From: Xiaoguang Wang Subject: Re: [PATCH] ext4: save goal group and offset in struct ext4_allocation_context.ac_g_ex Date: Fri, 1 Aug 2014 09:13:49 +0800 Message-ID: <53DAE9CD.5090308@cn.fujitsu.com> References: <1406108861-2581-1-git-send-email-wangxg.fnst@cn.fujitsu.com> <20140729135819.GA2256@thunk.org> <20140731153255.GE1566@thunk.org> Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: 7bit Cc: , , To: "Theodore Ts'o" Return-path: Received: from cn.fujitsu.com ([59.151.112.132]:41300 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1753421AbaHABOi (ORCPT ); Thu, 31 Jul 2014 21:14:38 -0400 In-Reply-To: <20140731153255.GE1566@thunk.org> Sender: linux-ext4-owner@vger.kernel.org List-ID: Hi Ted, On 07/31/2014 11:32 PM, Theodore Ts'o wrote: > On Tue, Jul 29, 2014 at 09:58:19AM -0400, Theodore Ts'o wrote: >> On Wed, Jul 23, 2014 at 05:47:41PM +0800, Xiaoguang Wang wrote: >>> In ext4_mb_normalize_request(), if ac_g_ex.fe_logical is adjacent to the closest logical >>> allocated block to the left or (ac_g_ex.fe_logical+len) adjacent to the closest logical >>> allocated block to the right, we'll attach EXT4_MB_HINT_TRY_GOAL flag taking the physical >>> block (ext4_allocation_request.lleft+1) or (ext4_allocation_request.pright-len) as a goal, >>> and put this information in ext4_allocation_context.ac_f_ex. >>> >>> But look at the ext4_mb_find_by_goal(), indeed it use ac_g_ex to look up, so this is wrong, >>> we should save goal group and offset in struct ext4_allocation_context.ac_g_ex. >>> >>> Signed-off-by: Xiaoguang Wang >> >> Nice catch! >> >> Thanks, applied. > > I've had to drop this patch, as it is causing xfstests failures for > generic/074. Before sending this patch, I didn't run the whole xfstests for ext4, sorry. Next time, if I'll send some new patches, I'll run the xfstests for ext4. > > I'm not sure why, and right now I don't have time to investigate. If > someone who has time and experience wading into fs/ext4/mballoc.c, it > would be great if someone could take a closer look. I'll investigate it and have full tests, thanks! Regards, Xiaoguang Wang > > - Ted > . >