The patch titled Subject: mm: fix race between __split_huge_pmd_locked() and GUP-fast has been added to the -mm mm-unstable branch. Its filename is mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast.patch

This patch will shortly appear at https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches...

This patch will later appear in the mm-unstable branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's

*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***

The -mm tree is included into linux-next via the mm-everything branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm and is updated there every 2-3 working days

------------------------------------------------------ From: Ryan Roberts ryan.roberts@arm.com Subject: mm: fix race between __split_huge_pmd_locked() and GUP-fast Date: Wed, 1 May 2024 15:33:10 +0100

__split_huge_pmd_locked() can be called for a present THP, devmap or (non-present) migration entry. It calls pmdp_invalidate() unconditionally on the pmdp and only determines if it is present or not based on the returned old pmd. This is a problem for the migration entry case because pmd_mkinvalid(), called by pmdp_invalidate() must only be called for a present pmd.

On arm64 at least, pmd_mkinvalid() will mark the pmd such that any future call to pmd_present() will return true. And therefore any lockless pgtable walker could see the migration entry pmd in this state and start interpretting the fields as if it were present, leading to BadThings (TM). GUP-fast appears to be one such lockless pgtable walker.

x86 does not suffer the above problem, but instead pmd_mkinvalid() will corrupt the offset field of the swap entry within the swap pte. See link below for discussion of that problem.

Fix all of this by only calling pmdp_invalidate() for a present pmd. And for good measure let's add a warning to all implementations of pmdp_invalidate[_ad](). I've manually reviewed all other pmdp_invalidate[_ad]() call sites and believe all others to be conformant.

This is a theoretical bug found during code review. I don't have any test case to trigger it in practice.

Link: https://lkml.kernel.org/r/20240501143310.1381675-1-ryan.roberts@arm.com Link: https://lore.kernel.org/all/0dd7827a-6334-439a-8fd0-43c98e6af22b@arm.com/ Fixes: 84c3fc4e9c56 ("mm: thp: check pmd migration entry in common path") Signed-off-by: Ryan Roberts ryan.roberts@arm.com Reviewed-by: Zi Yan ziy@nvidia.com Reviewed-by: Anshuman Khandual anshuman.khandual@arm.com Acked-by: David Hildenbrand david@redhat.com Cc: Andreas Larsson andreas@gaisler.com Cc: Andy Lutomirski luto@kernel.org Cc: Aneesh Kumar K.V aneesh.kumar@kernel.org Cc: Borislav Petkov (AMD) bp@alien8.de Cc: Catalin Marinas catalin.marinas@arm.com Cc: Christian Borntraeger borntraeger@linux.ibm.com Cc: Christophe Leroy christophe.leroy@csgroup.eu Cc: Dave Hansen dave.hansen@linux.intel.com Cc: "David S. Miller" davem@davemloft.net Cc: Ingo Molnar mingo@redhat.com Cc: Jonathan Corbet corbet@lwn.net Cc: Mark Rutland mark.rutland@arm.com Cc: Naveen N. Rao naveen.n.rao@linux.ibm.com Cc: Nicholas Piggin npiggin@gmail.com Cc: Peter Zijlstra peterz@infradead.org Cc: Sven Schnelle svens@linux.ibm.com Cc: Thomas Gleixner tglx@linutronix.de Cc: Will Deacon will@kernel.org Cc: stable@vger.kernel.org Signed-off-by: Andrew Morton akpm@linux-foundation.org ---

Documentation/mm/arch_pgtable_helpers.rst | 6 +- arch/powerpc/mm/book3s64/pgtable.c | 1 arch/s390/include/asm/pgtable.h | 4 + arch/sparc/mm/tlb.c | 1 arch/x86/mm/pgtable.c | 2 mm/huge_memory.c | 49 ++++++++++---------- mm/pgtable-generic.c | 2 7 files changed, 39 insertions(+), 26 deletions(-)

