From patchwork Fri Sep 29 05:37:39 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Darrick J. Wong" X-Patchwork-Id: 9977109 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id ECD3660329 for ; Fri, 29 Sep 2017 05:37:46 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id DDC0D295E2 for ; Fri, 29 Sep 2017 05:37:46 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id D24B8297E8; Fri, 29 Sep 2017 05:37:46 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 946DC297E4 for ; Fri, 29 Sep 2017 05:37:45 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751822AbdI2Fho (ORCPT ); Fri, 29 Sep 2017 01:37:44 -0400 Received: from aserp1040.oracle.com ([141.146.126.69]:43871 "EHLO aserp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750996AbdI2Fhn (ORCPT ); Fri, 29 Sep 2017 01:37:43 -0400 Received: from aserv0022.oracle.com (aserv0022.oracle.com [141.146.126.234]) by aserp1040.oracle.com (Sentrion-MTA-4.3.2/Sentrion-MTA-4.3.2) with ESMTP id v8T5bfFt023338 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Fri, 29 Sep 2017 05:37:41 GMT Received: from aserv0121.oracle.com (aserv0121.oracle.com [141.146.126.235]) by aserv0022.oracle.com (8.14.4/8.14.4) with ESMTP id v8T5bexU024862 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Fri, 29 Sep 2017 05:37:40 GMT Received: from abhmp0015.oracle.com (abhmp0015.oracle.com [141.146.116.21]) by aserv0121.oracle.com (8.14.4/8.13.8) with ESMTP id v8T5beFq013202; Fri, 29 Sep 2017 05:37:40 GMT Received: from localhost (/73.25.142.12) by default (Oracle Beehive Gateway v4.0) with ESMTP ; Thu, 28 Sep 2017 22:37:40 -0700 Date: Thu, 28 Sep 2017 22:37:39 -0700 From: "Darrick J. Wong" To: Eryu Guan Cc: Eric Sandeen , xfs , fstests Subject: [PATCH v2] xfs: test increased overlong directory extent discard threshold Message-ID: <20170929053739.GW5020@magnolia> References: <20170927010348.GK5020@magnolia> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <20170927010348.GK5020@magnolia> User-Agent: Mutt/1.5.24 (2015-08-30) X-Source-IP: aserv0022.oracle.com [141.146.126.234] Sender: linux-xfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-xfs@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP As of 2007, metadump has an interesting "feature" where it discards directory extents that are longer than 1000 (originally 20) blocks. This ostensibly was to protect metadump from corrupt bmbt records, but it also has the effect of omitting from the metadump valid long extents. The end result is that we create incomplete metadumps, which is exacerbated by the lack of warning unless -w is passed. So now that we've fixed the default threshold to MAXEXTLEN, check that the installed metadump no longer exhibits this behavior. Signed-off-by: Darrick J. Wong --- v2: work out the maths to capture why we're using mkfs options --- tests/xfs/707 | 117 +++++++++++++++++++++++++++++++++++++++++++++++++++++ tests/xfs/707.out | 6 +++ tests/xfs/group | 1 3 files changed, 124 insertions(+) create mode 100755 tests/xfs/707 create mode 100644 tests/xfs/707.out -- To unsubscribe from this list: send the line "unsubscribe linux-xfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/tests/xfs/707 b/tests/xfs/707 new file mode 100755 index 0000000..f7a90d0 --- /dev/null +++ b/tests/xfs/707 @@ -0,0 +1,117 @@ +#! /bin/bash +# FS QA Test No. 707 +# +# Ensure that metadump copies large directory extents +# +# Metadump helpfully discards directory (and xattr) extents that are +# longer than 1000 blocks. This is a little silly since a hardlink farm +# can easily create such a monster. +# +# Now that we've upped metadump's default too-long-extent discard +# threshold to 2^21 blocks, make sure we never do that again. +# +#----------------------------------------------------------------------- +# Copyright (c) 2017, Oracle and/or its affiliates. All Rights Reserved. +# +# 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" + +here=`pwd` +tmp=/tmp/$$ +status=1 # failure is the default! +trap "_cleanup; exit \$status" 0 1 2 3 15 + +_cleanup() +{ + cd / + rm -f "$tmp".