move per-vma lock into vm_area_struct
Subsystems (12)
Suren Baghdasaryan
maintainerv8 cover move per-vma lock into vm_area_struct
Back when per-vma locks were introduces, vm_lock was moved out of vm_area_struct in [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging the vm_lock back into vm_area_struct also allows vm_area_struct to use SLAB_TYPESAFE_BY_RCU later in this patchset. This patchset:
1. moves vm_lock back into vm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changes vm_area_struct initialization to mark new vma as detached until it is inserted into vma tree;
3. replaces vm_lock and vma->detached flag with a reference counter;
4. changes vm_area_struct cache to SLAB_TYPESAFE_BY_RCU to allow for their reuse and to minimize call_rcu() calls.
Pagefault microbenchmarks show performance improvement:
Hmean faults/cpu-1 507926.5547 ( 0.00%) 506519.3692 * -0.28%* Hmean faults/cpu-4 479119.7051 ( 0.00%) 481333.6802 * 0.46%* Hmean faults/cpu-7 452880.2961 ( 0.00%) 455845.6211 * 0.65%* Hmean faults/cpu-12 347639.1021 ( 0.00%) 352004.2254 * 1.26%* Hmean faults/cpu-21 200061.2238 ( 0.00%) 229597.0317 * 14.76%* Hmean faults/cpu-30 145251.2001 ( 0.00%) 164202.5067 * 13.05%* Hmean faults/cpu-48 106848.4434 ( 0.00%) 120641.5504 * 12.91%* Hmean faults/cpu-56 92472.3835 ( 0.00%) 103464.7916 * 11.89%* Hmean faults/sec-1 507566.1468 ( 0.00%) 506139.0811 * -0.28%* Hmean faults/sec-4 1880478.2402 ( 0.00%) 1886795.6329 * 0.34%* Hmean faults/sec-7 3106394.3438 ( 0.00%) 3140550.7485 * 1.10%* Hmean faults/sec-12 4061358.4795 ( 0.00%) 4112477.0206 * 1.26%* Hmean faults/sec-21 3988619.1169 ( 0.00%) 4577747.1436 * 14.77%* Hmean faults/sec-30 3909839.5449 ( 0.00%) 4311052.2787 * 10.26%* Hmean faults/sec-48 4761108.4691 ( 0.00%) 5283790.5026 * 10.98%* Hmean faults/sec-56 4885561.4590 ( 0.00%) 5415839.4045 * 10.85%*
Changes since v7 [4]:
- Removed additional parameter for vma_iter_store() and introduced vma_iter_store_attached() instead, per Vlastimil Babka and Liam R. Howlett
- Fixed coding style nits, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Liam R. Howlett
- Added Acked-by, per Davidlohr Bueso
- Removed unnecessary patch changeing nommu.c
- Folded a fixup patch [5] into the patch it was fixing
- Changed calculation in __refcount_add_not_zero_limited() to avoid overflow, to change the limit to be inclusive and to use INT_MAX to indicate no limits, per Vlastimil Babka and Matthew Wilcox
- Folded a fixup patch [6] into the patch it was fixing
- Added vm_refcnt rules summary in the changelog, per Liam R. Howlett
- Changed writers to not increment vm_refcnt and adjusted VMA_REF_LIMIT to not reserve one count for a writer, per Liam R. Howlett
- Changed vma_refcount_put() to wake up writers only when the last reader is leaving, per Liam R. Howlett
- Fixed rwsem_acquire_read() parameters when read-locking a vma to match the way down_read_trylock() does lockdep, per Vlastimil Babka
- Folded vma_lockdep_init() into vma_lock_init() for simplicity
- Brought back vma_copy() to keep vm_refcount at 0 during reuse, per Vlastimil Babka
What I did not include in this patchset:
- Liam's suggestion to change dump_vma() output since it's unclear to me how it should look like. The patch is for debug only and not critical for the rest of the series, we can change the output later or even drop it if necessary.
[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
[4] https://lore.kernel.org/all/[email protected]/
[5] https://lore.kernel.org/all/[email protected]/
[6] https://lore.kernel.org/all/[email protected]/
Patchset applies over mm-unstable after reverting v7 (current SHA range: 588f0086398e - fb2270654630)
Suren Baghdasaryan (16):
mm: introduce vma_start_read_locked{_nested} helpers
mm: move per-vma lock into vm_area_struct
mm: mark vma as detached until it's added into vma tree
mm: introduce vma_iter_store_attached() to use with attached vmas
mm: mark vmas detached upon exit
types: move struct rcuwait into types.h
mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
mm: move mmap_init_lock() out of the header file
mm: uninline the main body of vma_start_write()
refcount: introduce __refcount_{add|inc}_not_zero_limited
mm: replace vm_lock and detached flag with a reference count mm/debug: print vm_refcnt state when dumping the vma
mm: remove extra vma_numab_state_init() call
mm: prepare lock_vma_under_rcu() for vma reuse possibility
mm: make vma cache SLAB_TYPESAFE_BY_RCU
docs/mm: document latest changes to vm_lock
Documentation/mm/process_addrs.rst | 44 +++++---- include/linux/mm.h | 152 ++++++++++++++++++++++------- include/linux/mm_types.h | 36 ++++--- include/linux/mmap_lock.h | 6 -- include/linux/rcuwait.h | 13 +-- include/linux/refcount.h | 20 +++- include/linux/slab.h | 6 -- include/linux/types.h | 12 +++ kernel/fork.c | 128 +++++++++++------------- mm/debug.c | 12 +++ mm/init-mm.c | 1 + mm/memory.c | 94 +++++++++++++++--- mm/mmap.c | 3 +- mm/userfaultfd.c | 32 +++--- mm/vma.c | 23 ++--- mm/vma.h | 15 ++- tools/testing/vma/linux/atomic.h | 5 + tools/testing/vma/vma_internal.h | 93 ++++++++---------- 18 files changed, 435 insertions(+), 260 deletions(-)
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 1/16 mm: introduce vma_start_read_locked{_nested} helpers
Introduce helper functions which can be used to read-lock a VMA when holding mmap_lock for read. Replace direct accesses to vma->vm_lock with these new helpers.
--- include/linux/mm.h | 24 ++++++++++++++++++++++++ mm/userfaultfd.c | 22 +++++----------------- 2 files changed, 29 insertions(+), 17 deletions(-)
include/linux/mm.h +24 -0
@@ -735,6 +735,30 @@ static inline bool vma_start_read(struct vm_area_struct *vma)return true; } +/* +* Use only while holding mmap read lock which guarantees that locking will not +* fail (nobody can concurrently write-lock the vma). vma_start_read() should +* not be used in such cases because it might fail due to mm_lock_seq overflow. +* This functionality is used to obtain vma read lock and drop the mmap read lock. +*/ +static inline void vma_start_read_locked_nested (struct vm_area_struct *vma,int subclass) +{ +mmap_assert_locked (vma->vm_mm); +down_read_nested (&vma->vm_lock->lock, subclass); +} + +/* +* Use only while holding mmap read lock which guarantees that locking will not +* fail (nobody can concurrently write-lock the vma). vma_start_read() should +* not be used in such cases because it might fail due to mm_lock_seq overflow. +* This functionality is used to obtain vma read lock and drop the mmap read lock. +*/ +static inline void vma_start_read_locked (struct vm_area_struct *vma) +{ +mmap_assert_locked (vma->vm_mm); +down_read (&vma->vm_lock->lock); +} +static inline void vma_end_read (struct vm_area_struct *vma) {rcu_read_lock ();/* keeps vma alive till the end of up_read */
mm/userfaultfd.c +5 -17
@@ -84,16 +84,8 @@ static struct vm_area_struct *uffd_lock_vma(struct mm_struct *mm,mmap_read_lock (mm); vma =find_vma_and_prepare_anon (mm, address); -if (!IS_ERR (vma)) { -/* -* We cannot use vma_start_read() as it may fail due to -* false locked (see comment in vma_start_read()). We -* can avoid that by directly locking vm_lock under -* mmap_lock, which guarantees that nobody can lock the -* vma for write (vma_start_write()) under us. -*/ -down_read (&vma->vm_lock->lock); - } +if (!IS_ERR (vma)) +vma_start_read_locked (vma);mmap_read_unlock (mm);return vma;
@@ -1490,14 +1482,10 @@ static int uffd_move_lock(struct mm_struct *mm,mmap_read_lock (mm); err =find_vmas_mm_locked (mm, dst_start, src_start, dst_vmap, src_vmap);if (!err) { -/* -* See comment in uffd_lock_vma() as to why not using -* vma_start_read() here. -*/ -down_read (&(*dst_vmap)->vm_lock->lock); +vma_start_read_locked (*dst_vmap);if (*dst_vmap != *src_vmap) -down_read_nested (&(*src_vmap)->vm_lock->lock, -SINGLE_DEPTH_NESTING ); +vma_start_read_locked_nested (*src_vmap, +SINGLE_DEPTH_NESTING ); }mmap_read_unlock (mm);return err;
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 2/16 mm: move per-vma lock into vm_area_struct
In reply to Suren BaghdasaryanBack when per-vma locks were introduces, vm_lock was moved out of vm_area_struct in [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging the vm_lock back into vm_area_struct also allows vm_area_struct to use SLAB_TYPESAFE_BY_RCU later in this patchset. Move vm_lock back into vm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned as well. With kernel compiled using defconfig, this causes VMA memory consumption to grow from 160 (vm_area_struct) + 40 (vm_lock) bytes to 256 bytes:
slabinfo before:
<name> ... <objsize> <objperslab> <pagesperslab> : ...
vma_lock ... 40 102 1 : ...
vm_area_struct ... 160 51 2 : ...
slabinfo after moving vm_lock:
<name> ... <objsize> <objperslab> <pagesperslab> : ...
vm_area_struct ... 256 32 2 : ...
Aggregate VMA memory consumption per 1000 VMAs grows from 50 to 64 pages, which is 5.5MB per 100000 VMAs. Note that the size of this structure is dependent on the kernel configuration and typically the original size is higher than 160 bytes. Therefore these calculations are close to the worst case scenario. A more realistic vm_area_struct usage before this change is:
<name> ... <objsize> <objperslab> <pagesperslab> : ...
vma_lock ... 40 102 1 : ...
vm_area_struct ... 176 46 2 : ...Aggregate VMA memory consumption per 1000 VMAs grows from 54 to 64 pages, which is 3.9MB per 100000 VMAs. This memory consumption growth can be addressed later by optimizing the vm_lock.
