linux/tools/perf/tests/shell, branch v6.17

perf test: Checking BPF metadata collection fails on version string

2025-09-04T23:37:19Z

commit edf2cadf01e8f ("perf test: add test for BPF metadata collection") fails consistently on the version string check. The perf version string on some of the constant integration test machines contains characters with special meaning in grep's extended regular expression matching algorithm. The output of perf version is: # perf version perf version 6.17.0-20250814.rc1.git20.24ea63ea3877.63.fc42.s390x+git # and the '+' character has special meaning in egrep command. Also the use of egrep is deprecated. Change the perf version string check to fixed character matching and get rid of egrep's warning being deprecated. Use grep -F instead. Output before: # perf test -F 102 Checking BPF metadata collection egrep: warning: egrep is obsolescent; using grep -E Basic BPF metadata test [Failed invalid output] 102: BPF metadata collection test : FAILED! # Output after: # perf test -F 102 Checking BPF metadata collection Basic BPF metadata test [Success] 102: BPF metadata collection test : Ok # Fixes: edf2cadf01e8f ("perf test: add test for BPF metadata collection") Signed-off-by: Thomas Richter Reviewed-by: Ian Rogers Reviewed-by: Arnaldo Carvalho de Melo Acked-by: Sumanth Korikkar Cc: Blake Jones Link: https://lore.kernel.org/r/20250822122540.4104658-1-tmricht@linux.ibm.com Signed-off-by: Namhyung Kim

perf test: Ensure lock contention using pipe mode

2025-07-30T20:38:53Z

The 'kernel lock contention analysis test' requires reliable triggering of lock contention. On some systems, previous benchmark calls failed to generate sufficient contention due to low system activity or resource limits. This patch adds the -p (pipe) option to all calls of perf bench sched messaging, ensuring consistent lock contention without relying on socket-based communication. Suggested-by: Thomas Richter Signed-off-by: Jan Polensky Link: https://lore.kernel.org/r/20250725170801.3176678-1-japo@linux.ibm.com Signed-off-by: Namhyung Kim

perf test: Fix comment ordering

2025-07-26T22:34:48Z

The previous commit that introduced this test overlooked a behavior of "perf test list", causing it to print "SPDX-License-Identifier: GPL-2.0" as a description for that test. This reorders the comments to fix that issue. Fixes: edf2cadf01e8 ("perf test: add test for BPF metadata collection") Signed-off-by: Blake Jones Reviewed-by: Ian Rogers Link: https://lore.kernel.org/r/20250726004023.3466563-1-blakejones@google.com [ update the commit message a little bit ] Signed-off-by: Namhyung Kim

perf test: Add more test cases to sched test

2025-07-03T18:52:20Z

$ sudo ./perf test -vv 92 92: perf sched tests: --- start --- test child forked, pid 1360101 Sched record pid 1360105's current affinity list: 0-3 pid 1360105's new affinity list: 0 pid 1360107's current affinity list: 0-3 pid 1360107's new affinity list: 0 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 4.330 MB /tmp/__perf_test_sched.perf.data.b3319 (12246 samples) ] Sched latency Sched script Sched map Sched timehist Samples of sched_switch event do not have callchains. ---- end(0) ---- 92: perf sched tests : Ok Reviewed-by: Ian Rogers Tested-by: Ian Rogers Link: https://lore.kernel.org/r/20250703014942.1369397-9-namhyung@kernel.org Signed-off-by: Namhyung Kim

perf header: Fix pipe mode header dumping

2025-07-03T18:22:17Z

The pipe mode header dumping was accidentally removed when tracing of header feature events in pipe mode was added. Minor spelling tweak to header test failure message. Fixes: 61051f9a8452 ("perf header: In pipe mode dump features without --header/-I") Signed-off-by: Ian Rogers Tested-by: Namhyung Kim Link: https://lore.kernel.org/r/20250703042000.2740640-1-irogers@google.com Signed-off-by: Namhyung Kim

