Re: [PATCH V2] arm64/mm: Define PTDESC_ORDER

From: Mark Rutland
Date: Mon Mar 10 2025 - 06:29:11 EST

On Mon, Mar 10, 2025 at 09:31:15AM +0530, Anshuman Khandual wrote:
> diff --git a/arch/arm64/include/asm/kernel-pgtable.h b/arch/arm64/include/asm/kernel-pgtable.h
> index fd5a08450b12..78c7e03a0e35 100644
> --- a/arch/arm64/include/asm/kernel-pgtable.h
> +++ b/arch/arm64/include/asm/kernel-pgtable.h
> @@ -45,11 +45,14 @@
> #define SPAN_NR_ENTRIES(vstart, vend, shift) \
> ((((vend) - 1) >> (shift)) - ((vstart) >> (shift)) + 1)
>
> -#define EARLY_ENTRIES(vstart, vend, shift, add) \
> - (SPAN_NR_ENTRIES(vstart, vend, shift) + (add))
> +/* Number of VA bits resolved by a single translation table level */
> +#define PTDESC_TABLE_SHIFT (PAGE_SHIFT - PTDESC_ORDER)

To be clear, when I suggested adding PTDESC_TABLE_SHIFT, I expected that
it would be used consistently in place of (PAGE_SHIFT - PTDESC_ORDER),
and not only replacing that within the EARLY_ENTRIES() macro
specifically.

Mark.