[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
--- include/linux/mm.h | 28 ++++++++++-------- include/linux/mm_types.h | 6 ++-- kernel/fork.c | 49 ++++---------------------------- tools/testing/vma/vma_internal.h | 33 +++++---------------- 4 files changed, 32 insertions(+), 84 deletions(-)
include/linux/mm.h +16 -12
@@ -697,6 +697,12 @@ static inline void vma_numab_state_free(struct vm_area_struct *vma) {}#endif /* CONFIG_NUMA_BALANCING */ #ifdef CONFIG_PER_VMA_LOCK +static inline void vma_lock_init (struct vm_area_struct *vma) +{ +init_rwsem (&vma->vm_lock.lock); + vma->vm_lock_seq =UINT_MAX ; +} +/* * Try to read-lock a vma. The function is allowed to occasionally yield false * locked result to avoid performance overhead, in which case we fall back to
@@ -714,7 +720,7 @@ static inline bool vma_start_read(struct vm_area_struct *vma)if (READ_ONCE (vma->vm_lock_seq) ==READ_ONCE (vma->vm_mm->mm_lock_seq.sequence))return false; -if (unlikely (down_read_trylock (&vma->vm_lock->lock) ==0 )) +if (unlikely (down_read_trylock (&vma->vm_lock.lock) ==0 ))return false;/*
@@ -729,7 +735,7 @@ static inline bool vma_start_read(struct vm_area_struct *vma)* This pairs with RELEASE semantics in vma_end_write_all(). */ if (unlikely (vma->vm_lock_seq ==raw_read_seqcount (&vma->vm_mm->mm_lock_seq))) { -up_read (&vma->vm_lock->lock); +up_read (&vma->vm_lock.lock);return false; }return true;
@@ -744,7 +750,7 @@ static inline bool vma_start_read(struct vm_area_struct *vma)static inline void vma_start_read_locked_nested (struct vm_area_struct *vma,int subclass) {mmap_assert_locked (vma->vm_mm); -down_read_nested (&vma->vm_lock->lock, subclass); +down_read_nested (&vma->vm_lock.lock, subclass); }/*
@@ -756,13 +762,13 @@ static inline void vma_start_read_locked_nested(struct vm_area_struct *vma, intstatic inline void vma_start_read_locked (struct vm_area_struct *vma) {mmap_assert_locked (vma->vm_mm); -down_read (&vma->vm_lock->lock); +down_read (&vma->vm_lock.lock); }static inline void vma_end_read (struct vm_area_struct *vma) {rcu_read_lock ();/* keeps vma alive till the end of up_read */ -up_read (&vma->vm_lock->lock); +up_read (&vma->vm_lock.lock);rcu_read_unlock (); }
@@ -791,7 +797,7 @@ static inline void vma_start_write(struct vm_area_struct *vma)if (__is_vma_write_locked (vma, &mm_lock_seq))return ; -down_write (&vma->vm_lock->lock); +down_write (&vma->vm_lock.lock);/* * We should use WRITE_ONCE() here because we can have concurrent reads * from the early lockless pessimistic check in vma_start_read().
@@ -799,7 +805,7 @@ static inline void vma_start_write(struct vm_area_struct *vma)* we should use WRITE_ONCE() for cleanliness and to keep KCSAN happy. */ WRITE_ONCE (vma->vm_lock_seq, mm_lock_seq); -up_write (&vma->vm_lock->lock); +up_write (&vma->vm_lock.lock); }static inline void vma_assert_write_locked (struct vm_area_struct *vma)
@@ -811,7 +817,7 @@ static inline void vma_assert_write_locked(struct vm_area_struct *vma)static inline void vma_assert_locked (struct vm_area_struct *vma) { -if (!rwsem_is_locked (&vma->vm_lock->lock)) +if (!rwsem_is_locked (&vma->vm_lock.lock))vma_assert_write_locked (vma); }
@@ -844,6 +850,7 @@ struct vm_area_struct *lock_vma_under_rcu(struct mm_struct *mm,#else /* CONFIG_PER_VMA_LOCK */ +static inline void vma_lock_init (struct vm_area_struct *vma) {}static inline bool vma_start_read (struct vm_area_struct *vma) {return false; }static inline void vma_end_read (struct vm_area_struct *vma) {}
@@ -878,10 +885,6 @@ static inline void assert_fault_locked(struct vm_fault *vmf)extern const struct vm_operations_struct vma_dummy_vm_ops; -/* -* WARNING: vma_init does not initialize vma->vm_lock. -* Use vm_area_alloc()/vm_area_free() if vma needs locking. -*/ static inline void vma_init (struct vm_area_struct *vma,struct mm_struct *mm) {memset (vma,0 ,sizeof (*vma));
@@ -890,6 +893,7 @@ static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm)INIT_LIST_HEAD (&vma->anon_vma_chain);vma_mark_detached (vma, false);vma_numab_state_init (vma); +vma_lock_init (vma); }/* Use when VMA is not part of the VMA tree and needs no locking */
include/linux/mm_types.h +4 -2
@@ -738,8 +738,6 @@ struct vm_area_struct {* slowpath. */ unsigned int vm_lock_seq; -/* Unstable RCU readers are allowed to read this. */ -struct vma_lock *vm_lock;#endif /*
@@ -792,6 +790,10 @@ struct vm_area_struct {struct vma_numab_state *numab_state;/* NUMA Balancing state */ #endif struct vm_userfaultfd_ctx vm_userfaultfd_ctx; +#ifdef CONFIG_PER_VMA_LOCK +/* Unstable RCU readers are allowed to read this. */ +struct vma_lock vm_lock____cacheline_aligned_in_smp ; +#endif } __randomize_layout;#ifdef CONFIG_NUMA
kernel/fork.c +5 -44
85 lines not on this page · raw message
tools/testing/vma/vma_internal.h +7 -26
@@ -274,10 +274,10 @@ struct vm_area_struct {/* * Can only be written (using WRITE_ONCE()) while holding both: * - mmap_lock (in write mode) -* - vm_lock->lock (in write mode) +* - vm_lock.lock (in write mode) * Can be read reliably while holding one of: * - mmap_lock (in read or write mode) -* - vm_lock->lock (in read or write mode) +* - vm_lock.lock (in read or write mode) * Can be read unreliably (using READ_ONCE()) for pessimistic bailout * while holding nothing (except RCU to keep the VMA struct allocated). *
@@ -286,7 +286,7 @@ struct vm_area_struct {* slowpath. */ unsigned int vm_lock_seq; -struct vma_lock *vm_lock; +struct vma_lock vm_lock;#endif /*
@@ -463,17 +463,10 @@ static inline struct vm_area_struct *vma_next(struct vma_iterator *vmi)return mas_find (&vmi->mas,ULONG_MAX ); } -static inline bool vma_lock_alloc (struct vm_area_struct *vma) +static inline void vma_lock_init (struct vm_area_struct *vma) { - vma->vm_lock =calloc (1 ,sizeof (struct vma_lock )); - -if (!vma->vm_lock) -return false; - -init_rwsem (&vma->vm_lock->lock); +init_rwsem (&vma->vm_lock.lock); vma->vm_lock_seq =UINT_MAX ; - -return true; }static inline void vma_assert_write_locked (struct vm_area_struct *);
@@ -496,6 +489,7 @@ static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm) vma->vm_ops = &vma_dummy_vm_ops;INIT_LIST_HEAD (&vma->anon_vma_chain);vma_mark_detached (vma, false); +vma_lock_init (vma); }static inline struct vm_area_struct *vm_area_alloc (struct mm_struct *mm)
@@ -506,10 +500,6 @@ static inline struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return NULL ;vma_init (vma, mm); -if (!vma_lock_alloc (vma)) { -free (vma); -return NULL ; - }return vma; }
@@ -522,10 +512,7 @@ static inline struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)return NULL ;memcpy (new, orig,sizeof (*new)); -if (!vma_lock_alloc (new)) { -free (new); -return NULL ; - } +vma_lock_init (new);INIT_LIST_HEAD (&new->anon_vma_chain);return new;
@@ -695,14 +682,8 @@ static inline void mpol_put(struct mempolicy *) { } -static inline void vma_lock_free (struct vm_area_struct *vma) -{ -free (vma->vm_lock); -} -static inline void __vm_area_free (struct vm_area_struct *vma) { -vma_lock_free (vma);free (vma); }
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 3/16 mm: mark vma as detached until it's added into vma tree
In reply to Suren BaghdasaryanCurrent implementation does not set detached flag when a VMA is first allocated. This does not represent the real state of the VMA, which is detached until it is added into mm's VMA tree. Fix this by marking new VMAs as detached and resetting detached flag only after VMA is added into a tree.
Introduce vma_mark_attached() to make the API more readable and to simplify possible future cleanup when vma->vm_mm might be used to indicate detached vma and vma_mark_attached() will need an additional mm parameter.
