perf evlist: No need to setup affinities when enabling events for pid targets
When the target is a pid, not started by 'perf stat' we need to enable the events, and in that case there is no need to setup affinities as we use a dummy CPU map, with just one entry set to -1. So stop doing it to avoid this needless call to sched_getaffinity(): # strace -ke sched_getaffinity perf stat -e cycles -p 241957 sleep 1 <SNIP> sched_getaffinity(0, 512, [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31]) = 8 > /usr/lib64/libc-2.33.so(sched_getaffinity@@GLIBC_2.3.4+0x1a) [0xe6eea] > /var/home/acme/bin/perf(affinity__setup+0x6a) [0x5329ca] > /var/home/acme/bin/perf(__evlist__enable.constprop.0+0x23) [0x4b9693] > /var/home/acme/bin/perf(enable_counters+0x14d) [0x42de5d] > /var/home/acme/bin/perf(cmd_stat+0x2358) [0x4310c8] > /var/home/acme/bin/perf(run_builtin+0x6a) [0x4a2cfa] > /var/home/acme/bin/perf(main+0x612) [0x40f8c2] > /usr/lib64/libc-2.33.so(__libc_start_main+0xd4) [0x27b74] > /var/home/acme/bin/perf(_start+0x2d) [0x40fadd] <SNIP> Acked-by: Ian Rogers <irogers@google.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Andi Kleen <andi@firstfloor.org> Cc: Ian Rogers <irogers@google.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: https://lore.kernel.org/r/20220117160931.1191712-4-acme@kernel.org Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
This commit is contained in:
parent
49de179577
commit
f350ee9549
|
@ -487,12 +487,16 @@ static void __evlist__enable(struct evlist *evlist, char *evsel_name)
|
|||
{
|
||||
struct evsel *pos;
|
||||
struct evlist_cpu_iterator evlist_cpu_itr;
|
||||
struct affinity affinity;
|
||||
struct affinity saved_affinity, *affinity = NULL;
|
||||
|
||||
if (affinity__setup(&affinity) < 0)
|
||||
return;
|
||||
// See explanation in evlist__close()
|
||||
if (!cpu_map__is_dummy(evlist->core.cpus)) {
|
||||
if (affinity__setup(&saved_affinity) < 0)
|
||||
return;
|
||||
affinity = &saved_affinity;
|
||||
}
|
||||
|
||||
evlist__for_each_cpu(evlist_cpu_itr, evlist, &affinity) {
|
||||
evlist__for_each_cpu(evlist_cpu_itr, evlist, affinity) {
|
||||
pos = evlist_cpu_itr.evsel;
|
||||
if (evsel__strcmp(pos, evsel_name))
|
||||
continue;
|
||||
|
@ -500,7 +504,7 @@ static void __evlist__enable(struct evlist *evlist, char *evsel_name)
|
|||
continue;
|
||||
evsel__enable_cpu(pos, evlist_cpu_itr.cpu_map_idx);
|
||||
}
|
||||
affinity__cleanup(&affinity);
|
||||
affinity__cleanup(affinity);
|
||||
evlist__for_each_entry(evlist, pos) {
|
||||
if (evsel__strcmp(pos, evsel_name))
|
||||
continue;
|
||||
|
|
Loading…
Reference in New Issue