Re: [PATCH 1/2] mm: thp: relax __GFP_THISNODE for MADV_HUGEPAGE mappings

From: Andrew Morton
Date: Mon Oct 15 2018 - 18:45:04 EST

On Mon, 15 Oct 2018 15:30:17 -0700 (PDT) David Rientjes <rientjes@xxxxxxxxxx> wrote:

> At the risk of beating a dead horse that has already been beaten, what are
> the plans for this patch when the merge window opens?

I'll hold onto it until we've settled on something. Worst case,
Andrea's original is easily backportable.

> It would be rather
> unfortunate for us to start incurring a 14% increase in access latency and
> 40% increase in fault latency.


> Would it be possible to test with my
> patch[*] that does not try reclaim to address the thrashing issue?

Yes please.

And have you been able to test it with the sort of workloads which
Andrea is attempting to address?