On Tue, Jan 07, 2025 at 12:34:25PM -0800, Ian Rogers wrote: > On Tue, Jan 7, 2025 at 11:37 AM Namhyung Kim <namhyung@xxxxxxxxxx> wrote: > > > > Hi Ian, > > > > On Tue, Jan 07, 2025 at 10:08:53AM -0800, Ian Rogers wrote: > > > Whilst for many tools it is an expected behavior that failure to open > > > a perf event is a failure, ARM decided to name PMU events the same as > > > legacy events and then failed to rename such events on a server uncore > > > SLC PMU. As perf's default behavior when no PMU is specified is to > > > open the event on all PMUs that advertise/"have" the event, this > > > yielded failures when trying to make the priority of legacy and > > > sysfs/json events uniform - something requested by RISC-V and ARM. A > > > legacy event user on ARM hardware may find their event opened on an > > > uncore PMU which for perf record will fail. Arnaldo suggested skipping > > > such events which this patch implements. Rather than have the skipping > > > conditional on running on ARM, the skipping is done on all > > > architectures as such a fundamental behavioral difference could lead > > > to problems with tools built/depending on perf. > > > > > > An example of perf record failing to open events on x86 is: > > > ``` > > > $ perf record -e data_read,cycles,LLC-prefetch-read -a sleep 0.1 > > > Error: > > > Failure to open event 'data_read' on PMU 'uncore_imc_free_running_0' which will be removed. > > > The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (data_read). > > > "dmesg | grep -i perf" may provide additional information. > > > > > > Error: > > > Failure to open event 'data_read' on PMU 'uncore_imc_free_running_1' which will be removed. > > > The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (data_read). > > > "dmesg | grep -i perf" may provide additional information. > > > > > > Error: > > > Failure to open event 'LLC-prefetch-read' on PMU 'cpu' which will be removed. > > > The LLC-prefetch-read event is not supported. > > > [ perf record: Woken up 1 times to write data ] > > > [ perf record: Captured and wrote 2.188 MB perf.data (87 samples) ] > > > > > > $ perf report --stats > > > Aggregated stats: > > > TOTAL events: 17255 > > > MMAP events: 284 ( 1.6%) > > > COMM events: 1961 (11.4%) > > > EXIT events: 1 ( 0.0%) > > > FORK events: 1960 (11.4%) > > > SAMPLE events: 87 ( 0.5%) > > > MMAP2 events: 12836 (74.4%) > > > KSYMBOL events: 83 ( 0.5%) > > > BPF_EVENT events: 36 ( 0.2%) > > > FINISHED_ROUND events: 2 ( 0.0%) > > > ID_INDEX events: 1 ( 0.0%) > > > THREAD_MAP events: 1 ( 0.0%) > > > CPU_MAP events: 1 ( 0.0%) > > > TIME_CONV events: 1 ( 0.0%) > > > FINISHED_INIT events: 1 ( 0.0%) > > > cycles stats: > > > SAMPLE events: 87 > > > ``` > > > > > > If all events fail to open then the perf record will fail: > > > ``` > > > $ perf record -e LLC-prefetch-read true > > > Error: > > > Failure to open event 'LLC-prefetch-read' on PMU 'cpu' which will be removed. > > > The LLC-prefetch-read event is not supported. > > > Error: > > > Failure to open any events for recording > > > ``` > > > > > > This is done by detecting if dummy events were implicitly added by > > > perf and seeing if the evlist is empty without them. This allows the > > > dummy event still to be recorded: > > > ``` > > > $ perf record -e dummy true > > > [ perf record: Woken up 1 times to write data ] > > > [ perf record: Captured and wrote 0.046 MB perf.data ] > > > ``` > > > but fail when inserted: > > > ``` > > > $ perf record -e LLC-prefetch-read -a true > > > Error: > > > Failure to open event 'LLC-prefetch-read' on PMU 'cpu' which will be removed. > > > The LLC-prefetch-read event is not supported. > > > Error: > > > Failure to open any events for recording > > > ``` > > > > > > The issue with legacy events is that on RISC-V they want the driver to > > > not have mappings from legacy to non-legacy config encodings for each > > > vendor/model due to size, complexity and difficulty to update. It was > > > reported that on ARM Apple-M? CPUs the legacy mapping in the driver > > > was broken and the sysfs/json events should always take precedent, > > > however, it isn't clear this is still the case. It is the case that > > > without working around this issue a legacy event like cycles without a > > > PMU can encode differently than when specified with a PMU - the > > > non-PMU version favoring legacy encodings, the PMU one avoiding legacy > > > encodings. > > > > > > The patch removes events and then adjusts the idx value for each > > > evsel. This is done so that the dense xyarrays used for file > > > descriptors, etc. don't contain broken entries. As event opening > > > happens relatively late in the record process, use of the idx value > > > before the open will have become corrupted, so it is expected there > > > are latent bugs hidden behind this change - the change is best > > > effort. As the only vendor that has broken event names is ARM, this > > > will principally effect ARM users. They will also experience warning > > > messages like those above because of the uncore PMU advertising legacy > > > event names. > > > > > > Suggested-by: Arnaldo Carvalho de Melo <acme@xxxxxxxxxx> > > > Signed-off-by: Ian Rogers <irogers@xxxxxxxxxx> > > > Tested-by: James Clark <james.clark@xxxxxxxxxx> > > > Tested-by: Leo Yan <leo.yan@xxxxxxx> > > > Tested-by: Atish Patra <atishp@xxxxxxxxxxxx> > > > --- > > > tools/perf/builtin-record.c | 54 ++++++++++++++++++++++++++++++++----- > > > 1 file changed, 48 insertions(+), 6 deletions(-) > > > > > > diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c > > > index 5db1aedf48df..b3f06638f3c6 100644 > > > --- a/tools/perf/builtin-record.c > > > +++ b/tools/perf/builtin-record.c > > > @@ -161,6 +161,7 @@ struct record { > > > struct evlist *sb_evlist; > > > pthread_t thread_id; > > > int realtime_prio; > > > + int num_parsed_dummy_events; > > > bool switch_output_event_set; > > > bool no_buildid; > > > bool no_buildid_set; > > > @@ -961,7 +962,6 @@ static int record__config_tracking_events(struct record *rec) > > > */ > > > if (opts->target.initial_delay || target__has_cpu(&opts->target) || > > > perf_pmus__num_core_pmus() > 1) { > > > - > > > /* > > > * User space tasks can migrate between CPUs, so when tracing > > > * selected CPUs, sideband for all CPUs is still needed. > > > @@ -1366,6 +1366,7 @@ static int record__open(struct record *rec) > > > struct perf_session *session = rec->session; > > > struct record_opts *opts = &rec->opts; > > > int rc = 0; > > > + bool skipped = false; > > > > > > evlist__for_each_entry(evlist, pos) { > > > try_again: > > > @@ -1381,15 +1382,50 @@ static int record__open(struct record *rec) > > > pos = evlist__reset_weak_group(evlist, pos, true); > > > goto try_again; > > > } > > > - rc = -errno; > > > evsel__open_strerror(pos, &opts->target, errno, msg, sizeof(msg)); > > > - ui__error("%s\n", msg); > > > - goto out; > > > + ui__error("Failure to open event '%s' on PMU '%s' which will be removed.\n%s\n", > > > + evsel__name(pos), evsel__pmu_name(pos), msg); > > > + pos->skippable = true; > > > + skipped = true; > > > + } else { > > > + pos->supported = true; > > > } > > > - > > > - pos->supported = true; > > > } > > > > > > + if (skipped) { > > > + struct evsel *tmp; > > > + int idx = 0, num_dummy = 0, num_non_dummy = 0, > > > + removed_dummy = 0, removed_non_dummy = 0; > > > + > > > + /* Remove evsels that failed to open and update indices. */ > > > + evlist__for_each_entry_safe(evlist, tmp, pos) { > > > + if (evsel__is_dummy_event(pos)) > > > + num_dummy++; > > > + else > > > + num_non_dummy++; > > > + > > > + if (!pos->skippable) > > > + continue; > > > + > > > + if (evsel__is_dummy_event(pos)) > > > + removed_dummy++; > > > + else > > > + removed_non_dummy++; > > > + > > > + evlist__remove(evlist, pos); > > > + } > > > + evlist__for_each_entry(evlist, pos) { > > > + pos->core.idx = idx++; > > > + } > > > + /* If list is empty except implicitly added dummy events then fail. */ > > > + if ((num_non_dummy == removed_non_dummy) && > > > + ((rec->num_parsed_dummy_events == 0) || > > > + (removed_dummy >= (num_dummy - rec->num_parsed_dummy_events)))) { > > > + ui__error("Failure to open any events for recording.\n"); > > > + rc = -1; > > > + goto out; > > > + } > > > + } > > > > Instead of couting dummy events, I wonder if it could check any > > supported non-dummy events in the evlist. > > > > if (skipped) { > > bool found = false; > > > > evlist__for_each_entry_safe(evlist, tmp, pos) { > > if (pos->skippable) { > > evlist__remove(evlist, pos); > > continue; > > } > > if (evsel__is_dummy_event(pos)) > > continue; > > found = true; > > } > > if (!found) { > > ui__error("..."); > > rc = -1; > > goto out; > > } > > /* recalculate the index */ > > } > > > > Then it should do the same, no? The corner case would be when users > > specify dummy events in the command line (maybe to check permissions > > by the exit code). > > > > $ perf record -a -e dummy true > > > > If it fails to open, then 'skipped' set and 'found' not set so the > > command will fail. It it succeeds, then it doesn't set 'skipped' > > and the command will exit with 0. > > > > Do I miss something? > > Yep, but it depends on what we want the behavior to be. Consider an > event that doesn't support record and the dummy event which for the > sake of argument will open here: > > $ perf record -a -e LLC-prefetch-read,dummy true > > In this case initially found is false, we then process > LLC-prefetch-read which didn't open and remove it from the evlist. > Next we process the dummy event that did open but because it was a > dummy event found won't be set to true. The code will then proceed to > fail (found == false) even though the dummy event did open. Were this > cycles and not dummy: > > $ perf record -a -e LLC-prefetch-read,cycles true > > Then the expectation would be for this to proceed with a warning on > LLC-prefetch-read and to record cycles. With your code dummy will > behave differently to cycles as it will terminate perf record early. > Does this matter? I'm not sure, but I was trying to make the events > (dummy vs non-dummy) consistent. I'm not sure too. Anyway it won't get any samples if all other non-dummy events failed. Then there are not much points to run the perf record IMHO. I think dummy event is special as you counted it specifically so we don't need to worry about the special case too much. Thanks, Namhyung