From: =?ISO-8859-15?Q?Luk=E1=A8_Czerner?= Subject: Re: [PATCH] generic: test data corruption on ext4 caused by written/delayed extent Date: Fri, 24 Apr 2015 12:46:10 +0200 (CEST) Message-ID: References: <1429714555-22133-1-git-send-email-lczerner@redhat.com> <20150423034909.GS17534@dhcp-13-216.nay.redhat.com> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Cc: fstests@vger.kernel.org, linux-ext4@vger.kernel.org To: Eryu Guan Return-path: In-Reply-To: <20150423034909.GS17534@dhcp-13-216.nay.redhat.com> Sender: fstests-owner@vger.kernel.org List-Id: linux-ext4.vger.kernel.org On Thu, 23 Apr 2015, Eryu Guan wrote: > Date: Thu, 23 Apr 2015 11:49:09 +0800 > From: Eryu Guan > To: Lukas Czerner > Cc: fstests@vger.kernel.org, linux-ext4@vger.kernel.org > Subject: Re: [PATCH] generic: test data corruption on ext4 caused by > written/delayed extent > > On Wed, Apr 22, 2015 at 04:55:55PM +0200, Lukas Czerner wrote: > > This test excercises the problem with unwritten and delayed extents > > in ext4 extent status tree where we might in some cases lose a block > > worth of data. Even though this was a ext4 specific problem the > > reproducer can be easily tun on any file system so let's do that just > > in case. > > > > This tests excercises the problem fixed in kernel with commit > > "ext4: Fix data corruption caused by unwritten and delayed extents" > > > > Signed-off-by: Lukas Czerner > > --- > > tests/generic/082 | 84 +++++++++++++++++++++++++++++++++++++++++++++++++++ > > tests/generic/082.out | 14 +++++++++ > > 2 files changed, 98 insertions(+) > > create mode 100755 tests/generic/082 > > create mode 100644 tests/generic/082.out > > update to tests/generic/group file is missing > > > > > diff --git a/tests/generic/082 b/tests/generic/082 > > new file mode 100755 > > index 0000000..2fa22c6 > > --- /dev/null > > +++ b/tests/generic/082 > > @@ -0,0 +1,84 @@ > > +#! /bin/bash > > +# FS QA Test 082 > > +# > > +# This test excercises the problem with unwritten and delayed extents > > +# in ext4 extent status tree where we might in some cases lose a block > > +# worth of data. Even though this was a ext4 specific problem the > > +# reproducer can be easily tun on any file system so let's do that just > > +# in case. > > +# > > +# This tests excercises the problem fixed in kernel with commit > > +# "ext4: Fix data corruption caused by unwritten and delayed extents" > > +# > > +#----------------------------------------------------------------------- > > +# Copyright (c) 2015 Red Hat, Inc., Lukas Czerner > > +# > > +# 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 $tmp.* > > +} > > + > > +# get standard environment, filters and checks > > +. ./common/rc > > +. ./common/filter > > + > > +# real QA test starts here > > + > > +# Modify as appropriate. > > +_supported_fs generic > > +_supported_os IRIX Linux > > +_require_test > > _require_xfs_io_command "falloc" too > > I saw failure on cifs because it doesn't support fallocate so the file > size is not extended, and hexdump output differs > > > + > > +test_file=${TEST_DIR}/testfile-082 > > use $seq not hardcoded 082? > > > + > > +rm -f $test_file > > rm -f $seqres.full too, otherwise the file keeps growing Will resend. Thanks! -Lukas > > Thanks, > Eryu > > + > > +# The first write creates a delayed extent, fallocate creates > > +# unwritten extent which will be marked as delayed in ext4 > > +# extent status tree. Second write will convert unwritten/delayed > > +# block into written/delayed. > > +$XFS_IO_PROG -f -c "pwrite -S 0xaa 4096 2048" \ > > + -c "falloc 0 131072" \ > > + -c "pwrite -S 0xbb 65536 2048" \ > > + -c "fsync" $test_file >> $seqres.full 2>&1 > > + > > +# Drop the caches to evict dirty buffers from memory > > +echo 3 > /proc/sys/vm/drop_caches > > + > > +# Write into the second part of the block with 0xbb write from before > > +# will create new empty! buffer because the block is still marked as > > +# delayed even though it's already written - resulting in > > +# overwriting previous data. > > +$XFS_IO_PROG -c "pwrite -S 0xdd 67584 2048" $test_file >> $seqres.full 2>&1 > > + > > +# On a faulty ext4 oxbb data will be missing, overwritten by zeroes. > > +hexdump -C $test_file > > + > > +# success, all done > > +status=0 > > +exit > > diff --git a/tests/generic/082.out b/tests/generic/082.out > > new file mode 100644 > > index 0000000..ef28b9a > > --- /dev/null > > +++ b/tests/generic/082.out > > @@ -0,0 +1,14 @@ > > +QA output created by 082 > > +00000000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| > > +* > > +00001000 aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa aa |................| > > +* > > +00001800 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| > > +* > > +00010000 bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb bb |................| > > +* > > +00010800 dd dd dd dd dd dd dd dd dd dd dd dd dd dd dd dd |................| > > +* > > +00011000 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 |................| > > +* > > +00020000 > > -- > > 1.8.3.1 > > > > -- > > To unsubscribe from this list: send the line "unsubscribe fstests" in > > the body of a message to majordomo@vger.kernel.org > > More majordomo info at http://vger.kernel.org/majordomo-info.html >