Re: [RFC][PATCH v3 1/6] mm: Don't unmap gup()ed page

[Nick Piggin]

On Wednesday 15 April 2009 18:22:32 Nick Piggin wrote:
> On Wednesday 15 April 2009 18:05:54 KOSAKI Motohiro wrote:
> > Hi
> > 
> > > On Wednesday 15 April 2009 00:32:52 Andrea Arcangeli wrote:
> > > > On Wed, Apr 15, 2009 at 12:26:34AM +1000, Nick Piggin wrote:
> > > > > Andrea: I didn't veto that set_bit change of yours as such. I just
> > > > 
> > > > I know you didn't ;)
> > > > 
> > > > > noted there could be more atomic operations. Actually I would
> > > > > welcome more comparison between our two approaches, but they seem
> > > > 
> > > > Agree about the welcome of comparison, it'd be nice to measure it the
> > > > enterprise workloads that showed the gup_fast gain in the first place.
> > > 
> > > I think we should be able to ask IBM to run some tests, provided
> > > they still have machines available to do so. Although I don't want
> > > to waste their time so we need to have something that has got past
> > > initial code review and has a chance of being merged.
> > > 
> > > If we get that far, then I can ask them to run tests definitely.
> > 
> > Oh, it seem very charming idea.
> > Nick, I hope to help your patch's rollup. It makes good comparision, I think.
> > Is there my doable thing?
> 
> Well, I guess review and testing. There are few possibilities for
> reducing the cases where we have to de-cow (or increasing the
> cases where we can WP-on-fork), which I'd like to experiment with,
> but I don't know how much it will help...

Oh, I forgot, it needs to be updated for hugepages as well, I'd say.
--
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/