From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH v3 0/3] provide sdhci driver for mmp2 and pxa910 Date: Fri, 3 Jun 2011 18:20:51 +0200 Message-ID: <201106031820.51922.arnd@arndb.de> References: Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from moutng.kundenserver.de ([212.227.17.9]:56598 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752716Ab1FCQVF (ORCPT ); Fri, 3 Jun 2011 12:21:05 -0400 In-Reply-To: Sender: linux-mmc-owner@vger.kernel.org List-Id: linux-mmc@vger.kernel.org To: zhangfei gao Cc: Nicolas Pitre , Philip Rakity , Wolfram Sang , Chris Ball , "linux-mmc@vger.kernel.org" , Jun Nie , Raymond Wu , Haojian Zhuang , Shawn Guo , Qiming Wu , Eric Miao , linux-arm-kernel On Friday 03 June 2011, zhangfei gao wrote: > Here are patches to provide separate sdhci driver for mmp2 and pxa910 > based on sdhci-platfm. > Rename to sdhci-pxav3.c and sdhci-pxav2.c > Remove sdhci-pxa.c, which is used to share among pxa serious, since > under this method, platform difference have to be put under arch/arm, > which is not easy to track. > As a result, mmp2 mmc resource should be updated accordingly. Looks all good content-wise. I do have a few comments regarding the style of submission, please follow that the next time: * Make all patches a reply to the [PATCH 0/x] email, so they show up as a single thread. This helps finding the emails when your mail client sorts threads by last reply. The option in git-send-email is '--thread --no-chain-reply'. * The changelog has extra spaces in it that shouldn't be there. I assume they come from using git-show instead of git-format-patch to create the emails. * Make sure a series is bisectable. This means that each patch results in a working kernel without regressions when you apply only part of the series. There are multiple ways to get there. In this case, you could fold the patch that updates the resources into the patch that changes the format. Alternatively, you could first create the new driver in the new format, then change the resources, and finally remove the now obsolete driver. Arnd