[PATCH V2 1/4] arm64/mm: Use pmdp_get() for PMD accesses
Anshuman Khandual
anshuman.khandual at arm.com
Wed Sep 23 05:22:50 PDT 2026
On Wed, Sep 23, 2026 at 12:31:42PM +0100, Ryan Roberts wrote:
> On 23/09/2026 12:01, Anshuman Khandual wrote:
> > On Wed, Sep 23, 2026 at 09:32:17AM +0100, Ryan Roberts wrote:
> >> On 22/09/2026 07:16, Anshuman Khandual wrote:
> >>> Replace READ_ONCE() with pmdp_get() for PMD accesses in preparation for
> >>> supporting both D64 and D128 translation table formats.
> >>>
> >>> READ_ONCE() cannot currently be used for 128-bit page table entries on
> >>> arm64 because it does not provide the required 128-bit single-copy
> >>> atomicity, causing builds to fail for accesses wider than 64 bits.
> >>>
> >>> Although LDP/STP provide the required atomicity when FEAT_LSE is
> >>> available (as required by FEAT_D128), extending READ_ONCE() to support
> >>> 128-bit accesses is undesirable. READ_ONCE() is a general-purpose API,
> >>> so doing so could encourage other 128-bit users that would either fail
> >>> to build in configurations without D128 support or, if D128 becomes a
> >>> runtime option, silently permit tearing on systems without the required
> >>> hardware support.
> >>>
> >>> Instead, standardize PMD accesses on the existing page-table helpers.
> >>> These can be overridden on arm64 to provide 128-bit single-copy
> >>> atomicity when required. No functional change intended.
> >>
> >> I notice you have an unconverted READ_ONCE(*pmdp) in
> >> pmdp_test_and_clear_young(). Is that intentional?
> >
> > Converting the above READ_ONCE() instances as pmdp_get() cuases a build
> > failure as pmpd_get() which is defined in generic pgtable header is not
> > visible inside platform specific pgtable header, even with a forward
> > declaration.
> >
> > In D128 V2 series these conversions were performed later in the series,
> > when platform specific pxdp_get() overrides were added via ptval_get().
> >
> > But to avoid this, we could just move pmdp_test_and_clear_young() inside
> > arch/arm64/mm/mmu.c file instead. Folding in the following change builds.
>
> I'm not sure if/how this might affect performance? Probably not significant
> since they are for higher levels, so not called as frequently as pte helpers.
Right, should not have much performance impact.
>
> Another option would be to implement arm64-specific pxdp_get() helpers in
> asm/pgtable.h to override the generic versions?
For now those will still be based on READ_ONCE() just like the default ones.
Although trying to drop similar redundancies from other platforms :)
https://lore.kernel.org/linux-mm/20260923043226.331880-1-anshuman.khandual@arm.com/
But given D128 is going to define pxdp_get() anyway it might make sense to
define them now and withdraw the above series :) Seems like s390 platform
also has similar situation.
