perf stat: Fix duplicate output with --for-each-cgroup
Namhyung Kim
maintainerPATCH v3 1/4 perf stat: Fix duplicate output with --for-each-cgroup
Currently it produces following output with duplicate events when
--for-each-cgroup option is used. It seems perf stat adds them when
it handles default events but didn't copy some fields in evsel__clone().
$ sudo perf stat -a --for-each-cgroup / true
Performance counter stats for 'system wide':
8,440,165 duration_time /
8,439,895 duration_time /
8,440,015 duration_time /
8,440,024 duration_time /
8,440,075 duration_time /
8,440,095 duration_time /
330 context-switches / # 679.4 cs/sec cs_per_second
485.69 msec cpu-clock / # 57.5 CPUs CPUs_utilized
70 cpu-migrations / # 144.1 migrations/sec migrations_per_second
71 page-faults / # 146.2 faults/sec page_faults_per_second
12,183,711 branch-misses / # 10.9 % branch_miss_rate (5.15%)
111,981,297 branches / (5.15%)
95,844,809 branches / # 197.3 M/sec branch_frequency (35.49%)
65,611,429 cpu-cycles / # 0.1 GHz cycles_frequency (98.32%)
24,170,987 cpu-cycles / (95.12%)
18,552,509 instructions / # 0.8 instructions insn_per_cycle (95.12%)
22,405,293 cpu-cycles / (64.78%)
6,840,383 stalled-cycles-frontend / # 0.31 frontend_cycles_idle (64.78%)
<not counted> cpu-cycles /
<not supported> stalled-cycles-backend / # nan backend_cycles_idle
<not supported> stalled-cycles-backend / # nan stalled_cycles_per_instruction
<not supported> instructions /
<not supported> stalled-cycles-frontend /0.006546057 seconds time elapsed
Some events weren't counted. Try disabling the NMI watchdog:
echo 0 > /proc/sys/kernel/nmi_watchdog
perf stat ...
echo 1 > /proc/sys/kernel/nmi_watchdog
But I'm worrying about opening same events multiple times. Probably due to grouping, but I'm not sure if it's beneficial in the end. Without duplication, it seems it won't cause multiplexing (assuming no other users at the same time).
---
v2 changes)
* add missing bpf_counter and retire_lat fields (Sashiko)
tools/perf/util/evsel.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
tools/perf/util/evsel.c +10 -0
@@ -617,7 +617,13 @@ struct evsel *evsel__clone(struct evsel *orig) evsel->sample_read = orig->sample_read; evsel->collect_stat = orig->collect_stat; evsel->weak_group = orig->weak_group; + evsel->bpf_counter = orig->bpf_counter; evsel->use_config_name = orig->use_config_name; + evsel->skippable = orig->skippable; + evsel->dont_regroup = orig->dont_regroup; + evsel->default_metricgroup = orig->default_metricgroup; + evsel->default_show_events = orig->default_show_events; + evsel->pmu = orig->pmu; evsel->first_wildcard_match = orig->first_wildcard_match;
@@ -626,6 +632,10 @@ struct evsel *evsel__clone(struct evsel *orig) evsel->alternate_hw_config = orig->alternate_hw_config; + evsel->retire_lat = orig->retire_lat; +if (evsel->retire_lat) + evsel->retirement_latency = orig->retirement_latency; +return evsel; out_err:
Signature · 1 line
2.55.0.795.g602f6c329a-goog
Namhyung Kim
maintainerPATCH v3 2/4 perf evsel: Remove unused BPF related fields
IIUC bpf_fd and bpf_obj fields are not used anymore. It seems like leftover from 3d6dfae889174340 ("perf parse-events: Remove BPF event support").
