From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail172.messagelabs.com (mail172.messagelabs.com [216.82.254.3]) by kanga.kvack.org (Postfix) with ESMTP id B7DC6900001 for ; Tue, 26 Apr 2011 02:07:20 -0400 (EDT) Received: by wwi36 with SMTP id 36so232983wwi.26 for ; Mon, 25 Apr 2011 23:07:17 -0700 (PDT) MIME-Version: 1.0 In-Reply-To: References: <20110426055521.GA18473@localhost> Date: Tue, 26 Apr 2011 14:07:17 +0800 Message-ID: Subject: Re: readahead and oom From: Dave Young Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Sender: owner-linux-mm@kvack.org List-ID: To: Wu Fengguang Cc: linux-mm , Linux Kernel Mailing List On Tue, Apr 26, 2011 at 2:05 PM, Dave Young wro= te: > On Tue, Apr 26, 2011 at 1:55 PM, Wu Fengguang wr= ote: >> On Tue, Apr 26, 2011 at 01:49:25PM +0800, Dave Young wrote: >>> Hi, >>> >>> When memory pressure is high, readahead could cause oom killing. >>> IMHO we should stop readaheading under such circumstances=E3=80=82If it= 's true >>> how to fix it? >> >> Good question. Before OOM there will be readahead thrashings, which >> can be addressed by this patch: >> >> http://lkml.org/lkml/2010/2/2/229 > > Hi, I'm not clear about the patch, could be regard as below cases? > 1) readahead alloc fail due to low memory such as other large allocation For example vm balloon allocate lots of memory, then readahead could fail immediately and then oom > 2) readahead thrashing caused by itself > >> >> However there seems no much interest on that feature.. I can separate >> that out and resubmit it standalone if necessary. >> >> Thanks, >> Fengguang >> > > > > -- > Regards > dave > --=20 Regards dave -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/ Don't email: email@kvack.org