[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CANp29Y7UFVPYtdd0jVUbkUjdkpXXrv7Z+sczmKvZytNmtgx1hA@mail.gmail.com>
Date: Fri, 18 Oct 2024 18:10:27 +0200
From: Aleksandr Nogikh <nogikh@...gle.com>
To: Jann Horn <jannh@...gle.com>
Cc: syzbot <syzbot+f24dd060c1911fe54c85@...kaller.appspotmail.com>,
andreyknvl@...il.com, daniel.lezcano@...aro.org, elver@...gle.com,
linux-kernel@...r.kernel.org, linux-pm@...r.kernel.org, lukasz.luba@....com,
rafael@...nel.org, rui.zhang@...el.com, syzkaller-bugs@...glegroups.com,
vbabka@...e.cz
Subject: Re: [syzbot] [pm?] WARNING in thermal_thresholds_flush
Hi Jann,
On Fri, Oct 18, 2024 at 5:43 PM 'Jann Horn' via syzkaller-bugs
<syzkaller-bugs@...glegroups.com> wrote:
>
> On Fri, Oct 18, 2024 at 3:21 PM syzbot
> <syzbot+f24dd060c1911fe54c85@...kaller.appspotmail.com> wrote:
> >
> > syzbot has bisected this issue to:
> >
> > commit b8c8ba73c68bb3c3e9dad22f488b86c540c839f9
> > Author: Jann Horn <jannh@...gle.com>
> > Date: Fri Aug 9 15:36:56 2024 +0000
> >
> > slub: Introduce CONFIG_SLUB_RCU_DEBUG
> >
> > bisection log: https://syzkaller.appspot.com/x/bisect.txt?x=14cc4c5f980000
> > start commit: 15e7d45e786a Add linux-next specific files for 20241016
> > git tree: linux-next
> > console output: https://syzkaller.appspot.com/x/log.txt?x=12cc4c5f980000
> > kernel config: https://syzkaller.appspot.com/x/.config?x=c36416f1c54640c0
> > dashboard link: https://syzkaller.appspot.com/bug?extid=f24dd060c1911fe54c85
> > syz repro: https://syzkaller.appspot.com/x/repro.syz?x=1192f887980000
> > C reproducer: https://syzkaller.appspot.com/x/repro.c?x=1417e830580000
> >
> > Reported-by: syzbot+f24dd060c1911fe54c85@...kaller.appspotmail.com
> > Fixes: b8c8ba73c68b ("slub: Introduce CONFIG_SLUB_RCU_DEBUG")
> >
> > For information about bisection process see: https://goo.gl/tpsmEJ#bisection
>
> This looks like a bogus bisection - the blamed commit was already in
> v6.12-rc1, while the file that contains this warning
> (drivers/thermal/thermal_thresholds.c) doesn't even exist in mainline
> yet.
>
Indeed, the bisection result is wrong :( At one of the steps, syzbot
has hit too many random "lost connection to test machine" crashes,
which derailed the process.
I've filed https://github.com/google/syzkaller/issues/5414. Once
implemented, that should get rid of this class of false positives.
--
Aleksandr
Powered by blists - more mailing lists