Message-ID: <c3c4395c-62eb-4862-9258-13ffc50bc060@redhat.com>
Date: Fri, 2 May 2025 17:39:09 +0200
From: David Hildenbrand <david@...hat.com>
To: Lorenzo Stoakes <lorenzo.stoakes@...cle.com>
Cc: Jann Horn <jannh@...gle.com>, Nico Pache <npache@...hat.com>,
 linux-mm@...ck.org, linux-doc@...r.kernel.org, linux-kernel@...r.kernel.org,
 linux-trace-kernel@...r.kernel.org, akpm@...ux-foundation.org,
 corbet@....net, rostedt@...dmis.org, mhiramat@...nel.org,
 mathieu.desnoyers@...icios.com, baohua@...nel.org,
 baolin.wang@...ux.alibaba.com, ryan.roberts@....com, willy@...radead.org,
 peterx@...hat.com, ziy@...dia.com, wangkefeng.wang@...wei.com,
 usamaarif642@...il.com, sunnanyong@...wei.com, vishal.moola@...il.com,
 thomas.hellstrom@...ux.intel.com, yang@...amperecomputing.com,
 kirill.shutemov@...ux.intel.com, aarcange@...hat.com, raquini@...hat.com,
 dev.jain@....com, anshuman.khandual@....com, catalin.marinas@....com,
 tiwai@...e.de, will@...nel.org, dave.hansen@...ux.intel.com, jack@...e.cz,
 cl@...two.org, jglisse@...gle.com, surenb@...gle.com, zokeefe@...gle.com,
 hannes@...xchg.org, rientjes@...gle.com, mhocko@...e.com,
 rdunlap@...radead.org, Liam.Howlett@...cle.com
Subject: Re: [PATCH v5 07/12] khugepaged: add mTHP support

