On Sun, May 16, 2021 at 11:54:30PM +0800, Eryu Guan wrote:
> On Tue, May 11, 2021 at 07:02:24PM -0700, Darrick J. Wong wrote:
> > From: Darrick J. Wong <djwong@xxxxxxxxxx>
> >
> > Fix the obviously incorrect code here that wants to fail the test if
> > mkfs doesn't succeed. The return value ("$?") is always the status of
> > the /last/ command in the pipe. Change the checker to _notrun so that
> > we don't leave the scratch check files around.
> >
> > Signed-off-by: Darrick J. Wong <djwong@xxxxxxxxxx>
> > ---
> > tests/xfs/178 | 4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >
> >
> > diff --git a/tests/xfs/178 b/tests/xfs/178
> > index a24ef50c..bf72e640 100755
> > --- a/tests/xfs/178
> > +++ b/tests/xfs/178
> > @@ -57,8 +57,8 @@ _supported_fs xfs
> > # fix filesystem, new mkfs.xfs will be fine.
> >
> > _require_scratch
> > -_scratch_mkfs_xfs | _filter_mkfs 2>$tmp.mkfs \
> > - || _fail "mkfs failed!"
> > +_scratch_mkfs_xfs | _filter_mkfs 2>$tmp.mkfs
> > +test "${PIPESTATUS[0]}" -eq 0 || _notrun "mkfs failed!"
>
> I still don't understand why changing this to _notrun, shouldn't creating a
> default filesystem should always pass?
It will, unless you've injected a malicious mkfs.xfs to make sure
constructions like these actually work.
> and fail the test if mkfs failed?
Yeah, you're right, it /should/ fail. I'll leave it alone then.
--D
>
> Thanks,
> Eryu
>
> >
> > # By executing the followint tmp file, will get on the mkfs options stored in
> > # variables
> >
