linux-kernel - Re: [PATCH] watchdog:alim1535_wdt: Fix data race in ali_settimer() concerning ali_timeout

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: <20190731192410.GA4935@roeck-us.net>
Date:   Wed, 31 Jul 2019 12:24:10 -0700
From:   Guenter Roeck <linux@...ck-us.net>
To:     Mark Balantzyan <mbalant3@...il.com>
Cc:     wim@...ux-watchdog.org, linux-watchdog@...r.kernel.org,
        linux-kernel@...r.kernel.org, Pavel Andrianov <andrianov@...ras.ru>
Subject: Re: [PATCH] watchdog:alim1535_wdt: Fix data race in ali_settimer()
 concerning ali_timeout_bits variable.

Mark,

On Wed, Jul 31, 2019 at 11:23:19AM -0700, Mark Balantzyan wrote:
> Hi Guenter, all,
> 
> It's alright if you still don't wish to review my patch on alim1535_wdt, but
> my employer and I, using our race condition analysis tool, detected it to
> contain a race condition warning. I believe any possible issues could be
> resolved if it were rewritten to use the watchdog subsystem as you've
> previously stated.
> 
> Now, I don't wish to bother you too much, but it seems I forgot to work
> mainly with my assigned mentor prior to submitting patches..sorry. So, after
> I have worked on rewriting the alim1535 driver into common watchdog
> subsystem with my mentor, may I submit it to you then?
> 

Similar to pc87413, this driver very likely has zero users left, there
won't be any hardware to test the patch, and we won't be able to accept
such a patch because it wasn't tested.

On top of that, the only race condition I can see in that driver is in
ali_settimer(), between ali_timeout_bits and timeout. Yet, that is not
really a race condition because the driver can only be opened once,
and thus there is no means for two threads entering ali_settimer()
at the same time.

I don't really understand this focus on fixing theoretic/irrelevant
race conditions in drivers which no one uses anymore. Maybe someone
can enlighten me ?

Thanks,
Guenter