Re: [PATCH v1] fs/proc/task_mmu: Fix loss of young/dirty bits during pagemap scan

[Ryan Roberts]

On 29/04/2024 12:40, Ryan Roberts wrote:
> make_uffd_wp_pte() was previously doing:
> 
>   pte = ptep_get(ptep);
>   ptep_modify_prot_start(ptep);
>   pte = pte_mkuffd_wp(pte);
>   ptep_modify_prot_commit(ptep, pte);
> 
> But if another thread accessed or dirtied the pte between the first 2
> calls, this could lead to loss of that information. Since
> ptep_modify_prot_start() gets and clears atomically, the following is
> the correct pattern and prevents any possible race. Any access after the
> first call would see an invalid pte and cause a fault:
> 
>   pte = ptep_modify_prot_start(ptep);
>   pte = pte_mkuffd_wp(pte);
>   ptep_modify_prot_commit(ptep, pte);
> 
> Fixes: 52526ca7fdb9 ("fs/proc/task_mmu: implement IOCTL to get and optionally clear info about PTEs")
> Signed-off-by: Ryan Roberts <ryan.roberts@xxxxxxx>

I guess this should have cc'ed stable but I forgot to add it. Are you able to
fix this up when you take it, Andrew, or do I need to repost?

> ---
>  fs/proc/task_mmu.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
> index 23fbab954c20..af4bc1da0c01 100644
> --- a/fs/proc/task_mmu.c
> +++ b/fs/proc/task_mmu.c
> @@ -1825,7 +1825,7 @@ static void make_uffd_wp_pte(struct vm_area_struct *vma,
>  		pte_t old_pte;
> 
>  		old_pte = ptep_modify_prot_start(vma, addr, pte);
> -		ptent = pte_mkuffd_wp(ptent);
> +		ptent = pte_mkuffd_wp(old_pte);
>  		ptep_modify_prot_commit(vma, addr, pte, old_pte, ptent);
>  	} else if (is_swap_pte(ptent)) {
>  		ptent = pte_swp_mkuffd_wp(ptent);
> --
> 2.25.1
>