Re: [RFC] [PATCH 2/4] dio: add page locking for direct I/O

[Christoph Lameter]

On Tue, 17 Aug 2010, Andi Kleen wrote:

> On Tue, Aug 17, 2010 at 09:46:56AM -0400, Jeff Moyer wrote:
> > Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx> writes:
> >
> > > BTW, from the discussion with Christoph I noticed my misunderstanding
> > > about the necessity of additional page locking. It would seem that
> > > without page locking there is no danger of racing between direct I/O and
> > > page migration. So I retract this additional locking patch-set.
> >
> > OK, great!  ;-)
>
> Well it sounds like we still may need something. It isn't good if O_DIRECT
> can starve (or DoS) migration.

Anything that increments a page refcount can interfere with migration and
cause migration to be aborted and retried later. Not something new. There
never was any guarantee that migration must succeed.


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