--- include/linux/mm.h | 27 ++++++++++++++++++++------- kernel/fork.c | 4 ++++ mm/memory.c | 2 +- mm/vma.c | 6 +++--- mm/vma.h | 2 ++ tools/testing/vma/vma_internal.h | 17 ++++++++++++----- 6 files changed, 42 insertions(+), 16 deletions(-)
include/linux/mm.h +20 -7
@@ -821,12 +821,21 @@ static inline void vma_assert_locked(struct vm_area_struct *vma)vma_assert_write_locked (vma); } -static inline void vma_mark_detached (struct vm_area_struct *vma,bool detached) +static inline void vma_mark_attached (struct vm_area_struct *vma) +{ + vma->detached = false; +} + +static inline void vma_mark_detached (struct vm_area_struct *vma) {/* When detaching vma should be write-locked */ -if (detached) -vma_assert_write_locked (vma); - vma->detached = detached; +vma_assert_write_locked (vma); + vma->detached = true; +} + +static inline bool is_vma_detached (struct vm_area_struct *vma) +{ +return vma->detached; }static inline void release_fault_lock (struct vm_fault *vmf)
@@ -857,8 +866,8 @@ static inline void vma_end_read(struct vm_area_struct *vma) {}static inline void vma_start_write (struct vm_area_struct *vma) {}static inline void vma_assert_write_locked (struct vm_area_struct *vma) {mmap_assert_write_locked (vma->vm_mm); } -static inline void vma_mark_detached (struct vm_area_struct *vma, -bool detached) {} +static inline void vma_mark_attached (struct vm_area_struct *vma) {} +static inline void vma_mark_detached (struct vm_area_struct *vma) {}static inline struct vm_area_struct *lock_vma_under_rcu (struct mm_struct *mm,unsigned long address)
@@ -891,7 +900,10 @@ static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm) vma->vm_mm = mm; vma->vm_ops = &vma_dummy_vm_ops;INIT_LIST_HEAD (&vma->anon_vma_chain); -vma_mark_detached (vma, false); +#ifdef CONFIG_PER_VMA_LOCK +/* vma is not locked, can't use vma_mark_detached() */ + vma->detached = true; +#endif vma_numab_state_init (vma);vma_lock_init (vma); }
@@ -1086,6 +1098,7 @@ static inline int vma_iter_bulk_store(struct vma_iterator *vmi,if (unlikely (mas_is_err (&vmi->mas)))return -ENOMEM ; +vma_mark_attached (vma);return 0 ; }
kernel/fork.c +4 -0
@@ -465,6 +465,10 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)data_race (memcpy (new, orig,sizeof (*new)));vma_lock_init (new);INIT_LIST_HEAD (&new->anon_vma_chain); +#ifdef CONFIG_PER_VMA_LOCK +/* vma is not locked, can't use vma_mark_detached() */ + new->detached = true; +#endif vma_numab_state_init (new);dup_anon_vma_name (orig, new);
mm/memory.c +1 -1
@@ -6391,7 +6391,7 @@ struct vm_area_struct *lock_vma_under_rcu(struct mm_struct *mm, goto inval;/* Check if the VMA got isolated after we found it */ -if (vma->detached) { +if (is_vma_detached (vma)) {vma_end_read (vma);count_vm_vma_lock_event (VMA_LOCK_MISS );/* The area was replaced with another one */
mm/vma.c +3 -3
@@ -327,7 +327,7 @@ static void vma_complete(struct vma_prepare *vp, struct vma_iterator *vmi,if (vp->remove) { again: -vma_mark_detached (vp->remove, true); +vma_mark_detached (vp->remove);if (vp->file) {uprobe_munmap (vp->remove, vp->remove->vm_start, vp->remove->vm_end);
@@ -1221,7 +1221,7 @@ static void reattach_vmas(struct ma_state *mas_detach)mas_set (mas_detach,0 );mas_for_each (mas_detach, vma,ULONG_MAX ) -vma_mark_detached (vma, false); +vma_mark_attached (vma);__mt_destroy (mas_detach->tree); }
@@ -1296,7 +1296,7 @@ static int vms_gather_munmap_vmas(struct vma_munmap_struct *vms,if (error) goto munmap_gather_failed; -vma_mark_detached (next, true); +vma_mark_detached (next); nrpages =vma_pages (next); vms->nr_pages += nrpages;
mm/vma.h +2 -0
@@ -157,6 +157,7 @@ static inline int vma_iter_store_gfp(struct vma_iterator *vmi,if (unlikely (mas_is_err (&vmi->mas)))return -ENOMEM ; +vma_mark_attached (vma);return 0 ; }
@@ -389,6 +390,7 @@ static inline void vma_iter_store(struct vma_iterator *vmi,__mas_set_range (&vmi->mas, vma->vm_start, vma->vm_end -1 );mas_store_prealloc (&vmi->mas, vma); +vma_mark_attached (vma); }static inline unsigned long vma_iter_addr (struct vma_iterator *vmi)
tools/testing/vma/vma_internal.h +12 -5
@@ -469,13 +469,17 @@ static inline void vma_lock_init(struct vm_area_struct *vma) vma->vm_lock_seq =UINT_MAX ; } +static inline void vma_mark_attached (struct vm_area_struct *vma) +{ + vma->detached = false; +} +static inline void vma_assert_write_locked (struct vm_area_struct *); -static inline void vma_mark_detached (struct vm_area_struct *vma,bool detached) +static inline void vma_mark_detached (struct vm_area_struct *vma) {/* When detaching vma should be write-locked */ -if (detached) -vma_assert_write_locked (vma); - vma->detached = detached; +vma_assert_write_locked (vma); + vma->detached = true; }extern const struct vm_operations_struct vma_dummy_vm_ops;
@@ -488,7 +492,8 @@ static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm) vma->vm_mm = mm; vma->vm_ops = &vma_dummy_vm_ops;INIT_LIST_HEAD (&vma->anon_vma_chain); -vma_mark_detached (vma, false); +/* vma is not locked, can't use vma_mark_detached() */ + vma->detached = true;vma_lock_init (vma); }
@@ -514,6 +519,8 @@ static inline struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)memcpy (new, orig,sizeof (*new));vma_lock_init (new);INIT_LIST_HEAD (&new->anon_vma_chain); +/* vma is not locked, can't use vma_mark_detached() */ + new->detached = true;return new; }
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 4/16 mm: introduce vma_iter_store_attached() to use with attached vmas
In reply to Suren Baghdasaryanvma_iter_store() functions can be used both when adding a new vma and when updating an existing one. However for existing ones we do not need to mark them attached as they are already marked that way. Introduce vma_iter_store_attached() to be used with already attached vmas.
--- include/linux/mm.h | 12 ++++++++++++ mm/vma.c | 8 ++++---- mm/vma.h | 11 +++++++++-- 3 files changed, 25 insertions(+), 6 deletions(-)
include/linux/mm.h +12 -0
@@ -821,6 +821,16 @@ static inline void vma_assert_locked(struct vm_area_struct *vma)vma_assert_write_locked (vma); } +static inline void vma_assert_attached (struct vm_area_struct *vma) +{ +VM_BUG_ON_VMA (vma->detached, vma); +} + +static inline void vma_assert_detached (struct vm_area_struct *vma) +{ +VM_BUG_ON_VMA (!vma->detached, vma); +} +static inline void vma_mark_attached (struct vm_area_struct *vma) { vma->detached = false;
@@ -866,6 +876,8 @@ static inline void vma_end_read(struct vm_area_struct *vma) {}static inline void vma_start_write (struct vm_area_struct *vma) {}static inline void vma_assert_write_locked (struct vm_area_struct *vma) {mmap_assert_write_locked (vma->vm_mm); } +static inline void vma_assert_attached (struct vm_area_struct *vma) {} +static inline void vma_assert_detached (struct vm_area_struct *vma) {}static inline void vma_mark_attached (struct vm_area_struct *vma) {}static inline void vma_mark_detached (struct vm_area_struct *vma) {}
mm/vma.c +4 -4
@@ -660,14 +660,14 @@ static int commit_merge(struct vma_merge_struct *vmg,vma_set_range (vmg->vma, vmg->start, vmg->end, vmg->pgoff);if (expanded) -vma_iter_store (vmg->vmi, vmg->vma); +vma_iter_store_attached (vmg->vmi, vmg->vma);if (adj_start) { adjust->vm_start += adj_start; adjust->vm_pgoff +=PHYS_PFN (adj_start);if (adj_start <0 ) {WARN_ON (expanded); -vma_iter_store (vmg->vmi, adjust); +vma_iter_store_attached (vmg->vmi, adjust); } }
@@ -2845,7 +2845,7 @@ int expand_upwards(struct vm_area_struct *vma, unsigned long address)anon_vma_interval_tree_pre_update_vma (vma); vma->vm_end = address;/* Overwrite old entry in mtree. */ -vma_iter_store (&vmi, vma); +vma_iter_store_attached (&vmi, vma);anon_vma_interval_tree_post_update_vma (vma);perf_event_mmap (vma);
@@ -2925,7 +2925,7 @@ int expand_downwards(struct vm_area_struct *vma, unsigned long address) vma->vm_start = address; vma->vm_pgoff -= grow;/* Overwrite old entry in mtree. */ -vma_iter_store (&vmi, vma); +vma_iter_store_attached (&vmi, vma);anon_vma_interval_tree_post_update_vma (vma);perf_event_mmap (vma);
mm/vma.h +9 -2
@@ -365,9 +365,10 @@ static inline struct vm_area_struct *vma_iter_load(struct vma_iterator *vmi) }/* Store a VMA with preallocated memory */ -static inline void vma_iter_store (struct vma_iterator *vmi, -struct vm_area_struct *vma) +static inline void vma_iter_store_attached (struct vma_iterator *vmi, +struct vm_area_struct *vma) { +vma_assert_attached (vma);#if defined(CONFIG_DEBUG_VM_MAPLE_TREE ) if (MAS_WARN_ON(&vmi->mas, vmi->mas.status != ma_start &&
@@ -390,7 +391,13 @@ static inline void vma_iter_store(struct vma_iterator *vmi,__mas_set_range (&vmi->mas, vma->vm_start, vma->vm_end -1 );mas_store_prealloc (&vmi->mas, vma); +} + +static inline void vma_iter_store (struct vma_iterator *vmi, +struct vm_area_struct *vma) +{vma_mark_attached (vma); +vma_iter_store_attached (vmi, vma); }static inline unsigned long vma_iter_addr (struct vma_iterator *vmi)
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Re: [PATCH v8 04/16] mm: introduce vma_iter_store_attached() to use with attached vmas
Quoted 6 lines · Suren Baghdasaryan
vma_iter_store()functions can be used both when adding a new vma and when updating an existing one. However for existing ones we do not need to mark them attached as they are already marked that way. Introducevma_iter_store_attached()to be used with already attached vmas.Signed-off-by Suren Baghdasaryan <[email protected]>
Suren Baghdasaryan
maintainerPATCH v8 5/16 mm: mark vmas detached upon exit
In reply to Suren BaghdasaryanWhen exit_mmap() removes vmas belonging to an exiting task, it does not mark them as detached since they can't be reached by other tasks and they will be freed shortly. Once we introduce vma reuse, all vmas will have to be in detached state before they are freed to ensure vma when reused is in a consistent state. Add missing vma_mark_detached() before freeing the vma.
--- mm/vma.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
mm/vma.c +4 -2
@@ -413,10 +413,12 @@ void remove_vma(struct vm_area_struct *vma, bool unreachable)if (vma->vm_file)fput (vma->vm_file);mpol_put (vma_policy (vma)); -if (unreachable) +if (unreachable) { +vma_mark_detached (vma);__vm_area_free (vma); -else + }else {vm_area_free (vma); + } }/*
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 6/16 types: move struct rcuwait into types.h
In reply to Suren BaghdasaryanMove rcuwait struct definition into types.h so that rcuwait can be used without including rcuwait.h which includes other headers. Without this change mm_types.h can't use rcuwait due to a the following circular dependency:
mm_types.h -> rcuwait.h -> signal.h -> mm_types.h
--- include/linux/rcuwait.h | 13 +------------ include/linux/types.h | 12 ++++++++++++ 2 files changed, 13 insertions(+), 12 deletions(-)
include/linux/rcuwait.h +1 -12
@@ -4,18 +4,7 @@#include <linux/rcupdate.h> #include <linux/sched/signal.h> - -/* -* rcuwait provides a way of blocking and waking up a single -* task in an rcu-safe manner. -* -* The only time @task is non-nil is when a user is blocked (or -* checking if it needs to) on a condition, and reset as soon as we -* know that the condition has succeeded and are awoken. -*/ -struct rcuwait { -struct task_struct __rcu *task; -}; +#include <linux/types.h> #define __RCUWAIT_INITIALIZER (name) \ { .task = NULL, }
include/linux/types.h +12 -0
@@ -248,5 +248,17 @@ typedef void (*swap_func_t)(void *a, void *b, int size);typedef int (*cmp_r_func_t )(const void *a,const void *b,const void *priv);typedef int (*cmp_func_t )(const void *a,const void *b); +/* +* rcuwait provides a way of blocking and waking up a single +* task in an rcu-safe manner. +* +* The only time @task is non-nil is when a user is blocked (or +* checking if it needs to) on a condition, and reset as soon as we +* know that the condition has succeeded and are awoken. +*/ +struct rcuwait { +struct task_struct __rcu *task; +}; +#endif /* __ASSEMBLY__ */ #endif /* _LINUX_TYPES_H */
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 7/16 mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
In reply to Suren BaghdasaryanWith upcoming replacement of vm_lock with vm_refcnt, we need to handle a possibility of vma_start_read_locked/vma_start_read_locked_nested failing due to refcount overflow. Prepare for such possibility by changing these APIs and adjusting their users.
