[PATCH] fstests: btrfs: Add test case to check if btrfs can handle full fs trim correctly

[PATCH] fstests: btrfs: Add test case to check if btrfs can handle full fs trim correctly

[Qu Wenruo]

Ancient commit f4c697e6406d ("btrfs: return EINVAL if start > total_bytes in
fitrim ioctl") introduced a regression where btrfs may fail to trim any
free space in existing block groups.

It's caused by confusion with btrfs_super_block->total_bytes and btrfs
logical address space.
Unlike physical address, any aligned bytenr in range [0, U64_MAX) is
valid in btrfs logical address space, and it's chunk mapping mechanism
of btrfs to handle the logical<->physical mapping.

The test case will craft a btrfs with the following features:
0) Single data/meta profile
   Make trimmed bytes reporting and chunk allocation more predictable.

1) All chunks start beyond super_block->total_bytes (1G)
   By relocating these blocks several times.

2) Unallocated space is less than 50% of the whole fs

3) Fragmented data chunks
   Data chunks will be full of fragments, 50% of data chunks will be
   free space.

So in theory fstrim should be able to trim over 50% space of the fs.
(after fix, 64% of the fs can be trimmed)
While the regression makes btrfs only able to trim unallocated space,
which is less than 50% of the total space.
(without fix, it's only 31%)

Fixed by patch named "btrfs: Ensure btrfs_trim_fs can trim the whole fs".

Signed-off-by: Qu Wenruo <wqu@suse.com>
---
 tests/btrfs/155     | 120 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 tests/btrfs/155.out |   2 +
 tests/btrfs/group   |   1 +
 3 files changed, 123 insertions(+)
 create mode 100755 tests/btrfs/155
 create mode 100644 tests/btrfs/155.out

diff --git a/tests/btrfs/155 b/tests/btrfs/155
new file mode 100755
index 00000000..6918f093
--- /dev/null
+++ b/tests/btrfs/155
@@ -0,0 +1,120 @@
+#! /bin/bash
+# FS QA Test 155
+#
+# Check if btrfs can correctly trim free space in block groups
+#
+# An ancient regression prevent btrfs from trimming free space inside
+# existing block groups, if bytenr of block group starts beyond
+# btrfs_super_block->total_bytes.
+# However all bytenr in btrfs is in btrfs logical address space,
+# where any bytenr in range [0, U64_MAX] is valid.
+#
+# Fixed by patch named "btrfs: Ensure btrfs_trim_fs can trim the whole fs".
+#
+#-----------------------------------------------------------------------
+# Copyright (c) 2017 SUSE Linux Products GmbH.  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.*
+}
+
+# get standard environment, filters and checks
+. ./common/rc
+. ./common/filter
+
+# remove previous $seqres.full before test
+rm -f $seqres.full
+
+# real QA test starts here
+
+# Modify as appropriate.
+_supported_fs btrfs
+_supported_os Linux
+_require_scratch
+_require_fstrim
+
+# 1024fs size
+fs_size=$((1024 * 1024 * 1024))
+
+# Use small files to fill half of the fs
+file_size=$(( 1024 * 1024 ))
+nr_files=$(( $fs_size / $file_size / 2))
+
+# Force to use single data and meta profile.
+# Since the test relies on fstrim output, which will differ for different
+# profiles
+_scratch_mkfs -b $fs_size -m single -d single > /dev/null
+_scratch_mount
+
+_require_batched_discard "$SCRATCH_MNT"
+
+for n in $(seq -w 0 $(( $nr_files - 1))); do
+	$XFS_IO_PROG -f -c "pwrite 0 $file_size" "$SCRATCH_MNT/file_$n" \
+		> /dev/null
+done
+
+# Flush all buffer data into disk, to trigger chunk allocation
+sync
+
+# Now we have take at least 50% of the filesystem, relocate all chunks twice
+# so all chunks will start after 1G in logical space.
+# (Btrfs chunk allocation will not rewind to reuse lower space)
+_run_btrfs_util_prog balance start --full-balance "$SCRATCH_MNT"
+
+# To avoid possible false ENOSPC alert on v4.15-rc1, seems to be a
+# reserved space related bug (maybe related to outstanding space rework?),
+# but that's another story.
+sync
+
+_run_btrfs_util_prog balance start --full-balance "$SCRATCH_MNT"
+
+# Now remove half of the files to make some holes for later trim.
+# While still keep the chunk space fragmented, so no chunk will be freed
+rm $SCRATCH_MNT/file_*[13579] -f
+
+# Make sure space is freed
+sync
+
+trimmed=$($FSTRIM_PROG -v "$SCRATCH_MNT" | _filter_fstrim)
+echo "Trimmed=$trimmed total_size=$fs_size ratio=$(($trimmed * 100 / $fs_size))%" \
+	>> $seqres.full
+
+# For correct full fs trim, both unallocated space (less than 50%)
+# and free space in existing block groups (about 25%) should be trimmed.
+# If less than 50% is trimmed, then only unallocated space is trimmed.
+# BTW, without fix only 31% can be trimmed, while after fix it's 64%.
+if [ $trimmed -lt $(( $fs_size / 2)) ]; then
+	echo "Free space in block groups not trimmed"
+	echo "Trimmed=$trimmed total_size=$fs_size ratio=$(($trimmed * 100 / $fs_size))%"
+fi
+
+echo "Silence is golden"
+# success, all done
+status=0
+exit
diff --git a/tests/btrfs/155.out b/tests/btrfs/155.out
new file mode 100644
index 00000000..d25dcd3b
--- /dev/null
+++ b/tests/btrfs/155.out
@@ -0,0 +1,2 @@
+QA output created by 155
+Silence is golden
diff --git a/tests/btrfs/group b/tests/btrfs/group
index c98cf823..86c7128e 100644
--- a/tests/btrfs/group
+++ b/tests/btrfs/group
@@ -157,3 +157,4 @@
 152 auto quick metadata qgroup send
 153 auto quick qgroup
 154 auto quick
+155 auto trim
-- 
2.14.2

