* [PATCH] xfs: skip test that want to mdrestore to block devices on zoned devices
@ 2025-05-13 5:19 Christoph Hellwig
2025-05-13 14:37 ` Darrick J. Wong
0 siblings, 1 reply; 2+ messages in thread
From: Christoph Hellwig @ 2025-05-13 5:19 UTC (permalink / raw)
To: zlang; +Cc: djwong, hans.holmberg, linux-xfs, fstests
mdrestore doesn't work on zoned device, so skip tests using to
pre-populate a file system image.
This was previously papered over by requiring fallocate, which got
removed in commit eff1baf42a79 ("common/populate: drop fallocate mode 0
requirement").
Signed-off-by: Christoph Hellwig <hch@lst.de>
---
common/populate | 2 +-
common/xfs | 14 ++++++++++++++
tests/xfs/284 | 1 +
tests/xfs/598 | 1 +
4 files changed, 17 insertions(+), 1 deletion(-)
diff --git a/common/populate b/common/populate
index 50dc75d35259..1c0dd03e4ac7 100644
--- a/common/populate
+++ b/common/populate
@@ -19,7 +19,7 @@ _require_populate_commands() {
"xfs")
_require_command "$XFS_DB_PROG" "xfs_db"
_require_command "$WIPEFS_PROG" "wipefs"
- _require_command "$XFS_MDRESTORE_PROG" "xfs_mdrestore"
+ _require_scratch_xfs_mdrestore
;;
ext*)
_require_command "$DUMPE2FS_PROG" "dumpe2fs"
diff --git a/common/xfs b/common/xfs
index 96c15f3c7bb0..4ac29a95812b 100644
--- a/common/xfs
+++ b/common/xfs
@@ -772,6 +772,20 @@ _scratch_xfs_mdrestore()
_xfs_mdrestore "$metadump" "$SCRATCH_DEV" "$logdev" "$rtdev" "$@"
}
+# Check if mdrestore to the scratch device is supported
+_require_scratch_xfs_mdrestore() {
+ _require_command "$XFS_MDRESTORE_PROG" "xfs_mdrestore"
+
+ # mdrestore can't restore to zoned devices
+ _require_non_zoned_device $SCRATCH_DEV
+ if [ "$USE_EXTERNAL" = yes -a ! -z "$SCRATCH_LOGDEV" ]; then
+ _require_non_zoned_device $SCRATCH_LOGDEV
+ fi
+ if [ "$USE_EXTERNAL" = yes -a ! -z "$SCRATCH_RTDEV" ]; then
+ _require_non_zoned_device $SCRATCH_RTDEV
+ fi
+}
+
# Do not use xfs_repair (offline fsck) to rebuild the filesystem
_xfs_skip_offline_rebuild() {
touch "$RESULT_DIR/.skip_rebuild"
diff --git a/tests/xfs/284 b/tests/xfs/284
index 91c17690cabe..79bf80842234 100755
--- a/tests/xfs/284
+++ b/tests/xfs/284
@@ -27,6 +27,7 @@ _require_xfs_copy
_require_test
_require_scratch
_require_no_large_scratch_dev
+_require_scratch_xfs_mdrestore
function filter_mounted()
{
diff --git a/tests/xfs/598 b/tests/xfs/598
index 20a80fcb6b91..82a9a79208ab 100755
--- a/tests/xfs/598
+++ b/tests/xfs/598
@@ -28,6 +28,7 @@ _require_test
_require_scratch
_require_xfs_mkfs_ciname
_require_xfs_ciname
+_require_scratch_xfs_mdrestore
_scratch_mkfs -n version=ci > $seqres.full
_scratch_mount
--
2.47.2
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH] xfs: skip test that want to mdrestore to block devices on zoned devices
2025-05-13 5:19 [PATCH] xfs: skip test that want to mdrestore to block devices on zoned devices Christoph Hellwig
@ 2025-05-13 14:37 ` Darrick J. Wong
0 siblings, 0 replies; 2+ messages in thread
From: Darrick J. Wong @ 2025-05-13 14:37 UTC (permalink / raw)
To: Christoph Hellwig; +Cc: zlang, hans.holmberg, linux-xfs, fstests
On Tue, May 13, 2025 at 07:19:33AM +0200, Christoph Hellwig wrote:
> mdrestore doesn't work on zoned device, so skip tests using to
> pre-populate a file system image.
