From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754950AbaHLUCc (ORCPT ); Tue, 12 Aug 2014 16:02:32 -0400 Received: from seketeli.net ([94.23.218.202]:37947 "EHLO ms.seketeli.fr" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754637AbaHLUCa (ORCPT ); Tue, 12 Aug 2014 16:02:30 -0400 Date: Tue, 12 Aug 2014 22:02:28 +0200 From: Apelete Seketeli To: Chris Ball Cc: Ulf Hansson , H Hartley Sweeten , Lars-Peter Clausen , Wei Yongjun , Alex Smith , Laurent Pinchart , linux-mmc@vger.kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH v6 2/2] mmc: jz4740: prepare next dma transfer in parallel with current transfer Message-ID: <20140812200228.GB23978@tao> References: <1405917465-18185-1-git-send-email-apelete@seketeli.net> <1405917465-18185-3-git-send-email-apelete@seketeli.net> <86y4ut60ha.fsf@void.printf.net> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <86y4ut60ha.fsf@void.printf.net> X-Operating-System: Debian GNU/Linux 7 X-URL: http://apelete.seketeli.net User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi Chris, On Tue, Aug-12-2014 at 07:12:33 PM +0100, Chris Ball wrote: > Hi Apelete, > > On Mon, Jul 21 2014, Apelete Seketeli wrote: > > It should help reduce the impact of DMA preparation overhead on the SD > > card performance. > > Did you do any benchmarking to check that this is true? (If so, it'd > be good to note what performance change you saw in the commit message.) I actually did a few quick benchmarks, posted the results in the cover letter: https://lkml.org/lkml/2014/7/21/11. As you can see the improvement is measurable, although not significant with zcat. Cheers. -- Apelete