[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CANpmjNPu876ApAQzODM3OCY0uitkw9SuucNcpW9gRq19rrjuGQ@mail.gmail.com>
Date: Tue, 26 Oct 2021 16:39:44 +0200
From: Marco Elver <elver@...gle.com>
To: paulmck@...nel.org
Cc: syzbot <syzbot+4dfb96a94317a78f44d9@...kaller.appspotmail.com>,
rcu@...r.kernel.org, linux-fsdevel@...r.kernel.org,
linux-kernel@...r.kernel.org, syzkaller-bugs@...glegroups.com,
viro@...iv.linux.org.uk
Subject: Re: [syzbot] KCSAN: data-race in call_rcu / rcu_gp_fqs_loop
On Tue, 26 Oct 2021 at 16:33, Paul E. McKenney <paulmck@...nel.org> wrote:
> On Tue, Oct 26, 2021 at 07:07:30AM -0700, Paul E. McKenney wrote:
> > On Mon, Oct 25, 2021 at 12:31:53PM +0200, Marco Elver wrote:
> > > +Cc Paul
> > >
> > > data race is in rcu code, presumably not yet discovered by rcutorture?
> >
> > Quite possibly, and I will take a look. Thank you for sending this
> > along.
>
> And this is (allegedly) fixed by commit 2431774f04d10 ("rcu: Mark accesses
> to rcu_state.n_force_qs"), which is in -rcu and slated for the upcoming
> merge window. But yes, still a bug in mainline.
Thanks for confirming, then the latest incarnation of it must then be
a dup, and I'll mark it fixed:
#syz fix: rcu: Mark accesses to rcu_state.n_force_qs
Apologies I missed there was a previous report and your fix.
Thanks,
-- Marco
Powered by blists - more mailing lists