The patch titled Subject: mm: pagewalk: allow walking without vma has been added to the -mm tree. Its filename is mm-pagewalk-allow-walking-without-vma.patch This patch should soon appear at http://ozlabs.org/~akpm/mmots/broken-out/mm-pagewalk-allow-walking-without-vma.patch and later at http://ozlabs.org/~akpm/mmotm/broken-out/mm-pagewalk-allow-walking-without-vma.patch 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 and is updated there every 3-4 working days ------------------------------------------------------ From: Steven Price <steven.price@xxxxxxx> Subject: mm: pagewalk: allow walking without vma Since 48684a65b4e3: "mm: pagewalk: fix misbehavior of walk_page_range for vma(VM_PFNMAP)", page_table_walk() will report any kernel area as a hole, because it lacks a vma. This means each arch has re-implemented page table walking when needed, for example in the per-arch ptdump walker. Remove the requirement to have a vma in the generic code and add a new function walk_page_range_novma() which ignores the VMAs and simply walks the page tables. Link: http://lkml.kernel.org/r/20191218162402.45610-13-steven.price@xxxxxxx Signed-off-by: Steven Price <steven.price@xxxxxxx> Cc: Albert Ou <aou@xxxxxxxxxxxxxxxxx> Cc: Alexandre Ghiti <alex@xxxxxxxx> Cc: Andy Lutomirski <luto@xxxxxxxxxx> Cc: Ard Biesheuvel <ard.biesheuvel@xxxxxxxxxx> Cc: Arnd Bergmann <arnd@xxxxxxxx> Cc: Benjamin Herrenschmidt <benh@xxxxxxxxxxxxxxxxxxx> Cc: Borislav Petkov <bp@xxxxxxxxx> Cc: Catalin Marinas <catalin.marinas@xxxxxxx> Cc: Christian Borntraeger <borntraeger@xxxxxxxxxx> Cc: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx> Cc: David S. Miller <davem@xxxxxxxxxxxxx> Cc: Heiko Carstens <heiko.carstens@xxxxxxxxxx> Cc: "H. Peter Anvin" <hpa@xxxxxxxxx> Cc: Ingo Molnar <mingo@xxxxxxxxxx> Cc: James Hogan <jhogan@xxxxxxxxxx> Cc: James Morse <james.morse@xxxxxxx> Cc: Jerome Glisse <jglisse@xxxxxxxxxx> Cc: "Liang, Kan" <kan.liang@xxxxxxxxxxxxxxx> Cc: Mark Rutland <mark.rutland@xxxxxxx> Cc: Michael Ellerman <mpe@xxxxxxxxxxxxxx> Cc: Palmer Dabbelt <palmer@xxxxxxxxxx> Cc: Paul Burton <paul.burton@xxxxxxxx> Cc: Paul Mackerras <paulus@xxxxxxxxx> Cc: Paul Walmsley <paul.walmsley@xxxxxxxxxx> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx> Cc: Ralf Baechle <ralf@xxxxxxxxxxxxxx> Cc: Russell King <linux@xxxxxxxxxxxxxxx> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx> Cc: Vasily Gorbik <gor@xxxxxxxxxxxxx> Cc: Vineet Gupta <vgupta@xxxxxxxxxxxx> Cc: Will Deacon <will@xxxxxxxxxx> Cc: Zong Li <zong.li@xxxxxxxxxx> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> --- include/linux/pagewalk.h | 5 ++++ mm/pagewalk.c | 40 +++++++++++++++++++++++++++++-------- 2 files changed, 37 insertions(+), 8 deletions(-) --- a/include/linux/pagewalk.h~mm-pagewalk-allow-walking-without-vma +++ a/include/linux/pagewalk.h @@ -73,6 +73,7 @@ enum page_walk_action { * @mm: mm_struct representing the target process of page table walk * @vma: vma currently walked (NULL if walking outside vmas) * @action: next action to perform (see enum page_walk_action) + * @no_vma: walk ignoring vmas (vma will always be NULL) * @private: private data for callbacks' usage * * (see the comment on walk_page_range() for more details) @@ -82,12 +83,16 @@ struct mm_walk { struct mm_struct *mm; struct vm_area_struct *vma; enum page_walk_action action; + bool no_vma; void *private; }; int walk_page_range(struct mm_struct *mm, unsigned long start, unsigned long end, const struct mm_walk_ops *ops, void *private); +int walk_page_range_novma(struct mm_struct *mm, unsigned long start, + unsigned long end, const struct mm_walk_ops *ops, + void *private); int walk_page_vma(struct vm_area_struct *vma, const struct mm_walk_ops *ops, void *private); int walk_page_mapping(struct address_space *mapping, pgoff_t first_index, --- a/mm/pagewalk.c~mm-pagewalk-allow-walking-without-vma +++ a/mm/pagewalk.c @@ -39,7 +39,7 @@ static int walk_pmd_range(pud_t *pud, un do { again: next = pmd_addr_end(addr, end); - if (pmd_none(*pmd) || !walk->vma) { + if (pmd_none(*pmd) || (!walk->vma && !walk->no_vma)) { if (ops->pte_hole) err = ops->pte_hole(addr, next, walk); if (err) @@ -65,13 +65,16 @@ again: * Check this here so we only break down trans_huge * pages when we _need_ to */ - if (walk->action == ACTION_CONTINUE || + if ((!walk->vma && (pmd_leaf(*pmd) || !pmd_present(*pmd))) || + walk->action == ACTION_CONTINUE || !(ops->pte_entry)) continue; - split_huge_pmd(walk->vma, pmd, addr); - if (pmd_trans_unstable(pmd)) - goto again; + if (walk->vma) { + split_huge_pmd(walk->vma, pmd, addr); + if (pmd_trans_unstable(pmd)) + goto again; + } err = walk_pte_range(pmd, addr, next, walk); if (err) @@ -93,7 +96,7 @@ static int walk_pud_range(p4d_t *p4d, un do { again: next = pud_addr_end(addr, end); - if (pud_none(*pud) || !walk->vma) { + if (pud_none(*pud) || (!walk->vma && !walk->no_vma)) { if (ops->pte_hole) err = ops->pte_hole(addr, next, walk); if (err) @@ -111,11 +114,13 @@ static int walk_pud_range(p4d_t *p4d, un if (walk->action == ACTION_AGAIN) goto again; - if (walk->action == ACTION_CONTINUE || + if ((!walk->vma && (pud_leaf(*pud) || !pud_present(*pud))) || + walk->action == ACTION_CONTINUE || !(ops->pmd_entry || ops->pte_entry)) continue; - split_huge_pud(walk->vma, pud, addr); + if (walk->vma) + split_huge_pud(walk->vma, pud, addr); if (pud_none(*pud)) goto again; @@ -389,6 +394,25 @@ int walk_page_range(struct mm_struct *mm return err; } +int walk_page_range_novma(struct mm_struct *mm, unsigned long start, + unsigned long end, const struct mm_walk_ops *ops, + void *private) +{ + struct mm_walk walk = { + .ops = ops, + .mm = mm, + .private = private, + .no_vma = true + }; + + if (start >= end || !walk.mm) + return -EINVAL; + + lockdep_assert_held(&walk.mm->mmap_sem); + + return __walk_page_range(start, end, &walk); +} + int walk_page_vma(struct vm_area_struct *vma, const struct mm_walk_ops *ops, void *private) { _ Patches currently in -mm which might be from steven.price@xxxxxxx are mm-add-generic-pd_leaf-macros.patch arc-mm-add-pd_leaf-definitions.patch arm-mm-add-pd_leaf-definitions.patch arm64-mm-add-pd_leaf-definitions.patch mips-mm-add-pd_leaf-definitions.patch powerpc-mm-add-pd_leaf-definitions.patch riscv-mm-add-pd_leaf-definitions.patch s390-mm-add-pd_leaf-definitions.patch sparc-mm-add-pd_leaf-definitions.patch x86-mm-add-pd_leaf-definitions.patch mm-pagewalk-add-p4d_entry-and-pgd_entry.patch mm-pagewalk-allow-walking-without-vma.patch mm-pagewalk-dont-lock-ptes-for-walk_page_range_novma.patch mm-pagewalk-fix-termination-condition-in-walk_pte_range.patch mm-pagewalk-add-depth-parameter-to-pte_hole.patch x86-mm-point-to-struct-seq_file-from-struct-pg_state.patch x86-mmefi-convert-ptdump_walk_pgd_level-to-take-a-mm_struct.patch x86-mm-convert-ptdump_walk_pgd_level_debugfs-to-take-an-mm_struct.patch mm-add-generic-ptdump.patch x86-mm-convert-dump_pagetables-to-use-walk_page_range.patch arm64-mm-convert-mm-dumpc-to-use-walk_page_range.patch arm64-mm-display-non-present-entries-in-ptdump.patch mm-ptdump-reduce-level-numbers-by-1-in-note_page.patch