On Wed 28-01-09 12:19:15, Christoph Hellwig wrote:
> On Tue, Jan 27, 2009 at 06:35:40PM +0100, Jan Kara wrote:
> > On Tue 27-01-09 12:13:23, Christoph Hellwig wrote:
> > > On Tue, Jan 27, 2009 at 05:48:38PM +0100, Jan Kara wrote:
> > > > Use lowercase names of quota functions instead of old uppercase ones.
> > >
> > > Looking over this again I think the vfs_ names are a really bad choice.
> > > The normal use for vfs_ are higher level functions that call into
> > > specific filesystems, which these aren't really. What about just
> > > striping out that vfs_ prefix first?
> > Actually, these vfs_dq_ functions are usually just wrappers like:
> >
> > static inline void vfs_dq_init(struct inode *inode)
> > {
> > BUG_ON(!inode->i_sb);
> > if (sb_any_quota_active(inode->i_sb) && !IS_NOQUOTA(inode))
> > inode->i_sb->dq_op->initialize(inode, -1);
> > }
> >
> > where ->initialize either leads to a filesystem specified callback or to
> > generic quota init function. So this seems to reasonably match your
> > description above...
> > But if this still has not convinced you ;), I can change all these quota
> > functions to have names starting with dq_. That should not clash with
> > anything else. But I'm not very enthusiastic about this change (cleanups are
> > so boring...).
>
> Hmm, seems you're right - I must have been confused when looking at it.
>
> Changing away from the weird uppercase names defintively is a good
> thing.
>
> Although for the ops inside the filesystem it might be better to just
> call into the lowlevel operation instead of the vfs_ helper as we know
> what we want, no?
Yes, filesystem knows what it wants but the vfs_ helpers are there so that
every filesystem does not have to implement the same checks (usually
checking whether quota is enabled etc)...
Honza
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
