From: Peter Zijlstra <peterz@infradead.org>
To: Jiri Olsa <jolsa@redhat.com>
Cc: Numfor Mbiziwo-Tiapo <nums@google.com>,
mingo@redhat.com, acme@kernel.org,
alexander.shishkin@linux.intel.com, namhyung@kernel.org,
songliubraving@fb.com, mbd@fb.com, linux-kernel@vger.kernel.org,
irogers@google.com, eranian@google.com
Subject: Re: [PATCH] Fix perf test data race for tsan build
Date: Tue, 16 Jul 2019 14:34:11 +0200 [thread overview]
Message-ID: <20190716123411.GY3402@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20190716091039.GE22317@krava>
On Tue, Jul 16, 2019 at 11:10:39AM +0200, Jiri Olsa wrote:
> On Wed, Jul 10, 2019 at 03:16:20PM -0700, Numfor Mbiziwo-Tiapo wrote:
> > The tsan build of perf gives a data race warning in
>
> hi,
> what's 'tsan build of perf' ?
Thread Sanitizer, from the group that brought us KASAN and UBSAN.
I'd much rather see the below use READ_ONCE() and WRITE_ONCE(). Using
atomic_t without RmW ops is confusing.
> > mmap-thread-lookup.c when perf test is run on our local machines.
> > This is because there is no lock around the "go_away" variable in
> > "mmap-thread-lookup.c".
> >
> > The warning is difficult to reproduce from the tip branch and we
> > suspect it has something to do with tsan working differently
> > depending on the clang version.
> >
> > The warning can be silenced by adding locks around the variable
> > but it is better practice to make the variable atomic and use
> > the atomic_set and atomic_read functions. This does not actually
> > silence the warning because tsan doesn't recognize the atomic
> > functions as thread safe, but in actuality it should prevent a
> > data race.
> >
> > Signed-off-by: Numfor Mbiziwo-Tiapo <nums@google.com>
> > ---
> > tools/perf/tests/mmap-thread-lookup.c | 8 ++++----
> > 1 file changed, 4 insertions(+), 4 deletions(-)
> >
> > diff --git a/tools/perf/tests/mmap-thread-lookup.c b/tools/perf/tests/mmap-thread-lookup.c
> > index 5ede9b561d32..f4cbb3d92a46 100644
> > --- a/tools/perf/tests/mmap-thread-lookup.c
> > +++ b/tools/perf/tests/mmap-thread-lookup.c
> > @@ -17,7 +17,7 @@
> >
> > #define THREADS 4
> >
> > -static int go_away;
> > +static atomic_t go_away;
> >
> > struct thread_data {
> > pthread_t pt;
> > @@ -64,7 +64,7 @@ static void *thread_fn(void *arg)
> > return NULL;
> > }
> >
> > - while (!go_away) {
> > + while (!atomic_read(&go_away)) {
> > /* Waiting for main thread to kill us. */
> > usleep(100);
> > }
> > @@ -98,7 +98,7 @@ static int threads_create(void)
> > struct thread_data *td0 = &threads[0];
> > int i, err = 0;
> >
> > - go_away = 0;
> > + atomic_set(&go_away, 0);
> >
> > /* 0 is main thread */
> > if (thread_init(td0))
> > @@ -118,7 +118,7 @@ static int threads_destroy(void)
> > /* cleanup the main thread */
> > munmap(td0->map, page_size);
> >
> > - go_away = 1;
> > + atomic_set(&go_away, 1);
> >
> > for (i = 1; !err && i < THREADS; i++)
> > err = pthread_join(threads[i].pt, NULL);
> > --
> > 2.22.0.410.gd8fdbe21b5-goog
> >
prev parent reply other threads:[~2019-07-16 12:34 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-07-10 22:16 Numfor Mbiziwo-Tiapo
2019-07-16 9:10 ` Jiri Olsa
2019-07-16 12:34 ` Peter Zijlstra [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190716123411.GY3402@hirez.programming.kicks-ass.net \
--to=peterz@infradead.org \
--cc=acme@kernel.org \
--cc=alexander.shishkin@linux.intel.com \
--cc=eranian@google.com \
--cc=irogers@google.com \
--cc=jolsa@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mbd@fb.com \
--cc=mingo@redhat.com \
--cc=namhyung@kernel.org \
--cc=nums@google.com \
--cc=songliubraving@fb.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®