| Message ID | 1497534338-11433-1-git-send-email-nborisov@suse.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Thu, Jun 15, 2017 at 04:45:38PM +0300, Nikolay Borisov wrote: > When changing the size of disks/filesystem we should always be > rounding down to a multiple of sectorsize. This is a test for the following > btrfs patche: > > btrfs: Round up values which are written for total_bytes_size > > Signed-off-by: Nikolay Borisov <nborisov@suse.com> [ cc btrfs list ] I'd like to use some help from other btrfs developers on this test :) I'll comment from fstests perspective of view. > --- > tests/btrfs/145 | 147 ++++++++++++++++++++++++++++++++++++++++++++++++++++ > tests/btrfs/145.out | 20 +++++++ > tests/btrfs/group | 1 + > 3 files changed, 168 insertions(+) > create mode 100755 tests/btrfs/145 > create mode 100644 tests/btrfs/145.out > > diff --git a/tests/btrfs/145 b/tests/btrfs/145 > new file mode 100755 > index 0000000..4461b37 > --- /dev/null > +++ b/tests/btrfs/145 > @@ -0,0 +1,147 @@ > +#! /bin/bash > +# FS QA Test No. btrfs/145 > +# Trailing whitespace issue in above line, and many other lines in this test, please remove them all. > +# Test that various code paths which deal with modifying the total size > +# of devices/superblock correctly round the value to a multiple of > +# sector size > +# > +#----------------------------------------------------------------------- > +# > +# Copyright (C) 2017 SUSE Linux Products GmbH. All Rights Reserved. > +# Author: Nikolay Borisov <nborisov@suse.com> > +# > +# 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() > +{ > + $UMOUNT_PROG $loop_mnt > + _destroy_loop_device $loop_dev1 > + _destroy_loop_device $loop_dev2 > + cd / > + 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_loop > +_require_block_device $SCRATCH_DEV I see 'btrfs inspect-internal dump-super' is used, I think we need _require_btrfs_command inspect-internal dump-super > + > +rm -f $seqres.full > + > +_scratch_mkfs >>$seqres.full 2>&1 > +_scratch_mount > + > + > +#Size of devices which are going to be half a page larger than > +#default sectorsize (4kb) Nitpick, please add a space after '#' for comments. > +expectedsize=$(( 1 * 1024 * 1024 * 1024 )) > +filesize=$(( $expectedsize + 2048 )) > +loop_mnt=$SCRATCH_MNT/mount > +fs_img1=$SCRATCH_MNT/disk1.img > +fs_img2=$SCRATCH_MNT/disk2.img > + > +mkdir $loop_mnt > + > +#create files to hold dummy devices > +$XFS_IO_PROG -f -c "falloc 0 $filesize" $fs_img1 &> /dev/null > +$XFS_IO_PROG -f -c "falloc 0 $filesize" $fs_img2 &> /dev/null Perhaps we need to make sure $SCRATCH_MNT has enough free space to hold the two fs imgs. _require_fs_space $SCRATCH_MNT <size_in_kb> > + > +loop_dev1=$(_create_loop_device $fs_img1) > +loop_dev2=$(_create_loop_device $fs_img2) > + > +#create fs only on the first device > +_mkfs_dev $loop_dev1 > +_mount $loop_dev1 $loop_mnt > + > +echo "Size from mkfs" > +$BTRFS_UTIL_PROG inspect-internal dump-super /dev/loop0 | grep total ^^^^^^^^^^ $loop_dev1 > + > +#resize down to 768mb + 2k > +$BTRFS_UTIL_PROG fi resize 824182784 $loop_mnt >>$seqres.full 2>&1 Better to use full subcommand name, i.e. s/fi/filesystem/, I remembered there was a patch to do this 2fd80ca btrfs: use full subcommand names > +sync > + > +echo "" > + > +echo "Size after resizing down" > +$BTRFS_UTIL_PROG inspect-internal dump-super $loop_dev1 | grep total > + > +echo "" > + > +#now resize back up to 1 gb > +$BTRFS_UTIL_PROG fi resize max $loop_mnt >>$seqres.full 2>&1 > +sync > + > +echo "Size after resizing up" > +$BTRFS_UTIL_PROG inspect-internal dump-super /dev/loop0 | grep total > + > +#Add fs load for later balance operations, we