--- include/linux/mm.h | 6 ++++-- mm/userfaultfd.c | 18 +++++++++++++----- 2 files changed, 17 insertions(+), 7 deletions(-)
include/linux/mm.h +4 -2
@@ -747,10 +747,11 @@ static inline bool vma_start_read(struct vm_area_struct *vma)* not be used in such cases because it might fail due to mm_lock_seq overflow. * This functionality is used to obtain vma read lock and drop the mmap read lock. */ -static inline void vma_start_read_locked_nested (struct vm_area_struct *vma,int subclass) +static inline bool vma_start_read_locked_nested (struct vm_area_struct *vma,int subclass) {mmap_assert_locked (vma->vm_mm);down_read_nested (&vma->vm_lock.lock, subclass); +return true; }/*
@@ -759,10 +760,11 @@ static inline void vma_start_read_locked_nested(struct vm_area_struct *vma, int* not be used in such cases because it might fail due to mm_lock_seq overflow. * This functionality is used to obtain vma read lock and drop the mmap read lock. */ -static inline void vma_start_read_locked (struct vm_area_struct *vma) +static inline bool vma_start_read_locked (struct vm_area_struct *vma) {mmap_assert_locked (vma->vm_mm);down_read (&vma->vm_lock.lock); +return true; }static inline void vma_end_read (struct vm_area_struct *vma)
mm/userfaultfd.c +13 -5
@@ -85,7 +85,8 @@ static struct vm_area_struct *uffd_lock_vma(struct mm_struct *mm,mmap_read_lock (mm); vma =find_vma_and_prepare_anon (mm, address);if (!IS_ERR (vma)) -vma_start_read_locked (vma); +if (!vma_start_read_locked (vma)) + vma =ERR_PTR (-EAGAIN );mmap_read_unlock (mm);return vma;
@@ -1482,10 +1483,17 @@ static int uffd_move_lock(struct mm_struct *mm,mmap_read_lock (mm); err =find_vmas_mm_locked (mm, dst_start, src_start, dst_vmap, src_vmap);if (!err) { -vma_start_read_locked (*dst_vmap); -if (*dst_vmap != *src_vmap) -vma_start_read_locked_nested (*src_vmap, -SINGLE_DEPTH_NESTING ); +if (vma_start_read_locked (*dst_vmap)) { +if (*dst_vmap != *src_vmap) { +if (!vma_start_read_locked_nested (*src_vmap, +SINGLE_DEPTH_NESTING )) { +vma_end_read (*dst_vmap); + err = -EAGAIN ; + } + } + }else { + err = -EAGAIN ; + } }mmap_read_unlock (mm);return err;
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 8/16 mm: move mmap_init_lock() out of the header file
In reply to Suren Baghdasaryanmmap_init_lock() is used only from mm_init() in fork.c, therefore it does not have to reside in the header file. This move lets us avoid including additional headers in mmap_lock.h later, when mmap_init_lock() needs to initialize rcuwait object.
--- include/linux/mmap_lock.h | 6 ------ kernel/fork.c | 6 ++++++ 2 files changed, 6 insertions(+), 6 deletions(-)
include/linux/mmap_lock.h +0 -6
@@ -122,12 +122,6 @@ static inline bool mmap_lock_speculate_retry(struct mm_struct *mm, unsigned int#endif /* CONFIG_PER_VMA_LOCK */ -static inline void mmap_init_lock (struct mm_struct *mm) -{ -init_rwsem (&mm->mmap_lock); -mm_lock_seqcount_init (mm); -} -static inline void mmap_write_lock (struct mm_struct *mm) {__mmap_lock_trace_start_locking (mm, true);
kernel/fork.c +6 -0
@@ -1219,6 +1219,12 @@ static void mm_init_uprobes_state(struct mm_struct *mm)#endif } +static inline void mmap_init_lock (struct mm_struct *mm) +{ +init_rwsem (&mm->mmap_lock); +mm_lock_seqcount_init (mm); +} +static struct mm_struct *mm_init (struct mm_struct *mm,struct task_struct *p,struct user_namespace *user_ns) {
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 9/16 mm: uninline the main body of vma_start_write()
In reply to Suren Baghdasaryanvma_start_write() is used in many places and will grow in size very soon. It is not used in performance critical paths and uninlining it should limit the future code size growth.
No functional changes.
--- include/linux/mm.h | 12 +++--------- mm/memory.c | 14 ++++++++++++++ 2 files changed, 17 insertions(+), 9 deletions(-)
include/linux/mm.h +3 -9
@@ -787,6 +787,8 @@ static bool __is_vma_write_locked(struct vm_area_struct *vma, unsigned int *mm_lreturn (vma->vm_lock_seq == *mm_lock_seq); } +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq); +/* * Begin writing to a VMA. * Exclude concurrent readers under the per-VMA lock until the currently
@@ -799,15 +801,7 @@ static inline void vma_start_write(struct vm_area_struct *vma)if (__is_vma_write_locked (vma, &mm_lock_seq))return ; -down_write (&vma->vm_lock.lock); -/* -* We should use WRITE_ONCE() here because we can have concurrent reads -* from the early lockless pessimistic check in vma_start_read(). -* We don't really care about the correctness of that early check, but -* we should use WRITE_ONCE() for cleanliness and to keep KCSAN happy. -*/ -WRITE_ONCE (vma->vm_lock_seq, mm_lock_seq); -up_write (&vma->vm_lock.lock); +__vma_start_write (vma, mm_lock_seq); }static inline void vma_assert_write_locked (struct vm_area_struct *vma)
mm/memory.c +14 -0
@@ -6370,6 +6370,20 @@ struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm,#endif #ifdef CONFIG_PER_VMA_LOCK +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq) +{ +down_write (&vma->vm_lock.lock); +/* +* We should use WRITE_ONCE() here because we can have concurrent reads +* from the early lockless pessimistic check in vma_start_read(). +* We don't really care about the correctness of that early check, but +* we should use WRITE_ONCE() for cleanliness and to keep KCSAN happy. +*/ +WRITE_ONCE (vma->vm_lock_seq, mm_lock_seq); +up_write (&vma->vm_lock.lock); +} +EXPORT_SYMBOL_GPL (__vma_start_write); +/* * Lookup and lock a VMA under RCU protection. Returned VMA is guaranteed to be * stable and not isolated. If the VMA is not found or is being modified the
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 10/16 refcount: introduce __refcount_{add|inc}_not_zero_limited
In reply to Suren BaghdasaryanIntroduce functions to increase refcount but with a top limit above which they will fail to increase (the limit is inclusive). Setting the limit to INT_MAX indicates no limit.
--- include/linux/refcount.h | 20 +++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-)
include/linux/refcount.h +19 -1
@@ -137,13 +137,19 @@ static inline unsigned int refcount_read(const refcount_t *r) }static inline __must_check __signed_wrap -bool __refcount_add_not_zero(int i, refcount_t *r, int *oldp) +bool __refcount_add_not_zero_limited(int i, refcount_t *r, int *oldp, +int limit) {int old =refcount_read (r);do {if (!old)break ; +if (i > limit - old) { +if (oldp) + *oldp = old; +return false; + } }while (!atomic_try_cmpxchg_relaxed (&r->refs, &old, old + i));if (oldp)
@@ -155,6 +161,12 @@ bool __refcount_add_not_zero(int i, refcount_t *r, int *oldp)return old; } +static inline __must_check __signed_wrap +bool __refcount_add_not_zero(int i, refcount_t *r, int *oldp) +{ +return __refcount_add_not_zero_limited (i, r, oldp,INT_MAX ); +} +/** * refcount_add_not_zero - add a value to a refcount unless it is 0 * @i: the value to add to the refcount
@@ -213,6 +225,12 @@ static inline void refcount_add(int i, refcount_t *r)__refcount_add (i, r,NULL ); } +static inline __must_check bool __refcount_inc_not_zero_limited (refcount_t *r, +int *oldp,int limit) +{ +return __refcount_add_not_zero_limited (1 , r, oldp, limit); +} +static inline __must_check bool __refcount_inc_not_zero (refcount_t *r,int *oldp) {return __refcount_add_not_zero (1 , r, oldp);
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Re: [PATCH v8 10/16] refcount: introduce __refcount_{add|inc}_not_zero_limited
Quoted 5 lines · Suren Baghdasaryan
Introduce functions to increase refcount but with a top limit above which they will fail to increase (the limit is inclusive). Setting the limit to
INT_MAXindicates no limit.Signed-off-by Suren Baghdasaryan <[email protected]>
Suren Baghdasaryan
maintainerRe: [PATCH v8 10/16] refcount: introduce __refcount_{add|inc}_not_zero_limited
Quoted 9 lines · Vlastimil Babka
Quoted 5 lines · Suren Baghdasaryan
Introduce functions to increase refcount but with a top limit above which they will fail to increase (the limit is inclusive). Setting the limit to
INT_MAXindicates no limit.Signed-off-by Suren Baghdasaryan <[email protected]>Acked-by Vlastimil Babka <[email protected]>
Thanks Vlastimil!
Since per David's suggestion in v7 I'll be adding a new static check in this patch, I won't add your Acked-by in v9 until you have a chance to look at it again.
Quoted 1 line
Suren Baghdasaryan
maintainerPATCH v8 11/16 mm: replace vm_lock and detached flag with a reference count
In reply to Suren Baghdasaryanrw_semaphore is a sizable structure of 40 bytes and consumes considerable space for each vm_area_struct. However vma_lock has two important specifics which can be used to replace rw_semaphore with a simpler structure:
1. Readers never wait. They try to take the vma_lock and fall back to mmap_lock if that fails.
2. Only one writer at a time will ever try to write-lock a vma_lock because writers first take mmap_lock in write mode.
Because of these requirements, full rw_semaphore functionality is not needed and we can replace rw_semaphore and the vma->detached flag with a refcount (vm_refcnt).
When vma is in detached state, vm_refcnt is 0 and only a call to vma_mark_attached() can take it out of this state. Note that unlike before, now we enforce both vma_mark_attached() and vma_mark_detached() to be done only after vma has been write-locked. vma_mark_attached() changes vm_refcnt to 1 to indicate that it has been attached to the vma tree. When a reader takes read lock, it increments vm_refcnt, unless the top usable bit of vm_refcnt (0x40000000) is set, indicating presence of a writer. When writer takes write lock, it sets the top usable bit to indicate its presence. If there are readers, writer will wait using newly introduced mm->vma_writer_wait. Since all writers take mmap_lock in write mode first, there can be only one writer at a time. The last reader to release the lock will signal the writer to wake up.
refcount might overflow if there are many competing readers, in which case read-locking will fail. Readers are expected to handle such failures. In summary:
1. all readers increment the vm_refcnt;
2. writer sets top usable (writer) bit of vm_refcnt;
3. readers cannot increment the vm_refcnt if the writer bit is set;
4. in the presence of readers, writer must wait for the vm_refcnt to drop to 1 (ignoring the writer bit), indicating an attached vma with no readers;
5. vm_refcnt overflow is handled by the readers.
