Re: [PATCH 5/9] xfs: add a xfs_rmap_inode_owner helper

[Christoph Hellwig <hch@lst.de>]

On Mon, May 18, 2026 at 03:23:58PM +0200, Carlos Maiolino wrote:
> On Mon, May 18, 2026 at 03:16:39PM +0200, Christoph Hellwig wrote:
> > On Mon, May 18, 2026 at 01:14:31PM +0200, Carlos Maiolino wrote:
> > > On Fri, May 15, 2026 at 03:50:26PM +0200, Christoph Hellwig wrote:
> > > > Add a small wrapper for initializing the rmap owner to i_ino.
> > > > 
> > > > Signed-off-by: Christoph Hellwig <hch@lst.de>
> > > 
> > > This and the previous one, provides very little gain IMHO, do you intend
> > > to get rid of xfs_rmap_ino_owner and xfs_rmap_inode_bmbt_owner? It seems
> > > to me they could simply be replaced by these new helpers instead of just
> > > wrapping them into these new ones.
> > > A little disclaimer: I didn't look into the next patches yet to see if
> > > it will all make sense later.
> > 
> > My direct aim was to reduce uses of i_ino so I looked for common
> > patterns.  I also generally prefer APIs that pass typesave objects
> > over those passing integers as handles wherever possible.
> 
> Agreed. both xfs_rmap_ino_owner and xfs_rmap_inode_bmbt_owner look good
> candidates to fully replace their 'old' versions giving all their
> callers have access to the xfs_inode (if I didn't miss any user).

We can't convert all, because repair needs to pass a different owner.