--- a/arch/powerpc/mm/book3s64/pgtable.c~mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast +++ a/arch/powerpc/mm/book3s64/pgtable.c @@ -170,6 +170,7 @@ pmd_t pmdp_invalidate(struct vm_area_str { unsigned long old_pmd;

+ VM_WARN_ON_ONCE(!pmd_present(*pmdp)); old_pmd = pmd_hugepage_update(vma->vm_mm, address, pmdp, _PAGE_PRESENT, _PAGE_INVALID); flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE); return __pmd(old_pmd); --- a/arch/s390/include/asm/pgtable.h~mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast +++ a/arch/s390/include/asm/pgtable.h @@ -1769,8 +1769,10 @@ static inline pmd_t pmdp_huge_clear_flus static inline pmd_t pmdp_invalidate(struct vm_area_struct *vma, unsigned long addr, pmd_t *pmdp) { - pmd_t pmd = __pmd(pmd_val(*pmdp) | _SEGMENT_ENTRY_INVALID); + pmd_t pmd;

+ VM_WARN_ON_ONCE(!pmd_present(*pmdp)); + pmd = __pmd(pmd_val(*pmdp) | _SEGMENT_ENTRY_INVALID); return pmdp_xchg_direct(vma->vm_mm, addr, pmdp, pmd); }

--- a/arch/sparc/mm/tlb.c~mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast +++ a/arch/sparc/mm/tlb.c @@ -249,6 +249,7 @@ pmd_t pmdp_invalidate(struct vm_area_str { pmd_t old, entry;

+ VM_WARN_ON_ONCE(!pmd_present(*pmdp)); entry = __pmd(pmd_val(*pmdp) & ~_PAGE_VALID); old = pmdp_establish(vma, address, pmdp, entry); flush_tlb_range(vma, address, address + HPAGE_PMD_SIZE); --- a/arch/x86/mm/pgtable.c~mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast +++ a/arch/x86/mm/pgtable.c @@ -631,6 +631,8 @@ int pmdp_clear_flush_young(struct vm_are pmd_t pmdp_invalidate_ad(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp) { + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); + /* * No flush is necessary. Once an invalid PTE is established, the PTE's * access and dirty bits cannot be updated. --- a/Documentation/mm/arch_pgtable_helpers.rst~mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast +++ a/Documentation/mm/arch_pgtable_helpers.rst @@ -140,7 +140,8 @@ PMD Page Table Helpers +---------------------------+--------------------------------------------------+ | pmd_swp_clear_soft_dirty | Clears a soft dirty swapped PMD | +---------------------------+--------------------------------------------------+ -| pmd_mkinvalid | Invalidates a mapped PMD [1] | +| pmd_mkinvalid | Invalidates a present PMD; do not call for | +| | non-present PMD [1] | +---------------------------+--------------------------------------------------+ | pmd_set_huge | Creates a PMD huge mapping | +---------------------------+--------------------------------------------------+ @@ -196,7 +197,8 @@ PUD Page Table Helpers +---------------------------+--------------------------------------------------+ | pud_mkdevmap | Creates a ZONE_DEVICE mapped PUD | +---------------------------+--------------------------------------------------+ -| pud_mkinvalid | Invalidates a mapped PUD [1] | +| pud_mkinvalid | Invalidates a present PUD; do not call for | +| | non-present PUD [1] | +---------------------------+--------------------------------------------------+ | pud_set_huge | Creates a PUD huge mapping | +---------------------------+--------------------------------------------------+ --- a/mm/huge_memory.c~mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast +++ a/mm/huge_memory.c @@ -2430,32 +2430,11 @@ static void __split_huge_pmd_locked(stru return __split_huge_zero_page_pmd(vma, haddr, pmd); }