Re: [PATCH] fstests: btrfs: Add test case to check if btrfs can handle full fs trim correctly

[Qu Wenruo]

[-- Attachment #1.1: Type: text/plain, Size: 6672 bytes --]

Ping.

Any comment on this?

Thanks,
Qu

On 2017年11月29日 14:14, Qu Wenruo wrote:
> Ancient commit f4c697e6406d ("btrfs: return EINVAL if start > total_bytes in
> fitrim ioctl") introduced a regression where btrfs may fail to trim any
> free space in existing block groups.
> 
> It's caused by confusion with btrfs_super_block->total_bytes and btrfs
> logical address space.
> Unlike physical address, any aligned bytenr in range [0, U64_MAX) is
> valid in btrfs logical address space, and it's chunk mapping mechanism
> of btrfs to handle the logical<->physical mapping.
> 
> The test case will craft a btrfs with the following features:
> 0) Single data/meta profile
>    Make trimmed bytes reporting and chunk allocation more predictable.
> 
> 1) All chunks start beyond super_block->total_bytes (1G)
>    By relocating these blocks several times.
> 
> 2) Unallocated space is less than 50% of the whole fs
> 
> 3) Fragmented data chunks
>    Data chunks will be full of fragments, 50% of data chunks will be
>    free space.
> 
> So in theory fstrim should be able to trim over 50% space of the fs.
> (after fix, 64% of the fs can be trimmed)
> While the regression makes btrfs only able to trim unallocated space,
> which is less than 50% of the total space.
> (without fix, it's only 31%)
> 
> Fixed by patch named "btrfs: Ensure btrfs_trim_fs can trim the whole fs".
> 
> Signed-off-by: Qu Wenruo <wqu@suse.com>
> ---
>  tests/btrfs/155     | 120 ++++++++++++++++++++++++++++++++++++++++++++++++++++
>  tests/btrfs/155.out |   2 +
>  tests/btrfs/group   |   1 +
>  3 files changed, 123 insertions(+)
>  create mode 100755 tests/btrfs/155
>  create mode 100644 tests/btrfs/155.out
> 
> diff --git a/tests/btrfs/155 b/tests/btrfs/155
> new file mode 100755
> index 00000000..6918f093
> --- /dev/null
> +++ b/tests/btrfs/155
> @@ -0,0 +1,120 @@
> +#! /bin/bash
> +# FS QA Test 155
> +#
> +# Check if btrfs can correctly trim free space in block groups
> +#
> +# An ancient regression prevent btrfs from trimming free space inside
> +# existing block groups, if bytenr of block group starts beyond
> +# btrfs_super_block->total_bytes.
> +# However all bytenr in btrfs is in btrfs logical address space,
> +# where any bytenr in range [0, U64_MAX] is valid.
> +#
> +# Fixed by patch named "btrfs: Ensure btrfs_trim_fs can trim the whole fs".
> +#
> +#-----------------------------------------------------------------------
> +# Copyright (c) 2017 SUSE Linux Products GmbH.  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.*
> +}
> +
> +# get standard environment, filters and checks
> +. ./common/rc
> +. ./common/filter
> +
> +# remove previous $seqres.full before test
> +rm -f $seqres.full
> +
> +# real QA test starts here
> +
> +# Modify as appropriate.
> +_supported_fs btrfs
> +_supported_os Linux
> +_require_scratch