On 02.05.25 17:24, Lorenzo Stoakes wrote:
> On Fri, May 02, 2025 at 05:18:54PM +0200, David Hildenbrand wrote:
>> On 02.05.25 14:50, Jann Horn wrote:
>>> On Fri, May 2, 2025 at 8:29 AM David Hildenbrand <david@...hat.com> wrote:
>>>> On 02.05.25 00:29, Nico Pache wrote:
>>>>> On Wed, Apr 30, 2025 at 2:53 PM Jann Horn <jannh@...gle.com> wrote:
>>>>>>
>>>>>> On Mon, Apr 28, 2025 at 8:12 PM Nico Pache <npache@...hat.com> wrote:
>>>>>>> Introduce the ability for khugepaged to collapse to different mTHP sizes.
>>>>>>> While scanning PMD ranges for potential collapse candidates, keep track
>>>>>>> of pages in KHUGEPAGED_MIN_MTHP_ORDER chunks via a bitmap. Each bit
>>>>>>> represents a utilized region of order KHUGEPAGED_MIN_MTHP_ORDER ptes. If
>>>>>>> mTHPs are enabled we remove the restriction of max_ptes_none during the
>>>>>>> scan phase so we dont bailout early and miss potential mTHP candidates.
>>>>>>>
>>>>>>> After the scan is complete we will perform binary recursion on the
>>>>>>> bitmap to determine which mTHP size would be most efficient to collapse
>>>>>>> to. max_ptes_none will be scaled by the attempted collapse order to
>>>>>>> determine how full a THP must be to be eligible.
>>>>>>>
>>>>>>> If a mTHP collapse is attempted, but contains swapped out, or shared
>>>>>>> pages, we dont perform the collapse.
>>>>>> [...]
>>>>>>> @@ -1208,11 +1211,12 @@ static int collapse_huge_page(struct mm_struct *mm, unsigned long address,
>>>>>>>            vma_start_write(vma);
>>>>>>>            anon_vma_lock_write(vma->anon_vma);
>>>>>>>
>>>>>>> -       mmu_notifier_range_init(&range, MMU_NOTIFY_CLEAR, 0, mm, address,
>>>>>>> -                               address + HPAGE_PMD_SIZE);
>>>>>>> +       mmu_notifier_range_init(&range, MMU_NOTIFY_CLEAR, 0, mm, _address,
>>>>>>> +                               _address + (PAGE_SIZE << order));
>>>>>>>            mmu_notifier_invalidate_range_start(&range);
>>>>>>>
>>>>>>>            pmd_ptl = pmd_lock(mm, pmd); /* probably unnecessary */
>>>>>>> +
>>>>>>>            /*
>>>>>>>             * This removes any huge TLB entry from the CPU so we won't allow
>>>>>>>             * huge and small TLB entries for the same virtual address to
>>>>>>
>>>>>> It's not visible in this diff, but we're about to do a
>>>>>> pmdp_collapse_flush() here. pmdp_collapse_flush() tears down the
>>>>>> entire page table, meaning it tears down 2MiB of address space; and it
>>>>>> assumes that the entire page table exclusively corresponds to the
>>>>>> current VMA.
>>>>>>
>>>>>> I think you'll need to ensure that the pmdp_collapse_flush() only
>>>>>> happens for full-size THP, and that mTHP only tears down individual
>>>>>> PTEs in the relevant range. (That code might get a bit messy, since
>>>>>> the existing THP code tears down PTEs in a detached page table, while
>>>>>> mTHP would have to do it in a still-attached page table.)
>>>>> Hi Jann!
>>>>>
>>>>> I was under the impression that this is needed to prevent GUP-fast
>>>>> races (and potentially others).
>>>
>>> Why would you need to touch the PMD entry to prevent GUP-fast races for mTHP?
>>>
>>>>> As you state here, conceptually the PMD case is, detach the PMD, do
>>>>> the collapse, then reinstall the PMD (similarly to how the system
>>>>> recovers from a failed PMD collapse). I tried to keep the current
>>>>> locking behavior as it seemed the easiest way to get it right (and not
>>>>> break anything). So I keep the PMD detaching and reinstalling for the
>>>>> mTHP case too. As Hugh points out I am releasing the anon lock too
>>>>> early. I will comment further on his response.
>>>
>>> As I see it, you're not "keeping" the current locking behavior; you're
>>> making a big implicit locking change by reusing a codepath designed
>>> for PMD THP for mTHP, where the page table may not be exclusively
>>> owned by one VMA.
>>
>> That is not the intention. The intention in this series (at least as we
>> discussed) was to not do it across VMAs; that is considered the next logical
>> step (which will be especially relevant on arm64 IMHO).
>>
>>>
>>>>> As I familiarize myself with the code more, I do see potential code
>>>>> improvements/cleanups and locking improvements, but I was going to
>>>>> leave those to a later series.
>>>>
>>>> Right, the simplest approach on top of the current PMD collapse is to do
>>>> exactly what we do in the PMD case, including the locking: which
>>>> apparently is no completely the same yet :).
>>>>
>>>> Instead of installing a PMD THP, we modify the page table and remap that.
>>>>
>>>> Moving from the PMD lock to the PTE lock will not make a big change in
>>>> practice for most cases: we already must disable essentially all page
>>>> table walkers (vma lock, mmap lock in write, rmap lock in write).
>>>>
>>>> The PMDP clear+flush is primarily to disable the last possible set of
>>>> page table walkers: (1) HW modifications and (2) GUP-fast.
>>>>
>>>> So after the PMDP clear+flush we know that (A) HW can not modify the
>>>> pages concurrently and (B) GUP-fast cannot succeed anymore.
>>>>
>>>> The issue with PTEP clear+flush is that we will have to remember all PTE
>>>> values, to reset them if anything goes wrong. Using a single PMD value
>>>> is arguably simpler. And then, the benefit vs. complexity is unclear.
>>>>
>>>> Certainly something to look into later, but not a requirement for the
>>>> first support,
>>>
>>> As I understand, one rule we currently have in MM is that an operation
>>> that logically operates on one VMA (VMA 1) does not touch the page
>>> tables of other VMAs (VMA 2) in any way, except that it may walk page
>>> tables that cover address space that intersects with both VMA 1 and
>>> VMA 2, and create such page tables if they are missing.
>>
>> Yes, absolutely. That must not happen. And I think I raised it as a problem
>> in reply to one of Dev's series.
>>
>> If this series does not rely on that it must be fixed.
>>
>>>
>>> This proposed patch changes that, without explicitly discussing this
>>> locking change.
>>
>> Yes, that must not happen. We must not zap a PMD to temporarily replace it
>> with a pmd_none() entry if any other sane page table walker could stumble
>> over it.
>>
>> This includes another VMA that is not write-locked that could span the PMD.
> 
> I feel like we should document these restrictions somewhere :)
> 
> Perhaps in a new page table walker doc, or on the
> https://origin.kernel.org/doc/html/latest/mm/process_addrs.html page.
> 
> Which sounds like I'm volunteering myself to do so doesn't it...
> 
> [adds to todo...]

:) that would be nice.

Yeah, I mean this is very subtle, but essentially: unless you exclude 
all page table walkers (well, okay, HW and GUP-fast are a bit special), 
temporarily replacing a present PTE by pte_none() will cause trouble. 
Same for PMDs.

That's also one of the problems of looking into only using the PTE table 
lock and not any other heavy-weight locking in khugepaged. As soon as we 
temporarily unlock the PTE table, but have to temporarily unmap the PTEs 
(->pte_none()) as well, we need something else (e.g., migration entry) 
to tell other page table walkers to back off and wait for us to re-take 
the PTL lock and finish. Zapping must still be able to continue, at 
which point it gets hairy ...

... which is why I'm hoping that we can play with that once we have the 
basics running. With the intend of reusing the existing locking scheme 
-> single VMA spans the PMD.

-- 
Cheers,

David / dhildenb

Powered by blists - more mailing lists