> -#define EARLY_LEVEL(lvl, lvls, vstart, vend, add) \
> - (lvls > lvl ? EARLY_ENTRIES(vstart, vend, SWAPPER_BLOCK_SHIFT + lvl * (PAGE_SHIFT - 3), add) : 0)
> +#define EARLY_ENTRIES(lvl, vstart, vend) \
> + SPAN_NR_ENTRIES(vstart, vend, SWAPPER_BLOCK_SHIFT + lvl * PTDESC_TABLE_SHIFT)
> +
> +#define EARLY_LEVEL(lvl, lvls, vstart, vend, add) \
> + ((lvls) > (lvl) ? EARLY_ENTRIES(lvl, vstart, vend) + (add) : 0)
>
> #define EARLY_PAGES(lvls, vstart, vend, add) (1 /* PGDIR page */ \
> + EARLY_LEVEL(3, (lvls), (vstart), (vend), add) /* each entry needs a next level page table */ \
> diff --git a/arch/arm64/include/asm/pgtable-hwdef.h b/arch/arm64/include/asm/pgtable-hwdef.h
> index a9136cc551cc..3c544edc3968 100644
> --- a/arch/arm64/include/asm/pgtable-hwdef.h
> +++ b/arch/arm64/include/asm/pgtable-hwdef.h
> @@ -7,40 +7,43 @@
>
> #include <asm/memory.h>
>
> +#define PTDESC_ORDER 3
> +
> /*
> * Number of page-table levels required to address 'va_bits' wide
> * address, without section mapping. We resolve the top (va_bits - PAGE_SHIFT)
> - * bits with (PAGE_SHIFT - 3) bits at each page table level. Hence:
> + * bits with (PAGE_SHIFT - PTDESC_ORDER) bits at each page table level. Hence:
> *
> - * levels = DIV_ROUND_UP((va_bits - PAGE_SHIFT), (PAGE_SHIFT - 3))
> + * levels = DIV_ROUND_UP((va_bits - PAGE_SHIFT), (PAGE_SHIFT - PTDESC_ORDER))
> *
> * where DIV_ROUND_UP(n, d) => (((n) + (d) - 1) / (d))
> *
> * We cannot include linux/kernel.h which defines DIV_ROUND_UP here
> * due to build issues. So we open code DIV_ROUND_UP here:
> *
> - * ((((va_bits) - PAGE_SHIFT) + (PAGE_SHIFT - 3) - 1) / (PAGE_SHIFT - 3))
> + * ((((va_bits) - PAGE_SHIFT) + (PAGE_SHIFT - PTDESC_ORDER) - 1) / (PAGE_SHIFT - PTDESC_ORDER))
> *
> * which gets simplified as :
> */
> -#define ARM64_HW_PGTABLE_LEVELS(va_bits) (((va_bits) - 4) / (PAGE_SHIFT - 3))
> +#define ARM64_HW_PGTABLE_LEVELS(va_bits) \
> + (((va_bits) - PTDESC_ORDER - 1) / (PAGE_SHIFT - PTDESC_ORDER))
>
> /*
> * Size mapped by an entry at level n ( -1 <= n <= 3)
> - * We map (PAGE_SHIFT - 3) at all translation levels and PAGE_SHIFT bits
> + * We map (PAGE_SHIFT - PTDESC_ORDER) at all translation levels and PAGE_SHIFT bits
> * in the final page. The maximum number of translation levels supported by
> * the architecture is 5. Hence, starting at level n, we have further
> * ((4 - n) - 1) levels of translation excluding the offset within the page.
> * So, the total number of bits mapped by an entry at level n is :
> *
> - * ((4 - n) - 1) * (PAGE_SHIFT - 3) + PAGE_SHIFT
> + * ((4 - n) - 1) * (PAGE_SHIFT - PTDESC_ORDER) + PAGE_SHIFT
> *
> * Rearranging it a bit we get :
> - * (4 - n) * (PAGE_SHIFT - 3) + 3
> + * (4 - n) * (PAGE_SHIFT - PTDESC_ORDER) + PTDESC_ORDER
> */
> -#define ARM64_HW_PGTABLE_LEVEL_SHIFT(n) ((PAGE_SHIFT - 3) * (4 - (n)) + 3)
> +#define ARM64_HW_PGTABLE_LEVEL_SHIFT(n) ((PAGE_SHIFT - PTDESC_ORDER) * (4 - (n)) + PTDESC_ORDER)
>
> -#define PTRS_PER_PTE (1 << (PAGE_SHIFT - 3))
> +#define PTRS_PER_PTE (1 << (PAGE_SHIFT - PTDESC_ORDER))
>
> /*
> * PMD_SHIFT determines the size a level 2 page table entry can map.
> @@ -49,7 +52,7 @@
> #define PMD_SHIFT ARM64_HW_PGTABLE_LEVEL_SHIFT(2)
> #define PMD_SIZE (_AC(1, UL) << PMD_SHIFT)
> #define PMD_MASK (~(PMD_SIZE-1))
> -#define PTRS_PER_PMD (1 << (PAGE_SHIFT - 3))
> +#define PTRS_PER_PMD (1 << (PAGE_SHIFT - PTDESC_ORDER))
> #endif
>
> /*
> @@ -59,14 +62,14 @@
> #define PUD_SHIFT ARM64_HW_PGTABLE_LEVEL_SHIFT(1)
> #define PUD_SIZE (_AC(1, UL) << PUD_SHIFT)
> #define PUD_MASK (~(PUD_SIZE-1))
> -#define PTRS_PER_PUD (1 << (PAGE_SHIFT - 3))
> +#define PTRS_PER_PUD (1 << (PAGE_SHIFT - PTDESC_ORDER))
> #endif
>
> #if CONFIG_PGTABLE_LEVELS > 4
> #define P4D_SHIFT ARM64_HW_PGTABLE_LEVEL_SHIFT(0)
> #define P4D_SIZE (_AC(1, UL) << P4D_SHIFT)
> #define P4D_MASK (~(P4D_SIZE-1))
> -#define PTRS_PER_P4D (1 << (PAGE_SHIFT - 3))
> +#define PTRS_PER_P4D (1 << (PAGE_SHIFT - PTDESC_ORDER))
> #endif
>
> /*
> diff --git a/arch/arm64/kernel/pi/map_range.c b/arch/arm64/kernel/pi/map_range.c
> index 2b69e3beeef8..f74335e13929 100644
> --- a/arch/arm64/kernel/pi/map_range.c
> +++ b/arch/arm64/kernel/pi/map_range.c
> @@ -31,7 +31,7 @@ void __init map_range(u64 *pte, u64 start, u64 end, u64 pa, pgprot_t prot,
> {
> u64 cmask = (level == 3) ? CONT_PTE_SIZE - 1 : U64_MAX;
> pteval_t protval = pgprot_val(prot) & ~PTE_TYPE_MASK;
> - int lshift = (3 - level) * (PAGE_SHIFT - 3);
> + int lshift = (3 - level) * (PAGE_SHIFT - PTDESC_ORDER);
> u64 lmask = (PAGE_SIZE << lshift) - 1;
>
> start &= PAGE_MASK;
> diff --git a/arch/arm64/mm/kasan_init.c b/arch/arm64/mm/kasan_init.c
> index b65a29440a0c..211821f80571 100644
> --- a/arch/arm64/mm/kasan_init.c
> +++ b/arch/arm64/mm/kasan_init.c
> @@ -190,7 +190,7 @@ static void __init kasan_pgd_populate(unsigned long addr, unsigned long end,
> */
> static bool __init root_level_aligned(u64 addr)
> {
> - int shift = (ARM64_HW_PGTABLE_LEVELS(vabits_actual) - 1) * (PAGE_SHIFT - 3);
> + int shift = (ARM64_HW_PGTABLE_LEVELS(vabits_actual) - 1) * (PAGE_SHIFT - PTDESC_ORDER);
>
> return (addr % (PAGE_SIZE << shift)) == 0;
> }
> @@ -245,7 +245,7 @@ static int __init root_level_idx(u64 addr)
> */
> u64 vabits = IS_ENABLED(CONFIG_ARM64_64K_PAGES) ? VA_BITS
> : vabits_actual;
> - int shift = (ARM64_HW_PGTABLE_LEVELS(vabits) - 1) * (PAGE_SHIFT - 3);
> + int shift = (ARM64_HW_PGTABLE_LEVELS(vabits) - 1) * (PAGE_SHIFT - PTDESC_ORDER);
>
> return (addr & ~_PAGE_OFFSET(vabits)) >> (shift + PAGE_SHIFT);
> }
> @@ -269,7 +269,7 @@ static void __init clone_next_level(u64 addr, pgd_t *tmp_pg_dir, pud_t *pud)
> */
> static int __init next_level_idx(u64 addr)
> {
> - int shift = (ARM64_HW_PGTABLE_LEVELS(vabits_actual) - 2) * (PAGE_SHIFT - 3);
> + int shift = (ARM64_HW_PGTABLE_LEVELS(vabits_actual) - 2) * (PAGE_SHIFT - PTDESC_ORDER);
>
> return (addr >> (shift + PAGE_SHIFT)) % PTRS_PER_PTE;
> }
> --
> 2.25.1
>