Received-SPF: pass (google.com: domain of linux-kernel+bounces-36199-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Precedence: bulk
MIME-Version: 1.0
References: <20240123153313.GA21832@redhat.com> <20240123153359.GA21866@redhat.com>
In-Reply-To: <20240123153359.GA21866@redhat.com>
From: Dylan Hatch <dylanbhatch@google.com>
Date: Tue, 23 Jan 2024 15:52:21 -0800
Message-ID: <CADBMgpwhH4VyADf7ajYee=FFhdfwSQF3iAjCHTHWPv5FOoYRPQ@mail.gmail.com>
Subject: Re: [PATCH 3/3] exit: wait_task_zombie: kill the no longer necessary spin_lock_irq(siglock)
To: Oleg Nesterov <oleg@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>, "Eric W. Biederman" <ebiederm@xmission.com>, 
	linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Tue, Jan 23, 2024 at 7:35=E2=80=AFAM Oleg Nesterov <oleg@redhat.com> wro=
te:
>
> After the recent changes nobody use siglock to read the values protected
> by stats_lock, we can kill spin_lock_irq(&current->sighand->siglock) and
> update the comment.
>
> With this patch only __exit_signal() and thread_group_start_cputime() tak=
e
> stats_lock under siglock.
>
> Signed-off-by: Oleg Nesterov <oleg@redhat.com>
> ---
>  kernel/exit.c | 10 +++-------
>  1 file changed, 3 insertions(+), 7 deletions(-)
>
> diff --git a/kernel/exit.c b/kernel/exit.c
> index 3988a02efaef..dfb963d2f862 100644
> --- a/kernel/exit.c
> +++ b/kernel/exit.c
> @@ -1127,17 +1127,14 @@ static int wait_task_zombie(struct wait_opts *wo,=
 struct task_struct *p)
>                  * and nobody can change them.
>                  *
>                  * psig->stats_lock also protects us from our sub-threads
> -                * which can reap other children at the same time. Until
> -                * we change k_getrusage()-like users to rely on this loc=
k
> -                * we have to take ->siglock as well.
> +                * which can reap other children at the same time.
>                  *
>                  * We use thread_group_cputime_adjusted() to get times fo=
r
>                  * the thread group, which consolidates times for all thr=
eads
>                  * in the group including the group leader.
>                  */
>                 thread_group_cputime_adjusted(p, &tgutime, &tgstime);
> -               spin_lock_irq(&current->sighand->siglock);
> -               write_seqlock(&psig->stats_lock);
> +               write_seqlock_irq(&psig->stats_lock);
>                 psig->cutime +=3D tgutime + sig->cutime;
>                 psig->cstime +=3D tgstime + sig->cstime;
>                 psig->cgtime +=3D task_gtime(p) + sig->gtime + sig->cgtim=
e;
> @@ -1160,8 +1157,7 @@ static int wait_task_zombie(struct wait_opts *wo, s=
truct task_struct *p)
>                         psig->cmaxrss =3D maxrss;
>                 task_io_accounting_add(&psig->ioac, &p->ioac);
>                 task_io_accounting_add(&psig->ioac, &sig->ioac);
> -               write_sequnlock(&psig->stats_lock);
> -               spin_unlock_irq(&current->sighand->siglock);
> +               write_sequnlock_irq(&psig->stats_lock);
>         }
>
>         if (wo->wo_rusage)
> --
> 2.25.1.362.g51ebf55
>

Signed-off-by: Dylan Hatch <dylanbhatch@google.com>