From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexander Viro Subject: Re: [PATCH 2.6.19-rc4] audit: support for descriptor pairs (fwd) Date: Tue, 7 Nov 2006 13:31:57 -0500 Message-ID: <20061107183157.GJ25857@devserv.devel.redhat.com> References: <200611070858.20720.sgrubb@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <200611070858.20720.sgrubb@redhat.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: linux-audit-bounces@redhat.com Errors-To: linux-audit-bounces@redhat.com To: Steve Grubb Cc: Mark Workman , linux-audit@redhat.com List-Id: linux-audit@redhat.com On Tue, Nov 07, 2006 at 08:58:20AM -0500, Steve Grubb wrote: > On Tuesday 07 November 2006 08:18, Mark Workman wrote: > > provide an audit record of the descriptor pair returned by pipe() and > > socketpair(). > > Patch looks OK to me. Al, what do you think? Ehh... a) why bother with exporting sock_map_... at all? b) failure exits in socketpair are, AFAICS, botched by the patch. At the very least, you can get sock_release() done on a socket that already went through it due to fput(). Please, fix. I would suggest taking that into sys_socketpair(), actually, instead of a new helper...