| Message ID | 20161102040727.5765-1-wangxg.fnst@cn.fujitsu.com (mailing list archive) |
|---|---|
| State | Not Applicable |
| Headers | show |
On Wed, Nov 02, 2016 at 12:07:26PM +0800, Wang Xiaoguang wrote: > In original common/populate codes, we put _require_xfs_io_command "falloc" and > _require_xfs_io_command "fpunch" in the begin of common/populate, but it's > not appropriate, for fs, which does not support falloc and punch, will not > be able to use other helper functions in common/populate, so here I choose > to put _require_xfs_io_command "falloc" or "punch" in helper function which > really use falloc and fpunch. > > And xfs/120 uses fpunch, add _require_xfs_io_command "fpunch". > > Signed-off-by: Wang Xiaoguang <wangxg.fnst@cn.fujitsu.com> Thanks for the patch! But I noticed that the new patchset posted by Darrick recently removed "_require"s from common/populate too ([PATCH 2/9] populate: add _require_populate_commands to check for tools), and his subsequent patches depend on this update. So I think it's better to take Darrick's patch, rather than taking this one and forcing Darrick to rebase his whole patchset again. And patch 2 requires the movement of the "_require"s in common/populate, so I'm going to merge it after reviewing & applying Darrick's patchset (otherwise generic/256 won't run on filesystems that don't support fallocate(2), e.g. ext2/3). That means patch 2 won't be in the pull request this week either. Hopefully we can see it in next pr. Thanks, Eryu -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" 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/common/populate b/common/populate index d0003c5..3b9b531 100644 --- a/common/populate +++ b/common/populate @@ -22,9 +22,6 @@ # Mountain View, CA 94043, USA, or: http://www.sgi.com #----------------------------------------------------------------------- -_require_xfs_io_command "falloc" -_require_xfs_io_command "fpunch" - _require_xfs_db_blocktrash_z_command() { test "${FSTYP}" = "xfs" || _notrun "cannot run xfs_db on ${FSTYP}" $XFS_DB_PROG -x -f -c 'blocktrash -z' "${TEST_DEV}" | grep -q 'nothing on stack' || _notrun "blocktrash -z not supported" @@ -90,6 +87,8 @@ __populate_fill_fs() { # types of metadata block _scratch_xfs_populate() { _scratch_mount + _require_xfs_io_command "fpunch" + blksz="$(stat -f -c '%s' "${SCRATCH_MNT}")" dblksz="$(xfs_info "${SCRATCH_MNT}" | grep naming.*bsize | sed -e 's/^.*bsize=//g' -e 's/\([0-9]*\).*$/\1/g')" leaf_lblk="$((32 * 1073741824 / blksz))" @@ -192,6 +191,8 @@ _scratch_xfs_populate() { # types of metadata block _scratch_ext4_populate() { _scratch_mount + _require_xfs_io_command "fpunch" + blksz="$(stat -f -c '%s' "${SCRATCH_MNT}")" dblksz="${blksz}" leaf_lblk="$((32 * 1073741824 / blksz))" diff --git a/tests/xfs/120 b/tests/xfs/120 index 3deece6..631e2f2 100755 --- a/tests/xfs/120 +++ b/tests/xfs/120 @@ -47,6 +47,7 @@ _cleanup() _supported_fs xfs _supported_os Linux +_require_xfs_io_command "fpunch" _require_scratch test -n "${FORCE_FUZZ}" || _require_scratch_xfs_crc _require_attrs
In original common/populate codes, we put _require_xfs_io_command "falloc" and _require_xfs_io_command "fpunch" in the begin of common/populate, but it's not appropriate, for fs, which does not support falloc and punch, will not be able to use other helper functions in common/populate, so here I choose to put _require_xfs_io_command "falloc" or "punch" in helper function which really use falloc and fpunch. And xfs/120 uses fpunch, add _require_xfs_io_command "fpunch". Signed-off-by: Wang Xiaoguang <wangxg.fnst@cn.fujitsu.com> --- common/populate | 7 ++++--- tests/xfs/120 | 1 + 2 files changed, 5 insertions(+), 3 deletions(-)