Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751521AbdGQWjw (ORCPT ); Mon, 17 Jul 2017 18:39:52 -0400 Received: from mga05.intel.com ([192.55.52.43]:48874 "EHLO mga05.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751341AbdGQWjv (ORCPT ); Mon, 17 Jul 2017 18:39:51 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.40,376,1496127600"; d="scan'208";a="1152482170" Date: Mon, 17 Jul 2017 18:45:51 -0400 From: Keith Busch To: Sinan Kaya Cc: linux-nvme@lists.infradead.org, timur@codeaurora.org, linux-arm-msm@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Jens Axboe , Christoph Hellwig , Sagi Grimberg , linux-kernel@vger.kernel.org Subject: Re: [PATCH] nvme: Acknowledge completion queue on each iteration Message-ID: <20170717224551.GA1496@localhost.localdomain> References: <1500330983-27501-1-git-send-email-okaya@codeaurora.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1500330983-27501-1-git-send-email-okaya@codeaurora.org> User-Agent: Mutt/1.7.1 (2016-10-04) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 506 Lines: 10 On Mon, Jul 17, 2017 at 06:36:23PM -0400, Sinan Kaya wrote: > Code is moving the completion queue doorbell after processing all completed > events and sending callbacks to the block layer on each iteration. > > This is causing a performance drop when a lot of jobs are queued towards > the HW. Move the completion queue doorbell on each loop instead and allow new > jobs to be queued by the HW. That doesn't make sense. Aggregating doorbell writes should be much more efficient for high depth workloads.