Re: [PATCH v1 10/11] mm/memory: ignore dirty/accessed/soft-dirty bits in folio_pte_batch()

[Ryan Roberts <ryan.roberts@arm.com>]

On 23/01/2024 14:13, David Hildenbrand wrote:
>>> Although now I'm wondering if there is a race here... What happens if a page in
>>> the parent becomes dirty after you have checked it but before you write protect
>>> it? Isn't that already a problem with the current non-batched version? Why do we
>>> even to preserve dirty in the child for private mappings?
>>
>> I suspect, because the parent could zap the anon folio. If the folio is
>> clean, but the PTE dirty, I suspect that we could lose data of the child
>> if we were to evict that clean folio (swapout).
>>
>> So I assume we simply copy the dirty PTE bit, so the system knows that
>> that folio is actually dirty, because one PTE is dirty.
> 
> Oh, and regarding your race concern: it's undefined which page state
> would see if some write is racing with fork, so it also doesn't matter
> if we would copy the PTE dirty bit or not, if it gets set in a racy fashion.

Ahh that makes sense. Thanks.

> 
> I'll not experiment with:

Looks good as long as its still performant.

> 
> From 14e83ff2a422a96ce5701f9c8454a49f9ed947e3 Mon Sep 17 00:00:00 2001
> From: David Hildenbrand <david@redhat.com>
> Date: Sat, 30 Dec 2023 12:54:35 +0100
> Subject: [PATCH] mm/memory: ignore dirty/accessed/soft-dirty bits in
>  folio_pte_batch()
> 
> Let's always ignore the accessed/young bit: we'll always mark the PTE
> as old in our child process during fork, and upcoming users will
> similarly not care.
> 
> Ignore the dirty bit only if we don't want to duplicate the dirty bit
> into the child process during fork. Maybe, we could just set all PTEs
> in the child dirty if any PTE is dirty. For now, let's keep the behavior
> unchanged.
> 
> Ignore the soft-dirty bit only if the bit doesn't have any meaning in
> the src vma.
> 
> Signed-off-by: David Hildenbrand <david@redhat.com>
> ---
>  mm/memory.c | 34 ++++++++++++++++++++++++++++++----
>  1 file changed, 30 insertions(+), 4 deletions(-)
> 
> diff --git a/mm/memory.c b/mm/memory.c
> index 7690994929d26..9aba1b0e871ca 100644
> --- a/mm/memory.c
> +++ b/mm/memory.c
> @@ -953,24 +953,44 @@ static __always_inline void __copy_present_ptes(struct
> vm_area_struct *dst_vma,
>      set_ptes(dst_vma->vm_mm, addr, dst_pte, pte, nr);
>  }
>  
> +/* Flags for folio_pte_batch(). */
> +typedef int __bitwise fpb_t;
> +
> +/* Compare PTEs after pte_mkclean(), ignoring the dirty bit. */
> +#define FPB_IGNORE_DIRTY        ((__force fpb_t)BIT(0))
> +
> +/* Compare PTEs after pte_clear_soft_dirty(), ignoring the soft-dirty bit. */
> +#define FPB_IGNORE_SOFT_DIRTY        ((__force fpb_t)BIT(1))
> +
> +static inline pte_t __pte_batch_clear_ignored(pte_t pte, fpb_t flags)
> +{
> +    if (flags & FPB_IGNORE_DIRTY)
> +        pte = pte_mkclean(pte);
> +    if (likely(flags & FPB_IGNORE_SOFT_DIRTY))
> +        pte = pte_clear_soft_dirty(pte);
> +    return pte_mkold(pte);
> +}
> +
>  /*
>   * Detect a PTE batch: consecutive (present) PTEs that map consecutive
>   * pages of the same folio.
>   *
>   * All PTEs inside a PTE batch have the same PTE bits set, excluding the PFN.
> + * the accessed bit, dirty bit (with FPB_IGNORE_DIRTY) and soft-dirty bit
> + * (with FPB_IGNORE_SOFT_DIRTY).
>   */
>  static inline int folio_pte_batch(struct folio *folio, unsigned long addr,
> -        pte_t *start_ptep, pte_t pte, int max_nr)
> +        pte_t *start_ptep, pte_t pte, int max_nr, fpb_t flags)
>  {
>      unsigned long folio_end_pfn = folio_pfn(folio) + folio_nr_pages(folio);
>      const pte_t *end_ptep = start_ptep + max_nr;
> -    pte_t expected_pte = pte_next_pfn(pte);
> +    pte_t expected_pte = __pte_batch_clear_ignored(pte_next_pfn(pte), flags);
>      pte_t *ptep = start_ptep + 1;
>  
>      VM_WARN_ON_FOLIO(!pte_present(pte), folio);
>  
>      while (ptep != end_ptep) {
> -        pte = ptep_get(ptep);
> +        pte = __pte_batch_clear_ignored(ptep_get(ptep), flags);
>  
>          if (!pte_same(pte, expected_pte))
>              break;
> @@ -1004,6 +1024,7 @@ copy_present_ptes(struct vm_area_struct *dst_vma, struct
> vm_area_struct *src_vma
>  {
>      struct page *page;
>      struct folio *folio;
> +    fpb_t flags = 0;
>      int err, nr;
>  
>      page = vm_normal_page(src_vma, addr, pte);
> @@ -1018,7 +1039,12 @@ copy_present_ptes(struct vm_area_struct *dst_vma, struct
> vm_area_struct *src_vma
>       * by keeping the batching logic separate.
>       */
>      if (unlikely(!*prealloc && folio_test_large(folio) && max_nr != 1)) {
> -        nr = folio_pte_batch(folio, addr, src_pte, pte, max_nr);
> +        if (src_vma->vm_flags & VM_SHARED)
> +            flags |= FPB_IGNORE_DIRTY;
> +        if (!vma_soft_dirty_enabled(src_vma))
> +            flags |= FPB_IGNORE_SOFT_DIRTY;
> +
> +        nr = folio_pte_batch(folio, addr, src_pte, pte, max_nr, flags);
>          folio_ref_add(folio, nr);
>          if (folio_test_anon(folio)) {
>              if (unlikely(folio_try_dup_anon_rmap_ptes(folio, page,