--- include/linux/mm.h | 98 ++++++++++++++++++++++---------- include/linux/mm_types.h | 22 ++++--- kernel/fork.c | 13 ++--- mm/init-mm.c | 1 + mm/memory.c | 77 +++++++++++++++++++++---- tools/testing/vma/linux/atomic.h | 5 ++ tools/testing/vma/vma_internal.h | 66 +++++++++++---------- 7 files changed, 193 insertions(+), 89 deletions(-)
include/linux/mm.h +67 -31
@@ -32,6 +32,7 @@#include <linux/memremap.h> #include <linux/slab.h> #include <linux/cacheinfo.h> +#include <linux/rcuwait.h> struct mempolicy ;struct anon_vma ;
@@ -697,12 +698,41 @@ static inline void vma_numab_state_free(struct vm_area_struct *vma) {}#endif /* CONFIG_NUMA_BALANCING */ #ifdef CONFIG_PER_VMA_LOCK -static inline void vma_lock_init (struct vm_area_struct *vma) +static inline void vma_lock_init (struct vm_area_struct *vma,bool reset_refcnt) { -init_rwsem (&vma->vm_lock.lock); +#ifdef CONFIG_DEBUG_LOCK_ALLOC +static struct lock_class_key lockdep_key; + +lockdep_init_map (&vma->vmlock_dep_map,"vm_lock" , &lockdep_key,0 ); +#endif +if (reset_refcnt) +refcount_set (&vma->vm_refcnt,0 ); vma->vm_lock_seq =UINT_MAX ; } +static inline bool is_vma_writer_only (int refcnt) +{ +/* +* With a writer and no readers, refcnt is VMA_LOCK_OFFSET if the vma +* is detached and (VMA_LOCK_OFFSET + 1) if it is attached. Waiting on +* a detached vma happens only in vma_mark_detached() and is a rare +* case, therefore most of the time there will be no unnecessary wakeup. +*/ +return refcnt &VMA_LOCK_OFFSET && refcnt <=VMA_LOCK_OFFSET +1 ; +} + +static inline void vma_refcount_put (struct vm_area_struct *vma) +{ +int oldcnt; + +if (!__refcount_dec_and_test (&vma->vm_refcnt, &oldcnt)) { +rwsem_release (&vma->vmlock_dep_map, _RET_IP_); + +if (is_vma_writer_only (oldcnt -1 )) +rcuwait_wake_up (&vma->vm_mm->vma_writer_wait); + } +} +/* * Try to read-lock a vma. The function is allowed to occasionally yield false * locked result to avoid performance overhead, in which case we fall back to
@@ -710,6 +740,8 @@ static inline void vma_lock_init(struct vm_area_struct *vma)*/ static inline bool vma_start_read (struct vm_area_struct *vma) { +int oldcnt; +/* * Check before locking. A race might cause false locked result. * We can use READ_ONCE() for the mm_lock_seq here, and don't need
@@ -720,13 +752,19 @@ static inline bool vma_start_read(struct vm_area_struct *vma)if (READ_ONCE (vma->vm_lock_seq) ==READ_ONCE (vma->vm_mm->mm_lock_seq.sequence))return false; -if (unlikely (down_read_trylock (&vma->vm_lock.lock) ==0 )) +/* +* If VMA_LOCK_OFFSET is set, __refcount_inc_not_zero_limited() will fail +* because VMA_REF_LIMIT is less than VMA_LOCK_OFFSET. +*/ +if (unlikely (!__refcount_inc_not_zero_limited (&vma->vm_refcnt, &oldcnt, +VMA_REF_LIMIT )))return false; +rwsem_acquire_read (&vma->vmlock_dep_map,0 ,1 , _RET_IP_);/* -* Overflow might produce false locked result. +* Overflow of vm_lock_seq/mm_lock_seq might produce false locked result. * False unlocked result is impossible because we modify and check -* vma->vm_lock_seq under vma->vm_lock protection and mm->mm_lock_seq +* vma->vm_lock_seq under vma->vm_refcnt protection and mm->mm_lock_seq * modification invalidates all existing locks. * * We must use ACQUIRE semantics for the mm_lock_seq so that if we are
@@ -735,9 +773,10 @@ static inline bool vma_start_read(struct vm_area_struct *vma)* This pairs with RELEASE semantics in vma_end_write_all(). */ if (unlikely (vma->vm_lock_seq ==raw_read_seqcount (&vma->vm_mm->mm_lock_seq))) { -up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);return false; } +return true; }
@@ -749,8 +788,14 @@ static inline bool vma_start_read(struct vm_area_struct *vma)*/ static inline bool vma_start_read_locked_nested (struct vm_area_struct *vma,int subclass) { +int oldcnt; +mmap_assert_locked (vma->vm_mm); -down_read_nested (&vma->vm_lock.lock, subclass); +if (unlikely (!__refcount_inc_not_zero_limited (&vma->vm_refcnt, &oldcnt, +VMA_REF_LIMIT ))) +return false; + +rwsem_acquire_read (&vma->vmlock_dep_map,0 ,1 , _RET_IP_);return true; }
@@ -762,15 +807,13 @@ static inline bool vma_start_read_locked_nested(struct vm_area_struct *vma, int*/ static inline bool vma_start_read_locked (struct vm_area_struct *vma) { -mmap_assert_locked (vma->vm_mm); -down_read (&vma->vm_lock.lock); -return true; +return vma_start_read_locked_nested (vma,0 ); }static inline void vma_end_read (struct vm_area_struct *vma) {rcu_read_lock ();/* keeps vma alive till the end of up_read */ -up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);rcu_read_unlock (); }
@@ -813,36 +856,33 @@ static inline void vma_assert_write_locked(struct vm_area_struct *vma)static inline void vma_assert_locked (struct vm_area_struct *vma) { -if (!rwsem_is_locked (&vma->vm_lock.lock)) +if (refcount_read (&vma->vm_refcnt) <=1 )vma_assert_write_locked (vma); } +/* +* WARNING: to avoid racing with vma_mark_attached()/vma_mark_detached(), these +* assertions should be made either under mmap_write_lock or when the object +* has been isolated under mmap_write_lock, ensuring no competing writers. +*/ static inline void vma_assert_attached (struct vm_area_struct *vma) { -VM_BUG_ON_VMA (vma->detached, vma); +VM_BUG_ON_VMA (!refcount_read (&vma->vm_refcnt), vma); }static inline void vma_assert_detached (struct vm_area_struct *vma) { -VM_BUG_ON_VMA (!vma->detached, vma); +VM_BUG_ON_VMA (refcount_read (&vma->vm_refcnt), vma); }static inline void vma_mark_attached (struct vm_area_struct *vma) { - vma->detached = false; -} - -static inline void vma_mark_detached (struct vm_area_struct *vma) -{ -/* When detaching vma should be write-locked */ vma_assert_write_locked (vma); - vma->detached = true; +vma_assert_detached (vma); +refcount_set (&vma->vm_refcnt,1 ); } -static inline bool is_vma_detached (struct vm_area_struct *vma) -{ -return vma->detached; -} +void vma_mark_detached (struct vm_area_struct *vma);static inline void release_fault_lock (struct vm_fault *vmf) {
@@ -865,7 +905,7 @@ struct vm_area_struct *lock_vma_under_rcu(struct mm_struct *mm,#else /* CONFIG_PER_VMA_LOCK */ -static inline void vma_lock_init (struct vm_area_struct *vma) {} +static inline void vma_lock_init (struct vm_area_struct *vma,bool reset_refcnt) {}static inline bool vma_start_read (struct vm_area_struct *vma) {return false; }static inline void vma_end_read (struct vm_area_struct *vma) {}
@@ -908,12 +948,8 @@ static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm) vma->vm_mm = mm; vma->vm_ops = &vma_dummy_vm_ops;INIT_LIST_HEAD (&vma->anon_vma_chain); -#ifdef CONFIG_PER_VMA_LOCK -/* vma is not locked, can't use vma_mark_detached() */ - vma->detached = true; -#endif vma_numab_state_init (vma); -vma_lock_init (vma); +vma_lock_init (vma, false); }/* Use when VMA is not part of the VMA tree and needs no locking */
include/linux/mm_types.h +10 -12
@@ -19,6 +19,7 @@#include <linux/workqueue.h> #include <linux/seqlock.h> #include <linux/percpu_counter.h> +#include <linux/types.h> #include <asm/mmu.h>
@@ -637,9 +638,8 @@ static inline struct anon_vma_name *anon_vma_name_alloc(const char *name) }#endif -struct vma_lock { -struct rw_semaphore lock; -}; +#define VMA_LOCK_OFFSET 0x40000000 +#define VMA_REF_LIMIT (VMA_LOCK_OFFSET - 1)struct vma_numab_state {/*
@@ -717,19 +717,13 @@ struct vm_area_struct { };#ifdef CONFIG_PER_VMA_LOCK -/* -* Flag to indicate areas detached from the mm->mm_mt tree. -* Unstable RCU readers are allowed to read this. -*/ -bool detached; -/* * Can only be written (using WRITE_ONCE()) while holding both: * - mmap_lock (in write mode) -* - vm_lock->lock (in write mode) +* - vm_refcnt bit at VMA_LOCK_OFFSET is set * Can be read reliably while holding one of: * - mmap_lock (in read or write mode) -* - vm_lock->lock (in read or write mode) +* - vm_refcnt bit at VMA_LOCK_OFFSET is set or vm_refcnt > 1 * Can be read unreliably (using READ_ONCE()) for pessimistic bailout * while holding nothing (except RCU to keep the VMA struct allocated). *
@@ -792,7 +786,10 @@ struct vm_area_struct {struct vm_userfaultfd_ctx vm_userfaultfd_ctx;#ifdef CONFIG_PER_VMA_LOCK /* Unstable RCU readers are allowed to read this. */ -struct vma_lock vm_lock____cacheline_aligned_in_smp ; +refcount_t vm_refcnt____cacheline_aligned_in_smp ; +#ifdef CONFIG_DEBUG_LOCK_ALLOC +struct lockdep_map vmlock_dep_map; +#endif #endif } __randomize_layout;
@@ -927,6 +924,7 @@ struct mm_struct {* by mmlist_lock */ #ifdef CONFIG_PER_VMA_LOCK +struct rcuwait vma_writer_wait;/* * This field has lock-like semantics, meaning it is sometimes * accessed with ACQUIRE/RELEASE semantics.
