From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cn.fujitsu.com ([59.151.112.132]:61760 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1752771AbaFYKG0 (ORCPT ); Wed, 25 Jun 2014 06:06:26 -0400 Message-ID: <53AA9EDE.8020001@cn.fujitsu.com> Date: Wed, 25 Jun 2014 18:05:18 +0800 From: "gux.fnst" MIME-Version: 1.0 Subject: Re: [PATCH] xfstests: add test for ext4 fallocate collapse range check References: <1403157004-25112-1-git-send-email-gux.fnst@cn.fujitsu.com> <20140619065644.GL4453@dastard> In-Reply-To: <20140619065644.GL4453@dastard> Content-Type: text/plain; charset="UTF-8"; format=flowed Content-Transfer-Encoding: 7bit Sender: xfstests-owner@vger.kernel.org To: Dave Chinner , =?UTF-8?B?THVrw6HFoSBDemVybmVy?= Cc: fstests@vger.kernel.org List-ID: On 06/19/2014 02:56 PM, Dave Chinner wrote: > On Thu, Jun 19, 2014 at 01:50:04PM +0800, Xing Gu wrote: >> This is a test to verify the procedure which checks whether >> offset + len is bigger than i_size is removed from do_fallocate >> in collapse range to the file system, and ensure that i_size is >> not going to change. >> >> Signed-off-by: Xing Gu > > Does the test really need 3 minutes? Actually, this case can easily reproduce the issue below 30s. > > We can't keep adding tests that take minutes to *maybe* hit a > regression. This isn't even testing a regression - the > COLLAPSE_RANGE feature was introduced in 3.15-rc1 and this bug fixed > in 3.15-rc3, so there are no kernels out there that have this > bug in them. > > IOWs, I really don't think we should be the runtime of a typical > test cycle by 5% to exercise this non-bug. Maybe 30s at most, but a > shorter, directly targeted test would be better. > > And FWIW, there's nothing ext4 specific about this test - it should > be in the generic group if we are going to add it. Indeed, it's not ext4 specific. What about modifying this case according to Lukas's comment? Thanks! Regards, Xing Gu > > Cheers, > > Dave. >