perf test: Add sched latency and script shell tests

2025-07-01T22:37:34Z

Add shell tests covering the `perf sched latency` and `perf sched script` commands. The test creates 2 noploop processes on the same forced CPU, it then checks that the process appears in the `perf sched` output. Reviewed-by: James Clark Signed-off-by: Ian Rogers Link: https://lore.kernel.org/r/20250628012302.1242532-2-irogers@google.com Signed-off-by: Namhyung Kim

perf test annotate: Use --percent-limit rather than head to reduce output

2025-06-30T16:12:55Z

The annotate test was sped up by Thomas Richter in commit 658a8805cb60 ("perf test: Speed up test case 70 annotate basic tests") by reducing the annotate output using head. This causes flakes on hybrid machines where the first event dumped may not have the samples for the test within it. Rather than reduce the output using `head` switch to `--percent-limit 10` which will stop annotate dumping functions that have an overhead of less than 10%, the noploop program should be using more. Add the missing objdump option for the pipe mode version of the objdump with a command test. Signed-off-by: Ian Rogers Tested-by: Thomas Richter Link: https://lore.kernel.org/r/20250628015832.1271229-1-irogers@google.com Signed-off-by: Namhyung Kim

perf test: Add basic callgraph test to record testing

2025-06-30T16:11:39Z

Give some basic perf record callgraph coverage. Signed-off-by: Ian Rogers Reviewed-by: James Clark Tested-by: Thomas Richter Link: https://lore.kernel.org/r/20250628015553.1270748-1-irogers@google.com Signed-off-by: Namhyung Kim

perf test: perf header test fails on s390

2025-06-30T15:59:13Z

commit 2d584688643fa ("perf test: Add header shell test") introduced a new test case for perf header. It fails on s390 because call graph option -g is not supported on s390. Also the option --call-graph dwarf is only supported for the event cpu-clock. Remove this option and the test succeeds. Output after: # ./perf test 76 76: perf header tests : Ok Fixes: 2d584688643fa ("perf test: Add header shell test") Signed-off-by: Thomas Richter Reviewed-by: Ian Rogers Acked-by: Sumanth Korikkar Link: https://lore.kernel.org/r/20250630091613.3061664-1-tmricht@linux.ibm.com Signed-off-by: Namhyung Kim

perf stat: Fix uncore aggregation number

2025-06-27T23:14:10Z

Follow up: lore.kernel.org/CAP-5=fVDF4-qYL1Lm7efgiHk7X=_nw_nEFMBZFMcsnOOJgX4Kg@mail.gmail.com/ The patch adds unit aggregation during evsel merge the aggregated uncore counters. Change the name of the column to `ctrs` and `counters` for json mode. Tested on a 2-socket machine with SNC3, uncore_imc_[0-11] and cpumask="0,120" Before: perf stat -e clockticks -I 1000 --per-socket # time socket cpus counts unit events 1.001085024 S0 1 9615386315 clockticks 1.001085024 S1 1 9614287448 clockticks perf stat -e clockticks -I 1000 --per-node # time node cpus counts unit events 1.001029867 N0 1 3205726984 clockticks 1.001029867 N1 1 3205444421 clockticks 1.001029867 N2 1 3205234018 clockticks 1.001029867 N3 1 3205224660 clockticks 1.001029867 N4 1 3205207213 clockticks 1.001029867 N5 1 3205528246 clockticks After: perf stat -e clockticks -I 1000 --per-socket # time socket ctrs counts unit events 1.001026071 S0 12 9619677996 clockticks 1.001026071 S1 12 9618612614 clockticks perf stat -e clockticks -I 1000 --per-node # time node ctrs counts unit events 1.001027449 N0 4 3207251859 clockticks 1.001027449 N1 4 3207315930 clockticks 1.001027449 N2 4 3206981828 clockticks 1.001027449 N3 4 3206566126 clockticks 1.001027449 N4 4 3206032609 clockticks 1.001027449 N5 4 3205651355 clockticks Tested with JSON output linter: perf test "perf stat JSON output linter" 94: perf stat JSON output linter : Ok Suggested-by: Ian Rogers Reviewed-by: Ian Rogers Signed-off-by: Chun-Tse Shao Link: https://lore.kernel.org/r/20250627201818.479421-1-ctshao@google.com Signed-off-by: Namhyung Kim