Re: [PATCH v4 04/19] perf test cs-etm: Test process attribution

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The test script hardcodes CPU 0 for recording and pinning, causing false test failures on systems where CPU 0 is offlined or lacks a CoreSight ETM.
--

commit c50b8d82ea4d5dd7b4d85c01619bc432a03ea379
Author: James Clark <james.clark@linaro.org>

perf test cs-etm: Test process attribution

This commit adds a test to verify that CoreSight context switch workload
traces attribute symbols to the correct process when traced on a single CPU.

> diff --git a/tools/perf/tests/shell/coresight/context_switch_thread.sh b/tools/perf/tests/shell/coresight/context_switch_thread.sh
> new file mode 100755
> index 0000000000000..2b9c44b86c591
> --- /dev/null
> +++ b/tools/perf/tests/shell/coresight/context_switch_thread.sh
[ ... ]
> +cf="$tmpdir/ctl"
> +af="$tmpdir/ack"
> +mkfifo "$cf" "$af"
> +
> +# Pin to one CPU so the two threads alternate running but record into the same
> +# trace buffer. Start disabled and use the control FIFO to only record the
> +# workload and not startup.
> +perf record -o "$tmpdir/data" -e cs_etm/timestamp=0/u -C 0 -D -1 --control fifo:"$cf","$af" -- \
> +	taskset --cpu-list 0 perf test --record-ctl fifo:"$cf","$af" \
> +	-w context_switch_loop > /dev/null 2>&1

[Severity: Medium]
Will this test fail on systems where CPU 0 lacks a CoreSight ETM or is
currently offlined?

On some ARM architectures, ETMs are not guaranteed to be present on all CPUs.
Could we dynamically discover a CPU with an ETM via the sysfs event_source
devices instead of hardcoding CPU 0 for the record and taskset commands?

[ ... ]

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260609-james-cs-context-tracking-fix-v4-0-44f9fb9e5c42@linaro.org?part=4