Re: [PATCH v2 0/9] Add support for shared PTEs across processes

[David Hildenbrand <david@redhat.com>]

On 13.07.22 19:58, Mike Kravetz wrote:
> On 07/13/22 16:00, David Hildenbrand wrote:
>> On 08.07.22 21:36, Khalid Aziz wrote:
>>> On 7/8/22 05:47, David Hildenbrand wrote:
>>>> On 02.07.22 06:24, Andrew Morton wrote:
>>>>> On Wed, 29 Jun 2022 16:53:51 -0600 Khalid Aziz <khalid.aziz@oracle.com> wrote:
>>
>>> suggestion to extend hugetlb PMD sharing was discussed briefly. Conclusion from that discussion and earlier discussion 
>>> on mailing list was hugetlb PMD sharing is built with special case code in too many places in the kernel and it is 
>>> better to replace it with something more general purpose than build even more on it. Mike can correct me if I got that 
>>> wrong.
>>
>> Yes, I pushed for the removal of that yet-another-hugetlb-special-stuff,
>> and asked the honest question if we can just remove it and replace it by
>> something generic in the future. And as I learned, we most probably
>> cannot rip that out without affecting existing user space. Even
>> replacing it by mshare() would degrade existing user space.
>>
>> So the natural thing to reduce page table consumption (again, what this
>> cover letter talks about) for user space (semi- ?)automatically for
>> MAP_SHARED files is to factor out what hugetlb has, and teach generic MM
>> code to cache and reuse page tables (PTE and PMD tables should be
>> sufficient) where suitable.
>>
>> For reasonably aligned mappings and mapping sizes, it shouldn't be too
>> hard (I know, locking ...), to cache and reuse page tables attached to
>> files -- similar to what hugetlb does, just in a generic way. We might
>> want a mechanism to enable/disable this for specific processes and/or
>> VMAs, but these are minor details.
>>
>> And that could come for free for existing user space, because page
>> tables, and how they are handled, would just be an implementation detail.
>>
>>
>> I'd be really interested into what the major roadblocks/downsides
>> file-based page table sharing has. Because I am not convinced that a
>> mechanism like mshare() -- that has to be explicitly implemented+used by
>> user space -- is required for that.
> 
> Perhaps this is an 'opportunity' for me to write up in detail how
> hugetlb pmd sharing works.  As you know, I have been struggling with
> keeping that working AND safe AND performant. 

Yes, and I have your locking-related changes in my inbox marked as "to
be reviewed" :D Sheding some light on that would be highly appreciated,
especially, how hugetlb-specific it currently is and for which reason.

-- 
Thanks,

David / dhildenb