On 03/05/2024 01:50, Barry Song wrote:
> From: Barry Song <v-songbaohua@xxxxxxxx>
>
> Should do_swap_page() have the capability to directly map a large folio,
> metadata restoration becomes necessary for a specified number of pages
> denoted as nr. It's important to highlight that metadata restoration is
> solely required by the SPARC platform, which, however, does not enable
> THP_SWAP. Consequently, in the present kernel configuration, there
> exists no practical scenario where users necessitate the restoration of
> nr metadata. Platforms implementing THP_SWAP might invoke this function
> with nr values exceeding 1, subsequent to do_swap_page() successfully
> mapping an entire large folio. Nonetheless, their arch_do_swap_page_nr()
> functions remain empty.
>
> Cc: Khalid Aziz <khalid.aziz@xxxxxxxxxx>
> Cc: "David S. Miller" <davem@xxxxxxxxxxxxx>
> Cc: Andreas Larsson <andreas@xxxxxxxxxxx>
> Signed-off-by: Barry Song <v-songbaohua@xxxxxxxx>
> ---
> include/linux/pgtable.h | 26 ++++++++++++++++++++------
> mm/memory.c | 3 ++-
> 2 files changed, 22 insertions(+), 7 deletions(-)
>
> diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
> index 18019f037bae..463e84c3de26 100644
> --- a/include/linux/pgtable.h
> +++ b/include/linux/pgtable.h
> @@ -1084,6 +1084,15 @@ static inline int pgd_same(pgd_t pgd_a, pgd_t pgd_b)
> })
>
> #ifndef __HAVE_ARCH_DO_SWAP_PAGE
> +static inline void arch_do_swap_page_nr(struct mm_struct *mm,
> + struct vm_area_struct *vma,
> + unsigned long addr,
> + pte_t pte, pte_t oldpte,
> + int nr)
> +{
> +
> +}
> +#else
> /*
> * Some architectures support metadata associated with a page. When a
> * page is being swapped out, this metadata must be saved so it can be
> @@ -1092,12 +1101,17 @@ static inline int pgd_same(pgd_t pgd_a, pgd_t pgd_b)
> * page as metadata for the page. arch_do_swap_page() can restore this
> * metadata when a page is swapped back in.
> */
> -static inline void arch_do_swap_page(struct mm_struct *mm,
> - struct vm_area_struct *vma,
> - unsigned long addr,
> - pte_t pte, pte_t oldpte)
This hook seems to be very similar to arch_swap_restore(), I wonder if it makes
sense to merge them. Out of scope for this patch series though.
> -{
> -
> +static inline void arch_do_swap_page_nr(struct mm_struct *mm,
> + struct vm_area_struct *vma,
> + unsigned long addr,
> + pte_t pte, pte_t oldpte,
> + int nr)
> +{
> + for (int i = 0; i < nr; i++) {
> + arch_do_swap_page(vma->vm_mm, vma, addr + i * PAGE_SIZE,
> + pte_advance_pfn(pte, i),
> + pte_advance_pfn(oldpte, i));
It seems a bit odd to create a batched version of this, but not allow arches to
take advantage. Although I guess your point is that only SPARC implements it and
on that platform nr will always be 1? So no point right now? So this is just a
convenience for do_swap_page()? Makes sense.
Reviewed-by: Ryan Roberts <ryan.roberts@xxxxxxx>
> + }
> }
> #endif
>
> diff --git a/mm/memory.c b/mm/memory.c
> index f033eb3528ba..74cdefd58f5f 100644
> --- a/mm/memory.c
> +++ b/mm/memory.c
> @@ -4266,7 +4266,8 @@ vm_fault_t do_swap_page(struct vm_fault *vmf)
> VM_BUG_ON(!folio_test_anon(folio) ||
> (pte_write(pte) && !PageAnonExclusive(page)));
> set_pte_at(vma->vm_mm, vmf->address, vmf->pte, pte);
> - arch_do_swap_page(vma->vm_mm, vma, vmf->address, pte, vmf->orig_pte);
> + arch_do_swap_page_nr(vma->vm_mm, vma, vmf->address,
> + pte, vmf->orig_pte, 1);
>
> folio_unlock(folio);
> if (folio != swapcache && swapcache) {
