From aad1668bd8948e1ce3b8c459f65ac8b427abe931 Mon Sep 17 00:00:00 2001 From: Michel Lespinasse Date: Mon, 8 Jun 2020 21:33:44 -0700 Subject: [PATCH] mmap locking API: add mmap_assert_locked() and mmap_assert_write_locked() Add new APIs to assert that mmap_sem is held. Using this instead of rwsem_is_locked and lockdep_assert_held[_write] makes the assertions more tolerant of future changes to the lock type. Signed-off-by: Michel Lespinasse Signed-off-by: Andrew Morton Reviewed-by: Vlastimil Babka Reviewed-by: Daniel Jordan Cc: Davidlohr Bueso Cc: David Rientjes Cc: Hugh Dickins Cc: Jason Gunthorpe Cc: Jerome Glisse Cc: John Hubbard Cc: Laurent Dufour Cc: Liam Howlett Cc: Matthew Wilcox Cc: Peter Zijlstra Cc: Ying Han Link: http://lkml.kernel.org/r/20200520052908.204642-10-walken@google.com Signed-off-by: Linus Torvalds --- arch/x86/events/core.c | 2 +- fs/userfaultfd.c | 6 +++--- include/linux/mmap_lock.h | 14 ++++++++++++++ mm/gup.c | 2 +- mm/hmm.c | 2 +- mm/memory.c | 2 +- mm/mmu_notifier.c | 6 +++--- mm/pagewalk.c | 6 +++--- mm/util.c | 2 +- 9 files changed, 28 insertions(+), 14 deletions(-) diff --git a/arch/x86/events/core.c b/arch/x86/events/core.c index ea2a3d151294a..081cdfdf11f45 100644 --- a/arch/x86/events/core.c +++ b/arch/x86/events/core.c @@ -2181,7 +2181,7 @@ static void x86_pmu_event_mapped(struct perf_event *event, struct mm_struct *mm) * For now, this can't happen because all callers hold mmap_sem * for write. If this changes, we'll need a different solution. */ - lockdep_assert_held_write(&mm->mmap_sem); + mmap_assert_write_locked(mm); if (atomic_inc_return(&mm->context.perf_rdpmc_allowed) == 1) on_each_cpu_mask(mm_cpumask(mm), cr4_update_pce, NULL, 1); diff --git a/fs/userfaultfd.c b/fs/userfaultfd.c index 9c645eee1a595..12b4924090406 100644 --- a/fs/userfaultfd.c +++ b/fs/userfaultfd.c @@ -234,7 +234,7 @@ static inline bool userfaultfd_huge_must_wait(struct userfaultfd_ctx *ctx, pte_t *ptep, pte; bool ret = true; - VM_BUG_ON(!rwsem_is_locked(&mm->mmap_sem)); + mmap_assert_locked(mm); ptep = huge_pte_offset(mm, address, vma_mmu_pagesize(vma)); @@ -286,7 +286,7 @@ static inline bool userfaultfd_must_wait(struct userfaultfd_ctx *ctx, pte_t *pte; bool ret = true; - VM_BUG_ON(!rwsem_is_locked(&mm->mmap_sem)); + mmap_assert_locked(mm); pgd = pgd_offset(mm, address); if (!pgd_present(*pgd)) @@ -405,7 +405,7 @@ vm_fault_t handle_userfault(struct vm_fault *vmf, unsigned long reason) * Coredumping runs without mmap_sem so we can only check that * the mmap_sem is held, if PF_DUMPCORE was not set. */ - WARN_ON_ONCE(!rwsem_is_locked(&mm->mmap_sem)); + mmap_assert_locked(mm); ctx = vmf->vma->vm_userfaultfd_ctx.ctx; if (!ctx) diff --git a/include/linux/mmap_lock.h b/include/linux/mmap_lock.h index acac1bf5ecd20..43ef914e6468d 100644 --- a/include/linux/mmap_lock.h +++ b/include/linux/mmap_lock.h @@ -1,6 +1,8 @@ #ifndef _LINUX_MMAP_LOCK_H #define _LINUX_MMAP_LOCK_H +#include + #define MMAP_LOCK_INITIALIZER(name) \ .mmap_sem = __RWSEM_INITIALIZER((name).mmap_sem), @@ -73,4 +75,16 @@ static inline void mmap_read_unlock_non_owner(struct mm_struct *mm) up_read_non_owner(&mm->mmap_sem); } +static inline void mmap_assert_locked(struct mm_struct *mm) +{ + lockdep_assert_held(&mm->mmap_sem); + VM_BUG_ON_MM(!rwsem_is_locked(&mm->mmap_sem), mm); +} + +static inline void mmap_assert_write_locked(struct mm_struct *mm) +{ + lockdep_assert_held_write(&mm->mmap_sem); + VM_BUG_ON_MM(!rwsem_is_locked(&mm->mmap_sem), mm); +} + #endif /* _LINUX_MMAP_LOCK_H */ diff --git a/mm/gup.c b/mm/gup.c