Re: [PATCH 00/10] mm/hugetlb: Make huge_pte_offset() thread-safe for pmd unshare

[Andrew Morton]

On Tue, 29 Nov 2022 14:35:16 -0500 Peter Xu <peterx@xxxxxxxxxx> wrote:

> Based on latest mm-unstable (9ed079378408).
> 
> This can be seen as a follow-up series to Mike's recent hugetlb vma lock
> series for pmd unsharing, but majorly covering safe use of huge_pte_offset.

We're at -rc7 (a -rc8 appears probable this time) and I'm looking to
settle down and stabilize things...

> 
> ...
>
> huge_pte_offset() is always called with mmap lock held with either read or
> write.  It was assumed to be safe but it's actually not.  One race
> condition can easily trigger by: (1) firstly trigger pmd share on a memory
> range, (2) do huge_pte_offset() on the range, then at the meantime, (3)
> another thread unshare the pmd range, and the pgtable page is prone to lost
> if the other shared process wants to free it completely (by either munmap
> or exit mm).

That sounds like a hard-to-hit memory leak, but what we have here is a
user-triggerable use-after-free and an oops.  Ugh.

Could people please prioritize the review and testing of this patchset?