Re: [PATCH] vfs: remove externs from fs.h on functions modified by i_ino widening

[Jeff Layton]

On Sat, 2026-03-07 at 05:31 +0000, Matthew Wilcox wrote:
> On Fri, Mar 06, 2026 at 08:27:01AM -0500, Jeff Layton wrote:
> > @@ -2951,26 +2951,26 @@ struct inode *iget5_locked(struct super_block *, u64,
> >  struct inode *iget5_locked_rcu(struct super_block *, u64,
> >  			       int (*test)(struct inode *, void *),
> >  			       int (*set)(struct inode *, void *), void *);
> > -extern struct inode *iget_locked(struct super_block *, u64);
> > +struct inode *iget_locked(struct super_block *, u64);
> 
> I think plain 'u64' deserves a name.  I know some people get very
> upset when they see any unnamed parameter, but I don't think that you
> need to put "sb" in the first parameter.  A u64 is non-obvious though;
> is it i_ino?  Or hashval?
> 
> > -extern struct inode *find_inode_nowait(struct super_block *,
> > +struct inode *find_inode_nowait(struct super_block *,
> >  				       u64,
> >  				       int (*match)(struct inode *,
> >  						    u64, void *),
> >  				       void *data);
> 
> I think these need to be reflowed.  Before they were aligned with the
> open bracket, and this demonstrates why that's a stupid convention.
> And the u64 needs a name.
> 
> (other occurrences snipped)

I toyed with changing those too, but there were quite a few. I'll plan
to send an updated patch that names all of the parameters in anything
that has them unnamed. That'll make checkpatch happy too.
-- 
Jeff Layton <jlayton@xxxxxxxxxx>