linux-kernel - Re: regulator: deadlock vs memory reclaim

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] [day] [month] [year] [list]

Message-ID: <20200810195123.GJ6438@sirena.org.uk>
Date:   Mon, 10 Aug 2020 20:51:23 +0100
From:   Mark Brown <broonie@...nel.org>
To:     Dmitry Osipenko <digetx@...il.com>
Cc:     Michał Mirosław <mirq-linux@...e.qmqm.pl>,
        Liam Girdwood <lgirdwood@...il.com>,
        linux-kernel@...r.kernel.org
Subject: Re: regulator: deadlock vs memory reclaim

On Mon, Aug 10, 2020 at 10:41:54PM +0300, Dmitry Osipenko wrote:
> 10.08.2020 22:25, Michał Mirosław пишет:

> >> The initialization cases that are the trigger are only done for coupled
> >> regulators though AFAICT, otherwise we're not doing allocations with the
> >> lock held and should be able to progress.

> > I caught a few lockdep complaints that suggest otherwise, but I'm still
> > looking into that.

> The problem looks obvious to me. The regulator_init_coupling() is
> protected with the list_mutex, the regulator_lock_dependent() also
> protected with the list_mutex. Hence if offending reclaim happens from
> init_coupling(), then there is a lockup.

We may also have problems if I/O triggers allocations for some reason,
though that's also going to be a limited set of cases.  Might be what
lockdep was showing though.

> It should be enough just to keep the regulator_find_coupler() under
> lock, or even completely remove the locking around init_coupling(). I
> think it should be better to keep the find_coupler() protected.

> Michał, does this fix yours problem?

That was the sort of thing I was thinking about here - it should at
least be an improvement if nothing else.

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)