Message ID | 20201113112704.28798-5-chandanrlinux@gmail.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | xfs: Tests to check for inode fork extent count overflow detection | expand |
On Fri, Nov 13, 2020 at 04:56:56PM +0530, Chandan Babu R wrote: > This test verifies that XFS does not cause inode fork's extent count to > overflow when punching out an extent. > > Signed-off-by: Chandan Babu R <chandanrlinux@gmail.com> > --- > tests/xfs/524 | 210 ++++++++++++++++++++++++++++++++++++++++++++++ > tests/xfs/524.out | 25 ++++++ > tests/xfs/group | 1 + > 3 files changed, 236 insertions(+) > create mode 100755 tests/xfs/524 > create mode 100644 tests/xfs/524.out > > diff --git a/tests/xfs/524 b/tests/xfs/524 > new file mode 100755 > index 00000000..9e140c99 > --- /dev/null > +++ b/tests/xfs/524 > @@ -0,0 +1,210 @@ > +#! /bin/bash > +# SPDX-License-Identifier: GPL-2.0 > +# Copyright (c) 2020 Chandan Babu R. All Rights Reserved. > +# > +# FS QA Test 524 > +# > +# Verify that XFS does not cause inode fork's extent count to overflow when > +# punching out an extent. > +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 > +. ./common/inject > + > +# remove previous $seqres.full before test > +rm -f $seqres.full > + > +# real QA test starts here > + > +_supported_fs xfs > +_require_scratch > +_require_xfs_debug > +_require_xfs_io_command "finsert" > +_require_xfs_io_command "fcollapse" > +_require_xfs_io_command "fzero" For completeness, should this also be testing funshare? > +_require_test_program "punch-alternating" > +_require_xfs_io_error_injection "reduce_max_iextents" > + > +punch_range() > +{ > + echo "* Fpunch regular file" > + > + echo "Format and mount fs" > + _scratch_mkfs >> $seqres.full > + _scratch_mount >> $seqres.full I don't think you need a fresh format for each functional test. > + > + testfile=$SCRATCH_MNT/testfile > + bsize=$(_get_block_size $SCRATCH_MNT) > + > + nr_blks=30 > + > + echo "Create \$testfile" > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > + -c sync $testfile >> $seqres.full > + > + echo "Inject reduce_max_iextents error tag" > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > + > + echo "fpunch alternating blocks" > + $here/src/punch-alternating $testfile >> $seqres.full 2>&1 > + > + testino=$(stat -c "%i" $testfile) > + > + _scratch_unmount >> $seqres.full > + > + echo "Verify \$testfile's extent count" > + > + nextents=$(_scratch_get_iext_count $testino data || ...and now that I keep seeing this unmount-getiextcount-mount dance, you probably should convert these to grab the extent count info online. --D > + _fail "Unable to obtain inode fork's extent count") > + if (( $nextents > 10 )); then > + echo "Extent count overflow check failed: nextents = $nextents" > + exit 1 > + fi > +} > + > +finsert_range() > +{ > + echo "* Finsert regular file" > + > + echo "Format and mount fs" > + _scratch_mkfs >> $seqres.full > + _scratch_mount >> $seqres.full > + > + testfile=$SCRATCH_MNT/testfile > + bsize=$(_get_block_size $SCRATCH_MNT) > + > + nr_blks=30 > + > + echo "Create \$testfile" > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > + -c sync $testfile >> $seqres.full > + > + echo "Inject reduce_max_iextents error tag" > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > + > + echo "Finsert at every other block offset" > + for i in $(seq 1 2 $((nr_blks - 1))); do > + xfs_io -f -c "finsert $((i * bsize)) $bsize" $testfile \ > + >> $seqres.full 2>&1 > + [[ $? != 0 ]] && break > + done > + > + testino=$(stat -c "%i" $testfile) > + > + _scratch_unmount >> $seqres.full > + > + echo "Verify \$testfile's extent count" > + > + nextents=$(_scratch_get_iext_count $testino data || \ > + _fail "Unable to obtain inode fork's extent count") > + if (( $nextents > 10 )); then > + echo "Extent count overflow check failed: nextents = ${nextents}" > + exit 1 > + fi > +} > + > +fcollapse_range() > +{ > + echo "* Fcollapse regular file" > + > + echo "Format and mount fs" > + _scratch_mkfs >> $seqres.full > + _scratch_mount >> $seqres.full > + > + testfile=$SCRATCH_MNT/testfile > + bsize=$(_get_block_size $SCRATCH_MNT) > + > + nr_blks=30 > + > + echo "Create \$testfile" > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > + -c sync $testfile >> $seqres.full > + > + echo "Inject reduce_max_iextents error tag" > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > + > + echo "Fcollapse at every other block offset" > + for i in $(seq 1 $((nr_blks / 2 - 1))); do > + xfs_io -f -c "fcollapse $((i * bsize)) $bsize" $testfile \ > + >> $seqres.full 2>&1 > + [[ $? != 0 ]] && break > + done > + > + testino=$(stat -c "%i" $testfile) > + > + _scratch_unmount >> $seqres.full > + > + echo "Verify \$testfile's extent count" > + > + nextents=$(_scratch_get_iext_count $testino data || \ > + _fail "Unable to obtain inode fork's extent count") > + if (( $nextents > 10 )); then > + echo "Extent count overflow check failed: nextents = ${nextents}" > + exit 1 > + fi > +} > + > +fzero_range() > +{ > + echo "* Fzero regular file" > + > + echo "Format and mount fs" > + _scratch_mkfs >> $seqres.full > + _scratch_mount >> $seqres.full > + > + testfile=$SCRATCH_MNT/testfile > + bsize=$(_get_block_size $SCRATCH_MNT) > + > + nr_blks=30 > + > + echo "Create \$testfile" > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > + -c sync $testfile >> $seqres.full > + > + echo "Inject reduce_max_iextents error tag" > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > + > + echo "Fzero at every other block offset" > + for i in $(seq 1 2 $((nr_blks - 1))); do > + xfs_io -f -c "fzero $((i * bsize)) $bsize" $testfile \ > + >> $seqres.full 2>&1 > + [[ $? != 0 ]] && break > + done > + > + testino=$(stat -c "%i" $testfile) > + > + _scratch_unmount >> $seqres.full > + > + echo "Verify \$testfile's extent count" > + > + nextents=$(_scratch_get_iext_count $testino data || \ > + _fail "Unable to obtain inode fork's extent count") > + if (( $nextents > 10 )); then > + echo "Extent count overflow check failed: nextents = ${nextents}" > + exit 1 > + fi > +} > + > +punch_range > +finsert_range > +fcollapse_range > +fzero_range > + > +# success, all done > +status=0 > +exit > diff --git a/tests/xfs/524.out b/tests/xfs/524.out > new file mode 100644 > index 00000000..58f7d7ae > --- /dev/null > +++ b/tests/xfs/524.out > @@ -0,0 +1,25 @@ > +QA output created by 524 > +* Fpunch regular file > +Format and mount fs > +Create $testfile > +Inject reduce_max_iextents error tag > +fpunch alternating blocks > +Verify $testfile's extent count > +* Finsert regular file > +Format and mount fs > +Create $testfile > +Inject reduce_max_iextents error tag > +Finsert at every other block offset > +Verify $testfile's extent count > +* Fcollapse regular file > +Format and mount fs > +Create $testfile > +Inject reduce_max_iextents error tag > +Fcollapse at every other block offset > +Verify $testfile's extent count > +* Fzero regular file > +Format and mount fs > +Create $testfile > +Inject reduce_max_iextents error tag > +Fzero at every other block offset > +Verify $testfile's extent count > diff --git a/tests/xfs/group b/tests/xfs/group > index 018c70ef..3fa38c36 100644 > --- a/tests/xfs/group > +++ b/tests/xfs/group > @@ -521,3 +521,4 @@ > 521 auto quick realtime growfs > 522 auto quick quota > 523 auto quick realtime growfs > +524 auto quick punch zero insert collapse > -- > 2.28.0 >