kernel/fork.c +6 -7
@@ -463,12 +463,8 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)* will be reinitialized. */ data_race (memcpy (new, orig,sizeof (*new))); -vma_lock_init (new); +vma_lock_init (new, true);INIT_LIST_HEAD (&new->anon_vma_chain); -#ifdef CONFIG_PER_VMA_LOCK -/* vma is not locked, can't use vma_mark_detached() */ - new->detached = true; -#endif vma_numab_state_init (new);dup_anon_vma_name (orig, new);
@@ -477,6 +473,8 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)void __vm_area_free (struct vm_area_struct *vma) { +/* The vma should be detached while being destroyed. */ +vma_assert_detached (vma);vma_numab_state_free (vma);free_anon_vma_name (vma);kmem_cache_free (vm_area_cachep, vma);
@@ -488,8 +486,6 @@ static void vm_area_free_rcu_cb(struct rcu_head *head) struct vm_area_struct *vma = container_of(head, struct vm_area_struct, vm_rcu); -/* The vma should not be locked while being destroyed. */ -VM_BUG_ON_VMA (rwsem_is_locked (&vma->vm_lock.lock), vma); __vm_area_free(vma); }#endif
@@ -1223,6 +1219,9 @@ static inline void mmap_init_lock(struct mm_struct *mm) {init_rwsem (&mm->mmap_lock);mm_lock_seqcount_init (mm); +#ifdef CONFIG_PER_VMA_LOCK +rcuwait_init (&mm->vma_writer_wait); +#endif } static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p,
mm/init-mm.c +1 -0
@@ -40,6 +40,7 @@ struct mm_struct init_mm = { .arg_lock = __SPIN_LOCK_UNLOCKED(init_mm.arg_lock), .mmlist =LIST_HEAD_INIT (init_mm.mmlist), #ifdef CONFIG_PER_VMA_LOCK + .vma_writer_wait =__RCUWAIT_INITIALIZER (init_mm.vma_writer_wait), .mm_lock_seq = SEQCNT_ZERO(init_mm.mm_lock_seq), #endif .user_ns = &init_user_ns,
mm/memory.c +67 -10
106 lines not on this page · raw message
tools/testing/vma/linux/atomic.h +5 -0
@@ -9,4 +9,9 @@#define atomic_set (x, y) uatomic_set(x, y)#define U8_MAX UCHAR_MAX +#ifndef atomic_cmpxchg_relaxed +#define atomic_cmpxchg_relaxed uatomic_cmpxchg +#define atomic_cmpxchg_release uatomic_cmpxchg +#endif /* atomic_cmpxchg_relaxed */ +#endif /* _LINUX_ATOMIC_H */
tools/testing/vma/vma_internal.h +37 -29
140 lines not on this page · raw message
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Re: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
On Wed, 8 Jan 2025 18:30:20 -0800 Suren Baghdasaryan <[email protected]>
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 729
Quoted 7 lines
include/linux/mm.h +7
-0+ +static inline void vma_refcount_put (struct vm_area_struct *vma) +{ +int oldcnt; + +if (!__refcount_dec_and_test (&vma->vm_refcnt, &oldcnt)) { +rwsem_release (&vma->vmlock_dep_map, _RET_IP_);
In up_read() rwsem is released reguardless wakeup, which is different than what is added here. Nit.
Quoted 5 lines
include/linux/mm.h +5
-0+ +if (is_vma_writer_only (oldcnt -1 )) +rcuwait_wake_up (&vma->vm_mm->vma_writer_wait); + } +}
...
include/linux/mm.h static inline bool vma_start_read(struct vm_area_struct *vma) · line 778
Quoted 8 lines
include/linux/mm.h +1
-1@@ -735,9 +773,10 @@ static inline bool vma_start_read(struct vm_area_struct *vma)* This pairs with RELEASE semantics in vma_end_write_all(). */ if (unlikely (vma->vm_lock_seq ==raw_read_seqcount (&vma->vm_mm->mm_lock_seq))) { -up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);return false; }
void up_read(struct rw_semaphore *sem)
{
rwsem_release(&sem->dep_map, _RET_IP_);
__up_read(sem);
}Suren Baghdasaryan
maintainerRe: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
Quoted 12 lines · Hillf Danton
On Wed, 8 Jan 2025 18:30:20 -0800 Suren Baghdasaryan <[email protected]>
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 729
Quoted 7 lines
include/linux/mm.h +7
-0+ +static inline void vma_refcount_put (struct vm_area_struct *vma) +{ +int oldcnt; + +if (!__refcount_dec_and_test (&vma->vm_refcnt, &oldcnt)) { +rwsem_release (&vma->vmlock_dep_map, _RET_IP_);In
up_read()rwsem is released reguardless wakeup, which is different than what is added here. Nit.
Good point. I'll send a fixup since it's a small change. Thanks!
Quoted 22 lines
Quoted 6 lines
include/linux/mm.h +5
-0+ +if (is_vma_writer_only (oldcnt -1 )) +rcuwait_wake_up (&vma->vm_mm->vma_writer_wait); + } +}...
include/linux/mm.h static inline bool vma_start_read(struct vm_area_struct *vma) · line 778
Quoted 8 lines
include/linux/mm.h +1
-1@@ -735,9 +773,10 @@ static inline bool vma_start_read(struct vm_area_struct *vma)* This pairs with RELEASE semantics in vma_end_write_all(). */ if (unlikely (vma->vm_lock_seq ==raw_read_seqcount (&vma->vm_mm->mm_lock_seq))) { -up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);return false; }void
up_read(struct rw_semaphore *sem){ rwsem_release(&sem->dep_map, _RET_IP_); __up_read(sem); }
Re: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
In reply to Suren BaghdasaryanQuoted 33 lines · Suren Baghdasaryan
rw_semaphoreis a sizable structure of 40 bytes and consumes considerable space for eachvm_area_struct. Howevervma_lockhas two important specifics which can be used to replacerw_semaphorewith a simpler structure:
1. Readers never wait. They try to take thevma_lockand fall back tommap_lockif that fails.
2. Only one writer at a time will ever try to write-lock avma_lockbecause writers first takemmap_lockin write mode.
Because of these requirements, fullrw_semaphorefunctionality is not needed and we can replacerw_semaphoreand thevma->detachedflag with a refcount (vm_refcnt).
When vma is in detached state,vm_refcntis 0 and only a call tovma_mark_attached()can take it out of this state. Note that unlike before, now we enforce bothvma_mark_attached()andvma_mark_detached()to be done only after vma has been write-locked.vma_mark_attached()changesvm_refcntto 1 to indicate that it has been attached to the vma tree. When a reader takes read lock, it incrementsvm_refcnt, unless the top usable bit ofvm_refcnt(0x40000000) is set, indicating presence of a writer. When writer takes write lock, it sets the top usable bit to indicate its presence. If there are readers, writer will wait using newly introducedmm->vma_writer_wait. Since all writers takemmap_lockin write mode first, there can be only one writer at a time. The last reader to release the lock will signal the writer to wake up.
refcount might overflow if there are many competing readers, in which case read-locking will fail. Readers are expected to handle such failures. In summary:
1. all readers increment thevm_refcnt;
2. writer sets top usable (writer) bit ofvm_refcnt;
3. readers cannot increment thevm_refcntif the writer bit is set;
4. in the presence of readers, writer must wait for thevm_refcntto drop to 1 (ignoring the writer bit), indicating an attached vma with no readers;
5.vm_refcntoverflow is handled by the readers.
Quoted 3 lines
Suggested-by Peter Zijlstra <[email protected]>Suggested-by Matthew Wilcox <[email protected]>Signed-off-by Suren Baghdasaryan <[email protected]>
But think there's a problem that will manifest after patch 15. Also I don't feel qualified enough about the lockdep parts though (although I think I spotted another issue with those, below) so best if PeterZ can review those.
Some nits below too.
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 729
Quoted 7 lines
include/linux/mm.h +7
-0+ +static inline void vma_refcount_put (struct vm_area_struct *vma) +{ +int oldcnt; + +if (!__refcount_dec_and_test (&vma->vm_refcnt, &oldcnt)) { +rwsem_release (&vma->vmlock_dep_map, _RET_IP_);
Shouldn't we rwsem_release always? And also shouldn't it precede the refcount operation itself?
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 732
Quoted 2 lines
include/linux/mm.h +2
-0+if (is_vma_writer_only (oldcnt -1 )) +rcuwait_wake_up (&vma->vm_mm->vma_writer_wait);
Hmm hmm we should maybe read the vm_mm pointer before dropping the refcount? In case this races in a way that is_vma_writer_only tests true but the writer meanwhile finishes and frees the vma. It's safe now but not after making the cache SLAB_TYPESAFE_BY_RCU ?
include/linux/mm.h static inline bool vma_start_read_locked_nested(struct vm_area_struct *vma, int · line 815
Quoted 7 lines
include/linux/mm.h +3
-0+ } +} +static inline void vma_end_read (struct vm_area_struct *vma) {rcu_read_lock ();/* keeps vma alive till the end of up_read */
This should refer to vma_refcount_put(). But after fixing it I think we could stop doing this altogether? It will no longer keep vma "alive" with SLAB_TYPESAFE_BY_RCU.
Quoted 5 lines
include/linux/mm.h +1
-1-up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);rcu_read_unlock (); }
<snip>
Quoted 31 lines
mm/memory.c +24
-0@@ -6370,9 +6370,41 @@ struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm,#endif #ifdef CONFIG_PER_VMA_LOCK +static inline bool __vma_enter_locked (struct vm_area_struct *vma,unsigned int tgt_refcnt) +{ +/* +* If vma is detached then only vma_mark_attached() can raise the +* vm_refcnt. mmap_write_lock prevents racing with vma_mark_attached(). +*/ +if (!refcount_add_not_zero (VMA_LOCK_OFFSET , &vma->vm_refcnt)) +return false; + +rwsem_acquire (&vma->vmlock_dep_map,0 ,0 , _RET_IP_); +rcuwait_wait_event (&vma->vm_mm->vma_writer_wait, +refcount_read (&vma->vm_refcnt) == tgt_refcnt, +TASK_UNINTERRUPTIBLE ); +lock_acquired (&vma->vmlock_dep_map, _RET_IP_); + +return true; +} + +static inline void __vma_exit_locked (struct vm_area_struct *vma,bool *detached) +{ + *detached =refcount_sub_and_test (VMA_LOCK_OFFSET , &vma->vm_refcnt); +rwsem_release (&vma->vmlock_dep_map, _RET_IP_); +} +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq)
mm/memory.c struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm, · line 6406
Quoted 10 lines
mm/memory.c +8
-1{ -down_write (&vma->vm_lock.lock); +bool locked; + +/* +* __vma_enter_locked() returns false immediately if the vma is not +* attached, otherwise it waits until refcnt is (VMA_LOCK_OFFSET + 1) +* indicating that vma is attached with no readers. +*/ + locked =__vma_enter_locked (vma,VMA_LOCK_OFFSET +1 );
Wonder if it would be slightly better if tgt_refcount was just 1 (or 0 below in vma_mark_detached()) and the VMA_LOCK_OFFSET added to it in __vma_enter_locked() itself as it's the one adding it in the first place.
Suren Baghdasaryan
maintainerRe: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
Quoted 47 lines · Vlastimil Babka
Quoted 33 lines · Suren Baghdasaryan
rw_semaphoreis a sizable structure of 40 bytes and consumes considerable space for eachvm_area_struct. Howevervma_lockhas two important specifics which can be used to replacerw_semaphorewith a simpler structure:
1. Readers never wait. They try to take thevma_lockand fall back tommap_lockif that fails.
2. Only one writer at a time will ever try to write-lock avma_lockbecause writers first takemmap_lockin write mode.
Because of these requirements, fullrw_semaphorefunctionality is not needed and we can replacerw_semaphoreand thevma->detachedflag with a refcount (vm_refcnt).
