Re: [PATCH v4 09/19] perf test: Add named_threads workload

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The thread work loop is completely optimized away by the compiler, introduces a C11 data race on a shared global variable, and performs addition instead of the documented multiplication.
--

commit af99f4d30611866956b89fbc6d9ae547214d6310
Author: James Clark <james.clark@linaro.org>

perf test: Add named_threads workload

This commit adds a new named_threads test workload to perf, intended to spawn
multiple concurrent threads to verify processor trace decoding handles them
properly.

>     Add a workload that runs X threads that run a unique function named
>     "named_threads_thread[x]" which performs a multiplication in a loop for
>     Y loops. Each thread sets its name to "thread[x]".

[Severity: Medium]
The commit message states the workload performs a multiplication in a loop,
but the code performs an addition. Should the implementation be updated to
match the description?

> diff --git a/tools/perf/tests/workloads/named_threads.c b/tools/perf/tests/workloads/named_threads.c
> new file mode 100644
> index 0000000000000..d051d41a3cfe4
> --- /dev/null
> +++ b/tools/perf/tests/workloads/named_threads.c
[ ... ]
> +static int iterations = 500;
> +int named_threads_work = 1234;
> +
> +typedef void *(*thread_fn_t)(void *);
> +
> +#define DEFINE_THREAD(n)						\
> +noinline void *named_threads_thread##n(void *arg __maybe_unused)	\
> +{									\
> +	pthread_setname_np(pthread_self(), "thread" #n);		\
> +	for (int i = 0; i < iterations; i++)				\
> +		named_threads_work += 3;				\
> +									\
> +	return NULL;							\
> +}

[Severity: Medium]
Does updating the shared global named_threads_work concurrently across
multiple threads without atomics or locks cause an unintended data race?

Additionally, because this loop lacks memory barriers or volatile qualifiers,
will the compiler optimize it away under -O3 compilation? 

If it gets optimized into a single operation, the threads might exit almost
instantaneously, preventing the workload from generating the overlapping
processor trace samples it was designed to test.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260609-james-cs-context-tracking-fix-v4-0-44f9fb9e5c42@linaro.org?part=9