diff options
author | Ian Rogers <irogers@google.com> | 2024-09-26 15:48:34 +0100 |
---|---|---|
committer | Namhyung Kim <namhyung@kernel.org> | 2024-09-26 13:26:11 -0700 |
commit | d38461e977baf93f9e01967caa9571ccd1694def (patch) | |
tree | 70f9c3d3b87dd0b26b1a4a245c8b27467b460fd3 /tools/perf/builtin-stat.c | |
parent | 057f8bfc6f7070577523d1e3081081bbf4229c1c (diff) |
perf stat: Remove evlist__add_default_attrs use strings
add_default_atttributes would add evsels by having pre-created
perf_event_attr, however, this needed fixing for hybrid as the
extended PMU type was necessary for each core PMU. The logic for this
was in an arch specific x86 function and wasn't present for ARM,
meaning that default events weren't being opened on all PMUs on
ARM. Change the creation of the default events to use parse_events and
strings as that will open the events on all PMUs.
Rather than try to detect events on PMUs before parsing, parse the
event but skip its output in stat-display.
The previous order of hardware events was: cycles,
stalled-cycles-frontend, stalled-cycles-backend, instructions. As
instructions is a more fundamental concept the order is changed to:
instructions, cycles, stalled-cycles-frontend, stalled-cycles-backend.
Closes: https://lore.kernel.org/lkml/CAP-5=fVABSBZnsmtRn1uF-k-G1GWM-L5SgiinhPTfHbQsKXb_g@mail.gmail.com/
Acked-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Ian Rogers <irogers@google.com>
[Don't display unsupported default events except 'cycles']
Acked-by: Kan Liang <kan.liang@linux.intel.com>
Signed-off-by: James Clark <james.clark@linaro.org>
Cc: Yang Jihong <yangjihong@bytedance.com>
Cc: Dominique Martinet <asmadeus@codewreck.org>
Cc: Colin Ian King <colin.i.king@gmail.com>
Cc: Howard Chu <howardchu95@gmail.com>
Cc: Ze Gao <zegao2021@gmail.com>
Cc: Yicong Yang <yangyicong@hisilicon.com>
Cc: Weilin Wang <weilin.wang@intel.com>
Cc: Will Deacon <will@kernel.org>
Cc: Mike Leach <mike.leach@linaro.org>
Cc: Jing Zhang <renyu.zj@linux.alibaba.com>
Cc: Yang Li <yang.lee@linux.alibaba.com>
Cc: Leo Yan <leo.yan@linux.dev>
Cc: ak@linux.intel.com
Cc: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Cc: linux-arm-kernel@lists.infradead.org
Cc: Sun Haiyong <sunhaiyong@loongson.cn>
Cc: John Garry <john.g.garry@oracle.com>
Link: https://lore.kernel.org/r/20240926144851.245903-4-james.clark@linaro.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Diffstat (limited to 'tools/perf/builtin-stat.c')
-rw-r--r-- | tools/perf/builtin-stat.c | 297 |
1 files changed, 118 insertions, 179 deletions
diff --git a/tools/perf/builtin-stat.c b/tools/perf/builtin-stat.c index 6cad3435b198..1521b6df2606 100644 --- a/tools/perf/builtin-stat.c +++ b/tools/perf/builtin-stat.c @@ -1842,130 +1842,25 @@ static int perf_stat_init_aggr_mode_file(struct perf_stat *st) } /* - * Add default attributes, if there were no attributes specified or + * Add default events, if there were no attributes specified or * if -d/--detailed, -d -d or -d -d -d is used: */ -static int add_default_attributes(void) +static int add_default_events(void) { - struct perf_event_attr default_attrs0[] = { - - { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_TASK_CLOCK }, - { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CONTEXT_SWITCHES }, - { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_CPU_MIGRATIONS }, - { .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS }, - - { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES }, -}; - struct perf_event_attr frontend_attrs[] = { - { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_STALLED_CYCLES_FRONTEND }, -}; - struct perf_event_attr backend_attrs[] = { - { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_STALLED_CYCLES_BACKEND }, -}; - struct perf_event_attr default_attrs1[] = { - { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS }, - { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS }, - { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_MISSES }, - -}; - -/* - * Detailed stats (-d), covering the L1 and last level data caches: - */ - struct perf_event_attr detailed_attrs[] = { - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_L1D << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_L1D << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_MISS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_LL << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_LL << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_MISS << 16) }, -}; - -/* - * Very detailed stats (-d -d), covering the instruction cache and the TLB caches: - */ - struct perf_event_attr very_detailed_attrs[] = { - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_L1I << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_L1I << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_MISS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_DTLB << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_DTLB << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_MISS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_ITLB << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_ITLB << 0 | - (PERF_COUNT_HW_CACHE_OP_READ << 8) | - (PERF_COUNT_HW_CACHE_RESULT_MISS << 16) }, - -}; + const char *pmu = parse_events_option_args.pmu_filter ?: "all"; + struct parse_events_error err; + struct evlist *evlist = evlist__new(); + struct evsel *evsel; + int ret = 0; -/* - * Very, very detailed stats (-d -d -d), adding prefetch events: - */ - struct perf_event_attr very_very_detailed_attrs[] = { - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_L1D << 0 | - (PERF_COUNT_HW_CACHE_OP_PREFETCH << 8) | - (PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) }, - - { .type = PERF_TYPE_HW_CACHE, - .config = - PERF_COUNT_HW_CACHE_L1D << 0 | - (PERF_COUNT_HW_CACHE_OP_PREFETCH << 8) | - (PERF_COUNT_HW_CACHE_RESULT_MISS << 16) }, -}; + if (!evlist) + return -ENOMEM; - struct perf_event_attr default_null_attrs[] = {}; - const char *pmu = parse_events_option_args.pmu_filter ?: "all"; + parse_events_error__init(&err); /* Set attrs if no event is selected and !null_run: */ if (stat_config.null_run) - return 0; + goto out; if (transaction_run) { /* Handle -T as -M transaction. Once platform specific metrics @@ -1975,9 +1870,10 @@ static int add_default_attributes(void) */ if (!metricgroup__has_metric(pmu, "transaction")) { pr_err("Missing transaction metrics\n"); - return -1; + ret = -1; + goto out; } - return metricgroup__parse_groups(evsel_list, pmu, "transaction", + ret = metricgroup__parse_groups(evlist, pmu, "transaction", stat_config.metric_no_group, stat_config.metric_no_merge, stat_config.metric_no_threshold, @@ -1985,6 +1881,7 @@ static int add_default_attributes(void) stat_config.system_wide, stat_config.hardware_aware_grouping, &stat_config.metric_events); + goto out; } if (smi_cost) { @@ -1992,26 +1889,29 @@ static int add_default_attributes(void) if (sysfs__read_int(FREEZE_ON_SMI_PATH, &smi) < 0) { pr_err("freeze_on_smi is not supported.\n"); - return -1; + ret = -1; + goto out; } if (!smi) { if (sysfs__write_int(FREEZE_ON_SMI_PATH, 1) < 0) { - fprintf(stderr, "Failed to set freeze_on_smi.\n"); - return -1; + pr_err("Failed to set freeze_on_smi.\n"); + ret = -1; + goto out; } smi_reset = true; } if (!metricgroup__has_metric(pmu, "smi")) { pr_err("Missing smi metrics\n"); - return -1; + ret = -1; + goto out; } if (!force_metric_only) stat_config.metric_only = true; - return metricgroup__parse_groups(evsel_list, pmu, "smi", + ret = metricgroup__parse_groups(evlist, pmu, "smi", stat_config.metric_no_group, stat_config.metric_no_merge, stat_config.metric_no_threshold, @@ -2019,6 +1919,7 @@ static int add_default_attributes(void) stat_config.system_wide, stat_config.hardware_aware_grouping, &stat_config.metric_events); + goto out; } if (topdown_run) { @@ -2031,21 +1932,23 @@ static int add_default_attributes(void) if (!max_level) { pr_err("Topdown requested but the topdown metric groups aren't present.\n" "(See perf list the metric groups have names like TopdownL1)\n"); - return -1; + ret = -1; + goto out; } if (stat_config.topdown_level > max_level) { pr_err("Invalid top-down metrics level. The max level is %u.\n", max_level); - return -1; - } else if (!stat_config.topdown_level) + ret = -1; + goto out; + } else if (!stat_config.topdown_level) { stat_config.topdown_level = 1; - + } if (!stat_config.interval && !stat_config.metric_only) { fprintf(stat_config.output, "Topdown accuracy may decrease when measuring long periods.