linux-kernel - Re: [PATCH v1 1/1] platform/x86: ideapad-laptop: Make the scope

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: <20240913233324.GA306935@thelio-3990X>
Date: Fri, 13 Sep 2024 16:33:24 -0700
From: Nathan Chancellor <nathan@...nel.org>
To: Josh Poimboeuf <jpoimboe@...nel.org>
Cc: Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
	Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>,
	Gergo Koteles <soyer@....hu>, Hans de Goede <hdegoede@...hat.com>,
	platform-driver-x86@...r.kernel.org, linux-kernel@...r.kernel.org,
	Ike Panhc <ike.pan@...onical.com>,
	Peter Zijlstra <peterz@...radead.org>,
	kernel test robot <lkp@...el.com>
Subject: Re: [PATCH v1 1/1] platform/x86: ideapad-laptop: Make the
 scope_guard() clear of its scope

On Thu, Sep 05, 2024 at 08:16:01PM -0700, Josh Poimboeuf wrote:
> On Wed, Sep 04, 2024 at 01:26:11PM +0300, Andy Shevchenko wrote:
> > On Tue, Sep 03, 2024 at 09:52:01PM -0700, Josh Poimboeuf wrote:
> > > I'm not sure I buy that, we should look closer to understand what the
> > > issue is.  Can you share the config and/or toolchain version(s) need to
> > > trigger the warning?
> > 
> > .config is from the original report [1], toolchain is
> > Debian clang version 18.1.8 (9)
> > 	Target: x86_64-pc-linux-gnu
> > 	Thread model: posix
> > 	InstalledDir: /usr/bin
> > 
> > (Just whatever Debian unstable provides)
> > 
> > [1]: https://lore.kernel.org/oe-kbuild-all/202408290219.BrPO8twi-lkp@intel.com/
> 
> The warning is due to a (minor?) Clang bug, almost like it tried to
> optimize but didn't quite finish.
...
> In this case, mutex is an unconditional guard, so the constructor just
> returns the original value of '&priv->vpc_mutex'.  So if the original
> '&priv->vpc_mutex' is never NULL, the condition would always be true.

Thanks a lot for that Josh. I have a somewhat trivial reproducer for the
clang folks to take a look at. I should have some time on Monday to get
that reported upstream and I will see if I can find anyone to take a
look at it.

For what it is worth, I don't think the workaround for this is too bad
and it seems like it only shows up with KCOV.

Cheers,
Nathan