* $metadump_file $metadump_img +} + +# get standard environment, filters and checks +. ./common/rc +. ./common/filter + +# real QA test starts here +_supported_os Linux +_supported_fs xfs +_require_scratch + +rm -f "$seqres.full" + +echo "Format and mount" +# We need to create a directory with a huuuge extent record. Normally +# a rapidly expanding directory gets its blocks allocated in lockstep -- +# physically we end up writing (a couple of dir data blocks) followed by +# (a da btree block) over and over. +# +# Therefore, we crank the directory block size up to maximum and the +# filesystem down to minimum so that we have to allocate 64 blocks at +# a time, trying to end up with the result that we have > 1000 blocks +# allocated in a single extent. +# +# In theory the math works out here -- ~65500 bytes for a da leaf block / +# 8 bytes per da leaf entry == ~8187 hash entries for a da node. 65500 +# bytes for a dir data block / 264 bytes per dirent == ~248 dirents per +# block. 8187 hashes/dablk / 248 dirents/dirblock = ~33 dirblocks per +# dablock. 33 dirblocks * 64k mean that we can expand a directory by +# 2112k before we have to allocate another da btree block. +_scratch_mkfs -b size=1k -n size=64k > "$seqres.full" 2>&1 +_scratch_mount >> "$seqres.full" 2>&1 + +metadump_file="$TEST_DIR/meta-$seq" +metadump_img="$TEST_DIR/img-$seq" +rm -f $metadump_file $metadump_img +testdir="$SCRATCH_MNT/test-$seq" +max_fname_len=255 +blksz=$(_get_block_size $SCRATCH_MNT) + +# Try to create a directory w/ extents +blocks=1050 +names=$((blocks * (blksz / max_fname_len))) +echo "Create huge dir" +mkdir -p $testdir +touch $SCRATCH_MNT/a +seq 0 $names | while read f; do + name="$testdir/$(printf "%0${max_fname_len}d" $f)" + ln $SCRATCH_MNT/a $name +done +dir_inum=$(stat -c %i $testdir) + +echo "Check for > 1000 block extent?" +_scratch_unmount +check_for_long_extent() { + inum=$1 + + _scratch_xfs_db -x -c "inode $dir_inum" -c bmap | \ + sed -e 's/^.*count \([0-9]*\) flag.*$/\1/g' | \ + awk '{if ($1 > 1000) { printf("yes, %d\n", $1); } }' +} +extlen="$(check_for_long_extent $dir_inum)" +echo "qualifying extent: $extlen blocks" >> $seqres.full +test -n "$extlen" || _notrun "could not create dir extent > 1000 blocks" + +echo "Try to metadump" +_scratch_metadump $metadump_file -w +xfs_mdrestore $metadump_file $metadump_img + +echo "Check restored metadump image" +$XFS_REPAIR_PROG -n $metadump_img >> $seqres.full 2>&1 + +# success, all done +status=0 +exit diff --git a/tests/xfs/707.out b/tests/xfs/707.out new file mode 100644 index 0000000..0d2a222 --- /dev/null +++ b/tests/xfs/707.out @@ -0,0 +1,6 @@ +QA output created by 707 +Format and mount +Create huge dir +Check for > 1000 block extent? +Try to metadump +Check restored metadump image diff --git a/tests/xfs/group b/tests/xfs/group index 7353b9c..a70c884 100644 --- a/tests/xfs/group +++ b/tests/xfs/group @@ -432,3 +432,4 @@ 703 auto quick clone fsr 704 auto quick clone 705 auto quick clone fsr +707 auto quick dir metadata