linux-kernel - Re: [PATCH 1/2] PM: s2idle: Drop redundant locks when entering s2idle

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAPDyKFqcCMGdZ9kfpkRi+6Y9wzf16WXAQjrio0-KwP9cZGJt_w@mail.gmail.com>
Date: Fri, 7 Mar 2025 11:22:45 +0100
From: Ulf Hansson <ulf.hansson@...aro.org>
To: "Rafael J. Wysocki" <rafael@...nel.org>
Cc: linux-pm@...r.kernel.org, Len Brown <len.brown@...el.com>, 
	Pavel Machek <pavel@...nel.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/2] PM: s2idle: Drop redundant locks when entering s2idle

On Thu, 6 Mar 2025 at 15:34, Rafael J. Wysocki <rafael@...nel.org> wrote:
>
> On Thu, Mar 6, 2025 at 12:36 PM Ulf Hansson <ulf.hansson@...aro.org> wrote:
> >
> > The calls to cpus_read_lock|unlock() protects us from getting CPUS
> > hotplugged, while entering suspend-to-idle. However, when s2idle_enter() is
> > called we should be far beyond the point when CPUs may be hotplugged.
> > Let's therefore simplify the code and drop the use of the lock.
> >
> > Signed-off-by: Ulf Hansson <ulf.hansson@...aro.org>
> > ---
> >  kernel/power/suspend.c | 4 ----
> >  1 file changed, 4 deletions(-)
> >
> > diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c
> > index 09f8397bae15..e7aca4e40561 100644
> > --- a/kernel/power/suspend.c
> > +++ b/kernel/power/suspend.c
> > @@ -98,8 +98,6 @@ static void s2idle_enter(void)
> >         s2idle_state = S2IDLE_STATE_ENTER;
> >         raw_spin_unlock_irq(&s2idle_lock);
> >
> > -       cpus_read_lock();
> > -
>
> As you said above, this is not expected to be contended, so it mostly
> serves as an annotation.
>
> The correctness of the code "protected" by it in fact depends on the
> number of CPUs not changing while it runs and this needs to be
> documented this way or another.
>
> I guess a comment to that effect can be used here instead of the locking.

Okay. I will update the patch and add a comment.

Thanks for reviewing!

[...]

Kind regards
Uffe