\n" "Please print the result regularly, e.g. -I1000\n"); } str[8] = stat_config.topdown_level + '0'; - if (metricgroup__parse_groups(evsel_list, + if (metricgroup__parse_groups(evlist, pmu, str, /*metric_no_group=*/false, /*metric_no_merge=*/false, @@ -2053,41 +1956,49 @@ static int add_default_attributes(void) stat_config.user_requested_cpu_list, stat_config.system_wide, stat_config.hardware_aware_grouping, - &stat_config.metric_events) < 0) - return -1; + &stat_config.metric_events) < 0) { + ret = -1; + goto out; + } } if (!stat_config.topdown_level) stat_config.topdown_level = 1; - if (!evsel_list->core.nr_entries) { + if (!evlist->core.nr_entries && !evsel_list->core.nr_entries) { /* No events so add defaults. */ if (target__has_cpu(&target)) - default_attrs0[0].config = PERF_COUNT_SW_CPU_CLOCK; + ret = parse_events(evlist, "cpu-clock", &err); + else + ret = parse_events(evlist, "task-clock", &err); + if (ret) + goto out; + + ret = parse_events(evlist, + "context-switches," + "cpu-migrations," + "page-faults," + "instructions," + "cycles," + "stalled-cycles-frontend," + "stalled-cycles-backend," + "branches," + "branch-misses", + &err); + if (ret) + goto out; - if (evlist__add_default_attrs(evsel_list, default_attrs0) < 0) - return -1; - if (perf_pmus__have_event("cpu", "stalled-cycles-frontend")) { - if (evlist__add_default_attrs(evsel_list, frontend_attrs) < 0) - return -1; - } - if (perf_pmus__have_event("cpu", "stalled-cycles-backend")) { - if (evlist__add_default_attrs(evsel_list, backend_attrs) < 0) - return -1; - } - if (evlist__add_default_attrs(evsel_list, default_attrs1) < 0) - return -1; /* * Add TopdownL1 metrics if they exist. To minimize * multiplexing, don't request threshold computation. */ if (metricgroup__has_metric(pmu, "Default")) { struct evlist *metric_evlist = evlist__new(); - struct evsel *metric_evsel; - - if (!metric_evlist) - return -1; + if (!metric_evlist) { + ret = -ENOMEM; + goto out; + } if (metricgroup__parse_groups(metric_evlist, pmu, "Default", /*metric_no_group=*/false, /*metric_no_merge=*/false, @@ -2095,43 +2006,71 @@ static int add_default_attributes(void) stat_config.user_requested_cpu_list, stat_config.system_wide, stat_config.hardware_aware_grouping, - &stat_config.metric_events) < 0) - return -1; - - evlist__for_each_entry(metric_evlist, metric_evsel) { - metric_evsel->skippable = true; - metric_evsel->default_metricgroup = true; + &stat_config.metric_events) < 0) { + ret = -1; + goto out; } - evlist__splice_list_tail(evsel_list, &metric_evlist->core.entries); + + evlist__for_each_entry(metric_evlist, evsel) + evsel->default_metricgroup = true; + + evlist__splice_list_tail(evlist, &metric_evlist->core.entries); evlist__delete(metric_evlist); } - - /* Platform specific attrs */ - if (evlist__add_default_attrs(evsel_list, default_null_attrs) < 0) - return -1; } /* Detailed events get appended to the event list: */ - if (detailed_run < 1) - return 0; - - /* Append detailed run extra attributes: */ - if (evlist__add_default_attrs(evsel_list, detailed_attrs) < 0) - return -1; - - if (detailed_run < 2) - return 0; - - /* Append very detailed run extra attributes: */ - if (evlist__add_default_attrs(evsel_list, very_detailed_attrs) < 0) - return -1; - - if (detailed_run < 3) - return 0; - - /* Append very, very detailed run extra attributes: */ - return evlist__add_default_attrs(evsel_list, very_very_detailed_attrs); + if (!ret && detailed_run >= 1) { + /* + * Detailed stats (-d), covering the L1 and last level data + * caches: + */ + ret = parse_events(evlist, + "L1-dcache-loads," + "L1-dcache-load-misses," + "LLC-loads," + "LLC-load-misses", + &err); + } + if (!ret && detailed_run >= 2) { + /* + * Very detailed stats (-d -d), covering the instruction cache + * and the TLB caches: + */ + ret = parse_events(evlist, + "L1-icache-loads," + "L1-icache-load-misses," + "dTLB-loads," + "dTLB-load-misses," + "iTLB-loads," + "iTLB-load-misses", + &err); + } + if (!ret && detailed_run >= 3) { + /* + * Very, very detailed stats (-d -d -d), adding prefetch events: + */ + ret = parse_events(evlist, + "L1-dcache-prefetches," + "L1-dcache-prefetch-misses", + &err); + } +out: + if (!ret) { + evlist__for_each_entry(evlist, evsel) { + /* + * Make at least one event non-skippable so fatal errors are visible. + * 'cycles' always used to be default and non-skippable, so use that. + */ + if (strcmp("cycles", evsel__name(evsel))) + evsel->skippable = true; + } + } + parse_events_error__exit(&err); + evlist__splice_list_tail(evsel_list, &evlist->core.entries); + evlist__delete(evlist); + return ret; } static const char * const stat_record_usage[] = { @@ -2788,7 +2727,7 @@ int cmd_stat(int argc, const char **argv) } } - if (add_default_attributes()) + if (add_default_events()) goto out; if (stat_config.cgroup_list) { |