--- tools/perf/util/evsel.c | 21 --------------------- tools/perf/util/evsel.h | 2 -- 2 files changed, 23 deletions(-)
tools/perf/util/evsel.c +0 -21
@@ -455,8 +455,6 @@ static void evsel__init(struct evsel *evsel, evsel->scale =1.0 ; evsel->max_events =ULONG_MAX ; evsel->evlist =NULL ; - evsel->bpf_obj =NULL ; - evsel->bpf_fd =-1 ;INIT_LIST_HEAD (&evsel->config_terms);INIT_LIST_HEAD (&evsel->bpf_counter_list);INIT_LIST_HEAD (&evsel->bpf_filters);
@@ -540,10 +538,6 @@ struct evsel *evsel__clone(struct evsel *orig)BUG_ON (orig->priv);BUG_ON (orig->per_pkg_mask); -/* cannot handle BPF objects for now */ -if (orig->bpf_obj) -return NULL ; - evsel =evsel__new (&orig->core.attr);if (evsel ==NULL )return NULL ;
@@ -3079,21 +3073,6 @@ static int evsel__open_cpu(struct evsel *evsel, struct perf_cpu_map *cpus,/* Debug message used by test scripts */ pr_debug2_peo (" = %d\n" , fd); -if (evsel->bpf_fd >=0 ) { -int evt_fd = fd; -int bpf_fd = evsel->bpf_fd; - - err =ioctl (evt_fd, -PERF_EVENT_IOC_SET_BPF , - bpf_fd); -if (err && errno !=EEXIST ) { -pr_err ("failed to attach bpf fd %d: %m\n" , - bpf_fd); - err = -EINVAL ; - goto out_close; - } - } - set_rlimit =NO_CHANGE ;/*
tools/perf/util/evsel.h +0 -2
@@ -94,8 +94,6 @@ struct evsel {bool skippable;bool retire_lat;bool dont_regroup; -int bpf_fd; -struct bpf_object *bpf_obj;struct list_head config_terms;u64 alternate_hw_config; };
Signature · 1 line
2.55.0.795.g602f6c329a-goog
Namhyung Kim
maintainerPATCH v3 3/4 perf evsel: Arrange some fields that should be cloned
In reply to Namhyung KimIn the evsel, there's an internal struct to put fields need copy when the evsel is cloned. This is purely to make it easier track those fields even if it sometimes failed to do so. :)
--- tools/perf/util/evsel.h | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-)
tools/perf/util/evsel.h +6 -5
@@ -74,6 +74,10 @@ struct evsel {const char *unit;struct cgroup *cgrp;const char *metric_id; + +/* The PMU the event is from. Used for missing_features, PMU name, etc. */ +struct perf_pmu *pmu; +/* * This point to the first evsel with the same name, intended to store the * aggregated counts in aggregation mode.
@@ -94,6 +98,8 @@ struct evsel {bool skippable;bool retire_lat;bool dont_regroup; +bool default_metricgroup;/* A member of the Default metricgroup */ +bool default_show_events;/* If a default group member, show the event */ struct list_head config_terms;u64 alternate_hw_config; };
@@ -123,8 +129,6 @@ struct evsel {bool cmdline_group_boundary;bool reset_group;bool needs_auxtrace_mmap; -bool default_metricgroup;/* A member of the Default metricgroup */ -bool default_show_events;/* If a default group member, show the event */ bool needs_uniquify;bool fallenback_eacces;bool fallenback_eopnotsupp;
@@ -180,9 +184,6 @@ struct evsel {unsigned long open_flags;int precise_ip_original; -/* The PMU the event is from. Used for missing_features, PMU name, etc. */ -struct perf_pmu *pmu; -/* For tool events */ /* Beginning time subtracted when the counter is read. */ union {
Signature · 1 line
2.55.0.795.g602f6c329a-goog
Namhyung Kim
maintainerPATCH v3 4/4 perf test: Update test for --for-each-cgroup option
In reply to Namhyung KimTo simply check the number of output lines with and without the option.
Before this series, it failed like below:
$ perf test -v 125 125: perf stat --bpf-counters --for-each-cgroup test: ---- start ---- test child forked, pid 1941516 Normal output has 22 lines, but it now has 54 ---- end(-1) ---- 125: perf stat --bpf-counters --for-each-cgroup test : FAILED!
---
v3 changes)
* use default events and metrics (Sashiko)
* update the commit description
tools/perf/tests/shell/stat_bpf_counters_cgrp.sh | 15 +++++++++++++++ 1 file changed, 15 insertions(+)
tools/perf/tests/shell/stat_bpf_counters_cgrp.sh +15 -0
@@ -58,9 +58,24 @@ check_system_wide_counted() fi } +# Missing flags in evlist__clone() resulted in different output. +# Just check the number of output lines for simple verification. +check_evlist_expand() +{ + normal_output=$(perf stat -a -x, true 2>&1 | wc -l) + expand_output=$(perf stat -a -x, --bpf-counters --for-each-cgroup / true 2>&1 | wc -l) + if [ "${normal_output}" != "${expand_output}" ]; then + if [ "${verbose}" = "1" ]; then + echo "Normal output has ${normal_output} lines, but it now has ${expand_output}" + fi + exit 1 + fi +} + check_bpf_counter find_cgroups check_system_wide_counted +check_evlist_expand exit 0
Signature · 1 line
2.55.0.795.g602f6c329a-goog
Quoted 9 lines · Namhyung Kim
Currently it produces following output with duplicate events when
--for-each-cgroup option is used. It seems perf stat adds them when
it handles default events but didn't copy some fields inevsel__clone().$ sudo perf stat -a --for-each-cgroup / truePerformance counter stats for 'system wide':
[...]
Applied to perf-tools-next, thanks!
Best regards,
Namhyung