>
> Thanks,
> Ryan
>
> >
> > diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
> > index ae41946c54f26..13ca85ec89f6a 100644
> > --- a/arch/arm64/include/asm/pgtable.h
> > +++ b/arch/arm64/include/asm/pgtable.h
> > @@ -1289,14 +1289,8 @@ static inline bool __ptep_clear_flush_young(struct vm_area_struct *vma,
> >
> > #if defined(CONFIG_TRANSPARENT_HUGEPAGE) || defined(CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG)
> > #define __HAVE_ARCH_PMDP_TEST_AND_CLEAR_YOUNG
> > -static inline bool pmdp_test_and_clear_young(struct vm_area_struct *vma,
> > - unsigned long address, pmd_t *pmdp)
> > -{
> > - /* Operation applies to PMD table entry only if FEAT_HAFT is enabled */
> > - VM_WARN_ON(pmd_table(READ_ONCE(*pmdp)) && !system_supports_haft());
> > - return __ptep_test_and_clear_young(vma, address, (pte_t *)pmdp);
> > -}
> > -#endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG */
> > +bool pmdp_test_and_clear_young(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp);
> > +#endif
> >
> > static inline pte_t __ptep_get_and_clear_anysz(struct mm_struct *mm,
> > unsigned long address,
> > diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c
> > index 76d8d320aeedf..b3fd06a623db5 100644
> > --- a/arch/arm64/mm/mmu.c
> > +++ b/arch/arm64/mm/mmu.c
> > @@ -2440,4 +2440,15 @@ void __check_safe_pte_update(struct mm_struct *mm, pte_t *ptep, pte_t pte)
> > __func__, pte_str_old, pte_str);
> > }
> > #endif /* CONFIG_DEBUG_VM */
> > +
> > +#if defined(CONFIG_TRANSPARENT_HUGEPAGE) || defined(CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG)
> > +#define __HAVE_ARCH_PMDP_TEST_AND_CLEAR_YOUNG
> > +bool pmdp_test_and_clear_young(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp)
> > +{
> > + /* Operation applies to PMD table entry only if FEAT_HAFT is enabled */
> > + VM_WARN_ON(pmd_table(pmdp_get(pmdp)) && !system_supports_haft());
> > + return __ptep_test_and_clear_young(vma, address, (pte_t *)pmdp);
> > +}
> > +#endif /* CONFIG_TRANSPARENT_HUGEPAGE || CONFIG_ARCH_HAS_NONLEAF_PMD_YOUNG */
> > +
> > #endif
> >
> >>
> >> Thanks,
> >> Ryan
> >>
> >>
> >>>
> >>> Cc: Catalin Marinas <catalin.marinas at arm.com>
> >>> Cc: Will Deacon <will at kernel.org>
> >>> Cc: Ryan Roberts <ryan.roberts at arm.com>
> >>> Cc: Mark Rtland <mark.rtland at arm.com>
> >>> Cc: linx-arm-kernel at lists.infradead.org
> >>> Cc: linx-kernel at vger.kernel.org
> >>> Cc: kasan-dev at googlegrops.com
> >>> Reviewed-by: Ryan Roberts <ryan.roberts at arm.com>
> >>> Signed-off-by: Anshuman Khandual <anshuman.khandual at arm.com>
> >>> ---
> >>> arch/arm64/include/asm/pgtable.h | 3 ++-
> >>> arch/arm64/mm/fault.c | 2 +-
> >>> arch/arm64/mm/fixmap.c | 2 +-
> >>> arch/arm64/mm/hugetlbpage.c | 2 +-
> >>> arch/arm64/mm/kasan_init.c | 4 ++--
> >>> arch/arm64/mm/mmu.c | 22 +++++++++++-----------
> >>> arch/arm64/mm/pageattr.c | 2 +-
