Message ID | 49e01810eff8d5ddd7d3c99796a66b997faaaf84.1674644814.git.fdmanana@suse.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | btrfs: test send optimal cloning behaviour | expand |
On Wed, 25 Jan 2023 11:07:54 +0000, fdmanana@kernel.org wrote: > From: Filipe Manana <fdmanana@suse.com> > > Test that send operations do the best cloning decisions when we have > extents that are shared but some files refer to the full extent while > others refer to only a section of the extent. > > This exercises an optimization that was added to kernel 6.2, by the > following commit: > > c7499a64dcf6 ("btrfs: send: optimize clone detection to increase extent sharing") > > Signed-off-by: Filipe Manana <fdmanana@suse.com> > --- > tests/btrfs/283 | 88 +++++++++++++++++++++++++++++++++++++++++++++ > tests/btrfs/283.out | 26 ++++++++++++++ > 2 files changed, 114 insertions(+) > create mode 100755 tests/btrfs/283 > create mode 100644 tests/btrfs/283.out > > diff --git a/tests/btrfs/283 b/tests/btrfs/283 > new file mode 100755 > index 00000000..c1f6007d > --- /dev/null > +++ b/tests/btrfs/283 > @@ -0,0 +1,88 @@ > +#! /bin/bash > +# SPDX-License-Identifier: GPL-2.0 > +# Copyright (C) 2023 SUSE Linux Products GmbH. All Rights Reserved. > +# > +# FS QA Test 283 > +# > +# Test that send operations do the best cloning decisions when we have extents > +# that are shared but some files refer to the full extent while others refer to > +# only a section of the extent. > +# > +. ./common/preamble > +_begin_fstest auto quick send clone fiemap > + > +. ./common/filter > +. ./common/reflink > +. ./common/punch # for _filter_fiemap_flags > + > +_supported_fs btrfs > +_require_test > +_require_scratch_reflink > +_require_cp_reflink > +_require_xfs_io_command "fiemap" > +_require_fssum > + > +_wants_kernel_commit c7499a64dcf6 \ > + "btrfs: send: optimize clone detection to increase extent sharing" > + > +send_files_dir=$TEST_DIR/btrfs-test-$seq > +send_stream=$send_files_dir/snap.stream > +snap_fssum=$send_files_dir/snap.fssum > + > +rm -fr $send_files_dir > +mkdir $send_files_dir I'm not sure what the rules are regarding TEST_DIR residuals, but it might be worth adding a custom _cleanup() for $send_files_dir . Anyhow, looks good as-is. Reviewed-by: David Disseldorp <ddiss@suse.de>
LGTM
Reviewed-by: Anand Jain <anand.jain@oracle.com>
Thx
On Wed, Jan 25, 2023 at 01:21:21PM +0100, David Disseldorp wrote: > On Wed, 25 Jan 2023 11:07:54 +0000, fdmanana@kernel.org wrote: > > > From: Filipe Manana <fdmanana@suse.com> > > > > Test that send operations do the best cloning decisions when we have > > extents that are shared but some files refer to the full extent while > > others refer to only a section of the extent. > > > > This exercises an optimization that was added to kernel 6.2, by the > > following commit: > > > > c7499a64dcf6 ("btrfs: send: optimize clone detection to increase extent sharing") > > > > Signed-off-by: Filipe Manana <fdmanana@suse.com> > > --- > > tests/btrfs/283 | 88 +++++++++++++++++++++++++++++++++++++++++++++ > > tests/btrfs/283.out | 26 ++++++++++++++ > > 2 files changed, 114 insertions(+) > > create mode 100755 tests/btrfs/283 > > create mode 100644 tests/btrfs/283.out > > > > diff --git a/tests/btrfs/283 b/tests/btrfs/283 > > new file mode 100755 > > index 00000000..c1f6007d > > --- /dev/null > > +++ b/tests/btrfs/283 > > @@ -0,0 +1,88 @@ > > +#! /bin/bash > > +# SPDX-License-Identifier: GPL-2.0 > > +# Copyright (C) 2023 SUSE Linux Products GmbH. All Rights Reserved. > > +# > > +# FS QA Test 283 > > +# > > +# Test that send operations do the best cloning decisions when we have extents > > +# that are shared but some files refer to the full extent while others refer to > > +# only a section of the extent. > > +# > > +. ./common/preamble > > +_begin_fstest auto quick send clone fiemap > > + > > +. ./common/filter > > +. ./common/reflink > > +. ./common/punch # for _filter_fiemap_flags > > + > > +_supported_fs btrfs > > +_require_test > > +_require_scratch_reflink > > +_require_cp_reflink > > +_require_xfs_io_command "fiemap" > > +_require_fssum > > + > > +_wants_kernel_commit c7499a64dcf6 \ > > + "btrfs: send: optimize clone detection to increase extent sharing" > > + > > +send_files_dir=$TEST_DIR/btrfs-test-$seq > > +send_stream=$send_files_dir/snap.stream > > +snap_fssum=$send_files_dir/snap.fssum > > + > > +rm -fr $send_files_dir > > +mkdir $send_files_dir > > I'm not sure what the rules are regarding TEST_DIR residuals, but it > might be worth adding a custom _cleanup() for $send_files_dir . > Anyhow, looks good as-is. Thanks for reviewing this patch. The $send_files_dir (and files under it) won't take big space (maybe < 1M). So I think we can keep it in TEST_DIR, don't need more changes :) Thanks, Zorro > Reviewed-by: David Disseldorp <ddiss@suse.de> >
