[Ocfs2-devel] [PATCH v6 00/28] fs: fixes for serious clone/dedupe problems

[Dave Chinner]

On Mon, Oct 22, 2018 at 05:52:49AM +0100, Al Viro wrote:
> On Mon, Oct 22, 2018 at 03:37:41PM +1100, Dave Chinner wrote:
> 
> > Ok, this is a bit of a mess. the patches do not merge cleanly to a
> > 4.19-rc1 base kernel because of all the changes to
> > include/linux/fs.h that have hit the tree after this. There's also
> > failures against Documentation/filesystems/fs.h
> > 
> > IOWs, it's not going to get merged through the main XFS tree because
> > I don't have the patience to resolve all the patch application
> > failures, then when it comes to merge make sure all the merge
> > failures end up being resolved correctly.
> > 
> > So if I take it through the XFS tree, it will being a standalone
> > branch based on 4.19-rc8 and won't hit linux-next until after the
> > first XFS merge when I can rebase the for-next branch...
> 
> How many conflicts does it have with XFS tree?  I can take it via
> vfs.git...

I gave up after 4 of the first 6 or 7 patches had conflicts in vfs
and documentation code.

There were changes that went into 4.19-rc7 that changed
{do|vfs}_clone_file_range() prototypes and this patchset hits
prototypes adjacent to that multiple times. There's also a conflicts
against a new f_ops->fadvise method. These all appear to be direct
fallout of fixes needed for all the overlay f_ops changes.

The XFS changes at the end of the patchset are based on
commits that were merged into -rc7 and -rc8, so if you are using
-rc8 as your base, then it all merges cleanly. There are no
conflicts with the current xfs/for-next branch.

I've just merged and built it into my test tree (-rc8, xfs/for-next,
djwong/devel) so I can test it properly, but if it merges cleanly
with the vfs tree you are building then that's probably the easiest
way to merge it all...

Cheers,

Dave.
-- 
Dave Chinner
david at fromorbit.com