Re: [PATCH v5 2/5] mm: Add generic set_ptes()

["Kirill A. Shutemov" <kirill@shutemov.name>]

On Tue, Feb 07, 2023 at 07:49:32PM +0000, Matthew Wilcox (Oracle) wrote:
> This set_ptes should work for most architectures.  Those that need to
> set a special bit in their PTEs or have their PFNs located at a different
> offset from PAGE_SHIFT will need to override it.
> 
> Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> ---
>  include/linux/pgtable.h | 27 +++++++++++++++++++++++++++
>  1 file changed, 27 insertions(+)
> 
> diff --git a/include/linux/pgtable.h b/include/linux/pgtable.h
> index c63cd44777ec..e1804d23e7c4 100644
> --- a/include/linux/pgtable.h
> +++ b/include/linux/pgtable.h
> @@ -1439,6 +1439,33 @@ static inline int pmd_protnone(pmd_t pmd)
>  
>  #endif /* CONFIG_MMU */
>  
> +#ifndef set_ptes
> +/**
> + * set_ptes - Map consecutive pages to a contiguous range of addresses.
> + * @mm: Address space to map the pages into.
> + * @addr: Address to map the first page at.
> + * @ptep: Page table pointer for the first entry.
> + * @pte: Page table entry for the first page.
> + * @nr: Number of pages to map.
> + *
> + * Context: The caller holds the page table lock.  The PTEs all lie
> + * within a single PMD (and VMA, and folio).
> + */
> +static inline void set_ptes(struct mm_struct *mm, unsigned long addr,
> +		pte_t *ptep, pte_t pte, unsigned int nr)
> +{
> +	for (;;) {
> +		set_pte_at(mm, addr, ptep, pte);
> +		if (--nr == 0)
> +			break;

Maybe do { ... } while (--nr); instead of the for()?

> +		ptep++;
> +		addr += PAGE_SIZE;
> +		/* This works for x86.  Check how PTEs are encoded */
> +		pte = __pte(pte_val(pte) + PAGE_SIZE);

Looks like it deserves own helper. Something like

		pte = pte_next(pte);

> +	}
> +}
> +#endif
> +
>  #ifdef CONFIG_HAVE_ARCH_HUGE_VMAP
>  
>  #ifndef __PAGETABLE_P4D_FOLDED
> -- 
> 2.35.1
> 
> 

-- 
  Kiryl Shutsemau / Kirill A. Shutemov