diff --git a/tests/btrfs/283 b/tests/btrfs/283 new file mode 100755 index 00000000..c1f6007d --- /dev/null +++ b/tests/btrfs/283 @@ -0,0 +1,88 @@ +#! /bin/bash +# SPDX-License-Identifier: GPL-2.0 +# Copyright (C) 2023 SUSE Linux Products GmbH. All Rights Reserved. +# +# FS QA Test 283 +# +# Test that send operations do the best cloning decisions when we have extents +# that are shared but some files refer to the full extent while others refer to +# only a section of the extent. +# +. ./common/preamble +_begin_fstest auto quick send clone fiemap + +. ./common/filter +. ./common/reflink +. ./common/punch # for _filter_fiemap_flags + +_supported_fs btrfs +_require_test +_require_scratch_reflink +_require_cp_reflink +_require_xfs_io_command "fiemap" +_require_fssum + +_wants_kernel_commit c7499a64dcf6 \ + "btrfs: send: optimize clone detection to increase extent sharing" + +send_files_dir=$TEST_DIR/btrfs-test-$seq +send_stream=$send_files_dir/snap.stream +snap_fssum=$send_files_dir/snap.fssum + +rm -fr $send_files_dir +mkdir $send_files_dir + +_scratch_mkfs >> $seqres.full 2>&1 +_scratch_mount + +# When using compression, btrfs limits the extent size to 128K, so do not do +# larger writes and then expect larger extents, as that would break the test +# if we are run with compression enabled through $MOUNT_OPTIONS (resulting in +# mismatch with the golden output). +$XFS_IO_PROG -f -c "pwrite -S 0xab -b 128K 0 128K" $SCRATCH_MNT/foo | _filter_xfs_io + +# Now clone file foo twice, which will make the 128K extent shared 3 times. +_cp_reflink $SCRATCH_MNT/foo $SCRATCH_MNT/bar +_cp_reflink $SCRATCH_MNT/foo $SCRATCH_MNT/baz + +# Overwrite the second half of file foo. +$XFS_IO_PROG -c "pwrite -S 0xcd -b 64K 64K 64K" $SCRATCH_MNT/foo | _filter_xfs_io + +echo "Creating snapshot and a send stream for it..." +$BTRFS_UTIL_PROG subvolume snapshot -r $SCRATCH_MNT $SCRATCH_MNT/snap \ + | _filter_scratch + +$BTRFS_UTIL_PROG send -f $send_stream $SCRATCH_MNT/snap 2>&1 | _filter_scratch + +$FSSUM_PROG -A -f -w $snap_fssum $SCRATCH_MNT/snap + +echo "Creating a new filesystem to receive the send stream..." +_scratch_unmount +_scratch_mkfs >> $seqres.full 2>&1 +_scratch_mount + +$BTRFS_UTIL_PROG receive -f $send_stream $SCRATCH_MNT + +echo "Verifying data matches the original filesystem..." +$FSSUM_PROG -r $snap_fssum $SCRATCH_MNT/snap + +# Now verify that all extents, for all files, are shared. + +# File 'foo' should have a single 128K extent, which is shared because its first +# half is referred by files 'bar' and 'baz'. +echo -e "\nfiemap of file foo:\n" +$XFS_IO_PROG -r -c "fiemap -v" $SCRATCH_MNT/snap/foo | _filter_fiemap_flags + +# File 'bar' should have two 64K shared extents. The first one is shared with +# files 'foo' and 'baz', while the second one is only shared with file 'baz'. +echo -e "\nfiemap of file bar:\n" +$XFS_IO_PROG -r -c "fiemap -v" $SCRATCH_MNT/snap/bar | _filter_fiemap_flags + +# File 'baz' should have two 64K shared extents. The first one is shared with +# files 'foo' and 'bar', while the second one is only shared with file 'bar'. +echo -e "\nfiemap of file baz:\n" +$XFS_IO_PROG -r -c "fiemap -v" $SCRATCH_MNT/snap/baz | _filter_fiemap_flags + +# success, all done +status=0 +exit diff --git a/tests/btrfs/283.out b/tests/btrfs/283.out new file mode 100644 index 00000000..286dae33 --- /dev/null +++ b/tests/btrfs/283.out @@ -0,0 +1,26 @@ +QA output created by 283 +wrote 131072/131072 bytes at offset 0 +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +wrote 65536/65536 bytes at offset 65536 +XXX Bytes, X ops; XX:XX:XX.X (XXX YYY/sec and XXX ops/sec) +Creating snapshot and a send stream for it... +Create a readonly snapshot of 'SCRATCH_MNT' in 'SCRATCH_MNT/snap' +At subvol SCRATCH_MNT/snap +Creating a new filesystem to receive the send stream... +At subvol snap +Verifying data matches the original filesystem... +OK + +fiemap of file foo: + +0: [0..255]: shared|last + +fiemap of file bar: + +0: [0..127]: shared +1: [128..255]: shared|last + +fiemap of file baz: + +0: [0..127]: shared +1: [128..255]: shared|last