles-t' for pmu 'cpu'
Thanks,
Namhyung
Quoted 33 lines · Namhyung Kim
Hi Ian,
Quoted 15 lines · Ian Rogers
Add a new
has_eventfunction for metrics so that events that can be disabled by the kernel/firmware don't cause metrics to fail. Use this function for Intel transaction metrics fixing "perf all metrics test" on systems with TSX disabled. The update conversion script is posted in:
https://github.com/intel/perfmon/pull/90Re-generate Intel vendor events using:
https://github.com/intel/perfmon/blob/main/scripts/create_perf_json.py
Adding rocketlake support, uncore and many core events for meteorlake, and smaller updates for cascakelakex, icelake, icelakex, sapphirerapids, skylake, skylakex and tigerlake.v2. Handle failed memory allocated for evlist, John Garry.
Quoted 13 lines
Ian Rogers (12):
perf expr: Addhas_eventfunction
perf jevents: Support forhas_eventfunction
perf vendor metrics intel: Make transaction metrics conditional perf vendor events intel: Add rocketlake events/metrics perf vendor events intel: Update meteorlake to 1.03 perf vendor events intel: Update cascadelakex to 1.19 perf vendor events intel: Update icelake to 1.19
perf vendor events intel: Update icelakex to 1.21 perf vendor events intel: Update sapphirerapids to 1.14 perf vendor events intel: Update skylake to 57
perf vendor events intel: Update skylakex to 1.31 perf vendor events intel: Update tigerlake to 1.13
Quoted 8 lines
My tigerlake laptop now passes the all metrics test with this. It used to fail like below:
event syntax error:
'{cpu/cycles-t,metric-id=cpu!3cycles!1t!3/,cpu/tx-start,m..'
\___ unknown term 'cycles-t' for pmu 'cpu'Tested-by Namhyung Kim <[email protected]>
Applied to perf-tools-next, thanks!