Thanks Jan, for checking the merge.
I didn't realize the ext4 fixups were going in via the xfs tree. I
didn't see any merge conflicts from the linux-next tree (or maybe I
missed it) so I had assumed it wasn't in Dave's xfs.git tree. My plan
had been to deal with it after the prereq patch went in via the xfs
tree --- in fact I was thinking about checking to see if the xfs.git
tree had been merged so I could get the ext4 commit merged and sent to
Linus.
Sorry I lost track of things. If I had known I would have indeed
manually merged in the branch, resolved the merge, and done a
regression test cycle before pushing to Linus.
- Ted
On Mon, Mar 21, 2016 at 10:16:28AM +0100, Jan Kara wrote:
> On Mon 21-03-16 13:05:05, Dave Chinner wrote:
> The resolution of conflict in fs/ext4/inode.c is fine except for one minor
> issue:
>
> > diff --cc fs/ext4/inode.c
> > index b2e9576,2b98171..e5ba3b0
> > --- a/fs/ext4/inode.c
> > +++ b/fs/ext4/inode.c
> > @@@ -3289,10 -3161,10 +3289,10 @@@ out
> > }
> > #endif
> >
> > - static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
> > + static int ext4_end_io_dio(struct kiocb *iocb, loff_t offset,
> > ssize_t size, void *private)
> > {
> > - ext4_io_end_t *io_end = iocb->private;
> > + ext4_io_end_t *io_end = private;
> >
> > /* if not async direct IO just return */
> > if (!io_end)
> > @@@ -3300,8 -3172,18 +3300,17 @@@
> >
> > ext_debug("ext4_end_io_dio(): io_end 0x%p "
> > "for inode %lu, iocb 0x%p, offset %llu, size %zd\n",
> > - iocb->private, io_end->inode->i_ino, iocb, offset,
> > - size);
> > + io_end, io_end->inode->i_ino, iocb, offset, size);
> >
> > + iocb->private = NULL;
>
> The line above should not exist in the result. It does no harm but is
> unnecessary.
>
> Honza
> --
> Jan Kara <[email protected]>
> SUSE Labs, CR
