From: Usama Arif <usamaarif642@gmail.com>
To: David Hildenbrand <david@redhat.com>,
akpm@linux-foundation.org, linux-mm@kvack.org
Cc: hannes@cmpxchg.org, riel@surriel.com, shakeel.butt@linux.dev,
roman.gushchin@linux.dev, yuzhao@google.com, 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: Tue, 30 Jul 2024 18:22:46 +0100 [thread overview]
Message-ID: <01899bc3-1920-4ff2-a470-decd1c282e38@gmail.com> (raw)
In-Reply-To: <95ed1631-ff62-4627-8dc6-332096e673b4@redhat.com>
On 30/07/2024 17:11, David Hildenbrand wrote:
> On 30.07.24 17:19, Usama Arif wrote:
>>
>>
>> On 30/07/2024 16:14, Usama Arif wrote:
>>>
>>>
>>> On 30/07/2024 15:35, David Hildenbrand wrote:
>>>> On 30.07.24 14:45, Usama Arif 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).)
>>>>
>>>> As shared previously, there is one issue with uffd (even when currently not active for a VMA!), where we must not zap present page table entries.
>>>>
>>>> Something that is always possible (assuming no GUP pins of course, which) is replacing the zero-filled subpages by shared zeropages.
>>>>
>>>> Is that being done in this patch set already, or are we creating pte_none() entries?
>>>>
>>>
>>> I think thats done in Patch 4/6. In function try_to_unmap_unused, we have below which I think does what you are suggesting? i.e. point to shared zeropage and not clear pte for uffd armed vma.
>>>
>>> if (userfaultfd_armed(pvmw->vma)) {
>>> newpte = pte_mkspecial(pfn_pte(page_to_pfn(ZERO_PAGE(pvmw->address)),
>>> pvmw->vma->vm_page_prot));
>>> ptep_clear_flush(pvmw->vma, pvmw->address, pvmw->pte);
>>> set_pte_at(pvmw->vma->vm_mm, pvmw->address, pvmw->pte, newpte);
>>> }
>>
>>
>> Ah are you suggesting userfaultfd_armed(pvmw->vma) will evaluate to false even if its uffd? I think something like below would work in that case.
>
> I remember one ugly case in QEMU with postcopy live-migration where we must not zap zero-filled pages. I am not 100% regarding THP (if it could be enabled at that point), but imagine the following
>
> 1) mmap(), enable THP
> 2) Migrate a bunch of pages from the source during precopy (writing to
> the memory). Might end up creating THPs (during fault/khugepaged)
> 3) Register UFFD on the VMA
> 4) Disable new THPs from forming via MADV_NOHUGEPAGE on the VMA
> 5) Discard any pages that have been re-dirtied or not migrated yet
> 6) Migrate-on-demand any holes using uffd
>
>
> If we discard zero-filled pages between 2) and 3) we might get wrong uffd notifications in 6 for pages that have already been migrated).
>
> I'll have to check if that actually happens in that sequence in QEMU: if QEMU would disable THP right before 2) we would be safe. But I recall that it is not the case :/
>
>
Thanks for the example!
Just to understand the issue better, as I am not very familiar with live-migration code, the problem is only for zero-filled pages that were migrated, right? If a THP is created and a subpage of it was a zero-page that was migrated and its split before VMA is armed with uffd, userfaultfd_armed(pvmw->vma) will return false when splitting and it will become pte_none. And afterwards when the destination faults on it, uffd will see that its pte_clear and will request the zero-page back from source. Uffd will then have to get the page again from source.
If I understand the example correctly, the below diff over patch 6 should be good? i.e. just point to the empty_zero_page instead of doing pte_clear. This should still use the same amount of memory, although ptep_clear_flush means it might be slighly more expensive.
diff --git a/mm/migrate.c b/mm/migrate.c
index 2731ac20ff33..52aa4770fbed 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -206,14 +206,10 @@ static bool try_to_unmap_unused(struct page_vma_mapped_walk *pvmw,
if (dirty)
return false;
- pte_clear_not_present_full(pvmw->vma->vm_mm, pvmw->address, pvmw->pte, false);
-
- if (userfaultfd_armed(pvmw->vma)) {
- newpte = pte_mkspecial(pfn_pte(page_to_pfn(ZERO_PAGE(pvmw->address)),
- pvmw->vma->vm_page_prot));
- ptep_clear_flush(pvmw->vma, pvmw->address, pvmw->pte);
- set_pte_at(pvmw->vma->vm_mm, pvmw->address, pvmw->pte, newpte);
- }
+ newpte = pte_mkspecial(pfn_pte(page_to_pfn(ZERO_PAGE(pvmw->address)),
+ pvmw->vma->vm_page_prot));
+ ptep_clear_flush(pvmw->vma, pvmw->address, pvmw->pte);
+ set_pte_at(pvmw->vma->vm_mm, pvmw->address, pvmw->pte, newpte);
dec_mm_counter(pvmw->vma->vm_mm, mm_counter(folio));
return true;
next prev parent reply other threads:[~2024-07-30 17:22 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 [this message]
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
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=01899bc3-1920-4ff2-a470-decd1c282e38@gmail.com \
--to=usamaarif642@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=baohua@kernel.org \
--cc=cerasuolodomenico@gmail.com \
--cc=corbet@lwn.net \
--cc=david@redhat.com \
--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=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).