From patchwork Sun May 3 00:53:34 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Filipe Manana X-Patchwork-Id: 6327091 Return-Path: X-Original-To: patchwork-linux-btrfs@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork1.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork1.web.kernel.org (Postfix) with ESMTP id 35A869F32E for ; Mon, 4 May 2015 14:26:46 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 12FFE20274 for ; Mon, 4 May 2015 14:26:45 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id B737520272 for ; Mon, 4 May 2015 14:26:43 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751272AbbEDO0i (ORCPT ); Mon, 4 May 2015 10:26:38 -0400 Received: from victor.provo.novell.com ([137.65.250.26]:48608 "EHLO prv3-mh.provo.novell.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1750820AbbEDO0h (ORCPT ); Mon, 4 May 2015 10:26:37 -0400 Received: from debian3.lan (prv-ext-foundry1int.gns.novell.com [137.65.251.240]) by prv3-mh.provo.novell.com with ESMTP (NOT encrypted); Mon, 04 May 2015 08:26:27 -0600 From: Filipe Manana To: fstests@vger.kernel.org Cc: linux-btrfs@vger.kernel.org, Filipe Manana Subject: [PATCH RESEND] fstests: regression test for btrfs file range cloning Date: Sun, 3 May 2015 01:53:34 +0100 Message-Id: <1430614414-2752-1-git-send-email-fdmanana@suse.com> X-Mailer: git-send-email 2.1.3 Sender: linux-btrfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-btrfs@vger.kernel.org X-Spam-Status: No, score=-5.6 required=5.0 tests=BAYES_00, DATE_IN_PAST_24_48, RCVD_IN_DNSWL_HI, T_RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP Test btrfs file range cloning with the same file as a source and destination. This tests a specific scenario where the extent layout of the file confused the clone ioctl implementation making it return -EEXIST to userspace. This issue was fixed by the following linux kernel patch: Btrfs: fix range cloning when same inode used as source and destination Signed-off-by: Filipe Manana Reviewed-by: Josef Bacik Reviewed-by: David Sterba Tested-by: David Sterba --- Resending as it was missed during last merge. Rebased on latest master branch, renumbered the test accordingly and included Josef's reviewed-by tag. No changes otherwise. tests/btrfs/093 | 120 ++++++++++++++++++++++++++++++++++++++++++++++++++++ tests/btrfs/093.out | 20 +++++++++ tests/btrfs/group | 1 + 3 files changed, 141 insertions(+) create mode 100755 tests/btrfs/093 create mode 100644 tests/btrfs/093.out diff --git a/tests/btrfs/093 b/tests/btrfs/093 new file mode 100755 index 0000000..643c6be --- /dev/null +++ b/tests/btrfs/093 @@ -0,0 +1,120 @@ +#! /bin/bash +# FS QA Test No. btrfs/093 +# +# Test btrfs file range cloning with the same file as a source and destination. +# +# This tests a specific scenario where the extent layout of the file confused +# the clone ioctl implementation making it return -EEXIST to userspace. +# This issue was fixed by the following linux kernel patch: +# +# Btrfs: fix range cloning when same inode used as source and destination +# +#----------------------------------------------------------------------- +# Copyright (C) 2015 SUSE Linux Products GmbH. All Rights Reserved. +# Author: Filipe Manana +# +# This program is free software; you can redistribute it and/or +# modify it under the terms of the GNU General Public License as +# published by the Free Software Foundation. +# +# This program is distributed in the hope that it would be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write the Free Software Foundation, +# Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA +#----------------------------------------------------------------------- +# + +seq=`basename $0` +seqres=$RESULT_DIR/$seq +echo "QA output created by $seq" + +tmp=/tmp/$$ +status=1 # failure is the default! +trap "_cleanup; exit \$status" 0 1 2 3 15 + +_cleanup() +{ + rm -f $tmp.