From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755937Ab2LNLPN (ORCPT ); Fri, 14 Dec 2012 06:15:13 -0500 Received: from mail-vc0-f174.google.com ([209.85.220.174]:42850 "EHLO mail-vc0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754871Ab2LNLPL (ORCPT ); Fri, 14 Dec 2012 06:15:11 -0500 MIME-Version: 1.0 In-Reply-To: <20121214072755.GR4939@ZenIV.linux.org.uk> References: <3b624af48f4ba4affd78466b73b6afe0e2f66549.1355463438.git.luto@amacapital.net> <20121214072755.GR4939@ZenIV.linux.org.uk> From: Andy Lutomirski Date: Fri, 14 Dec 2012 03:14:50 -0800 Message-ID: Subject: Re: [PATCH] mm: Downgrade mmap_sem before locking or populating on mmap To: Al Viro Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org, Andrew Morton , Ingo Molnar , Michel Lespinasse , Hugh Dickins , "J??rn Engel" Content-Type: text/plain; charset=ISO-8859-1 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Dec 13, 2012 at 11:27 PM, Al Viro wrote: > On Thu, Dec 13, 2012 at 09:49:43PM -0800, Andy Lutomirski wrote: >> This is a serious cause of mmap_sem contention. MAP_POPULATE >> and MCL_FUTURE, in particular, are disastrous in multithreaded programs. >> >> Signed-off-by: Andy Lutomirski >> --- >> >> Sensible people use anonymous mappings. I write kernel patches :) >> >> I'm not entirely thrilled by the aesthetics of this patch. The MAP_POPULATE case >> could also be improved by doing it without any lock at all. This is still a big >> improvement, though. > > Wait a minute. get_user_pages() relies on ->mmap_sem being held. Unless > I'm seriously misreading your patch it removes that protection. And yes, > I'm aware of execve-related exception; it's in special circumstances - > bprm->mm is guaranteed to be not shared (and we need to rearchitect that > area anyway, but that's a separate story). Unless I completely screwed up the patch, ->mmap_sem is still held for read (it's downgraded from write). It's just not held for write anymore. --Andy