mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Ian Rogers <irogers@google.com>
To: irogers@google.com, acme@kernel.org, namhyung@kernel.org
Cc: adrian.hunter@intel.com, james.clark@linaro.org,
	jolsa@kernel.org,  linux-kernel@vger.kernel.org,
	linux-perf-users@vger.kernel.org,  mingo@redhat.com,
	nigro.fra@gmail.com, peterz@infradead.org,
	 tmricht@linux.ibm.com
Subject: [PATCH v3 0/2] perf tool_pmu: Support enable/disable for tool PMU events
Date: Mon, 18 May 2026 15:37:31 -0700	[thread overview]
Message-ID: <20260518223733.3034897-1-irogers@google.com> (raw)
In-Reply-To: <20260518201419.2943132-1-irogers@google.com>

A regression in perf stat was reported where tool PMU events (like
duration_time used in CPUs_utilized metric) incorrectly included the
delay period when using the delay option (-D).

This series fixes the regression by making tool PMU events
(duration_time, user_time, system_time) behave more like regular
counters by implementing proper enable and disable support. They now
correctly accumulate values only when enabled.

The first patch implements the core enable/disable support for tool PMU
events, and the second patch adds a shell test to verify that
duration_time correctly excludes the delay period.

Changes in v3:
- Refine group handling: only manually enable/disable group members when
  the leader or member is a non-perf-event open PMU, as the kernel allows
  grouping of software and hardware PMUs.
- Fix a file descriptor leak in evsel__tool_pmu_open() on error paths by
  explicitly closing the successfully opened fd before exiting.
- Synchronize the 'disabled' flag for all group members in enable/disable
  paths (both per-CPU and batch loops) to prevent stale disabled flags.
- Add explicit early exits to evsel__tool_pmu_enable() and disable() based
  on evsel->disabled to protect internal metric state.
- Add an upper bound check in test_stat_delay to verify that the delay was
  actually excluded.

Changes in v2:
- Implement evsel__tool_pmu_enable() and evsel__tool_pmu_disable() to
  avoid ioctl failures in batch evsel__enable() and evsel__disable()
  functions.
- Correctly iterate and enable/disable tool PMU events configured as
  non-leader members of event groups.
- Correct the lseek() arguments order in the read_stat helper:
  lseek(fd, 0, SEEK_SET) instead of lseek(fd, SEEK_SET, 0).
- Introduce INVALID_START_TIME (~0ULL) to prevent erroneous large delta
  accumulation in evsel__tool_pmu_read() if /proc/<pid>/stat fails to
  read in enable_cpu (e.g., process exited).
- Improve test parsing to use LC_ALL=C and cut to be robust against
  different locales, and use awk to dynamically compare duration_time to
  time elapsed with a 200ms tolerance (avoiding loaded CI false failures).
  Also added a lower-bound check.
- Fix style warnings from checkpatch.pl (line lengths, braces, and blank
  lines).

Ian Rogers (2):
  perf tool_pmu: Make tool PMU events respect enable/disable
  perf tests: Add test for stat delay option with duration_time

 tools/perf/tests/shell/stat.sh |  53 +++++++
 tools/perf/util/evlist.c       |  10 +-
 tools/perf/util/evsel.c        | 196 +++++++++++++++++++------
 tools/perf/util/evsel.h        |  15 +-
 tools/perf/util/tool_pmu.c     | 253 ++++++++++++++++++++++++++-------
 tools/perf/util/tool_pmu.h     |   4 +
 6 files changed, 432 insertions(+), 99 deletions(-)

-- 
2.54.0.631.ge1b05301d1-goog


  parent reply	other threads:[~2026-05-18 22:37 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-18 18:39 [PATCH v1 " Ian Rogers
2026-05-18 18:39 ` [PATCH v1 1/2] perf tool_pmu: Make tool PMU events respect enable/disable Ian Rogers
2026-05-18 18:39 ` [PATCH v1 2/2] perf tests: Add test for stat delay option with duration_time Ian Rogers
2026-05-18 20:14 ` [PATCH v2 0/2] perf tool_pmu: Support enable/disable for tool PMU events Ian Rogers
2026-05-18 20:14   ` [PATCH v2 1/2] perf tool_pmu: Make tool PMU events respect enable/disable Ian Rogers
2026-05-18 20:14   ` [PATCH v2 2/2] perf tests: Add test for stat delay option with duration_time Ian Rogers
2026-05-18 22:37   ` Ian Rogers [this message]
2026-05-18 22:37     ` [PATCH v3 1/2] perf tool_pmu: Make tool PMU events respect enable/disable Ian Rogers
2026-05-18 22:37     ` [PATCH v3 2/2] perf tests: Add test for stat delay option with duration_time Ian Rogers
2026-05-19  1:41     ` [PATCH v4 0/2] perf tool_pmu: Support enable/disable for tool PMU events Ian Rogers
2026-05-19  1:41       ` [PATCH v4 1/2] perf tool_pmu: Make tool PMU events respect enable/disable Ian Rogers
2026-05-19  6:43         ` Namhyung Kim
2026-05-19  8:13           ` Ian Rogers
2026-05-20 23:06             ` Namhyung Kim
2026-05-23  0:40         ` Arnaldo Carvalho de Melo
2026-05-19  1:41       ` [PATCH v4 2/2] perf tests: Add test for stat delay option with duration_time Ian Rogers
2026-05-22 22:10       ` [PATCH v4 0/2] perf tool_pmu: Support enable/disable for tool PMU events Ian Rogers

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260518223733.3034897-1-irogers@google.com \
    --to=irogers@google.com \
    --cc=acme@kernel.org \
    --cc=adrian.hunter@intel.com \
    --cc=james.clark@linaro.org \
    --cc=jolsa@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-perf-users@vger.kernel.org \
    --cc=mingo@redhat.com \
    --cc=namhyung@kernel.org \
    --cc=nigro.fra@gmail.com \
    --cc=peterz@infradead.org \
    --cc=tmricht@linux.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®