* +} + +# get standard environment, filters and checks +. ./common/rc +. ./common/filter + +# real QA test starts here +_supported_fs btrfs +_supported_os Linux +_require_scratch +_require_cloner +_need_to_be_root + +rm -f $seqres.full + +# Create a file with an extent layout that confused the btrfs clone ioctl +# implementation. The first extent item that is cloned by the second call +# to the cloner program will have only a trailing part of it referenced by +# a new extent item, since the source offset starts in the middle of that +# extent. This confused the clone ioctl because after inserting this new +# extent item it would immediately after process it again thinking it +# corresponded to an extent that existed before - this made it attempt to +# insert a duplicated extent item pointing to the same extent again, which +# made it return an -EEXIST error to userspace and turn the filesystem to +# readonly mode (since the current transaction got aborted). +test_clone() +{ + local bs=$1 + + $XFS_IO_PROG -f -c "pwrite -S 0xaa $((2 * $bs)) $((2 * $bs))" \ + $SCRATCH_MNT/foo | _filter_xfs_io + + $CLONER_PROG -s $((3 * $bs)) -d $((267 * $bs)) -l 0 $SCRATCH_MNT/foo \ + $SCRATCH_MNT/foo + $CLONER_PROG -s $((217 * $bs)) -d $((95 * $bs)) -l 0 $SCRATCH_MNT/foo \ + $SCRATCH_MNT/foo + + echo "File digest after clone operations using same file as source and destination" + md5sum $SCRATCH_MNT/foo | _filter_scratch + + # Test cloning using different source and destination files for the + # same exact data - it must produce the exact same result as the case + # before. + $XFS_IO_PROG -f -c "pwrite -S 0xaa $((2 * $bs)) $((2 * $bs))" \ + $SCRATCH_MNT/a | _filter_xfs_io + cp $SCRATCH_MNT/a $SCRATCH_MNT/b + + $CLONER_PROG -s $((3 * $bs)) -d $((267 * $bs)) -l 0 $SCRATCH_MNT/a \ + $SCRATCH_MNT/b + + cp $SCRATCH_MNT/b $SCRATCH_MNT/foo2 + $CLONER_PROG -s $((217 * $bs)) -d $((95 * $bs)) -l 0 $SCRATCH_MNT/b \ + $SCRATCH_MNT/foo2 + + echo "File digest after clone operations using different files as source and destination" + md5sum $SCRATCH_MNT/foo2 | _filter_scratch + +} + +# Make sure the test passes offsets and lengths to the btrfs clone ioctl that +# are multiples of the fs block size. Currently the block size on btrfs must +# be a multiple of the page size, so use a 64Kb fs block size in order to be +# able to test on every platform supported by linux. +bs=$((64 * 1024)) + +echo "Testing without the no-holes feature" +_scratch_mkfs "-O ^no-holes -n $bs" >>$seqres.full 2>&1 +_scratch_mount +test_clone $bs +_check_scratch_fs + +echo -e "\nTesting with the no-holes feature" +_scratch_unmount +_scratch_mkfs "-O no-holes -n $bs" >>$seqres.full 2>&1 +_scratch_mount +test_clone $bs + +status=0 +exit diff --git a/tests/btrfs/093.out b/tests/btrfs/093.out new file mode 100644 index 0000000..dd51cfd --- /dev/null +++ b/tests/btrfs/093.out @@ -0,0 +1,20 @@ +QA output created by 093 +Testing without the no-holes feature +wrote 131072/131072 bytes at offset 131072 +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +File digest after clone operations using same file as source and destination +cc3cc722ad761bd10488a0f1232ead19 SCRATCH_MNT/foo +wrote 131072/131072 bytes at offset 131072 +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +File digest after clone operations using different files as source and destination +cc3cc722ad761bd10488a0f1232ead19 SCRATCH_MNT/foo2 + +Testing with the no-holes feature +wrote 131072/131072 bytes at offset 131072 +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +File digest after clone operations using same file as source and destination +cc3cc722ad761bd10488a0f1232ead19 SCRATCH_MNT/foo +wrote 131072/131072 bytes at offset 131072 +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +File digest after clone operations using different files as source and destination +cc3cc722ad761bd10488a0f1232ead19 SCRATCH_MNT/foo2 diff --git a/tests/btrfs/group b/tests/btrfs/group index 8e32b20..a2112ab 100644 --- a/tests/btrfs/group +++ b/tests/btrfs/group @@ -94,3 +94,4 @@ 090 auto quick metadata 091 auto quick qgroup 092 auto quick send +093 auto quick clone