Re: [patch for-5.3 0/4] revert immediate fallback to remote hugepages

From: Vlastimil Babka
Date: Tue Oct 29 2019 - 17:49:42 EST


On 10/29/19 10:33 PM, Andrew Morton wrote:
> On Tue, 29 Oct 2019 16:15:49 +0100 Michal Hocko <mhocko@xxxxxxxxxx> wrote:
>
>>>
>>> 1. local node only THP allocation with no reclaim, just compaction.
>>> 2. for madvised VMA's or when synchronous compaction is enabled always - THP
>>> allocation from any node with effort determined by global defrag setting
>>> and VMA madvise
>>> 3. fallback to base pages on any node
>>>
>>> Signed-off-by: Vlastimil Babka <vbabka@xxxxxxx>
>>
>> I've given this a try and here are the results of my previous testcase
>> (memory full of page cache).
>
> Thanks, I'll queue this for some more testing. At some point we should
> decide on a suitable set of Fixes: tags and a backporting strategy, if any?

I guess this below, as 3f36d8669457 had and this patch is similar kind
of refinement.

Fixes: b39d0ee2632d ("mm, page_alloc: avoid expensive reclaim when
compaction may not succeed")

If accepted, would be nice if it made it to 5.4, or at least 5.4.y (it
will be a LTSS AFAIK). Not sure if we should backport all 5.4 changes to
5.3.y as that will be EOL rather soon after 5.4 final? I guess that's
for David. And for going to older LTSS's we would at least need more
longer-term confidence and tests on real workloads - Andrea?