>
> This was previously papered over by requiring fallocate, which got
> removed in commit eff1baf42a79 ("common/populate: drop fallocate mode 0
> requirement").
>
> Signed-off-by: Christoph Hellwig <hch@lst.de>
> ---
Looks good,
Reviewed-by: "Darrick J. Wong" <djwong@kernel.org>
--D
> common/populate | 2 +-
> common/xfs | 14 ++++++++++++++
> tests/xfs/284 | 1 +
> tests/xfs/598 | 1 +
> 4 files changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/common/populate b/common/populate
> index 50dc75d35259..1c0dd03e4ac7 100644
> --- a/common/populate
> +++ b/common/populate
> @@ -19,7 +19,7 @@ _require_populate_commands() {
> "xfs")
> _require_command "$XFS_DB_PROG" "xfs_db"
> _require_command "$WIPEFS_PROG" "wipefs"
> - _require_command "$XFS_MDRESTORE_PROG" "xfs_mdrestore"
> + _require_scratch_xfs_mdrestore
> ;;
> ext*)
> _require_command "$DUMPE2FS_PROG" "dumpe2fs"
> diff --git a/common/xfs b/common/xfs
> index 96c15f3c7bb0..4ac29a95812b 100644
> --- a/common/xfs
> +++ b/common/xfs
> @@ -772,6 +772,20 @@ _scratch_xfs_mdrestore()
> _xfs_mdrestore "$metadump" "$SCRATCH_DEV" "$logdev" "$rtdev" "$@"
> }
>
> +# Check if mdrestore to the scratch device is supported
> +_require_scratch_xfs_mdrestore() {
> + _require_command "$XFS_MDRESTORE_PROG" "xfs_mdrestore"
> +
> + # mdrestore can't restore to zoned devices
> + _require_non_zoned_device $SCRATCH_DEV
> + if [ "$USE_EXTERNAL" = yes -a ! -z "$SCRATCH_LOGDEV" ]; then
> + _require_non_zoned_device $SCRATCH_LOGDEV
> + fi
> + if [ "$USE_EXTERNAL" = yes -a ! -z "$SCRATCH_RTDEV" ]; then
> + _require_non_zoned_device $SCRATCH_RTDEV
> + fi
> +}
> +
> # Do not use xfs_repair (offline fsck) to rebuild the filesystem
> _xfs_skip_offline_rebuild() {
> touch "$RESULT_DIR/.skip_rebuild"
> diff --git a/tests/xfs/284 b/tests/xfs/284
> index 91c17690cabe..79bf80842234 100755
> --- a/tests/xfs/284
> +++ b/tests/xfs/284
> @@ -27,6 +27,7 @@ _require_xfs_copy
> _require_test
> _require_scratch
> _require_no_large_scratch_dev
> +_require_scratch_xfs_mdrestore
>
> function filter_mounted()
> {
> diff --git a/tests/xfs/598 b/tests/xfs/598
> index 20a80fcb6b91..82a9a79208ab 100755
> --- a/tests/xfs/598
> +++ b/tests/xfs/598
> @@ -28,6 +28,7 @@ _require_test
> _require_scratch
> _require_xfs_mkfs_ciname
> _require_xfs_ciname
> +_require_scratch_xfs_mdrestore
>
> _scratch_mkfs -n version=ci > $seqres.full
> _scratch_mount
> --
> 2.47.2
>
>
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2025-05-13 14:37 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-05-13 5:19 [PATCH] xfs: skip test that want to mdrestore to block devices on zoned devices Christoph Hellwig
2025-05-13 14:37 ` Darrick J. Wong
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).