On Saturday 14 November 2020 5:58:12 AM IST Darrick J. Wong wrote: > On Fri, Nov 13, 2020 at 04:56:56PM +0530, Chandan Babu R wrote: > > This test verifies that XFS does not cause inode fork's extent count to > > overflow when punching out an extent. > > > > Signed-off-by: Chandan Babu R <chandanrlinux@gmail.com> > > --- > > tests/xfs/524 | 210 ++++++++++++++++++++++++++++++++++++++++++++++ > > tests/xfs/524.out | 25 ++++++ > > tests/xfs/group | 1 + > > 3 files changed, 236 insertions(+) > > create mode 100755 tests/xfs/524 > > create mode 100644 tests/xfs/524.out > > > > diff --git a/tests/xfs/524 b/tests/xfs/524 > > new file mode 100755 > > index 00000000..9e140c99 > > --- /dev/null > > +++ b/tests/xfs/524 > > @@ -0,0 +1,210 @@ > > +#! /bin/bash > > +# SPDX-License-Identifier: GPL-2.0 > > +# Copyright (c) 2020 Chandan Babu R. All Rights Reserved. > > +# > > +# FS QA Test 524 > > +# > > +# Verify that XFS does not cause inode fork's extent count to overflow when > > +# punching out an extent. > > +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 > > +. ./common/inject > > + > > +# remove previous $seqres.full before test > > +rm -f $seqres.full > > + > > +# real QA test starts here > > + > > +_supported_fs xfs > > +_require_scratch > > +_require_xfs_debug > > +_require_xfs_io_command "finsert" > > +_require_xfs_io_command "fcollapse" > > +_require_xfs_io_command "fzero" > > For completeness, should this also be testing funshare? This script tests the limits imposed by XFS_IEXT_PUNCH_HOLE_CNT. funshare causes xfs_reflink_end_cow_extent() to be invoked. Hence I will add that to the script testing XFS_IEXT_REFLINK_END_COW_CNT. > > > +_require_test_program "punch-alternating" > > +_require_xfs_io_error_injection "reduce_max_iextents" > > + > > +punch_range() > > +{ > > + echo "* Fpunch regular file" > > + > > + echo "Format and mount fs" > > + _scratch_mkfs >> $seqres.full > > + _scratch_mount >> $seqres.full > > I don't think you need a fresh format for each functional test. Yes, you are right. We could have just one mkfs and mount followed by all test operations. > > > + > > + testfile=$SCRATCH_MNT/testfile > > + bsize=$(_get_block_size $SCRATCH_MNT) > > + > > + nr_blks=30 > > + > > + echo "Create \$testfile" > > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > > + -c sync $testfile >> $seqres.full > > + > > + echo "Inject reduce_max_iextents error tag" > > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > > + > > + echo "fpunch alternating blocks" > > + $here/src/punch-alternating $testfile >> $seqres.full 2>&1 > > + > > + testino=$(stat -c "%i" $testfile) > > + > > + _scratch_unmount >> $seqres.full > > + > > + echo "Verify \$testfile's extent count" > > + > > + nextents=$(_scratch_get_iext_count $testino data || > > ...and now that I keep seeing this unmount-getiextcount-mount dance, you > probably should convert these to grab the extent count info online. I agree. I will fix this up. > > --D > > > + _fail "Unable to obtain inode fork's extent count") > > + if (( $nextents > 10 )); then > > + echo "Extent count overflow check failed: nextents = $nextents" > > + exit 1 > > + fi > > +} > > + > > +finsert_range() > > +{ > > + echo "* Finsert regular file" > > + > > + echo "Format and mount fs" > > + _scratch_mkfs >> $seqres.full > > + _scratch_mount >> $seqres.full > > + > > + testfile=$SCRATCH_MNT/testfile > > + bsize=$(_get_block_size $SCRATCH_MNT) > > + > > + nr_blks=30 > > + > > + echo "Create \$testfile" > > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > > + -c sync $testfile >> $seqres.full > > + > > + echo "Inject reduce_max_iextents error tag" > > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > > + > > + echo "Finsert at every other block offset" > > + for i in $(seq 1 2 $((nr_blks - 1))); do > > + xfs_io -f -c "finsert $((i * bsize)) $bsize" $testfile \ > > + >> $seqres.full 2>&1 > > + [[ $? != 0 ]] && break > > + done > > + > > + testino=$(stat -c "%i" $testfile) > > + > > + _scratch_unmount >> $seqres.full > > + > > + echo "Verify \$testfile's extent count" > > + > > + nextents=$(_scratch_get_iext_count $testino data || \ > > + _fail "Unable to obtain inode fork's extent count") > > + if (( $nextents > 10 )); then > > + echo "Extent count overflow check failed: nextents = ${nextents}" > > + exit 1 > > + fi > > +} > > + > > +fcollapse_range() > > +{ > > + echo "* Fcollapse regular file" > > + > > + echo "Format and mount fs" > > + _scratch_mkfs >> $seqres.full > > + _scratch_mount >> $seqres.full > > + > > + testfile=$SCRATCH_MNT/testfile > > + bsize=$(_get_block_size $SCRATCH_MNT) > > + > > + nr_blks=30 > > + > > + echo "Create \$testfile" > > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > > + -c sync $testfile >> $seqres.full > > + > > + echo "Inject reduce_max_iextents error tag" > > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > > + > > + echo "Fcollapse at every other block offset" > > + for i in $(seq 1 $((nr_blks / 2 - 1))); do > > + xfs_io -f -c "fcollapse $((i * bsize)) $bsize" $testfile \ > > + >> $seqres.full 2>&1 > > + [[ $? != 0 ]] && break > > + done > > + > > + testino=$(stat -c "%i" $testfile) > > + > > + _scratch_unmount >> $seqres.full > > + > > + echo "Verify \$testfile's extent count" > > + > > + nextents=$(_scratch_get_iext_count $testino data || \ > > + _fail "Unable to obtain inode fork's extent count") > > + if (( $nextents > 10 )); then > > + echo "Extent count overflow check failed: nextents = ${nextents}" > > + exit 1 > > + fi > > +} > > + > > +fzero_range() > > +{ > > + echo "* Fzero regular file" > > + > > + echo "Format and mount fs" > > + _scratch_mkfs >> $seqres.full > > + _scratch_mount >> $seqres.full > > + > > + testfile=$SCRATCH_MNT/testfile > > + bsize=$(_get_block_size $SCRATCH_MNT) > > + > > + nr_blks=30 > > + > > + echo "Create \$testfile" > > + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ > > + -c sync $testfile >> $seqres.full > > + > > + echo "Inject reduce_max_iextents error tag" > > + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT > > + > > + echo "Fzero at every other block offset" > > + for i in $(seq 1 2 $((nr_blks - 1))); do > > + xfs_io -f -c "fzero $((i * bsize)) $bsize" $testfile \ > > + >> $seqres.full 2>&1 > > + [[ $? != 0 ]] && break > > + done > > + > > + testino=$(stat -c "%i" $testfile) > > + > > + _scratch_unmount >> $seqres.full > > + > > + echo "Verify \$testfile's extent count" > > + > > + nextents=$(_scratch_get_iext_count $testino data || \ > > + _fail "Unable to obtain inode fork's extent count") > > + if (( $nextents > 10 )); then > > + echo "Extent count overflow check failed: nextents = ${nextents}" > > + exit 1 > > + fi > > +} > > + > > +punch_range > > +finsert_range > > +fcollapse_range > > +fzero_range > > + > > +# success, all done > > +status=0 > > +exit > > diff --git a/tests/xfs/524.out b/tests/xfs/524.out > > new file mode 100644 > > index 00000000..58f7d7ae > > --- /dev/null > > +++ b/tests/xfs/524.out > > @@ -0,0 +1,25 @@ > > +QA output created by 524 > > +* Fpunch regular file > > +Format and mount fs > > +Create $testfile > > +Inject reduce_max_iextents error tag > > +fpunch alternating blocks > > +Verify $testfile's extent