Re: [PATCH v10 00/13] khugepaged: mTHP support

[David Hildenbrand <david@redhat.com>]

On 05.09.25 14:31, Usama Arif wrote:
> 
> 
> On 05/09/2025 12:55, David Hildenbrand wrote:
>> On 05.09.25 13:48, Lorenzo Stoakes wrote:
>>> On Wed, Sep 03, 2025 at 08:54:39PM -0600, Nico Pache wrote:
>>>> On Tue, Sep 2, 2025 at 2:23 PM Usama Arif <usamaarif642@gmail.com> wrote:
>>>>>>>> So I question the utility of max_ptes_none. If you can't tame page faults, then there is only
>>>>>>>> limited sense in taming khugepaged. I think there is vale in setting max_ptes_none=0 for some
>>>>>>>> corner cases, but I am yet to learn why max_ptes_none=123 would make any sense.
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> For PMD mapped THPs with THP shrinker, this has changed. You can basically tame pagefaults, as when you encounter
>>>>>>> memory pressure, the shrinker kicks in if the value is less than HPAGE_PMD_NR -1 (i.e. 511 for x86), and
>>>>>>> will break down those hugepages and free up zero-filled memory.
>>>>>>
>>>>>> You are not really taming page faults, though, you are undoing what page faults might have messed up :)
>>>>>>
>>>>>> I have seen in our prod workloads where
>>>>>>> the memory usage and THP usage can spike (usually when the workload starts), but with memory pressure,
>>>>>>> the memory usage is lower compared to with max_ptes_none = 511, while still still keeping the benefits
>>>>>>> of THPs like lower TLB misses.
>>>>>>
>>>>>> Thanks for raising that: I think the current behavior is in place such that you don't bounce back-and-forth between khugepaged collapse and shrinker-split.
>>>>>>
>>>>>
>>>>> Yes, both collapse and shrinker split hinge on max_ptes_none to prevent one of these things thrashing the effect of the other.
>>>> I believe with mTHP support in khugepaged, the max_ptes_none value in
>>>> the shrinker must also leverage the 'order' scaling to properly
>>>> prevent thrashing.
>>>
>>> No please do not extend this 'scalling' stuff somewhere else, it's really horrid.
>>>
>>> We have to find an alternative to that, it's extremely confusing in what is
>>> already extremely confusing THP code.
>>>
>>> As I said before, if we can't have a boolean we need another interface, which
>>> makes most sense to be a ratio or in practice, a percentage sysctl.
>>>
>>> Speaking with David off-list, maybe the answer - if we must have this - is to
>>> add a new percentage interface and have this in lock-step with the existing
>>> max_ptes_none interface. One updates the other, but internally we're just using
>>> the percentage value.
>>
>> Yes, I'll try hacking something up and sending it as an RFC.
>>
>>>
>>>> I've been testing a patch for this that I might include in the V11.
>>>>>
>>>>>> There are likely other ways to achieve that, when we have in mind that the thp shrinker will install zero pages and max_ptes_none includes
>>>>>> zero pages.
>>>>>>
>>>>>>>
>>>>>>> I do agree that the value of max_ptes_none is magical and different workloads can react very differently
>>>>>>> to it. The relationship is definitely not linear. i.e. if I use max_ptes_none = 256, it does not mean
>>>>>>> that the memory regression of using THP=always vs THP=madvise is halved.
>>>>>>
>>>>>> To which value would you set it? Just 510? 0?
>>
>> Sorry, I missed Usama's reply. Thanks Usama!
>>
>>>>>>
>>>>>
>>>>> There are some very large workloads in the meta fleet that I experimented with and found that having
>>>>> a small value works out. I experimented with 0, 51 (10%) and 256 (50%). 51 was found to be an optimal
>>>>> comprimise in terms of application metrics improving, having an acceptable amount of memory regression and
>>>>> improved system level metrics (lower TLB misses, lower page faults). I am sure there was a better value out
>>>>> there for these workloads, but not possible to experiment with every value.
>>>
>>> (->Usama) It's a pity that such workloads exist. But then the percentage solution should work.
>>
>> Good. So if there is no strong case for > 255, that's already valuable for mTHP.
>>
> 
> tbh the default value of 511 is horrible. I have thought about sending a patch to change it to 0 as default
> in upstream for sometime, but it might mean that people who upgrade their kernel might suddenly see
> their memory not getting hugified and it could be confusing for them?

511 is just what a page fault would have done, so I think it makes 
perfect sense. More than anything else, actually.

It's just not optimal in many cases.

-- 
Cheers

David / dhildenb