Alexander Viro wrote:
>
> On Wed, 13 Sep 2000, Daniel Phillips wrote:
>
> > "Juan J. Quintela" wrote:
> > >
> > > if everybody agrees, here is the patch against test8 using the second
> > > alternative.
> >
> > How about letting the world see it:
> >
> > -static void create_empty_buffers(struct page *page, struct inode
> > *inode, unsigned long blocksize)
> > +void create_empty_buffers(struct page *page, kdev_t dev, unsigned
> > long blocksize)
>
> Definitely not. Reason: interfaces using kdev_t to indicate block device
> are asking for trouble. Especially when you export them. At least with
> inode we have ->i_bdev and that allows to localize the changes during the
> potential switch. Bare kdev_t that is supplied by caller means additional
> (and large) amount of future changes.
Fine with me, I have to hack ksyms anyway, which I didn't even bother
asking about. But does that mean that Juan's patch is otherwise ok?
> That's one of the reasons why I moved cont_... stuff into fs/buffer.c,
> BTW. I suspect that your stuff also belongs there, but I'll need to look
> at it to comment.
Soon...
--
Daniel
-
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to [EMAIL PROTECTED]