count > > +* Finsert regular file > > +Format and mount fs > > +Create $testfile > > +Inject reduce_max_iextents error tag > > +Finsert at every other block offset > > +Verify $testfile's extent count > > +* Fcollapse regular file > > +Format and mount fs > > +Create $testfile > > +Inject reduce_max_iextents error tag > > +Fcollapse at every other block offset > > +Verify $testfile's extent count > > +* Fzero regular file > > +Format and mount fs > > +Create $testfile > > +Inject reduce_max_iextents error tag > > +Fzero at every other block offset > > +Verify $testfile's extent count > > diff --git a/tests/xfs/group b/tests/xfs/group > > index 018c70ef..3fa38c36 100644 > > --- a/tests/xfs/group > > +++ b/tests/xfs/group > > @@ -521,3 +521,4 @@ > > 521 auto quick realtime growfs > > 522 auto quick quota > > 523 auto quick realtime growfs > > +524 auto quick punch zero insert collapse >
diff --git a/tests/xfs/524 b/tests/xfs/524 new file mode 100755 index 00000000..9e140c99 --- /dev/null +++ b/tests/xfs/524 @@ -0,0 +1,210 @@ +#! /bin/bash +# SPDX-License-Identifier: GPL-2.0 +# Copyright (c) 2020 Chandan Babu R. All Rights Reserved. +# +# FS QA Test 524 +# +# Verify that XFS does not cause inode fork's extent count to overflow when +# punching out an extent. +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 +. ./common/inject + +# remove previous $seqres.full before test +rm -f $seqres.full + +# real QA test starts here + +_supported_fs xfs +_require_scratch +_require_xfs_debug +_require_xfs_io_command "finsert" +_require_xfs_io_command "fcollapse" +_require_xfs_io_command "fzero" +_require_test_program "punch-alternating" +_require_xfs_io_error_injection "reduce_max_iextents" + +punch_range() +{ + echo "* Fpunch regular file" + + echo "Format and mount fs" + _scratch_mkfs >> $seqres.full + _scratch_mount >> $seqres.full + + testfile=$SCRATCH_MNT/testfile + bsize=$(_get_block_size $SCRATCH_MNT) + + nr_blks=30 + + echo "Create \$testfile" + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ + -c sync $testfile >> $seqres.full + + echo "Inject reduce_max_iextents error tag" + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT + + echo "fpunch alternating blocks" + $here/src/punch-alternating $testfile >> $seqres.full 2>&1 + + testino=$(stat -c "%i" $testfile) + + _scratch_unmount >> $seqres.full + + echo "Verify \$testfile's extent count" + + nextents=$(_scratch_get_iext_count $testino data || + _fail "Unable to obtain inode fork's extent count") + if (( $nextents > 10 )); then + echo "Extent count overflow check failed: nextents = $nextents" + exit 1 + fi +} + +finsert_range() +{ + echo "* Finsert regular file" + + echo "Format and mount fs" + _scratch_mkfs >> $seqres.full + _scratch_mount >> $seqres.full + + testfile=$SCRATCH_MNT/testfile + bsize=$(_get_block_size $SCRATCH_MNT) + + nr_blks=30 + + echo "Create \$testfile" + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ + -c sync $testfile >> $seqres.full + + echo "Inject reduce_max_iextents error tag" + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT + + echo "Finsert at every other block offset" + for i in $(seq 1 2 $((nr_blks - 1))); do + xfs_io -f -c "finsert $((i * bsize)) $bsize" $testfile \ + >> $seqres.full 2>&1 + [[ $? != 0 ]] && break + done + + testino=$(stat -c "%i" $testfile) + + _scratch_unmount >> $seqres.full + + echo "Verify \$testfile's extent count" + + nextents=$(_scratch_get_iext_count $testino data || \ + _fail "Unable to obtain inode fork's extent count") + if (( $nextents > 10 )); then + echo "Extent count overflow check failed: nextents = ${nextents}" + exit 1 + fi +} + +fcollapse_range() +{ + echo "* Fcollapse regular file" + + echo "Format and mount fs" + _scratch_mkfs >> $seqres.full + _scratch_mount >> $seqres.full + + testfile=$SCRATCH_MNT/testfile + bsize=$(_get_block_size $SCRATCH_MNT) + + nr_blks=30 + + echo "Create \$testfile" + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ + -c sync $testfile >> $seqres.full + + echo "Inject reduce_max_iextents error tag" + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT + + echo "Fcollapse at every other block offset" + for i in $(seq 1 $((nr_blks / 2 - 1))); do + xfs_io -f -c "fcollapse $((i * bsize)) $bsize" $testfile \ + >> $seqres.full 2>&1 + [[ $? != 0 ]] && break + done + + testino=$(stat -c "%i" $testfile) + + _scratch_unmount >> $seqres.full + + echo "Verify \$testfile's extent count" + + nextents=$(_scratch_get_iext_count $testino data || \ + _fail "Unable to obtain inode fork's extent count") + if (( $nextents > 10 )); then + echo "Extent count overflow check failed: nextents = ${nextents}" + exit 1 + fi +} + +fzero_range() +{ + echo "* Fzero regular file" + + echo "Format and mount fs" + _scratch_mkfs >> $seqres.full + _scratch_mount >> $seqres.full + + testfile=$SCRATCH_MNT/testfile + bsize=$(_get_block_size $SCRATCH_MNT) + + nr_blks=30 + + echo "Create \$testfile" + xfs_io -f -c "pwrite -b $((nr_blks * bsize)) 0 $((nr_blks * bsize))" \ + -c sync $testfile >> $seqres.full + + echo "Inject reduce_max_iextents error tag" + xfs_io -x -c 'inject reduce_max_iextents' $SCRATCH_MNT + + echo "Fzero at every other block offset" + for i in $(seq 1 2 $((nr_blks - 1))); do + xfs_io -f -c "fzero $((i * bsize)) $bsize" $testfile \ + >> $seqres.full 2>&1 + [[ $? != 0 ]] && break + done + + testino=$(stat -c "%i" $testfile) + + _scratch_unmount >> $seqres.full + + echo "Verify \$testfile's extent count" + + nextents=$(_scratch_get_iext_count $testino data || \ + _fail "Unable to obtain inode fork's extent count") + if (( $nextents > 10 )); then + echo "Extent count overflow check failed: nextents = ${nextents}" + exit 1 + fi +} + +punch_range +finsert_range +fcollapse_range +fzero_range + +# success, all done +status=0 +exit diff --git a/tests/xfs/524.out b/tests/xfs/524.out new file mode 100644 index 00000000..58f7d7ae --- /dev/null +++ b/tests/xfs/524.out @@ -0,0 +1,25 @@ +QA output created by 524 +* Fpunch regular file +Format and mount fs +Create $testfile +Inject reduce_max_iextents error tag +fpunch alternating blocks +Verify $testfile's extent count +* Finsert regular file +Format and mount fs +Create $testfile +Inject reduce_max_iextents error tag +Finsert at every other block offset +Verify $testfile's extent count +* Fcollapse regular file +Format and mount fs +Create $testfile +Inject reduce_max_iextents error tag +Fcollapse at every other block offset +Verify $testfile's extent count +* Fzero regular file +Format and mount fs +Create $testfile +Inject reduce_max_iextents error tag +Fzero at every other block offset +Verify $testfile's extent count diff --git a/tests/xfs/group b/tests/xfs/group index 018c70ef..3fa38c36 100644 --- a/tests/xfs/group +++ b/tests/xfs/group @@ -521,3 +521,4 @@ 521 auto quick realtime growfs 522 auto quick quota 523 auto quick realtime growfs +524 auto quick punch zero insert collapse
This test verifies that XFS does not cause inode fork's extent count to overflow when punching out an extent. Signed-off-by: Chandan Babu R <chandanrlinux@gmail.com> --- tests/xfs/524 | 210 ++++++++++++++++++++++++++++++++++++++++++++++ tests/xfs/524.out | 25 ++++++ tests/xfs/group | 1 + 3 files changed, 236 insertions(+) create mode 100755 tests/xfs/524 create mode 100644 tests/xfs/524.out