From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754199AbaDYOXx (ORCPT ); Fri, 25 Apr 2014 10:23:53 -0400 Received: from mail-pd0-f177.google.com ([209.85.192.177]:38119 "EHLO mail-pd0-f177.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752861AbaDYOXt (ORCPT ); Fri, 25 Apr 2014 10:23:49 -0400 Message-ID: <535A6FF1.6070805@kernel.dk> Date: Fri, 25 Apr 2014 08:23:45 -0600 From: Jens Axboe User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.4.0 MIME-Version: 1.0 To: Christoph Hellwig CC: "Elliott, Robert (Server Storage)" , "scameron@beardog.cce.hp.com" , Bart Van Assche , linux-kernel@vger.kernel.org Subject: Re: [PATCH] blk-mq: respect rq_affinity References: <94D0CD8314A33A4D9D801C0FE68B402956EDDA26@G9W0745.americas.hpqcorp.net> <53592665.8050406@kernel.dk> <20140425053018.GA30957@infradead.org> <94D0CD8314A33A4D9D801C0FE68B402956EDE0D8@G9W0745.americas.hpqcorp.net> <20140425071659.GA15384@infradead.org> <20140425093253.GA18345@infradead.org> In-Reply-To: <20140425093253.GA18345@infradead.org> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 04/25/2014 03:32 AM, Christoph Hellwig wrote: > From: Christoph Hellwig > > The blk-mq code is using it's own version of the I/O completion affinity > tunables, which causes a few issues: > > - the rq_affinity sysfs file doesn't work for blk-mq devices, even if it > still is present, thus breaking existing tuning setups. > - the rq_affinity = 1 mode, which is the defauly for legacy request based > drivers isn't implemented at all. > - blk-mq drivers don't implement any completion affinity with the default > flag settings. > > This patches removes the blk-mq ipi_redirect flag and sysfs file, as well > as the internal BLK_MQ_F_SHOULD_IPI flag and replaces it with code that > respects the queue-wide rq_affinity flags and also implements the > rq_affinity = 1 mode. > > This means I/O completion affinity can now only be tuned block-queue wide > instead of per context, which seems more sensible to me anyway. Agree, this is a good cleanup, and there's no reason blk-mq needs its own sysfs file for this. -- Jens Axboe