Received-SPF: pass (google.com: domain of linux-kernel+bounces-84284-linux.lists.archive=gmail.com@vger.kernel.org designates 139.178.88.99 as permitted sender) client-ip=139.178.88.99;
Precedence: bulk
MIME-Version: 1.0
References: <20231108215322.2845536-1-namhyung@kernel.org> <CAM9d7chyJun57UV-6qRzgTzDEmUu5Z0mStgjRbrg2dcjUkMQzw@mail.gmail.com>
 <CAM9d7cjQv=RiOkW5=7vXUSwQn5v1XQNiJyL9egGy2VgmKWO69Q@mail.gmail.com> <c29d648c-451a-42af-81d3-e1660e3af46f@redhat.com>
In-Reply-To: <c29d648c-451a-42af-81d3-e1660e3af46f@redhat.com>
From: Namhyung Kim <namhyung@kernel.org>
Date: Tue, 27 Feb 2024 16:49:45 -0800
Message-ID: <CAM9d7cip9xBfx+ZW06TO7BhKvTkQgQ=UZfrc+UOb3wOUrt8VYg@mail.gmail.com>
Subject: Re: [PATCH] locking/percpu-rwsem: Trigger contention tracepoints only
 if contended
To: Waiman Long <longman@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>, Ingo Molnar <mingo@kernel.org>, 
	Will Deacon <will@kernel.org>, Boqun Feng <boqun.feng@gmail.com>, 
	LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Tue, Feb 27, 2024 at 4:19=E2=80=AFPM Waiman Long <longman@redhat.com> wr=
ote:
>
>
> On 2/27/24 18:02, Namhyung Kim wrote:
> > Hello,
> >
> > On Mon, Nov 20, 2023 at 12:28=E2=80=AFPM Namhyung Kim <namhyung@kernel.=
org> wrote:
> >> Ping!
> >>
> >> On Wed, Nov 8, 2023 at 1:53=E2=80=AFPM Namhyung Kim <namhyung@kernel.o=
rg> wrote:
> >>> It mistakenly fires lock contention tracepoints always in the writer =
path.
> >>> It should be conditional on the try lock result.
> > Can anybody take a look at this?  This makes a large noise
> > in the lock contention result.
> >
> > Thanks,
> > Namhyung
> >
> >>> Signed-off-by: Namhyung Kim <namhyung@kernel.org>
> >>> ---
> >>>   kernel/locking/percpu-rwsem.c | 11 ++++++++---
> >>>   1 file changed, 8 insertions(+), 3 deletions(-)
> >>>
> >>> diff --git a/kernel/locking/percpu-rwsem.c b/kernel/locking/percpu-rw=
sem.c
> >>> index 185bd1c906b0..6083883c4fe0 100644
> >>> --- a/kernel/locking/percpu-rwsem.c
> >>> +++ b/kernel/locking/percpu-rwsem.c
> >>> @@ -223,9 +223,10 @@ static bool readers_active_check(struct percpu_r=
w_semaphore *sem)
> >>>
> >>>   void __sched percpu_down_write(struct percpu_rw_semaphore *sem)
> >>>   {
> >>> +       bool contended =3D false;
> >>> +
> >>>          might_sleep();
> >>>          rwsem_acquire(&sem->dep_map, 0, 0, _RET_IP_);
> >>> -       trace_contention_begin(sem, LCB_F_PERCPU | LCB_F_WRITE);
> >>>
> >>>          /* Notify readers to take the slow path. */
> >>>          rcu_sync_enter(&sem->rss);
> >>> @@ -234,8 +235,11 @@ void __sched percpu_down_write(struct percpu_rw_=
semaphore *sem)
> >>>           * Try set sem->block; this provides writer-writer exclusion=
.
> >>>           * Having sem->block set makes new readers block.
> >>>           */
> >>> -       if (!__percpu_down_write_trylock(sem))
> >>> +       if (!__percpu_down_write_trylock(sem)) {
> >>> +               trace_contention_begin(sem, LCB_F_PERCPU | LCB_F_WRIT=
E);
> >>>                  percpu_rwsem_wait(sem, /* .reader =3D */ false);
> >>> +               contended =3D true;
> >>> +       }
> >>>
> >>>          /* smp_mb() implied by __percpu_down_write_trylock() on succ=
ess -- D matches A */
> >>>
> >>> @@ -247,7 +251,8 @@ void __sched percpu_down_write(struct percpu_rw_s=
emaphore *sem)
> >>>
> >>>          /* Wait for all active readers to complete. */
> >>>          rcuwait_wait_event(&sem->writer, readers_active_check(sem), =
TASK_UNINTERRUPTIBLE);
> >>> -       trace_contention_end(sem, 0);
> >>> +       if (contended)
> >>> +               trace_contention_end(sem, 0);
> >>>   }
> >>>   EXPORT_SYMBOL_GPL(percpu_down_write);
> >>>
> >>> --
> >>> 2.42.0.869.gea05f2083d-goog
>
> Yes, that makes sense. Sorry for missing this patch.
>
> Reviewed-by: Waiman Long <longman@redhat.com>

Thanks for your review.
Namhyung