From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from userp1040.oracle.com ([156.151.31.81]:16443 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751690AbbHMDWX (ORCPT ); Wed, 12 Aug 2015 23:22:23 -0400 Date: Thu, 13 Aug 2015 11:22:04 +0800 From: Liu Bo To: Dave Chinner Cc: fstests@vger.kernel.org, linux-btrfs@vger.kernel.org Subject: Re: [PATCH] fstests: btrfs regression test for defrag tail extents Message-ID: <20150813032203.GB920@localhost.localdomain> Reply-To: bo.li.liu@oracle.com References: <1439194379-29290-1-git-send-email-bo.li.liu@oracle.com> <20150811023201.GL3902@dastard> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <20150811023201.GL3902@dastard> Sender: linux-btrfs-owner@vger.kernel.org List-ID: On Tue, Aug 11, 2015 at 12:32:01PM +1000, Dave Chinner wrote: > On Mon, Aug 10, 2015 at 04:12:59PM +0800, Liu Bo wrote: > > Regression test for btrfs defragment tool, it's aimed to verify > > that tail extents won't be skipped as a separate extent while the previous > > extents have been defrag'ed into a whole extent. > > > > Signed-off-by: Liu Bo > > I don't see anything btrfs specific in this test, so it belongs in > tests/generic/ > > Also, how is this different to generic/018 testing fragmented files > defrag back to one extent? Thanks for reminding me of this, it's in fact quite similar to one of the cases in generic/018, and the difference is that btrfs has an extent length thresh (256K) for deciding if the extent is big enough (no need to defrag). That said, I'd like to add this into generic/018 to keep it simple. Thanks, -liubo