index 41c648947fbef..279739fbd09aa 100644 --- a/mm/gup.c +++ b/mm/gup.c @@ -1425,7 +1425,7 @@ long populate_vma_page_range(struct vm_area_struct *vma, VM_BUG_ON(end & ~PAGE_MASK); VM_BUG_ON_VMA(start < vma->vm_start, vma); VM_BUG_ON_VMA(end > vma->vm_end, vma); - VM_BUG_ON_MM(!rwsem_is_locked(&mm->mmap_sem), mm); + mmap_assert_locked(mm); gup_flags = FOLL_TOUCH | FOLL_POPULATE | FOLL_MLOCK; if (vma->vm_flags & VM_LOCKONFAULT) diff --git a/mm/hmm.c b/mm/hmm.c index 41673a6d8d46d..e9a545751108c 100644 --- a/mm/hmm.c +++ b/mm/hmm.c @@ -563,7 +563,7 @@ int hmm_range_fault(struct hmm_range *range) struct mm_struct *mm = range->notifier->mm; int ret; - lockdep_assert_held(&mm->mmap_sem); + mmap_assert_locked(mm); do { /* If range is no longer valid force retry. */ diff --git a/mm/memory.c b/mm/memory.c index 4615d321d77b4..533293faaaf55 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -1211,7 +1211,7 @@ static inline unsigned long zap_pud_range(struct mmu_gather *tlb, next = pud_addr_end(addr, end); if (pud_trans_huge(*pud) || pud_devmap(*pud)) { if (next - addr != HPAGE_PUD_SIZE) { - VM_BUG_ON_VMA(!rwsem_is_locked(&tlb->mm->mmap_sem), vma); + mmap_assert_locked(tlb->mm); split_huge_pud(vma, pud, addr); } else if (zap_huge_pud(tlb, vma, pud, addr)) goto next; diff --git a/mm/mmu_notifier.c b/mm/mmu_notifier.c index cfd0a03bf5ccc..24eb9d1ed0a73 100644 --- a/mm/mmu_notifier.c +++ b/mm/mmu_notifier.c @@ -609,7 +609,7 @@ int __mmu_notifier_register(struct mmu_notifier *subscription, struct mmu_notifier_subscriptions *subscriptions = NULL; int ret; - lockdep_assert_held_write(&mm->mmap_sem); + mmap_assert_write_locked(mm); BUG_ON(atomic_read(&mm->mm_users) <= 0); if (IS_ENABLED(CONFIG_LOCKDEP)) { @@ -761,7 +761,7 @@ struct mmu_notifier *mmu_notifier_get_locked(const struct mmu_notifier_ops *ops, struct mmu_notifier *subscription; int ret; - lockdep_assert_held_write(&mm->mmap_sem); + mmap_assert_write_locked(mm); if (mm->notifier_subscriptions) { subscription = find_get_mmu_notifier(mm, ops); @@ -1006,7 +1006,7 @@ int mmu_interval_notifier_insert_locked( mm->notifier_subscriptions; int ret; - lockdep_assert_held_write(&mm->mmap_sem); + mmap_assert_write_locked(mm); if (!subscriptions || !subscriptions->has_itree) { ret = __mmu_notifier_register(NULL, mm); diff --git a/mm/pagewalk.c b/mm/pagewalk.c index 928df1638c30d..3b452aa05cd08 100644 --- a/mm/pagewalk.c +++ b/mm/pagewalk.c @@ -395,7 +395,7 @@ int walk_page_range(struct mm_struct *mm, unsigned long start, if (!walk.mm) return -EINVAL; - lockdep_assert_held(&walk.mm->mmap_sem); + mmap_assert_locked(walk.mm); vma = find_vma(walk.mm, start); do { @@ -453,7 +453,7 @@ int walk_page_range_novma(struct mm_struct *mm, unsigned long start, if (start >= end || !walk.mm) return -EINVAL; - lockdep_assert_held(&walk.mm->mmap_sem); + mmap_assert_locked(walk.mm); return __walk_page_range(start, end, &walk); } @@ -472,7 +472,7 @@ int walk_page_vma(struct vm_area_struct *vma, const struct mm_walk_ops *ops, if (!walk.mm) return -EINVAL; - lockdep_assert_held(&walk.mm->mmap_sem); + mmap_assert_locked(walk.mm); err = walk_page_test(vma->vm_start, vma->vm_end, &walk); if (err > 0) diff --git a/mm/util.c b/mm/util.c index e7e8647fa205e..09f62d7d6e3e6 100644 --- a/mm/util.c +++ b/mm/util.c @@ -437,7 +437,7 @@ int __account_locked_vm(struct mm_struct *mm, unsigned long pages, bool inc, unsigned long locked_vm, limit; int ret = 0; - lockdep_assert_held_write(&mm->mmap_sem); + mmap_assert_write_locked(mm); locked_vm = mm->locked_vm; if (inc) { -- 2.39.5