Re: [PATCH 08/11] xfs: widen ondisk timestamps to deal with y2038 problem

["Darrick J. Wong" <darrick.wong@oracle.com>]

On Tue, Aug 18, 2020 at 03:00:49PM +0300, Amir Goldstein wrote:
> On Tue, Aug 18, 2020 at 1:57 AM Darrick J. Wong <darrick.wong@oracle.com> wrote:
> >
> > From: Darrick J. Wong <darrick.wong@oracle.com>
> >
> > Redesign the ondisk timestamps to be a simple unsigned 64-bit counter of
> > nanoseconds since 14 Dec 1901 (i.e. the minimum time in the 32-bit unix
> > time epoch).  This enables us to handle dates up to 2486, which solves
> > the y2038 problem.
> >
> > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> > ---
> ...
> 
> > diff --git a/fs/xfs/scrub/inode.c b/fs/xfs/scrub/inode.c
> > index 9f036053fdb7..b354825f4e51 100644
> > --- a/fs/xfs/scrub/inode.c
> > +++ b/fs/xfs/scrub/inode.c
> > @@ -190,6 +190,11 @@ xchk_inode_flags2(
> >         if ((flags2 & XFS_DIFLAG2_DAX) && (flags2 & XFS_DIFLAG2_REFLINK))
> >                 goto bad;
> >
> > +       /* the incore bigtime iflag always follows the feature flag */
> > +       if (!!xfs_sb_version_hasbigtime(&mp->m_sb) ^
> > +           !!(flags2 & XFS_DIFLAG2_BIGTIME))
> > +               goto bad;
> > +
> 
> Seems like flags2 are not the incore iflags and that the dinode iflags
> can very well
> have no bigtime on fs with bigtime support:
> 
> xchk_dinode(...
> ...
>                 flags2 = be64_to_cpu(dip->di_flags2);
> 
> What am I missing?

Nothing.  That chunk is just plain wrong and needs to reworked.  Repair
gets it right, the only failure case is if the inode flag is set but the
feature isn't.

I probably wrote this before I had the thought of letting people upgrade
existing filesystems.

Will fix, thanks for catching this.

--D

> 
> Thanks,
> Amir.