From: David Hildenbrand <david@redhat.com>
To: Usama Arif <usamaarif642@gmail.com>, Yu Zhao <yuzhao@google.com>
Cc: akpm@linux-foundation.org, linux-mm@kvack.org,
hannes@cmpxchg.org, riel@surriel.com, shakeel.butt@linux.dev,
roman.gushchin@linux.dev, baohua@kernel.org,
ryan.roberts@arm.com, rppt@kernel.org, willy@infradead.org,
cerasuolodomenico@gmail.com, corbet@lwn.net,
linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
kernel-team@meta.com
Subject: Re: [PATCH 0/6] mm: split underutilized THPs
Date: Thu, 1 Aug 2024 18:27:21 +0200 [thread overview]
Message-ID: <524fb638-73b2-45e7-ae2c-7445d394ed50@redhat.com> (raw)
In-Reply-To: <f5c5f602-718a-4408-95d3-ed114398ac26@gmail.com>
On 01.08.24 18:22, Usama Arif wrote:
>
>
> On 01/08/2024 07:09, Yu Zhao wrote:
>> On Tue, Jul 30, 2024 at 6:54 AM Usama Arif <usamaarif642@gmail.com> wrote:
>>>
>>> The current upstream default policy for THP is always. However, Meta
>>> uses madvise in production as the current THP=always policy vastly
>>> overprovisions THPs in sparsely accessed memory areas, resulting in
>>> excessive memory pressure and premature OOM killing.
>>> Using madvise + relying on khugepaged has certain drawbacks over
>>> THP=always. Using madvise hints mean THPs aren't "transparent" and
>>> require userspace changes. Waiting for khugepaged to scan memory and
>>> collapse pages into THP can be slow and unpredictable in terms of performance
>>> (i.e. you dont know when the collapse will happen), while production
>>> environments require predictable performance. If there is enough memory
>>> available, its better for both performance and predictability to have
>>> a THP from fault time, i.e. THP=always rather than wait for khugepaged
>>> to collapse it, and deal with sparsely populated THPs when the system is
>>> running out of memory.
>>>
>>> This patch-series is an attempt to mitigate the issue of running out of
>>> memory when THP is always enabled. During runtime whenever a THP is being
>>> faulted in or collapsed by khugepaged, the THP is added to a list.
>>> Whenever memory reclaim happens, the kernel runs the deferred_split
>>> shrinker which goes through the list and checks if the THP was underutilized,
>>> i.e. how many of the base 4K pages of the entire THP were zero-filled.
>>> If this number goes above a certain threshold, the shrinker will attempt
>>> to split that THP. Then at remap time, the pages that were zero-filled are
>>> not remapped, hence saving memory. This method avoids the downside of
>>> wasting memory in areas where THP is sparsely filled when THP is always
>>> enabled, while still providing the upside THPs like reduced TLB misses without
>>> having to use madvise.
>>>
>>> Meta production workloads that were CPU bound (>99% CPU utilzation) were
>>> tested with THP shrinker. The results after 2 hours are as follows:
>>>
>>> | THP=madvise | THP=always | THP=always
>>> | | | + shrinker series
>>> | | | + max_ptes_none=409
>>> -----------------------------------------------------------------------------
>>> Performance improvement | - | +1.8% | +1.7%
>>> (over THP=madvise) | | |
>>> -----------------------------------------------------------------------------
>>> Memory usage | 54.6G | 58.8G (+7.7%) | 55.9G (+2.4%)
>>> -----------------------------------------------------------------------------
>>> max_ptes_none=409 means that any THP that has more than 409 out of 512
>>> (80%) zero filled filled pages will be split.
>>>
>>> To test out the patches, the below commands without the shrinker will
>>> invoke OOM killer immediately and kill stress, but will not fail with
>>> the shrinker:
>>>
>>> echo 450 > /sys/kernel/mm/transparent_hugepage/khugepaged/max_ptes_none
>>> mkdir /sys/fs/cgroup/test
>>> echo $$ > /sys/fs/cgroup/test/cgroup.procs
>>> echo 20M > /sys/fs/cgroup/test/memory.max
>>> echo 0 > /sys/fs/cgroup/test/memory.swap.max
>>> # allocate twice memory.max for each stress worker and touch 40/512 of
>>> # each THP, i.e. vm-stride 50K.
>>> # With the shrinker, max_ptes_none of 470 and below won't invoke OOM
>>> # killer.
>>> # Without the shrinker, OOM killer is invoked immediately irrespective
>>> # of max_ptes_none value and kill stress.
>>> stress --vm 1 --vm-bytes 40M --vm-stride 50K
>>>
>>> Patches 1-2 add back helper functions that were previously removed
>>> to operate on page lists (needed by patch 3).
>>> Patch 3 is an optimization to free zapped tail pages rather than
>>> waiting for page reclaim or migration.
>>> Patch 4 is a prerequisite for THP shrinker to not remap zero-filled
>>> subpages when splitting THP.
>>> Patches 6 adds support for THP shrinker.
>>>
>>> (This patch-series restarts the work on having a THP shrinker in kernel
>>> originally done in
>>> https://lore.kernel.org/all/cover.1667454613.git.alexlzhu@fb.com/.
