On 10/17/13 11:48 PM, Eryu Guan wrote: > On Thu, Oct 17, 2013 at 11:16:45PM -0500, Eric Sandeen wrote: >> In some cases selinux's creation of an xattr on the temporary >> fd creates a local xattr, but the file we are trying to >> defragment has attrs in extent format, and the forkoff mismatch >> will cause xfs_fsr to fail. This test demonstrates it; I >> have old patches sent to the list long ago that should fix >> it. I'll resend them soon. >> >> Signed-off-by: Eric Sandeen <sandeen@xxxxxxxxxx> >> --- >> >> diff --git a/tests/xfs/351 b/tests/xfs/351 >> new file mode 100755 >> index 0000000..b21ba7a >> --- /dev/null >> +++ b/tests/xfs/351 >> @@ -0,0 +1,89 @@ >> +#! /bin/bash >> +# FS QA Test No. 350 >> +# >> +# Test xfs_fsr / swapext management of di_forkoff w/ selinux >> +# >> +#----------------------------------------------------------------------- >> +# Copyright (c) 2013 Red Hat, Inc. 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 >> +#----------------------------------------------------------------------- >> +# >> +# creator > > Seems we don't need "creator" in comment now. yup >> + >> +seq=`basename $0` > > Forgot about "seqres=$RESULT_DIR/$seq" ? crud. :) ... >> +# Create an extents-format attribute on a file >> +xfs_io -f -c "pwrite -S 0x63 0 4096" $SCRATCH_MNT/attrvals 2>&1 &> $seqres.full > > Use $XFS_IO_PROG instead? yup > And seems "2>&1" is redundant, "&>" can do the work. heh, yeah. Ok, will tidy up and resend tomorrow. Thanks! -Eric > Thanks, > Eryu Guan > >> +cat $SCRATCH_MNT/attrvals | attr -s name $SCRATCH_MNT/$seq.test 2>&1 &> $seqres.full >> + >> +# Fragment the file by writing backwards >> +for I in `seq 6 -1 0`; do >> + dd if=/dev/zero of=$SCRATCH_MNT/$seq.test seek=$I bs=4k \ >> + oflag=direct count=1 conv=notrunc 2>&1 &> $seqres.full >> +done >> + >> +# See if xfs_fsr works. >> + >> +# The target file now has a non-local attribute, but the >> +# fsr temp file gets assigned a local attribute on creation, >> +# due to selinux. >> + >> +# In this case, the large-ish selinux attr will create a forkoff >> +# in the temp file smaller than the forkoff in the target file, >> +# because the FMT_EXTENTS attr takes up less space. If there is >> +# no mechanism to grow the temp forkoff to match, we can end up >> +# failing to swap these 2 inodes if the result is not enough >> +# data space in the temp inode as a result. >> + >> +$XFS_FSR_PROG -v $SCRATCH_MNT/$seq.test | _filter_scratch >> + >> +# success, all done >> +status=0 >> +exit >> diff --git a/tests/xfs/351.out b/tests/xfs/351.out >> new file mode 100644 >> index 0000000..86cbe5d >> --- /dev/null >> +++ b/tests/xfs/351.out >> @@ -0,0 +1,3 @@ >> +QA output created by 351 >> +SCRATCH_MNT/351.test >> +extents before:6 after:1 DONE SCRATCH_MNT/351.test >> diff --git a/tests/xfs/group b/tests/xfs/group >> index acb92ac..da76770 100644 >> --- a/tests/xfs/group >> +++ b/tests/xfs/group >> @@ -180,4 +180,4 @@ >> 298 auto attr symlink quick >> 299 auto quota >> 350 auto dump quota >> - >> +351 auto fsr >> >> >> _______________________________________________ >> xfs mailing list >> xfs@xxxxxxxxxxx >> http://oss.sgi.com/mailman/listinfo/xfs > > _______________________________________________ > xfs mailing list > xfs@xxxxxxxxxxx > http://oss.sgi.com/mailman/listinfo/xfs > _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
