Skip to contentSkip to conversation

perf tools fixes for v5.9

GIT PULL Merged Arnaldo Carvalho de Melo 3 participants 4 messages Sep 1, 2020 – Sep 2, 2020

lore
Share & tools
lore
#0 lore raw

Hi Linus,

Please consider pulling,

Best regards,

- Arnaldo

The following changes since commit da2968ff879b9e74688cdc658f646971991d2c56:

Merge tag 'pci-v5.9-fixes-1' of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci (2020-08-20 14:17:03 -0700)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git tags/perf-tools-fixes-for-v5.9-2020-09-01

for you to fetch changes up to 977f739b7126bf98b5202e243f60cbc0a1ec2c3b:

perf report: Disable ordered_events for raw dump (2020-09-01 12:20:25 -0300)

----------------------------------------------------------------
perf tools fixes for v5.9:

- Fix infinite loop in the TUI for grouped events in 'perf top/record', for instance when using "perf top -e '{cycles,instructions,cache-misses}'.

- Fix segfault by skipping side-band event setup if HAVE_LIBBPF_SUPPORT is not set.

- Fix synthesized branch stacks generated from CoreSight ETM trace and Intel PT hardware traces.

- Fix error when synthesizing events from ARM SPE hardware trace.

- The SNOOPX and REMOTE offsets in the data_src bitmask in perf records were were both 37, SNOOPX is 38, fix it.

- Fix use of CPU list with summary option in 'perf sched timehist'.

- Avoid an uninitialized read when using fake PMUs.

- Set perf_event_attr.exclude_guest=1 for user-space counting.

- Don't order events when doing a 'perf report -D' raw dump of perf.data records.

- Set NULL sentinel in pmu_events table in "Parse and process metrics" 'perf test'

- Fix basic bpf filtering 'perf test' on s390x.

- Fix out of bounds array access in the 'perf stat' print_counters() evlist method.

- Add mwait_idle_with_hints.constprop.0 to the list of idle symbols.

- Use %zd for size_t printf formats on 32-bit.

- Correct the help info of "perf record --no-bpf-event" option.

- Add entries for CoreSight and Arm SPE tooling to MAINTAINERS.

Al Grant (3):
perf cs-etm: Fix corrupt data after perf inject from perf intel-pt: Fix corrupt data after perf inject from perf tools: Correct SNOOPX field offset

Arnaldo Carvalho de Melo (2):
perf symbols: Add mwait_idle_with_hints.constprop.0 to the list of idle symbols perf top/report: Fix infinite loop in the TUI for grouped events

Chris Wilson (1):
perf tools: Use %zd for size_t printf formats on 32-bit

David Ahern (1):
perf sched timehist: Fix use of CPU list with summary option

Ian Rogers (1):
perf parse-events: Avoid an uninitialized read when using fake PMUs

Jin Yao (1):
perf parse-events: Set exclude_guest=1 for user-space counting

Jiri Olsa (1):
perf report: Disable ordered_events for raw dump

Mathieu Poirier (1):
MAINTAINERS: Add entries for CoreSight and Arm SPE tooling

Sumanth Korikkar (1):
perf test: Fix basic bpf filtering test

Thomas Richter (2):
perf test: Set NULL sentinel in pmu_events table in "Parse and process metrics" test perf stat: Fix out of bounds array access in the print_counters() evlist method

Tiezhu Yang (1):
perf top: Skip side-band event setup if HAVE_LIBBPF_SUPPORT is not set

Wei Li (2):
perf: arm-spe: Fix check error when synthesizing events perf record: Correct the help info of option "--no-bpf-event"

 MAINTAINERS                                   |  8 ++++-
 tools/include/uapi/linux/perf_event.h         |  2 +-
 tools/perf/builtin-record.c                   |  2 +-
 tools/perf/builtin-report.c                   |  3 ++
 tools/perf/builtin-sched.c                    |  6 +++-
 tools/perf/builtin-top.c                      |  2 ++
 tools/perf/tests/bpf.c                        |  2 +-
 tools/perf/tests/parse-events.c               |  4 +--
 tools/perf/tests/parse-metric.c               |  3 ++
 tools/perf/ui/browsers/hists.c                |  3 +-
 .../util/arm-spe-decoder/arm-spe-decoder.c    |  6 ++--
 tools/perf/util/cs-etm.c                      |  9 +++++-
 tools/perf/util/intel-pt.c                    |  9 +++++-
 tools/perf/util/parse-events.c                | 31 ++++++++++++-------
 tools/perf/util/session.c                     |  2 +-
 tools/perf/util/stat-display.c                |  2 +-
 tools/perf/util/symbol.c                      |  1 +
 tools/perf/util/zstd.c                        |  2 +-
 18 files changed, 69 insertions(+), 28 deletions(-)
Signature · 1 line
2.26.2
#1 lore raw
Quoted 3 lines · Arnaldo Carvalho de Melo

- The SNOOPX and REMOTE offsets in the data_src bitmask in perf records were were both 37, SNOOPX is 38, fix it.

The kernel uapi side thinks it's 37. What's up?

  $ git grep PERF_MEM_SNOOPX_SHIFT

results in

include/uapi/linux/perf_event.h:#define PERF_MEM_SNOOPX_SHIFT 37
tools/include/uapi/linux/perf_event.h:#define PERF_MEM_SNOOPX_SHIFT 38

Hmm?

Linus

#2 lore raw

Em Tue, Sep 01, 2020 at 07:43:33PM -0700, Linus Torvalds escreveu:

Quoted 5 lines · Arnaldo Carvalho de Melo
Quoted 3 lines

- The SNOOPX and REMOTE offsets in the data_src bitmask in perf records were were both 37, SNOOPX is 38, fix it.

Quoted 8 lines

The kernel uapi side thinks it's 37. What's up?

  $ git grep PERF_MEM_SNOOPX_SHIFT

results in

include/uapi/linux/perf_event.h:#define PERF_MEM_SNOOPX_SHIFT 37
tools/include/uapi/linux/perf_event.h:#define PERF_MEM_SNOOPX_SHIFT 38

Its going in a separate patch, via PeterZ/Ingo,

- Arnaldo

merge bot Merged
#3 lore raw
In reply to Arnaldo Carvalho de Melo
The pull request you sent on Tue, 1 Sep 2020 19:41:54 -0300:

The pull request you sent on Tue, 1 Sep 2020 19:41:54 -0300:

Quoted 1 line

git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux.git tags/perf-tools-fixes-for-v5.9-2020-09-01

has been merged into torvalds/linux.git:
https://git.kernel.org/torvalds/c/9c7d619be5a002ea29c172df5e3c1227c22cbb41

Thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/prtracker.html