linux-kernel - Re: [patch V2 22/29] lockup_detector: Make watchdog_nmi

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <alpine.DEB.2.20.1710031411440.2281@nanos>
Date:   Tue, 3 Oct 2017 14:13:54 +0200 (CEST)
From:   Thomas Gleixner <tglx@...utronix.de>
To:     Michael Ellerman <mpe@...erman.id.au>
cc:     LKML <linux-kernel@...r.kernel.org>,
        Ingo Molnar <mingo@...nel.org>,
        Peter Zijlstra <peterz@...radead.org>,
        Borislav Petkov <bp@...en8.de>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Sebastian Siewior <bigeasy@...utronix.de>,
        Nicholas Piggin <npiggin@...il.com>,
        Don Zickus <dzickus@...hat.com>,
        Chris Metcalf <cmetcalf@...lanox.com>,
        Ulrich Obergfell <uobergfe@...hat.com>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        linuxppc-dev@...ts.ozlabs.org
Subject: Re: [patch V2 22/29] lockup_detector: Make watchdog_nmi_reconfigure()
 two stage

On Tue, 3 Oct 2017, Michael Ellerman wrote:
> Thomas Gleixner <tglx@...utronix.de> writes:
> >> The first call is new because previously watchdog_nmi_reconfigure()
> >> wasn't called from softlockup_reconfigure_threads().
> >
> > Hmm, don't you have the same problem with CPU hotplug or do you just get
> > lucky because the hotplug callback in your code is ordered vs. the
> > softlockup thread hotplug callback in a way that this does not hit?
> 
> I don't see it with CPU hotplug.
> 
> AFAICS that's because softlockup_reconfigure_threads() isn't called for
> CPU hotplug. Unless there's a path I'm missing?

As I said in the other reply, I assumed that its called via
watchdog_nmi_enable(cpu), but that's a weak function which is not
implemented on power. So no issue.

> >> I'm not sure what the easiest fix is. One option would be to just drop
> >> the WARN_ON, it's just there for paranoia AFAICS.
> >
> > The straight forward way is to make use of the new probe function. Patch
> > below.
> 
> Hmm, I tried that patch, it makes the warning go away. But then I
> triggered a deliberate hard lockup and got nothing.
> 
> Then I went back to the existing code (in linux-next), and I still get
> no warning from a deliberate hard lockup.
> 
> So seems there may be some more gremlins. Will test more in the morning.

Hrm. That's weird. I'll have a look and send a proper patch series on top
of next.

Thanks,

	tglx