> >>> arch/arm64/mm/trans_pgd.c | 2 +-
> >>> 8 files changed, 20 insertions(+), 19 deletions(-)
> >>>
> >>> diff --git a/arch/arm64/include/asm/pgtable.h b/arch/arm64/include/asm/pgtable.h
> >>> index 763c5a411d64..b953ea0ba3ed 100644
> >>> --- a/arch/arm64/include/asm/pgtable.h
> >>> +++ b/arch/arm64/include/asm/pgtable.h
> >>> @@ -811,7 +811,8 @@ static inline unsigned long pmd_page_vaddr(pmd_t pmd)
> >>> }
> >>>
> >>> /* Find an entry in the third-level page table. */
> >>> -#define pte_offset_phys(dir,addr) (pmd_page_paddr(READ_ONCE(*(dir))) + pte_index(addr) * sizeof(pte_t))
> >>> +#define pte_offset_phys(dir, addr) (pmd_page_paddr(pmdp_get(dir)) + \
> >>> + pte_index(addr) * sizeof(pte_t))
> >>>
> >>> #define pte_set_fixmap(addr) ((pte_t *)set_fixmap_offset(FIX_PTE, addr))
> >>> #define pte_set_fixmap_offset(pmd, addr) pte_set_fixmap(pte_offset_phys(pmd, addr))
> >>> diff --git a/arch/arm64/mm/fault.c b/arch/arm64/mm/fault.c
> >>> index 2cecf6ba6df7..2757ee0c4300 100644
> >>> --- a/arch/arm64/mm/fault.c
> >>> +++ b/arch/arm64/mm/fault.c
> >>> @@ -188,7 +188,7 @@ static void show_pte(unsigned long addr)
> >>> break;
> >>>
> >>> pmdp = pmd_offset_lockless(pudp, pud, addr);
> >>> - pmd = READ_ONCE(*pmdp);
> >>> + pmd = pmdp_get(pmdp);
> >>> ptval_to_str(pxd_str, pmd_val(pmd));
> >>> pr_cont(", pmd=%s", pxd_str);
> >>> if (pmd_none(pmd) || pmd_bad(pmd))
> >>> diff --git a/arch/arm64/mm/fixmap.c b/arch/arm64/mm/fixmap.c
> >>> index f66a0016dd02..3cdac8021d4f 100644
> >>> --- a/arch/arm64/mm/fixmap.c
> >>> +++ b/arch/arm64/mm/fixmap.c
> >>> @@ -42,7 +42,7 @@ static inline pte_t *fixmap_pte(unsigned long addr)
> >>>
> >>> static void __init early_fixmap_init_pte(pmd_t *pmdp, unsigned long addr)
> >>> {
> >>> - pmd_t pmd = READ_ONCE(*pmdp);
> >>> + pmd_t pmd = pmdp_get(pmdp);
> >>> pte_t *ptep;
> >>>
> >>> if (pmd_none(pmd)) {
> >>> diff --git a/arch/arm64/mm/hugetlbpage.c b/arch/arm64/mm/hugetlbpage.c
> >>> index 8e799c1fe0aa..cdaa4500faf9 100644
> >>> --- a/arch/arm64/mm/hugetlbpage.c
> >>> +++ b/arch/arm64/mm/hugetlbpage.c
> >>> @@ -304,7 +304,7 @@ pte_t *huge_pte_offset(struct mm_struct *mm,
> >>> addr &= CONT_PMD_MASK;
> >>>
> >>> pmdp = pmd_offset(pudp, addr);
> >>> - pmd = READ_ONCE(*pmdp);
> >>> + pmd = pmdp_get(pmdp);
> >>> if (!(sz == PMD_SIZE || sz == CONT_PMD_SIZE) &&
> >>> pmd_none(pmd))
> >>> return NULL;
> >>> diff --git a/arch/arm64/mm/kasan_init.c b/arch/arm64/mm/kasan_init.c
> >>> index 45fbdce684c8..7ca833c5de5e 100644
> >>> --- a/arch/arm64/mm/kasan_init.c
> >>> +++ b/arch/arm64/mm/kasan_init.c
> >>> @@ -62,7 +62,7 @@ static phys_addr_t __init kasan_alloc_raw_page(int node)
> >>> static pte_t *__init kasan_pte_offset(pmd_t *pmdp, unsigned long addr, int node,
> >>> bool early)
> >>> {
> >>> - if (pmd_none(READ_ONCE(*pmdp))) {
> >>> + if (pmd_none(pmdp_get(pmdp))) {
> >>> phys_addr_t pte_phys = early ?
