From mboxrd@z Thu Jan  1 00:00:00 1970
From: Steve Grubb <sgrubb@redhat.com>
Subject: Re: race in audit_log_untrusted_string for task_struct::comm
Date: Mon, 17 Mar 2014 09:01:45 -0400
Message-ID: <4829575.y6dA3xFjkS@x2>
References: <20140315232846.GE27244@madcap2.tricolour.ca>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7Bit
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <20140315232846.GE27244@madcap2.tricolour.ca>
Sender: linux-kernel-owner@vger.kernel.org
To: Richard Guy Briggs <rgb@redhat.com>
Cc: linux-audit@redhat.com, LKML <linux-kernel@vger.kernel.org>, Eric Paris <eparis@redhat.com>
List-Id: linux-audit@redhat.com

On Saturday, March 15, 2014 07:28:46 PM Richard Guy Briggs wrote:
> I'm inclined to go get_task_comm() in all 5 locations, but if we care
> more about locking overhead, I'll switch to memcpy().
> 
> Steve, do we care about the integrity of the comm field?

In the case of interpreters, its about the only thing we know about the 
application being executed. For example, a shell script will have exe=/bin/sh, 
so comm= is our only clue.

-Steve