Re: [PATCH 1/1] mm/nommu: remove unnecessary VMA locking

[David Hildenbrand]

On 01.03.23 20:04, Suren Baghdasaryan wrote:
> Since CONFIG_PER_VMA_LOCK depends on CONFIG_MMU, the changes in nommu
> are not needed. Remove them.
>
> Fixes: bad94decd6a4 ("mm: write-lock VMAs before removing them from VMA tree")
> Reported-by: Hyeonggon Yoo <42.hyeyoo@xxxxxxxxx>
> Link: https://lore.kernel.org/all/Y%2F8CJQGNuMUTdLwP@localhost/
> Signed-off-by: Suren Baghdasaryan <surenb@xxxxxxxxxx>
> ---
> Fix cleanly applies over mm-unstable, SHA in "Fixes" is from that tree.
>
>   mm/nommu.c | 5 -----
>   1 file changed, 5 deletions(-)
>
> diff --git a/mm/nommu.c b/mm/nommu.c
> index 2ab162d773e2..57ba243c6a37 100644
> --- a/mm/nommu.c
> +++ b/mm/nommu.c
> @@ -588,7 +588,6 @@ static int delete_vma_from_mm(struct vm_area_struct *vma)
>   		       current->pid);
>   		return -ENOMEM;
>   	}
> -	vma_start_write(vma);
>   	cleanup_vma_from_mm(vma);
>   
>   	/* remove from the MM's tree and list */
> @@ -1520,10 +1519,6 @@ void exit_mmap(struct mm_struct *mm)
>   	 */
>   	mmap_write_lock(mm);
>   	for_each_vma(vmi, vma) {
> -		/*
> -		 * No need to lock VMA because this is the only mm user and no
> -		 * page fault handled can race with it.
> -		 */
>   		cleanup_vma_from_mm(vma);
>   		delete_vma(mm, vma);
>   		cond_resched();

So, i assume this should be squashed.

Reviewed-by: David Hildenbrand <david@xxxxxxxxxx>


Just a general comment: usually, if review of the original series is 
still going on, it makes a lot more sense to raise such things in the 
original series so the author can fixup while things are still in 
mm-unstable. Once the series is in mm-stable, it's a different story. In 
that case, it is usually good to have the mail subjects be something 
like  "[PATCH mm-stable 1/1] ...".

--
Thanks,

David / dhildenb