> >>> __pa_symbol(kasan_early_shadow_pte)
> >>> : kasan_alloc_zeroed_page(node);
> >>> @@ -138,7 +138,7 @@ static void __init kasan_pmd_populate(pud_t *pudp, unsigned long addr,
> >>> do {
> >>> next = pmd_addr_end(addr, end);
> >>> kasan_pte_populate(pmdp, addr, next, node, early);
> >>> - } while (pmdp++, addr = next, addr != end && pmd_none(READ_ONCE(*pmdp)));
> >>> + } while (pmdp++, addr = next, addr != end && pmd_none(pmdp_get(pmdp)));
> >>> }
> >>>
> >>> static void __init kasan_pud_populate(p4d_t *p4dp, unsigned long addr,
> >>> diff --git a/arch/arm64/mm/mmu.c b/arch/arm64/mm/mmu.c
> >>> index e589fb00d1e3..95621913679e 100644
> >>> --- a/arch/arm64/mm/mmu.c
> >>> +++ b/arch/arm64/mm/mmu.c
> >>> @@ -200,7 +200,7 @@ static int alloc_init_cont_pte(pmd_t *pmdp, unsigned long addr,
> >>> int flags)
> >>> {
> >>> unsigned long next;
> >>> - pmd_t pmd = READ_ONCE(*pmdp);
> >>> + pmd_t pmd = pmdp_get(pmdp);
> >>> pte_t *ptep;
> >>>
> >>> BUG_ON(pmd_leaf(pmd));
> >>> @@ -257,7 +257,7 @@ static int init_pmd(pmd_t *pmdp, unsigned long addr, unsigned long end,
> >>> unsigned long next;
> >>>
> >>> do {
> >>> - pmd_t old_pmd = READ_ONCE(*pmdp);
> >>> + pmd_t old_pmd = pmdp_get(pmdp);
> >>>
> >>> next = pmd_addr_end(addr, end);
> >>>
> >>> @@ -272,7 +272,7 @@ static int init_pmd(pmd_t *pmdp, unsigned long addr, unsigned long end,
> >>> * only allow updates to the permission attributes.
> >>> */
> >>> BUG_ON(!pgattr_change_is_safe(pmd_val(old_pmd),
> >>> - READ_ONCE(pmd_val(*pmdp))));
> >>> + pmd_val(pmdp_get(pmdp))));
> >>> } else {
> >>> int ret;
> >>>
> >>> @@ -282,7 +282,7 @@ static int init_pmd(pmd_t *pmdp, unsigned long addr, unsigned long end,
> >>> return ret;
> >>>
> >>> VM_WARN_ON_ONCE(pmd_val(old_pmd) != 0 &&
> >>> - pmd_val(old_pmd) != READ_ONCE(pmd_val(*pmdp)));
> >>> + pmd_val(old_pmd) != pmd_val(pmdp_get(pmdp)));
> >>> }
> >>> phys += next - addr;
> >>> } while (pmdp++, addr = next, addr != end);
> >>> @@ -293,7 +293,7 @@ static int init_pmd(pmd_t *pmdp, unsigned long addr, unsigned long end,
> >>> static bool pmd_range_has_valid_noncont(pmd_t *pmdp)
> >>> {
> >>> for (int i = 0; i < CONT_PMDS; i++) {
> >>> - pte_t pte = pmd_pte(READ_ONCE(pmdp[i]));
> >>> + pte_t pte = pmd_pte(pmdp_get(pmdp + i));
> >>>
> >>> if (pte_valid(pte) && !pte_cont(pte))
> >>> return true;
> >>> @@ -1553,7 +1553,7 @@ static void unmap_hotplug_pmd_range(pud_t *pudp, unsigned long addr,
> >>> do {
> >>> next = pmd_addr_end(addr, end);
> >>> pmdp = pmd_offset(pudp, addr);
> >>> - pmd = READ_ONCE(*pmdp);
> >>> + pmd = pmdp_get(pmdp);