>>> The THP shrinker in this series is significantly different than the
>>> original one, hence its labelled v1 (although the prerequisite to not
>>> remap clean subpages is the same).)
>>>
>>> Alexander Zhu (1):
>>> mm: add selftests to split_huge_page() to verify unmap/zap of zero
>>> pages
>>>
>>> Usama Arif (3):
>>> Revert "memcg: remove mem_cgroup_uncharge_list()"
>>> Revert "mm: remove free_unref_page_list()"
>>> mm: split underutilized THPs
>>>
>>> Yu Zhao (2):
>>> mm: free zapped tail pages when splitting isolated thp
>>> mm: don't remap unused subpages when splitting isolated thp
>>
>> I would recommend shatter [1] instead of splitting so that
>> 1) whoever underutilized their THPs get punished for the overhead;
>> 2) underutilized THPs are kept intact and can be reused by others.
>>
>> [1] https://lore.kernel.org/20240229183436.4110845-3-yuzhao@google.com/
>
> The objective of this series is to reduce memory usage, while trying to keep the performance benefits you get of using THP=always. Punishing any applications performance is the opposite of what I am trying to do here.
> For e.g. if there is only one main application running in production, and its using majority of the THPs, then reducing its performance doesn't make sense.
>
I'm not sure if there would really be a performance degradation
regarding the THP, after all we zap PTEs either way.
Shattering will take longer because real migration is involved IIUC.
> Also, just going through the commit, and found the line "The advantage of shattering is that it keeps the original THP intact" a bit confusing. I am guessing the THP is freed? i.e. if a 2M THP has 10 non-zero filled base pages and the rest are zero-filled, then after shattering we will have 10*4K memory and not 2M+10*4K? Is it the case the THP is reused at next fault?
The idea is (as I understand it) to free the full THP abck to the buddy,
replacing the individual pieces that are kept to freshly allocated
order-0 pages from the buddy.
--
Cheers,
David / dhildenb
next prev parent reply other threads:[~2024-08-01 16:27 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-07-30 12:45 [PATCH 0/6] mm: split underutilized THPs Usama Arif
2024-07-30 12:45 ` [PATCH 1/6] Revert "memcg: remove mem_cgroup_uncharge_list()" Usama Arif
2024-07-30 12:45 ` [PATCH 2/6] Revert "mm: remove free_unref_page_list()" Usama Arif
2024-07-30 12:46 ` [PATCH 3/6] mm: free zapped tail pages when splitting isolated thp Usama Arif
2024-07-30 15:14 ` David Hildenbrand
2024-08-04 19:02 ` Usama Arif
2024-08-05 9:00 ` David Hildenbrand
2024-08-06 9:58 ` Usama Arif
2024-07-30 12:46 ` [PATCH 4/6] mm: don't remap unused subpages " Usama Arif
2024-07-30 18:07 ` Rik van Riel
2024-07-31 17:08 ` Usama Arif
2024-07-30 12:46 ` [PATCH 5/6] mm: add selftests to split_huge_page() to verify unmap/zap of zero pages Usama Arif
2024-07-30 18:10 ` Rik van Riel
2024-08-01 4:45 ` kernel test robot
2024-07-30 12:46 ` [PATCH 6/6] mm: split underutilized THPs Usama Arif
2024-07-30 13:59 ` Randy Dunlap
2024-07-30 14:35 ` [PATCH 0/6] " David Hildenbrand
2024-07-30 15:14 ` Usama Arif
2024-07-30 15:19 ` Usama Arif
2024-07-30 16:11 ` David Hildenbrand
2024-07-30 17:22 ` Usama Arif
2024-07-30 20:25 ` David Hildenbrand
2024-07-31 17:01 ` Usama Arif
2024-07-31 17:51 ` David Hildenbrand
2024-07-31 20:41 ` Usama Arif
2024-08-01 6:36 ` David Hildenbrand
2024-08-04 23:04 ` Usama Arif
[not found] ` <20240806172830.GD322282@cmpxchg.org>
2024-08-06 17:33 ` David Hildenbrand
2024-08-01 6:09 ` Yu Zhao
2024-08-01 15:47 ` David Hildenbrand
2024-08-04 21:54 ` Yu Zhao
2024-08-05 1:32 ` Rik van Riel
2024-08-05 19:51 ` Yu Zhao
2024-08-01 16:22 ` Usama Arif
2024-08-01 16:27 ` David Hildenbrand [this message]
2024-08-04 19:10 ` Usama Arif
2024-08-04 23:32 ` Yu Zhao
2024-08-04 23:23 ` Yu Zhao
2024-08-06 11:18 ` Usama Arif
2024-08-06 17:38 ` Johannes Weiner
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=524fb638-73b2-45e7-ae2c-7445d394ed50@redhat.com \
--to=david@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=baohua@kernel.org \
--cc=cerasuolodomenico@gmail.com \
--cc=corbet@lwn.net \
--cc=hannes@cmpxchg.org \
--cc=kernel-team@meta.com \
--cc=linux-doc@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=riel@surriel.com \
--cc=roman.gushchin@linux.dev \
--cc=rppt@kernel.org \
--cc=ryan.roberts@arm.com \
--cc=shakeel.butt@linux.dev \
--cc=usamaarif642@gmail.com \
--cc=willy@infradead.org \
--cc=yuzhao@google.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).