When vma is in detached state,vm_refcntis 0 and only a call tovma_mark_attached()can take it out of this state. Note that unlike before, now we enforce bothvma_mark_attached()andvma_mark_detached()to be done only after vma has been write-locked.vma_mark_attached()changesvm_refcntto 1 to indicate that it has been attached to the vma tree. When a reader takes read lock, it incrementsvm_refcnt, unless the top usable bit ofvm_refcnt(0x40000000) is set, indicating presence of a writer. When writer takes write lock, it sets the top usable bit to indicate its presence. If there are readers, writer will wait using newly introducedmm->vma_writer_wait. Since all writers takemmap_lockin write mode first, there can be only one writer at a time. The last reader to release the lock will signal the writer to wake up.
refcount might overflow if there are many competing readers, in which case read-locking will fail. Readers are expected to handle such failures. In summary:
1. all readers increment thevm_refcnt;
2. writer sets top usable (writer) bit ofvm_refcnt;
3. readers cannot increment thevm_refcntif the writer bit is set;
4. in the presence of readers, writer must wait for thevm_refcntto drop to 1 (ignoring the writer bit), indicating an attached vma with no readers;
5.vm_refcntoverflow is handled by the readers.Quoted 3 lines
Suggested-by Peter Zijlstra <[email protected]>Suggested-by Matthew Wilcox <[email protected]>Signed-off-by Suren Baghdasaryan <[email protected]>Reviewed-by Vlastimil Babka <[email protected]>But think there's a problem that will manifest after patch 15. Also I don't feel qualified enough about the lockdep parts though (although I think I spotted another issue with those, below) so best if PeterZ can review those.
Some nits below too.
Quoted 10 lines
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 729
Quoted 7 lines
include/linux/mm.h +7
-0+ +static inline void vma_refcount_put (struct vm_area_struct *vma) +{ +int oldcnt; + +if (!__refcount_dec_and_test (&vma->vm_refcnt, &oldcnt)) { +rwsem_release (&vma->vmlock_dep_map, _RET_IP_);Shouldn't we
rwsem_releasealways? And also shouldn't it precede the refcount operation itself?
Yes. Hillf pointed to the same issue. It will be fixed in the next version.
Quoted 8 lines
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 732
Quoted 3 lines
include/linux/mm.h +2
-0+if (is_vma_writer_only (oldcnt -1 )) +rcuwait_wake_up (&vma->vm_mm->vma_writer_wait);Hmm hmm we should maybe read the
vm_mmpointer before dropping the refcount? In case this races in a way thatis_vma_writer_onlytests true but the writer meanwhile finishes and frees the vma. It's safe now but not after making the cacheSLAB_TYPESAFE_BY_RCU?
Hmm. But if is_vma_writer_only() is true that means the writed is blocked and is waiting for the reader to drop the vm_refcnt. IOW, it won't proceed and free the vma until the reader calls rcuwait_wake_up(). Your suggested change is trivial and I can do it but I want to make sure I'm not missing something. Am I?
Quoted 5 lines
Quoted 4 lines
include/linux/mm.h +3
-0+ } +} +
Quoted 7 lines
Quoted 3 lines
static inline void
vma_end_read(struct vm_area_struct *vma){ rcu_read_lock(); /* keeps vma alive till the end of up_read */This should refer to
vma_refcount_put(). But after fixing it I think we could stop doing this altogether? It will no longer keep vma "alive" withSLAB_TYPESAFE_BY_RCU.
Yeah, I think the comment along with rcu_read_lock()/rcu_read_unlock() here can be safely removed.
Quoted 46 lines
Quoted 6 lines
include/linux/mm.h +1
-1-up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);rcu_read_unlock (); }<snip>
Quoted 37 lines
mm/memory.c +28
-1@@ -6370,9 +6370,41 @@ struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm,#endif #ifdef CONFIG_PER_VMA_LOCK +static inline bool __vma_enter_locked (struct vm_area_struct *vma,unsigned int tgt_refcnt) +{ +/* +* If vma is detached then only vma_mark_attached() can raise the +* vm_refcnt. mmap_write_lock prevents racing with vma_mark_attached(). +*/ +if (!refcount_add_not_zero (VMA_LOCK_OFFSET , &vma->vm_refcnt)) +return false; + +rwsem_acquire (&vma->vmlock_dep_map,0 ,0 , _RET_IP_); +rcuwait_wait_event (&vma->vm_mm->vma_writer_wait, +refcount_read (&vma->vm_refcnt) == tgt_refcnt, +TASK_UNINTERRUPTIBLE ); +lock_acquired (&vma->vmlock_dep_map, _RET_IP_); + +return true; +} + +static inline void __vma_exit_locked (struct vm_area_struct *vma,bool *detached) +{ + *detached =refcount_sub_and_test (VMA_LOCK_OFFSET , &vma->vm_refcnt); +rwsem_release (&vma->vmlock_dep_map, _RET_IP_); +} +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq) { -down_write (&vma->vm_lock.lock); +bool locked; + +/* +* __vma_enter_locked() returns false immediately if the vma is not
Quoted 8 lines
mm/memory.c struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm, · line 6406
Quoted 4 lines
mm/memory.c +4
-0+* attached, otherwise it waits until refcnt is (VMA_LOCK_OFFSET + 1) +* indicating that vma is attached with no readers. +*/ + locked =__vma_enter_locked (vma,VMA_LOCK_OFFSET +1 );Wonder if it would be slightly better if
tgt_refcountwas just 1 (or 0 below invma_mark_detached()) and theVMA_LOCK_OFFSETadded to it in__vma_enter_locked()itself as it's the one adding it in the first place.
Well, it won't be called tgt_refcount then. Maybe "bool vma_attached" and inside __vma_enter_locked() we do:
unsigned int tgt_refcnt = VMA_LOCK_OFFSET + vma_attached ? 1 : 0;
Is that better?
Quoted 1 line
Suren Baghdasaryan
maintainerRe: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
Quoted 71 lines · Suren Baghdasaryan
Quoted 47 lines · Vlastimil Babka
Quoted 33 lines · Suren Baghdasaryan
rw_semaphoreis a sizable structure of 40 bytes and consumes considerable space for eachvm_area_struct. Howevervma_lockhas two important specifics which can be used to replacerw_semaphorewith a simpler structure:
1. Readers never wait. They try to take thevma_lockand fall back tommap_lockif that fails.
2. Only one writer at a time will ever try to write-lock avma_lockbecause writers first takemmap_lockin write mode.
Because of these requirements, fullrw_semaphorefunctionality is not needed and we can replacerw_semaphoreand thevma->detachedflag with a refcount (vm_refcnt).
When vma is in detached state,vm_refcntis 0 and only a call tovma_mark_attached()can take it out of this state. Note that unlike before, now we enforce bothvma_mark_attached()andvma_mark_detached()to be done only after vma has been write-locked.vma_mark_attached()changesvm_refcntto 1 to indicate that it has been attached to the vma tree. When a reader takes read lock, it incrementsvm_refcnt, unless the top usable bit ofvm_refcnt(0x40000000) is set, indicating presence of a writer. When writer takes write lock, it sets the top usable bit to indicate its presence. If there are readers, writer will wait using newly introducedmm->vma_writer_wait. Since all writers takemmap_lockin write mode first, there can be only one writer at a time. The last reader to release the lock will signal the writer to wake up.
refcount might overflow if there are many competing readers, in which case read-locking will fail. Readers are expected to handle such failures. In summary:
1. all readers increment thevm_refcnt;
2. writer sets top usable (writer) bit ofvm_refcnt;
3. readers cannot increment thevm_refcntif the writer bit is set;
4. in the presence of readers, writer must wait for thevm_refcntto drop to 1 (ignoring the writer bit), indicating an attached vma with no readers;
5.vm_refcntoverflow is handled by the readers.Quoted 3 lines
Suggested-by Peter Zijlstra <[email protected]>Suggested-by Matthew Wilcox <[email protected]>Signed-off-by Suren Baghdasaryan <[email protected]>Reviewed-by Vlastimil Babka <[email protected]>But think there's a problem that will manifest after patch 15. Also I don't feel qualified enough about the lockdep parts though (although I think I spotted another issue with those, below) so best if PeterZ can review those.
Some nits below too.Quoted 10 lines
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 729
Quoted 7 lines
include/linux/mm.h +7
-0+ +static inline void vma_refcount_put (struct vm_area_struct *vma) +{ +int oldcnt; + +if (!__refcount_dec_and_test (&vma->vm_refcnt, &oldcnt)) { +rwsem_release (&vma->vmlock_dep_map, _RET_IP_);Shouldn't we
rwsem_releasealways? And also shouldn't it precede the refcount operation itself?Yes. Hillf pointed to the same issue. It will be fixed in the next version.
Quoted 8 lines
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 732
Quoted 3 lines
include/linux/mm.h +2
-0+if (is_vma_writer_only (oldcnt -1 )) +rcuwait_wake_up (&vma->vm_mm->vma_writer_wait);Hmm hmm we should maybe read the
vm_mmpointer before dropping the refcount? In case this races in a way thatis_vma_writer_onlytests true but the writer meanwhile finishes and frees the vma. It's safe now but not after making the cacheSLAB_TYPESAFE_BY_RCU?
Quoted 5 lines
Hmm. But if
is_vma_writer_only()is true that means the writed is blocked and is waiting for the reader to drop thevm_refcnt. IOW, it won't proceed and free the vma until the reader callsrcuwait_wake_up(). Your suggested change is trivial and I can do it but I want to make sure I'm not missing something. Am I?
Ok, after thinking some more, I think the race you might be referring to is this:
writer reader
__vma_enter_locked
refcount_add_not_zero(VMA_LOCK_OFFSET, ...)
vma_refcount_put
__refcount_dec_and_test()
if
(is_vma_writer_only())
rcuwait_wait_event(&vma->vm_mm->vma_writer_wait, ...)
__vma_exit_locked
refcount_sub_and_test(VMA_LOCK_OFFSET, ...)
free the vmarcuwait_wake_up(&vma->vm_mm->vma_writer_wait);
I think it's possible and your suggestion of storing the mm before doing __refcount_dec_and_test() should work. Thanks for pointing this out! I'll fix it in the next version.
Quoted 80 lines
Quoted 6 lines
Quoted 5 lines
include/linux/mm.h +3
-0+ } +} +Quoted 7 lines
Quoted 3 lines
static inline void
vma_end_read(struct vm_area_struct *vma){ rcu_read_lock(); /* keeps vma alive till the end of up_read */This should refer to
vma_refcount_put(). But after fixing it I think we could stop doing this altogether? It will no longer keep vma "alive" withSLAB_TYPESAFE_BY_RCU.Yeah, I think the comment along with rcu_read_lock()/rcu_read_unlock() here can be safely removed.
