Re: [PATCH 1/4] xfs: Add helper for inode size

[Andrei Borzenkov <arvidjaar@gmail.com>]

В Mon, 11 May 2015 14:15:48 +0200
Jan Kara <jack@suse.cz> пишет:

> On Mon 11-05-15 14:53:57, Andrei Borzenkov wrote:
> > В Mon, 14 Jul 2014 17:21:28 +0200
> > Jan Kara <jack@suse.cz> пишет:
> > 
> > > Signed-off-by: Jan Kara <jack@suse.cz>
> > > ---
> > >  grub-core/fs/xfs.c | 17 +++++++++++------
> > >  1 file changed, 11 insertions(+), 6 deletions(-)
> > > 
> > > diff --git a/grub-core/fs/xfs.c b/grub-core/fs/xfs.c
> > > index 16ffd3f1ebd9..a2fc942707c1 100644
> > > --- a/grub-core/fs/xfs.c
> > > +++ b/grub-core/fs/xfs.c
> > > @@ -255,6 +255,11 @@ grub_xfs_inode_offset (struct grub_xfs_data *data,
> > >  	  data->sblock.log2_inode);
> > >  }
> > >  
> > > +static inline int
> > > +grub_xfs_inode_size(struct grub_xfs_data *data)
> > 
> > This should be grub_size_t.
>   OK.
> 
> > What is the reason to add it? It does not look like subsequent
> > patches modify this function like making it conditional on XFS version.
>   This is just a cleanup so that it's clearer that 1 <<
> data->sblock.log2_inode is actually inode size. But if you don't like it,
> you can just skip it. So should I send update patch? Thanks for having a
> look!

Yes, send updated patch, it is less risky as series had been tested for
quite some time. But actually helper to compute allocation size with
comment would be quite useful; I stared for quite some time at

                  sizeof (struct grub_fshelp_node)
                   - sizeof (struct grub_xfs_inode)
                   + (1 << data->sblock.log2_inode));

trying to understand what it's about :)