Re: invalidate_inode_pages in 2.5.32/3

Daniel Phillips (phillips@arcor.de)
Tue, 24 Sep 2002 07:09:20 +0200


On Tuesday 24 September 2002 00:43, Trond Myklebust wrote:
> >>>>> " " == Daniel Phillips <phillips@arcor.de> writes:
>
> >> Note that in doing so, we do not want to invalidate any reads
> >> or writes that may have been already scheduled. The existing
> >> mapping still would need to hang around long enough to permit
> >> them to complete.
>
> > With the mechanism I described above, that would just work.
> > The fault path would do lock_page, thus waiting for the IO to
> > complete.
>
> NFS writes do not hold the page lock until completion. How would you
> expect to be able to coalesce writes to the same page if they did?

Coalesce before initiating writeout? I don't see why NFS should be special
in this regard, or why it should not leave a page locked until IO has
completed, like other filesystems. Could you please explain?

-- 
Daniel
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/