On 21/08/2026 11:01, sashiko-bot@kernel.org wrote:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Unconditionally enabling timestamps in per-thread mode breaks recording out-of-the-box on hardware without timestamp support (e.g., ETMv3 or ETMv4 without TRCIDR0.TSSIZE).
--
--- Patch [3]: [PATCH v2 03/14] perf cs-etm: Turn on context packet timestamps in per-thread mode --- commit 23dc37429523715cdb26c5647b22b1a9dc0c89ab Author: James Clark james.clark@linaro.org
perf cs-etm: Turn on context packet timestamps in per-thread mode
This enables context packet timestamps in per-thread mode, replacing the previous unset default. It helps keep samples in order when supporting multiple per-thread mode threads.
diff --git a/tools/perf/arch/arm/util/cs-etm.c b/tools/perf/arch/arm/util/cs-etm.c index acfdb3f53821e..d0486a164a700 100644 --- a/tools/perf/arch/arm/util/cs-etm.c +++ b/tools/perf/arch/arm/util/cs-etm.c @@ -453,13 +453,22 @@ static int cs_etm_recording_options(struct auxtrace_record *itr,
[ ... ]
if (!perf_cpu_map__is_any_cpu_or_is_empty(cpus)) {
/** Timestamps are required to interleave samples from different* CPUs.*//* Context IDs are required to associate trace to a process */- } else {
/** Enable context packet timestamps only (no periodic* timestamps). Even in per-thread mode with a single process we* still need to correlate trace to a specific mmap around* execs, which can be done with just context packet timestamps.*/evsel__set_config_if_unset(cs_etm_evsel, "timestamp", UINT64_MAX);[Severity: Medium] Will unconditionally setting this timestamp break recording out-of-the-box on hardware without timestamp support, like ETMv3?
Looking at cs_etm_validate_timestamp(), if the timestamp config is set but unsupported, it aborts the trace:
tools/perf/arch/arm/util/cs-etm.c:cs_etm_validate_timestamp() { if (!val) return 0;
if (cs_etm_get_version(cs_etm_pmu, cpu) == CS_ETMV3) { pr_err("%s: timestamp not supported in ETMv3...\n"); return -EINVAL; }}
Does this mean users running perf record --per-thread on such hardware will now get an error by default unless they explicitly pass timestamp=0?
Yes but the behavior is the same in per-CPU mode so this isn't a new issue. It's easily worked around and nobody has ever complained.
} /*