- /* - * Up to this point the pmd is present and huge and userland has the - * whole access to the hugepage during the split (which happens in - * place). If we overwrite the pmd with the not-huge version pointing - * to the pte here (which of course we could if all CPUs were bug - * free), userland could trigger a small page size TLB miss on the - * small sized TLB while the hugepage TLB entry is still established in - * the huge TLB. Some CPU doesn't like that. - * See http://support.amd.com/TechDocs/41322_10h_Rev_Gd.pdf, Erratum - * 383 on page 105. Intel should be safe but is also warns that it's - * only safe if the permission and cache attributes of the two entries - * loaded in the two TLB is identical (which should be the case here). - * But it is generally safer to never allow small and huge TLB entries - * for the same virtual address to be loaded simultaneously. So instead - * of doing "pmd_populate(); flush_pmd_tlb_range();" we first mark the - * current pmd notpresent (atomically because here the pmd_trans_huge - * must remain set at all times on the pmd until the split is complete - * for this pmd), then we flush the SMP TLB and finally we write the - * non-huge version of the pmd entry with pmd_populate. - */ - old_pmd = pmdp_invalidate(vma, haddr, pmd); - - pmd_migration = is_pmd_migration_entry(old_pmd); + pmd_migration = is_pmd_migration_entry(*pmd); if (unlikely(pmd_migration)) { swp_entry_t entry;

+ old_pmd = *pmd; entry = pmd_to_swp_entry(old_pmd); page = pfn_swap_entry_to_page(entry); write = is_writable_migration_entry(entry); @@ -2466,6 +2445,30 @@ static void __split_huge_pmd_locked(stru soft_dirty = pmd_swp_soft_dirty(old_pmd); uffd_wp = pmd_swp_uffd_wp(old_pmd); } else { + /* + * Up to this point the pmd is present and huge and userland has + * the whole access to the hugepage during the split (which + * happens in place). If we overwrite the pmd with the not-huge + * version pointing to the pte here (which of course we could if + * all CPUs were bug free), userland could trigger a small page + * size TLB miss on the small sized TLB while the hugepage TLB + * entry is still established in the huge TLB. Some CPU doesn't + * like that. See + * http://support.amd.com/TechDocs/41322_10h_Rev_Gd.pdf, Erratum + * 383 on page 105. Intel should be safe but is also warns that + * it's only safe if the permission and cache attributes of the + * two entries loaded in the two TLB is identical (which should + * be the case here). But it is generally safer to never allow + * small and huge TLB entries for the same virtual address to be + * loaded simultaneously. So instead of doing "pmd_populate(); + * flush_pmd_tlb_range();" we first mark the current pmd + * notpresent (atomically because here the pmd_trans_huge must + * remain set at all times on the pmd until the split is + * complete for this pmd), then we flush the SMP TLB and finally + * we write the non-huge version of the pmd entry with + * pmd_populate. + */ + old_pmd = pmdp_invalidate(vma, haddr, pmd); page = pmd_page(old_pmd); folio = page_folio(page); if (pmd_dirty(old_pmd)) { --- a/mm/pgtable-generic.c~mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast +++ a/mm/pgtable-generic.c @@ -198,6 +198,7 @@ pgtable_t pgtable_trans_huge_withdraw(st pmd_t pmdp_invalidate(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp) { + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); pmd_t old = pmdp_establish(vma, address, pmdp, pmd_mkinvalid(*pmdp)); flush_pmd_tlb_range(vma, address, address + HPAGE_PMD_SIZE); return old; @@ -208,6 +209,7 @@ pmd_t pmdp_invalidate(struct vm_area_str pmd_t pmdp_invalidate_ad(struct vm_area_struct *vma, unsigned long address, pmd_t *pmdp) { + VM_WARN_ON_ONCE(!pmd_present(*pmdp)); return pmdp_invalidate(vma, address, pmdp); } #endif _

Patches currently in -mm which might be from ryan.roberts@arm.com are

fs-proc-task_mmu-fix-loss-of-young-dirty-bits-during-pagemap-scan.patch fs-proc-task_mmu-fix-uffd-wp-confusion-in-pagemap_scan_pmd_entry.patch selftests-mm-soft-dirty-should-fail-if-a-testcase-fails.patch mm-debug_vm_pgtable-test-pmd_leaf-behavior-with-pmd_mkinvalid.patch mm-fix-race-between-__split_huge_pmd_locked-and-gup-fast.patch