On Sun, Oct 08, 2017 at 05:51:10PM -0400, Theodore Ts'o wrote: > On Mon, Oct 09, 2017 at 08:41:17AM +1100, Dave Chinner wrote: > > On Sun, Oct 08, 2017 at 03:23:24PM -0400, Theodore Ts'o wrote: > > > Signed-off-by: Theodore Ts'o <tytso@xxxxxxx> > > > --- > > > tests/ext4/026 | 3 ++- > > > 1 file changed, 2 insertions(+), 1 deletion(-) > > > > > > diff --git a/tests/ext4/026 b/tests/ext4/026 > > > index 94a737ce..ba77f36e 100755 > > > --- a/tests/ext4/026 > > > +++ b/tests/ext4/026 > > > @@ -53,7 +53,8 @@ _require_attrs > > > _require_ext4_mkfs_feature ea_inode > > > > > > _scratch_mkfs_ext4 -O ea_inode >/dev/null 2>&1 > > > -_scratch_mount > > > +_scratch_mount >/dev/null 2>&1 \ > > > + || _notrun "kernel doesn't support ea_inode feature on $FSTYP" > > > > Needs to be in a _requires check? > > There's only one test currently testing ea_inode at the moment, but > sure, I can make it into a _requires check. I note there's a _require_ext4_mkfs_feature check above. Perhaps just turn that into a _require_ext4_feature that checks both mkfs and whether the kernel supports the functionality? That's typically what we've done in the past - one _require rule to check all aspects of support needed to use the feature... Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx
