From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx144.netapp.com ([216.240.21.25]:65262 "EHLO mx144.netapp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753793AbbLCUng (ORCPT ); Thu, 3 Dec 2015 15:43:36 -0500 Subject: Re: sys_copy_file_range tests To: Christoph Hellwig References: <20151201181011.GA6202@infradead.org> CC: , From: Anna Schumaker Message-ID: <5660A974.3010200@Netapp.com> Date: Thu, 3 Dec 2015 15:43:32 -0500 MIME-Version: 1.0 In-Reply-To: <20151201181011.GA6202@infradead.org> Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Sender: linux-fsdevel-owner@vger.kernel.org List-ID: Hi Christoph, On 12/01/2015 01:10 PM, Christoph Hellwig wrote: > Hi Anna, > > I recently looked into allowing sys_copy_file_range to use the > clone_file_range operation so that file systems don't have to support > two methods if they only support clone-like features, and one thing > I noticed is that unlike for clone there are no tests for > clone_file_range in xfstests, and I couldn't find a testsuite anywhere > else either. Do you have any good tests available to verify the > implementation? > Sadly, I don't have very good tests and I was mostly using the example program from the man page under various filesystems and setups. At LPC, Darrick made it sound like he was going to work on adding vfs_copy_file_range() tests to xfstests but that probably turned into the reflink/dedupe ioctl tests. I can try to find time to write something for xfstests if that would be useful! Anna