> +_require_fstrim
> +
> +# 1024fs size
> +fs_size=$((1024 * 1024 * 1024))
> +
> +# Use small files to fill half of the fs
> +file_size=$(( 1024 * 1024 ))
> +nr_files=$(( $fs_size / $file_size / 2))
> +
> +# Force to use single data and meta profile.
> +# Since the test relies on fstrim output, which will differ for different
> +# profiles
> +_scratch_mkfs -b $fs_size -m single -d single > /dev/null
> +_scratch_mount
> +
> +_require_batched_discard "$SCRATCH_MNT"
> +
> +for n in $(seq -w 0 $(( $nr_files - 1))); do
> +	$XFS_IO_PROG -f -c "pwrite 0 $file_size" "$SCRATCH_MNT/file_$n" \
> +		> /dev/null
> +done
> +
> +# Flush all buffer data into disk, to trigger chunk allocation
> +sync
> +
> +# Now we have take at least 50% of the filesystem, relocate all chunks twice
> +# so all chunks will start after 1G in logical space.
> +# (Btrfs chunk allocation will not rewind to reuse lower space)
> +_run_btrfs_util_prog balance start --full-balance "$SCRATCH_MNT"
> +
> +# To avoid possible false ENOSPC alert on v4.15-rc1, seems to be a
> +# reserved space related bug (maybe related to outstanding space rework?),
> +# but that's another story.
> +sync
> +
> +_run_btrfs_util_prog balance start --full-balance "$SCRATCH_MNT"
> +
> +# Now remove half of the files to make some holes for later trim.
> +# While still keep the chunk space fragmented, so no chunk will be freed
> +rm $SCRATCH_MNT/file_*[13579] -f
> +
> +# Make sure space is freed
> +sync
> +
> +trimmed=$($FSTRIM_PROG -v "$SCRATCH_MNT" | _filter_fstrim)
> +echo "Trimmed=$trimmed total_size=$fs_size ratio=$(($trimmed * 100 / $fs_size))%" \
> +	>> $seqres.full
> +
> +# For correct full fs trim, both unallocated space (less than 50%)
> +# and free space in existing block groups (about 25%) should be trimmed.
> +# If less than 50% is trimmed, then only unallocated space is trimmed.
> +# BTW, without fix only 31% can be trimmed, while after fix it's 64%.
> +if [ $trimmed -lt $(( $fs_size / 2)) ]; then
> +	echo "Free space in block groups not trimmed"
> +	echo "Trimmed=$trimmed total_size=$fs_size ratio=$(($trimmed * 100 / $fs_size))%"
> +fi
> +
> +echo "Silence is golden"
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/btrfs/155.out b/tests/btrfs/155.out
> new file mode 100644
> index 00000000..d25dcd3b
> --- /dev/null
> +++ b/tests/btrfs/155.out
> @@ -0,0 +1,2 @@
> +QA output created by 155
> +Silence is golden
> diff --git a/tests/btrfs/group b/tests/btrfs/group
> index c98cf823..86c7128e 100644
> --- a/tests/btrfs/group
> +++ b/tests/btrfs/group
> @@ -157,3 +157,4 @@
>  152 auto quick metadata qgroup send
>  153 auto quick qgroup
>  154 auto quick
> +155 auto trim
> 


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 520 bytes --]

Re: [PATCH] fstests: btrfs: Add test case to check if btrfs can handle full fs trim correctly

[Eryu Guan]

On Thu, Dec 07, 2017 at 08:43:43AM +0800, Qu Wenruo wrote:
> Ping.
> 
> Any comment on this?

It's been pushed out to upstream, see commit 88231c0c0b9d

Thanks,
Eryu