Message-ID: <18BEDC9A-77D2-4E9B-BF5A-90F7C789D535@nvidia.com>
Date: Mon, 09 Jun 2025 09:19:08 -0400
From: Zi Yan <ziy@...dia.com>
To: Usama Arif <usamaarif642@...il.com>
Cc: Andrew Morton <akpm@...ux-foundation.org>, david@...hat.com,
 linux-mm@...ck.org, hannes@...xchg.org, shakeel.butt@...ux.dev,
 riel@...riel.com, baolin.wang@...ux.alibaba.com, lorenzo.stoakes@...cle.com,
 Liam.Howlett@...cle.com, npache@...hat.com, ryan.roberts@....com,
 dev.jain@....com, hughd@...gle.com, linux-kernel@...r.kernel.org,
 linux-doc@...r.kernel.org, kernel-team@...a.com,
 Juan Yescas <jyescas@...gle.com>, Breno Leitao <leitao@...ian.org>
Subject: Re: [RFC] mm: khugepaged: use largest enabled hugepage order for
 min_free_kbytes

On 9 Jun 2025, at 7:13, Usama Arif wrote:

> On 06/06/2025 17:10, Zi Yan wrote:
>> On 6 Jun 2025, at 11:38, Usama Arif wrote:
>>
>>> On 06/06/2025 16:18, Zi Yan wrote:
>>>> On 6 Jun 2025, at 10:37, Usama Arif wrote:
>>>>
>>>>> On arm64 machines with 64K PAGE_SIZE, the min_free_kbytes and hence the
>>>>> watermarks are evaluated to extremely high values, for e.g. a server with
>>>>> 480G of memory, only 2M mTHP hugepage size set to madvise, with the rest
>>>>> of the sizes set to never, the min, low and high watermarks evaluate to
>>>>> 11.2G, 14G and 16.8G respectively.
>>>>> In contrast for 4K PAGE_SIZE of the same machine, with only 2M THP hugepage
>>>>> size set to madvise, the min, low and high watermarks evaluate to 86M, 566M
>>>>> and 1G respectively.
>>>>> This is because set_recommended_min_free_kbytes is designed for PMD
>>>>> hugepages (pageblock_order = min(HPAGE_PMD_ORDER, PAGE_BLOCK_ORDER)).
>>>>> Such high watermark values can cause performance and latency issues in
>>>>> memory bound applications on arm servers that use 64K PAGE_SIZE, eventhough
>>>>> most of them would never actually use a 512M PMD THP.
>>>>>
>>>>> Instead of using HPAGE_PMD_ORDER for pageblock_order use the highest large
>>>>> folio order enabled in set_recommended_min_free_kbytes.
>>>>> With this patch, when only 2M THP hugepage size is set to madvise for the
>>>>> same machine with 64K page size, with the rest of the sizes set to never,
>>>>> the min, low and high watermarks evaluate to 2.08G, 2.6G and 3.1G
>>>>> respectively. When 512M THP hugepage size is set to madvise for the same
>>>>> machine with 64K page size, the min, low and high watermarks evaluate to
>>>>> 11.2G, 14G and 16.8G respectively, the same as without this patch.
>>>>
>>>> Getting pageblock_order involved here might be confusing. I think you just
>>>> want to adjust min, low and high watermarks to reasonable values.
>>>> Is it OK to rename min_thp_pageblock_nr_pages to min_nr_free_pages_per_zone
>>>> and move MIGRATE_PCPTYPES * MIGRATE_PCPTYPES inside? Otherwise, the changes
>>>> look reasonable to me.
>>>
>>> Hi Zi,
>>>
>>> Thanks for the review!
>>>
>>> I forgot to change it in another place, sorry about that! So can't move
>>> MIGRATE_PCPTYPES * MIGRATE_PCPTYPES into the combined function.
>>> Have added the additional place where min_thp_pageblock_nr_pages() is called
>>> as a fixlet here:
>>> https://lore.kernel.org/all/a179fd65-dc3f-4769-9916-3033497188ba@gmail.com/
>>>
>>> I think atleast in this context the orginal name pageblock_nr_pages isn't
>>> correct as its min(HPAGE_PMD_ORDER, PAGE_BLOCK_ORDER).
>>> The new name min_thp_pageblock_nr_pages is also not really good, so happy
>>> to change it to something appropriate.
>>
>> Got it. pageblock is the defragmentation granularity. If user only wants
>> 2MB mTHP, maybe pageblock order should be adjusted. Otherwise,
>> kernel will defragment at 512MB granularity, which might not be efficient.
>> Maybe make pageblock_order a boot time parameter?
>>
>> In addition, we are mixing two things together:
>> 1. min, low, and high watermarks: they affect when memory reclaim and compaction
>>    will be triggered;
>> 2. pageblock order: it is the granularity of defragmentation for creating
>>    mTHP/THP.
>>
>> In your use case, you want to lower watermarks, right? Considering what you
>> said below, I wonder if we want a way of enforcing vm.min_free_kbytes,
>> like a new sysctl knob, vm.force_min_free_kbytes (yeah the suggestion
>> is lame, sorry).
>>
>> I think for 2, we might want to decouple pageblock order from defragmentation
>> granularity.
>>
>
> This is a good point. I only did it for the watermarks in the RFC, but there
> is no reason that the defrag granularity is done in 512M chunks and is probably
> very inefficient to do so?
>
> Instead of replacing the pageblock_nr_pages for just set_recommended_min_free_kbytes,
> maybe we just need to change the definition of pageblock_order in [1] to take into
> account the highest large folio order enabled instead of HPAGE_PMD_ORDER?

Ideally, yes. But pageblock migratetypes are stored in a fixed size array
determined by pageblock_order at boot time (see usemap_size() in mm/mm_init.c).
Changing pageblock_order at runtime means we will need to resize pageblock
migratetypes array, which is a little unrealistic. In a system with GBs or TBs
memory, reducing pageblock_order by 1 means doubling pageblock migratetypes
array and replicating one pageblock migratetypes to two; increasing pageblock
order by 1 means halving the array and splitting a pageblock into two.
The former, if memory is enough, might be easy, but the latter is a little
involved, since for a pageblock with both movable and unmovable pages,
you will need to check all pages to decide the migratetypes of the after-split
pageblocks to make sure pageblock migratetype matches the pages inside that
pageblock.


>
> [1] https://elixir.bootlin.com/linux/v6.15.1/source/include/linux/pageblock-flags.h#L50
>
> I really want to avoid coming up with a solution that requires changing a Kconfig or needs
> kernel commandline to change. It would mean a reboot whenever a different workload
> runs on a server that works optimally with a different THP size, and that would make
> workload orchestration a nightmare.
>

As I said above, changing pageblock order at runtime might not be easy. But
changing defragmentation granularity should be fine, since it just changes
the range of memory compaction. That is the reason of my proposal,
decoupling pageblock order from defragmentation granularity. We probably
need to do some experiments to see the impact of the decoupling, as I
imagine defragmenting a range smaller than pageblock order is fine, but
defragmenting a range larger than pageblock order might cause issues
if there is any unmovable pageblock within that range. Since it is very likely
unmovable pages reside in an unmovable pageblock and lead to a defragmentation
failure.


--
Best Regards,
Yan, Zi

Powered by blists - more mailing lists