Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S270956AbTGPQmX (ORCPT ); Wed, 16 Jul 2003 12:42:23 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S270955AbTGPQmX (ORCPT ); Wed, 16 Jul 2003 12:42:23 -0400 Received: from fw.osdl.org ([65.172.181.6]:4527 "EHLO mail.osdl.org") by vger.kernel.org with ESMTP id S270956AbTGPQmT (ORCPT ); Wed, 16 Jul 2003 12:42:19 -0400 Date: Wed, 16 Jul 2003 09:49:31 -0700 From: Andrew Morton To: Jens Axboe Cc: andrea@suse.de, alan@lxorguk.ukuu.org.uk, marcelo@conectiva.com.br, mason@suse.com, linux-kernel@vger.kernel.org, sct@redhat.com, jgarzik@pobox.com, viro@math.psu.edu Subject: Re: RFC on io-stalls patch Message-Id: <20030716094931.0a5015a5.akpm@osdl.org> In-Reply-To: <20030716130442.GZ833@suse.de> References: <20030714131206.GJ833@suse.de> <20030714195138.GX833@suse.de> <20030714201637.GQ16313@dualathlon.random> <20030715052640.GY833@suse.de> <1058268126.3857.25.camel@dhcp22.swansea.linux.org.uk> <20030715112737.GQ833@suse.de> <20030716124355.GE4978@dualathlon.random> <20030716124656.GY833@suse.de> <20030716125933.GF4978@dualathlon.random> <20030716130442.GZ833@suse.de> X-Mailer: Sylpheed version 0.9.0pre1 (GTK+ 1.2.10; i686-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1143 Lines: 28 Jens Axboe wrote: > > On Wed, Jul 16 2003, Andrea Arcangeli wrote: > > On Wed, Jul 16, 2003 at 02:46:56PM +0200, Jens Axboe wrote: > > > Well it's a combined problem. Threshold too high on dirty memory, > > > someone doing a read well get stuck flushing out as well. > > > > a pure read not. the write throttling should be per-process, then there > > will be little risk. > > A read from user space, dirtying data along the way. Actually it's a read from userspace which allocates a page which goes into direct reclaim which discovers a locked buffer on the tail of the LRU and then waits on it. And if he's especially unlucky: while he waits, some other process continues to pound more writes into the queue which get merged ahead of the one he's waiting on, up to a point. (I don't know if 2.6 does much better in this regard. It is supposed to. Has anyone tested for it?) - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/