From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753143Ab3KSQRJ (ORCPT ); Tue, 19 Nov 2013 11:17:09 -0500 Received: from merlin.infradead.org ([205.233.59.134]:41890 "EHLO merlin.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751301Ab3KSQRH (ORCPT ); Tue, 19 Nov 2013 11:17:07 -0500 Message-ID: <528B8EE6.7030005@kernel.dk> Date: Tue, 19 Nov 2013 09:16:38 -0700 From: Jens Axboe User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.1.0 MIME-Version: 1.0 To: Christoph Hellwig CC: Dave Chinner , linux-kernel@vger.kernel.org Subject: Re: [Regression x2, 3.13-git] virtio block mq hang, iostat busted on virtio devices References: <20131119080218.GJ11434@dastard> <20131119103635.GA9223@infradead.org> <20131119160509.GB21343@kernel.dk> <20131119160949.GA24191@infradead.org> In-Reply-To: <20131119160949.GA24191@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 11/19/2013 09:09 AM, Christoph Hellwig wrote: > On Tue, Nov 19, 2013 at 09:05:09AM -0700, Jens Axboe wrote: >> I have queued up the below. > > I don't think that will work. We only set QUEUE_FLAG_IO_STAT in > blk_init_allocated_queue, which doesn't get called from > blk_mq_init_queue. That being said I really don't understand why we > need to have the QUEUE_FLAG_IO_STAT flag at all. Hmm good point, we need to init those flags there too. I'll get this properly tested. Back when the accounting was initially made optional, there was about a 1% reduction in overhead with it off. That's why the tunable is there. -- Jens Axboe