Re: [PATCH 2/4] repair: add canonical names for the XR_INO_ constants

[Christoph Hellwig <hch@lst.de>]

On Fri, Nov 28, 2025 at 09:00:59AM +0100, Carlos Maiolino wrote:
> On Fri, Nov 28, 2025 at 07:37:00AM +0100, Christoph Hellwig wrote:
> > Add an array with the canonical name for each inode type so that code
> > doesn't have to implement switch statements for that, and remove the now
> > trivial process_misc_ino_types and process_misc_ino_types_blocks
> > functions.
> > 
> > Signed-off-by: Christoph Hellwig <hch@lst.de>
> > -
> 
> 
> >  static inline int
> >  dinode_fmt(
> >  	struct xfs_dinode *dino)
> > @@ -2261,16 +2180,20 @@ _("directory inode %" PRIu64 " has bad size %" PRId64 "\n"),
> >  	case XR_INO_BLKDEV:
> >  	case XR_INO_SOCK:
> >  	case XR_INO_FIFO:
> > -		if (process_misc_ino_types(mp, dino, lino, type))
> > -			return 1;
> > -		break;
> > -
> >  	case XR_INO_UQUOTA:
> >  	case XR_INO_GQUOTA:
> >  	case XR_INO_PQUOTA:
> > -		/* Quota inodes have same restrictions as above types */
> > -		if (process_misc_ino_types(mp, dino, lino, type))
> > +		/*
> > +		 * Misc inode types that have no associated data storage (fifos,
> > +		 * pipes, devices, etc.) mad thus must also have a zero size.
> 
> 		Perhaps is my non-native English, but this sentence
> 		doesn't make much sense to me. Not sure if 'mad' has
> 		some special meaning in this context?

that should be an "and"