| Message ID | 20250128071315.676272-3-hch@lst.de (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [1/3] generic/363: remove _supported_fs xfs | expand |
On Tue, Jan 28, 2025 at 08:12:59AM +0100, Christoph Hellwig wrote: > Despite the comment, _cleanup_dump is only called from xfs specific > tests, so this is superfluous. > > Signed-off-by: Christoph Hellwig <hch@lst.de> This patch looks ok, but I have a few extra[neous] comments... Reviewed-by: "Darrick J. Wong" <djwong@kernel.org> > --- > common/dump | 6 ------ > 1 file changed, 6 deletions(-) > > diff --git a/common/dump b/common/dump > index 50b2ba03c670..3761c16100d8 100644 > --- a/common/dump > +++ b/common/dump > @@ -223,12 +223,6 @@ _require_tape() > # > _cleanup_dump() > { > - # Some tests include this before checking _supported_fs xfs > - # and the sleeps & checks here get annoying > - if [ "$FSTYP" != "xfs" ]; then > - return > - fi > - I went looking in common/dump and I noticed the lack of an explicit setup routine. Instead, variables and _require calls are done when sourcing the file. Curiously there's no check for FSTYP==xfs, which I guess is reasonable for sourced stuff, but I think that should all get pulled up into _init_dump() or something. The other thing I noticed is that sourcing common/dump deletes $seqres.full, which seems like a real bug. --D > cd $here > > if [ -n "$DEBUGDUMP" ]; then > -- > 2.45.2 > >
On Tue, Jan 28, 2025 at 11:14:42AM -0800, Darrick J. Wong wrote: > I went looking in common/dump and I noticed the lack of an explicit > setup routine. Instead, variables and _require calls are done when > sourcing the file. Curiously there's no check for FSTYP==xfs, which I > guess is reasonable for sourced stuff, but I think that should all get > pulled up into _init_dump() or something. > > The other thing I noticed is that sourcing common/dump deletes > $seqres.full, which seems like a real bug. Yeah, dump testing is at least as messy as xfsdump itself :) Btw, did someone every spend some time analyzing why xfs/059 xfs/060 fail relatively frequently when using RT devices? This is one of the thing on my TODO list when I find a little more time..
diff --git a/common/dump b/common/dump index 50b2ba03c670..3761c16100d8 100644 --- a/common/dump +++ b/common/dump @@ -223,12 +223,6 @@ _require_tape() # _cleanup_dump() { - # Some tests include this before checking _supported_fs xfs - # and the sleeps & checks here get annoying - if [ "$FSTYP" != "xfs" ]; then - return - fi - cd $here if [ -n "$DEBUGDUMP" ]; then
Despite the comment, _cleanup_dump is only called from xfs specific tests, so this is superfluous. Signed-off-by: Christoph Hellwig <hch@lst.de> --- common/dump | 6 ------ 1 file changed, 6 deletions(-)