> >>> if (pmd_none(pmd))
> >>> continue;
> >>>
> >>> @@ -1708,7 +1708,7 @@ static void free_empty_pmd_table(pud_t *pudp, unsigned long addr,
> >>> do {
> >>> next = pmd_addr_end(addr, end);
> >>> pmdp = pmd_offset(pudp, addr);
> >>> - pmd = READ_ONCE(*pmdp);
> >>> + pmd = pmdp_get(pmdp);
> >>> if (pmd_none(pmd))
> >>> continue;
> >>>
> >>> @@ -1729,7 +1729,7 @@ static void free_empty_pmd_table(pud_t *pudp, unsigned long addr,
> >>> */
> >>> pmdp = pmd_offset(pudp, 0UL);
> >>> for (i = 0; i < PTRS_PER_PMD; i++) {
> >>> - if (!pmd_none(READ_ONCE(pmdp[i])))
> >>> + if (!pmd_none(pmdp_get(pmdp + i)))
> >>> return;
> >>> }
> >>>
> >>> @@ -1881,7 +1881,7 @@ int pmd_set_huge(pmd_t *pmdp, phys_addr_t phys, pgprot_t prot)
> >>> pmd_t new_pmd = pfn_pmd(__phys_to_pfn(phys), mk_pmd_sect_prot(prot));
> >>>
> >>> /* Only allow permission changes for now */
> >>> - if (!pgattr_change_is_safe(READ_ONCE(pmd_val(*pmdp)),
> >>> + if (!pgattr_change_is_safe(pmd_val(pmdp_get(pmdp)),
> >>> pmd_val(new_pmd)))
> >>> return 0;
> >>>
> >>> @@ -1906,7 +1906,7 @@ int pud_clear_huge(pud_t *pudp)
> >>>
> >>> int pmd_clear_huge(pmd_t *pmdp)
> >>> {
> >>> - if (!pmd_leaf(READ_ONCE(*pmdp)))
> >>> + if (!pmd_leaf(pmdp_get(pmdp)))
> >>> return 0;
> >>> pmd_clear(pmdp);
> >>> return 1;
> >>> @@ -1917,7 +1917,7 @@ int pmd_free_pte_page(pmd_t *pmdp, unsigned long addr)
> >>> pte_t *table;
> >>> pmd_t pmd;
> >>>
> >>> - pmd = READ_ONCE(*pmdp);
> >>> + pmd = pmdp_get(pmdp);
> >>>
> >>> if (!pmd_table(pmd)) {
> >>> VM_WARN_ON(1);
> >>> diff --git a/arch/arm64/mm/pageattr.c b/arch/arm64/mm/pageattr.c
> >>> index bbe98ac9ad8c..0ca07bd5ded9 100644
> >>> --- a/arch/arm64/mm/pageattr.c
> >>> +++ b/arch/arm64/mm/pageattr.c
> >>> @@ -414,7 +414,7 @@ bool kernel_page_present(struct page *page)
> >>> return pud_valid(pud);
> >>>
> >>> pmdp = pmd_offset(pudp, addr);
> >>> - pmd = READ_ONCE(*pmdp);
> >>> + pmd = pmdp_get(pmdp);
> >>> if (pmd_none(pmd))
> >>> return false;
> >>> if (pmd_leaf(pmd))
> >>> diff --git a/arch/arm64/mm/trans_pgd.c b/arch/arm64/mm/trans_pgd.c
> >>> index cca9706a875c..b27b2d2c20c3 100644
> >>> --- a/arch/arm64/mm/trans_pgd.c
> >>> +++ b/arch/arm64/mm/trans_pgd.c
> >>> @@ -74,7 +74,7 @@ static int copy_pmd(struct trans_pgd_info *info, pud_t *dst_pudp,
> >>>
> >>> src_pmdp = pmd_offset(src_pudp, start);
> >>> do {
> >>> - pmd_t pmd = READ_ONCE(*src_pmdp);
> >>> + pmd_t pmd = pmdp_get(src_pmdp);
> >>>
> >>> next = pmd_addr_end(addr, end);
> >>> if (pmd_none(pmd))
> >>
>
More information about the linux-arm-kernel
mailing list