Re: [PATCH v3 13/47] mm/huge_memory: fix truncate racing with PMD zapping

[Baolin Wang]

On 10/31/26 5:39 PM, David Hildenbrand (Arm) wrote:
> From: Chris Mason <mason@xxxxxxxxxx>
>
> zap_huge_pmd_folio() removes the rmap before calling folio_mark_dirty().
> If this was the last mapping, truncation can skip unmapping the folio
> and proceed without waiting for the PMD lock, racing with dirtying.
>
> ifs_set_range_dirty() can oops if MADV_DONTNEED races with
> ftruncate() on XFS, when we get the folio set dirty after it was found
> to not be mapped anymore.
>
> Move folio_mark_dirty() and folio_mark_accessed() before rmap removal,
> matching the PTE zap path. This keeps the folio mapped while updating
> its state, so truncation must wait for the page table lock.
>
> Fix generated by an LLM and sent by Chris Mason in private.
>
> Fixes: fe6cf984939d ("mm/huge_memory: transfer the pmd dirty bit to the folio on zap")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Chris Mason <mason@xxxxxxxxxx>
> [ david: simplify comment and description ]
> Reviewed-by: Gregory Price (Meta) <gourry@xxxxxxxxxx>
> Tested-by: Muhammad Usama Anjum <usama.anjum@xxxxxxx>
> Signed-off-by: David Hildenbrand (Arm) <david@xxxxxxxxxx>
> ---

I checked the memory ordering issues, and it looks like the implicit 
barriers provided by the fully ordered atomic operations already ensure 
the required ordering. So LGTM.
Reviewed-by: Baolin Wang <baolin.wang@xxxxxxxxxxxxxxxxx>