Date: Thu, 3 Nov 2016 15:06:31 +0000
From: Jakub Kicinski <kubakici@wp.pl>
To: ebiederm@xmission.com (Eric W. Biederman)
Cc: Dmitry Torokhov <dmitry.torokhov@gmail.com>,
        linux-kernel@vger.kernel.org, Kees Cook <keescook@chromium.org>,
        Johannes Berg <johannes@sipsolutions.net>,
        Maciej =?UTF-8?B?xbtlbmN6eWtvd3NraQ==?= <zenczykowski@gmail.com>
Subject: Re: [PATCH] userns: suppress kmemleak message
Message-ID: <20161103150631.0c655722@jkicinski-Precision-T1700>
In-Reply-To: <87a8dgvcke.fsf@xmission.com>
References: <20161103053948.GA40936@dtor-ws>
        <87a8dgvcke.fsf@xmission.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1014
Lines: 24

On Thu, 03 Nov 2016 09:54:25 -0500, Eric W. Biederman wrote:
> Dmitry Torokhov <dmitry.torokhov@gmail.com> writes:
> 
> > We do not ever intend to unregister "user" sysctl table, unfortunately
> > it leads kmemleak to believe that we are leaking memory:  
> 
> Sounds like an issue with kmemleak because we do retain references.
> 
> So no we don't intend to unregister the table.
> 
> As for the patch.
> 
> Nacked-by: "Eric W. Biederman" <ebiederm@xmission.com>
> 
> I can't see the using kmemleak_not_leak is possibly good form.  I
> would much rather have suggestions about constructs that won't confuse
> kmemleak and won't need ugly annotations that serve no purpose but to
> appease a tool.  Perhaps the user_header variable needs to be moved out
> of user_namespace_sysctl_init.

FWIW the problem now is that the compiler is clever enough to never
write the pointer to memory so kmemleak can't find it.  user_header is
just held in a register for as long as it's needed even though the
variable is static.