On Fri, Nov 26, 2021 at 10:26 AM Qu Wenruo <wqu@xxxxxxxx> wrote:
>
> Since kernel commit d4088803f511 ("btrfs: subpage: make lzo_compress_pages()
> compatible"), lzo compression no longer respects the max compressed page
> limit, and can cause kernel crash.
>
> The upstream fix is 6f019c0e0193 ("btrfs: fix a out-of-bound access in
> copy_compressed_data_to_page()").
>
> This patch will add such regression test for all possible compress-force
> mount options, including lzo, zstd and zlib.
>
> And since we're here, also make sure the content of the file matches
> after a mount cycle.
>
> Signed-off-by: Qu Wenruo <wqu@xxxxxxxx>
> ---
> Changelog:
> v2:
> - Also test zlib and zstd
> - Add file content verification check
> v3:
> - Use $tmp.good as a known good file source
> - Also make sure we didn't get short read for the good copy
> ---
> tests/btrfs/049 | 67 +++++++++++++++++++++++++++++++++++++++++++++
> tests/btrfs/049.out | 6 ++++
> 2 files changed, 73 insertions(+)
> create mode 100755 tests/btrfs/049
>
> diff --git a/tests/btrfs/049 b/tests/btrfs/049
> new file mode 100755
> index 00000000..c1f35dc9
> --- /dev/null
> +++ b/tests/btrfs/049
> @@ -0,0 +1,67 @@
> +#! /bin/bash
> +# SPDX-License-Identifier: GPL-2.0
> +# Copyright (C) 2021 SUSE Linux Products GmbH. All Rights Reserved.
> +#
> +# FS QA Test 049
> +#
> +# Test if btrfs will crash when using compress-force mount option against
> +# incompressible data
> +#
> +. ./common/preamble
> +_begin_fstest auto quick compress dangerous
> +
> +# Override the default cleanup function.
> +_cleanup()
> +{
> + cd /
> + rm -r -f $tmp.*
> +}
> +
> +# Import common functions.
> +. ./common/filter
> +
> +# real QA test starts here
> +
> +# Modify as appropriate.
> +_supported_fs btrfs
> +_require_scratch
> +
> +pagesize=$(get_page_size)
> +
> +# Read the content from urandom to a known safe location
> +$XFS_IO_PROG -f -c "pwrite -i /dev/urandom 0 $pagesize" "$tmp.good" > /dev/null
> +
> +# Make sure we didn't get short read
s/read/write/
> +if [ $(_get_filesize "$tmp.good") != "$pagesize" ]; then
> + _fail "Got a short read from /dev/urandom"
> +fi
> +
> +workload()
> +{
> + local compression=$1
> +
> + echo "=== Testing compress-force=$compression ==="
> + _scratch_mkfs -s "$pagesize">> $seqres.full
> + _scratch_mount -o compress-force="$compression"
> + cp "$tmp.good" "$SCRATCH_MNT/$compression"
> +
> + # When unpatched, compress-force=lzo would crash at data writeback
> + _scratch_cycle_mount
> +
> + # Make sure the content matches
> + if [ "$(_md5_checksum $tmp.good)" != \
> + "$(_md5_checksum $SCRATCH_MNT/$compression)" ]; then
> + _fail "Content of '$SCRATCH_MNT/file' mismatch with known good copy"
We could just 'echo' instead of '_fail' and let the test continue with
the remaining compression algorithms.
That's usually the approach preferred for fstests, _fail is for cases
where the test can't proceed and must stop immediately.
Anyway, those are two minor things that maybe Eryu can fixup when he
picks the patch.
So I won't ask you to send yet another version.
Thanks!
Reviewed-by: Filipe Manana <fdmanana@xxxxxxxx>
> + else
> + echo "OK"
> + fi
> + _scratch_unmount
> +}
> +
> +workload lzo
> +workload zstd
> +workload zlib
> +
> +# success, all done
> +status=0
> +exit
> diff --git a/tests/btrfs/049.out b/tests/btrfs/049.out
> index cb0061b3..41bffeaa 100644
> --- a/tests/btrfs/049.out
> +++ b/tests/btrfs/049.out
> @@ -1 +1,7 @@
> QA output created by 049
> +=== Testing compress-force=lzo ===
> +OK
> +=== Testing compress-force=zstd ===
> +OK
> +=== Testing compress-force=zlib ===
> +OK
> --
> 2.34.0
>
--
Filipe David Manana,
“Whether you think you can, or you think you can't — you're right.”