need to do this > +#before adding a second device > +$FSSTRESS_PROG -w -n 15000 -p 2 -d $loop_mnt >> $seqres.full 2>&1 > + > +#add second unaligned device, this checks the btrfs_init_new_device codepath > +#device should be aligned > +$BTRFS_UTIL_PROG device add $loop_dev2 $loop_mnt >>$seqres.full 2>&1 > +sync > + > +echo "" > + > +#Ensure that adding a device doesn't cause the superbock to be > +#unaligned > +echo "Size of superblock after device addition" > +$BTRFS_UTIL_PROG inspect-internal dump-super $loop_dev1 | grep total > + > +echo "" > + > +#The newly added device must also be aligned > +echo "Size of superblock/device for second device" > +$BTRFS_UTIL_PROG inspect-internal dump-super $loop_dev2 | grep total > + > +#now start a balance, this might produce warnings while > +#it's relocating chunks and updating the size of devices > +$BTRFS_UTIL_PROG balance start $loop_mnt >> $seqres.full 2>&1 > +sync > + > +#delete everything and then run balance > +rm -rf $loop_mnt/* > + > +#do a bit more stress testing until we fill the fs > +$FSSTRESS_PROG -w -n 15000 -p 4 -d $loop_mnt >> $seqres.full 2>&1 > +rm -rf $loop_mnt/p0 $loop_mnt/p2 > + > +#run balance again, this time we will be freeing chunks > +$BTRFS_UTIL_PROG balance start $loop_mnt >> $seqres.full 2>&1 > +sync > + > +#Warnings are produced if we try to store unaligned values > +_check_dmesg By default, the test harness will do this after each test, so there's no need to call _check_dmesg explicitly in test. Thanks, Eryu > + > +status=$? > +exit > diff --git a/tests/btrfs/145.out b/tests/btrfs/145.out > new file mode 100644 > index 0000000..7340db2 > --- /dev/null > +++ b/tests/btrfs/145.out > @@ -0,0 +1,20 @@ > +QA output created by 145 > +Size from mkfs > +total_bytes 1073741824 > +dev_item.total_bytes 1073741824 > + > +Size after resizing down > +total_bytes 824180736 > +dev_item.total_bytes 824180736 > + > +Size after resizing up > +total_bytes 1073741824 > +dev_item.total_bytes 1073741824 > + > +Size of superblock after device addition > +total_bytes 2147483648 > +dev_item.total_bytes 1073741824 > + > +Size of superblock/device for second device > +total_bytes 2147483648 > +dev_item.total_bytes 1073741824 > diff --git a/tests/btrfs/group b/tests/btrfs/group > index be05487..6c930d4 100644 > --- a/tests/btrfs/group > +++ b/tests/btrfs/group > @@ -147,3 +147,4 @@ > 142 auto quick > 143 auto quick > 144 auto quick send > +145 auto quick > -- > 2.7.4 > > -- > 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 -- 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
diff --git a/tests/btrfs/145 b/tests/btrfs/145 new file mode 100755 index 0000000..4461b37 --- /dev/null +++ b/tests/btrfs/145 @@ -0,0 +1,147 @@ +#! /bin/bash +# FS QA Test No. btrfs/145 +# +# Test that various code paths which deal with modifying the total size +# of devices/superblock correctly round the value to a multiple of +# sector size +# +#----------------------------------------------------------------------- +# +# Copyright (C) 2017 SUSE Linux Products GmbH. All Rights Reserved. +# Author: Nikolay Borisov <nborisov@suse.com> +# +# 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() +{ + $UMOUNT_PROG $loop_mnt + _destroy_loop_device $loop_dev1 + _destroy_loop_device $loop_dev2 + cd / + 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_loop +_require_block_device $SCRATCH_DEV + +rm -f $seqres.full + +_scratch_mkfs >>$seqres.full 2>&1 +_scratch_mount + + +#Size of devices which are going to be half a page larger than +#default sectorsize (4kb) +expectedsize=$(( 1 * 1024 * 1024 * 1024 )) +filesize=$(( $expectedsize + 2048 )) +loop_mnt=$SCRATCH_MNT/mount +fs_img1=$SCRATCH_MNT/disk1.img +fs_img2=$SCRATCH_MNT/disk2.img + +mkdir $loop_mnt + +#create files to hold dummy devices +$XFS_IO_PROG -f -c "falloc 0 $filesize" $fs_img1 &> /dev/null +$XFS_IO_PROG -f -c "falloc 0 $filesize" $fs_img2 &> /dev/null + +loop_dev1=$(_create_loop_device $fs_img1) +loop_dev2=$(_create_loop_device $fs_img2) + +#create fs only on the first device +_mkfs_dev $loop_dev1 +_mount $loop_dev1 $loop_mnt + +echo "Size from mkfs" +$BTRFS_UTIL_PROG inspect-internal dump-super /dev/loop0 | grep total + +#resize down to 768mb + 2k +$BTRFS_UTIL_PROG fi resize 824182784 $loop_mnt >>$seqres.full 2>&1 +sync + +echo "" + +echo "Size after resizing down" +$BTRFS_UTIL_PROG inspect-internal dump-super $loop_dev1 | grep total + +echo "" + +#now resize back up to 1 gb +$BTRFS_UTIL_PROG fi resize max $loop_mnt >>$seqres.full 2>&1 +sync + +echo "Size after resizing up" +$BTRFS_UTIL_PROG inspect-internal dump-super /dev/loop0 | grep total + +#Add fs load for later balance operations, we need to do this +#before adding a second device +$FSSTRESS_PROG -w -n 15000 -p 2 -d $loop_mnt >> $seqres.full 2>&1 + +#add second unaligned device, this checks the btrfs_init_new_device codepath +#device should be aligned +$BTRFS_UTIL_PROG device add $loop_dev2 $loop_mnt >>$seqres.full 2>&1 +sync + +echo "" + +#Ensure that adding a device doesn't cause the superbock to be +#unaligned +echo "Size of superblock after device addition" +$BTRFS_UTIL_PROG inspect-internal dump-super $loop_dev1 | grep total + +echo "" + +#The newly added device must also be aligned +echo "Size of superblock/device for second device" +$BTRFS_UTIL_PROG inspect-internal dump-super $loop_dev2 | grep total + +#now start a balance, this might produce warnings while +#it's relocating chunks and updating the size of devices +$BTRFS_UTIL_PROG balance start $loop_mnt >> $seqres.full 2>&1 +sync + +#delete everything and then run balance +rm -rf $loop_mnt/* + +#do a bit more stress testing until we fill the fs +$FSSTRESS_PROG -w -n 15000 -p 4 -d $loop_mnt >> $seqres.full 2>&1 +rm -rf $loop_mnt/p0 $loop_mnt/p2 + +#run balance again, this time we will be freeing chunks +$BTRFS_UTIL_PROG balance start $loop_mnt >> $seqres.full 2>&1 +sync + +#Warnings are produced if we try to store unaligned values +_check_dmesg + +status=$? +exit diff --git a/tests/btrfs/145.out b/tests/btrfs/145.out new file mode 100644 index 0000000..7340db2 --- /dev/null +++ b/tests/btrfs/145.out @@ -0,0 +1,20 @@ +QA output created by 145 +Size from mkfs +total_bytes 1073741824 +dev_item.total_bytes 1073741824 + +Size after resizing down +total_bytes 824180736 +dev_item.total_bytes 824180736 + +Size after resizing up +total_bytes 1073741824 +dev_item.total_bytes 1073741824 + +Size of superblock after device addition +total_bytes 2147483648 +dev_item.total_bytes 1073741824 + +Size of superblock/device for second device +total_bytes 2147483648 +dev_item.total_bytes 1073741824 diff --git a/tests/btrfs/group b/tests/btrfs/group index be05487..6c930d4 100644 --- a/tests/btrfs/group +++ b/tests/btrfs/group @@ -147,3 +147,4 @@ 142 auto quick 143 auto quick 144 auto quick send +145 auto quick
When changing the size of disks/filesystem we should always be rounding down to a multiple of sectorsize. This is a test for the following btrfs patche: btrfs: Round up values which are written for total_bytes_size Signed-off-by: Nikolay Borisov <nborisov@suse.com> --- tests/btrfs/145 | 147 ++++++++++++++++++++++++++++++++++++++++++++++++++++ tests/btrfs/145.out | 20 +++++++ tests/btrfs/group | 1 + 3 files changed, 168 insertions(+) create mode 100755 tests/btrfs/145 create mode 100644 tests/btrfs/145.out