linux-kernel - Re: [PATCH] platform/x86: dell-laptop: Guard SMBIOS calls with a mutex

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: <20180131185448.GE8676@fury>
Date:   Wed, 31 Jan 2018 10:54:48 -0800
From:   Darren Hart <dvhart@...radead.org>
To:     Andy Shevchenko <andy.shevchenko@...il.com>
Cc:     Mario Limonciello <Mario.Limonciello@...l.com>,
        Pali Rohár <pali.rohar@...il.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Platform Driver <platform-driver-x86@...r.kernel.org>
Subject: Re: [PATCH] platform/x86: dell-laptop: Guard SMBIOS calls with a
 mutex

On Tue, Jan 30, 2018 at 05:39:58PM +0200, Andy Shevchenko wrote:
> On Tue, Jan 30, 2018 at 5:35 PM,  <Mario.Limonciello@...l.com> wrote:
> 
> >> >     dell_set_arguments(0x2, 0, 0, 0);
> >> >     ret = dell_send_request(CLASS_INFO, SELECT_RFKILL);
> >>
> >> Hi! I'm looking at this code, and do we need shared global buffer with
> >> mutex protection at all? Is not buffer allocated on stack enough?
> >
> > Oh you mean rather than create buffer mutex to just remove global
> > buffer and allocate in each function?  That seems like a workable
> > approach to me too.
> >
> > I'm fine with either way.
> >
> > Andy or Darren, what's your preference in this area?
> 
> It reminds me USB stuff where buffer for transfer is allocated on heap
> before performing communication.
> So, it looks similar to some extent and I have no objection on that
> kind of approach.

Late to the party it seems, but FWIW:

I don't see a significant advantage of a global buffer. It doesn't *need* to be
global, and the locking just adds complexity. The heap solution seems much
preferable to me.

-- 
Darren Hart
VMware Open Source Technology Center