Re: [PATCH 07/12] xfs/148: fix failure from bad shortform size assumptions

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]



On Fri, May 20, 2022 at 03:34:38PM +0800, Zorro Lang wrote:
> On Tue, May 17, 2022 at 05:01:06PM +1000, Dave Chinner wrote:
> > From: Dave Chinner <dchinner@xxxxxxxxxx>
> > @@ -41,9 +41,12 @@ test_names=("something" "$nullstr" "$slashstr" "another")
> >  rm -f $imgfile $imgfile.old
> >  
> >  # Format image file w/o crcs so we can sed the image file
> > +# We need to use 512 byte inodes to ensure the attr forks remain in short form
> > +# even when security xattrs are present so we are always doing name matches on
> > +# lookup and not name hash compares as leaf/node forms will do.
> >  $XFS_IO_PROG -f -c 'truncate 40m' $imgfile
> >  loopdev=$(_create_loop_device $imgfile)
> > -MKFS_OPTIONS="-m crc=0" _mkfs_dev $loopdev >> $seqres.full
> > +MKFS_OPTIONS="-m crc=0 -i size=512" _mkfs_dev $loopdev >> $seqres.full
> 
> It's good to me, although I'm wondering why you didn't choose to use
> $SELINUX_MOUNT_OPTIONS [1]. Anyway, extending inode size can help too.

Because if we are running under any other context that adds xattrs
to the files by defaults (e.g. default acls, other LSMs, etc) will
cause it to fail the same way.

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx



[Index of Archives]     [Linux Filesystems Development]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux