Re: [PATCH] mm/rmap: synchronize lock and unlock target in anon_vma_clone

From: Lorenzo Stoakes (ARM)

Date: Fri Aug 14 2026 - 02:53:20 EST


On Fri, Aug 14, 2026 at 03:30:51PM +0900, Eric Kim wrote:
> Currently, in anon_vma_clone(), src vma's anon_vma is assigned to
> active_anon_vma and is used when unlocking anon_vma after linking new
> AVCs. However, the anon_vma is locked using src->anon_vma, instead of
> active_anon_vma, making the lock and unlock target inconsistent.

IOW, I added active_anon_vma as an alias to src->anon_vma but didn't use it
consistently in the function :)

>
> Use active_anon_vma for both locking and unlocking.
>
> Signed-off-by: Eric Kim <seohyun.kim@xxxxxxxxxx>

LGTM, so:

Reviewed-by: Lorenzo Stoakes (ARM) <ljs@xxxxxxxxxx>

> ---
> mm/rmap.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/mm/rmap.c b/mm/rmap.c
> index 1c77d5dc06e9..f3fadfb69c7f 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -350,7 +350,7 @@ int anon_vma_clone(struct vm_area_struct *dst, struct vm_area_struct *src,
> * Now link the anon_vma's back to the newly inserted AVCs.
> * Note that all anon_vma's share the same root.
> */
> - anon_vma_lock_write(src->anon_vma);
> + anon_vma_lock_write(active_anon_vma);
> list_for_each_entry_reverse(avc, &dst->anon_vma_chain, same_vma) {
> struct anon_vma *anon_vma = avc->anon_vma;
>
> --
> 2.55.0
>

--
Cheers, Lorenzo