Re: [PATCH] mm: Re-allow pinning of zero pfns

[Alistair Popple <apopple@nvidia.com>]

David Hildenbrand <david@redhat.com> writes:

> On 28.07.22 10:45, Alistair Popple wrote:
>>
>> Looks like the original patch might need rebasing. I am about to post a
>> clean-up for the tortured logic in check_and_migrate_movable_pages() so
>> can incorporate it there, but I'm wondering what the consensus was for
>> pinning of zero pfn?
>
> We have to keep it working right now, but in most cases (inside
> MAP_PRIVATE regions) it's shaky and undesired.

Ok. Well I've looked at this now so may as well stick

Reviewed-by: Alistair Popple <apopple@nvidia.com>

on it. However I think it needs rebasing, should I send an updated
version?

>>
>> Currently my clean-up will result in PUP returning an error for the zero
>> pfn rather than looping indefinitely in the kernel. However it wasn't
>> clear from this thread if returning an error is ok, or if R/O pinning
>> of the zero pfn should succeed?
>
> I'm working on proper COW breaking in MAP_PRIVATE mappings, which will,
> for example, unshare the shared zeropage and properly replace it by
> exclusive anon pages first in the FOLL_LONGTERM case.