Quoted 46 lines
Quoted 6 lines
include/linux/mm.h +1
-1-up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);rcu_read_unlock (); }<snip>
Quoted 37 lines
mm/memory.c +28
-1@@ -6370,9 +6370,41 @@ struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm,#endif #ifdef CONFIG_PER_VMA_LOCK +static inline bool __vma_enter_locked (struct vm_area_struct *vma,unsigned int tgt_refcnt) +{ +/* +* If vma is detached then only vma_mark_attached() can raise the +* vm_refcnt. mmap_write_lock prevents racing with vma_mark_attached(). +*/ +if (!refcount_add_not_zero (VMA_LOCK_OFFSET , &vma->vm_refcnt)) +return false; + +rwsem_acquire (&vma->vmlock_dep_map,0 ,0 , _RET_IP_); +rcuwait_wait_event (&vma->vm_mm->vma_writer_wait, +refcount_read (&vma->vm_refcnt) == tgt_refcnt, +TASK_UNINTERRUPTIBLE ); +lock_acquired (&vma->vmlock_dep_map, _RET_IP_); + +return true; +} + +static inline void __vma_exit_locked (struct vm_area_struct *vma,bool *detached) +{ + *detached =refcount_sub_and_test (VMA_LOCK_OFFSET , &vma->vm_refcnt); +rwsem_release (&vma->vmlock_dep_map, _RET_IP_); +} +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq) { -down_write (&vma->vm_lock.lock); +bool locked; + +/* +* __vma_enter_locked() returns false immediately if the vma is not Quoted 8 lines
mm/memory.c struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm, · line 6406
Quoted 4 lines
mm/memory.c +4
-0+* attached, otherwise it waits until refcnt is (VMA_LOCK_OFFSET + 1) +* indicating that vma is attached with no readers. +*/ + locked =__vma_enter_locked (vma,VMA_LOCK_OFFSET +1 );Wonder if it would be slightly better if
tgt_refcountwas just 1 (or 0 below invma_mark_detached()) and theVMA_LOCK_OFFSETadded to it in__vma_enter_locked()itself as it's the one adding it in the first place.Well, it won't be called
tgt_refcountthen. Maybe "boolvma_attached" and inside__vma_enter_locked()we do:unsigned int
tgt_refcnt=VMA_LOCK_OFFSET+vma_attached? 1 : 0;Is that better?
Quoted 1 line
Suren Baghdasaryan
maintainerRe: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
Quoted 95 lines · Suren Baghdasaryan
Quoted 71 lines · Suren Baghdasaryan
Quoted 47 lines · Vlastimil Babka
Quoted 33 lines · Suren Baghdasaryan
rw_semaphoreis a sizable structure of 40 bytes and consumes considerable space for eachvm_area_struct. Howevervma_lockhas two important specifics which can be used to replacerw_semaphorewith a simpler structure:
1. Readers never wait. They try to take thevma_lockand fall back tommap_lockif that fails.
2. Only one writer at a time will ever try to write-lock avma_lockbecause writers first takemmap_lockin write mode.
Because of these requirements, fullrw_semaphorefunctionality is not needed and we can replacerw_semaphoreand thevma->detachedflag with a refcount (vm_refcnt).
When vma is in detached state,vm_refcntis 0 and only a call tovma_mark_attached()can take it out of this state. Note that unlike before, now we enforce bothvma_mark_attached()andvma_mark_detached()to be done only after vma has been write-locked.vma_mark_attached()changesvm_refcntto 1 to indicate that it has been attached to the vma tree. When a reader takes read lock, it incrementsvm_refcnt, unless the top usable bit ofvm_refcnt(0x40000000) is set, indicating presence of a writer. When writer takes write lock, it sets the top usable bit to indicate its presence. If there are readers, writer will wait using newly introducedmm->vma_writer_wait. Since all writers takemmap_lockin write mode first, there can be only one writer at a time. The last reader to release the lock will signal the writer to wake up.
refcount might overflow if there are many competing readers, in which case read-locking will fail. Readers are expected to handle such failures. In summary:
1. all readers increment thevm_refcnt;
2. writer sets top usable (writer) bit ofvm_refcnt;
3. readers cannot increment thevm_refcntif the writer bit is set;
4. in the presence of readers, writer must wait for thevm_refcntto drop to 1 (ignoring the writer bit), indicating an attached vma with no readers;
5.vm_refcntoverflow is handled by the readers.Quoted 3 lines
Suggested-by Peter Zijlstra <[email protected]>Suggested-by Matthew Wilcox <[email protected]>Signed-off-by Suren Baghdasaryan <[email protected]>Reviewed-by Vlastimil Babka <[email protected]>But think there's a problem that will manifest after patch 15. Also I don't feel qualified enough about the lockdep parts though (although I think I spotted another issue with those, below) so best if PeterZ can review those.
Some nits below too.Quoted 10 lines
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 729
Quoted 7 lines
include/linux/mm.h +7
-0+ +static inline void vma_refcount_put (struct vm_area_struct *vma) +{ +int oldcnt; + +if (!__refcount_dec_and_test (&vma->vm_refcnt, &oldcnt)) { +rwsem_release (&vma->vmlock_dep_map, _RET_IP_);Shouldn't we
rwsem_releasealways? And also shouldn't it precede the refcount operation itself?Yes. Hillf pointed to the same issue. It will be fixed in the next version.
Quoted 8 lines
include/linux/mm.h static inline void vma_numab_state_free(struct vm_area_struct *vma) {} · line 732
Quoted 3 lines
include/linux/mm.h +2
-0+if (is_vma_writer_only (oldcnt -1 )) +rcuwait_wake_up (&vma->vm_mm->vma_writer_wait);Hmm hmm we should maybe read the
vm_mmpointer before dropping the refcount? In case this races in a way thatis_vma_writer_onlytests true but the writer meanwhile finishes and frees the vma. It's safe now but not after making the cacheSLAB_TYPESAFE_BY_RCU?Quoted 5 lines
Hmm. But if
is_vma_writer_only()is true that means the writed is blocked and is waiting for the reader to drop thevm_refcnt. IOW, it won't proceed and free the vma until the reader callsrcuwait_wake_up(). Your suggested change is trivial and I can do it but I want to make sure I'm not missing something. Am I?Ok, after thinking some more, I think the race you might be referring to is this:
writer reader
__vma_enter_locked refcount_add_not_zero(VMA_LOCK_OFFSET, ...) vma_refcount_put __refcount_dec_and_test() if (is_vma_writer_only()) rcuwait_wait_event(&vma->vm_mm->vma_writer_wait, ...) __vma_exit_locked refcount_sub_and_test(VMA_LOCK_OFFSET, ...) free the vma
Quoted 1 line
rcuwait_wake_up(&vma->vm_mm->vma_writer_wait);
Sorry, this should be more readable:
writer reader
__vma_enter_locked
refcount_add_not_zero(VMA_LOCK_OFFSET, ...)
vma_refcount_put
__refcount_dec_and_test()
if (is_vma_writer_only())
rcuwait_wait_event()
__vma_exit_locked
refcount_sub_and_test(VMA_LOCK_OFFSET, ...)free the vma
rcuwait_wake_up(); <-- access to vma->vm_mm
Quoted 85 lines
I think it's possible and your suggestion of storing the mm before doing
__refcount_dec_and_test()should work. Thanks for pointing this out! I'll fix it in the next version.Quoted 80 lines
Quoted 6 lines
Quoted 5 lines
include/linux/mm.h +3
-0+ } +} +Quoted 7 lines
Quoted 3 lines
static inline void
vma_end_read(struct vm_area_struct *vma){ rcu_read_lock(); /* keeps vma alive till the end of up_read */This should refer to
vma_refcount_put(). But after fixing it I think we could stop doing this altogether? It will no longer keep vma "alive" withSLAB_TYPESAFE_BY_RCU.Yeah, I think the comment along with rcu_read_lock()/rcu_read_unlock() here can be safely removed.
Quoted 46 lines
Quoted 6 lines
include/linux/mm.h +1
-1-up_read (&vma->vm_lock.lock); +vma_refcount_put (vma);rcu_read_unlock (); }<snip>
Quoted 37 lines
mm/memory.c +28
-1@@ -6370,9 +6370,41 @@ struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm,#endif #ifdef CONFIG_PER_VMA_LOCK +static inline bool __vma_enter_locked (struct vm_area_struct *vma,unsigned int tgt_refcnt) +{ +/* +* If vma is detached then only vma_mark_attached() can raise the +* vm_refcnt. mmap_write_lock prevents racing with vma_mark_attached(). +*/ +if (!refcount_add_not_zero (VMA_LOCK_OFFSET , &vma->vm_refcnt)) +return false; + +rwsem_acquire (&vma->vmlock_dep_map,0 ,0 , _RET_IP_); +rcuwait_wait_event (&vma->vm_mm->vma_writer_wait, +refcount_read (&vma->vm_refcnt) == tgt_refcnt, +TASK_UNINTERRUPTIBLE ); +lock_acquired (&vma->vmlock_dep_map, _RET_IP_); + +return true; +} + +static inline void __vma_exit_locked (struct vm_area_struct *vma,bool *detached) +{ + *detached =refcount_sub_and_test (VMA_LOCK_OFFSET , &vma->vm_refcnt); +rwsem_release (&vma->vmlock_dep_map, _RET_IP_); +} +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq) { -down_write (&vma->vm_lock.lock); +bool locked; + +/* +* __vma_enter_locked() returns false immediately if the vma is not Quoted 8 lines
mm/memory.c struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm, · line 6406
Quoted 4 lines
mm/memory.c +4
-0+* attached, otherwise it waits until refcnt is (VMA_LOCK_OFFSET + 1) +* indicating that vma is attached with no readers. +*/ + locked =__vma_enter_locked (vma,VMA_LOCK_OFFSET +1 );Wonder if it would be slightly better if
tgt_refcountwas just 1 (or 0 below invma_mark_detached()) and theVMA_LOCK_OFFSETadded to it in__vma_enter_locked()itself as it's the one adding it in the first place.Well, it won't be called
tgt_refcountthen. Maybe "boolvma_attached" and inside__vma_enter_locked()we do:unsigned int
tgt_refcnt=VMA_LOCK_OFFSET+vma_attached? 1 : 0;Is that better?
Quoted 1 line
Re: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
In reply to Suren BaghdasaryanQuoted 46 lines · Suren Baghdasaryan
Quoted 37 lines
Quoted 37 lines
mm/memory.c +28
-1@@ -6370,9 +6370,41 @@ struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm,#endif #ifdef CONFIG_PER_VMA_LOCK +static inline bool __vma_enter_locked (struct vm_area_struct *vma,unsigned int tgt_refcnt) +{ +/* +* If vma is detached then only vma_mark_attached() can raise the +* vm_refcnt. mmap_write_lock prevents racing with vma_mark_attached(). +*/ +if (!refcount_add_not_zero (VMA_LOCK_OFFSET , &vma->vm_refcnt)) +return false; + +rwsem_acquire (&vma->vmlock_dep_map,0 ,0 , _RET_IP_); +rcuwait_wait_event (&vma->vm_mm->vma_writer_wait, +refcount_read (&vma->vm_refcnt) == tgt_refcnt, +TASK_UNINTERRUPTIBLE ); +lock_acquired (&vma->vmlock_dep_map, _RET_IP_); + +return true; +} + +static inline void __vma_exit_locked (struct vm_area_struct *vma,bool *detached) +{ + *detached =refcount_sub_and_test (VMA_LOCK_OFFSET , &vma->vm_refcnt); +rwsem_release (&vma->vmlock_dep_map, _RET_IP_); +} +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq) { -down_write (&vma->vm_lock.lock); +bool locked; + +/* +* __vma_enter_locked() returns false immediately if the vma is not Quoted 8 lines
mm/memory.c struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm, · line 6406
Quoted 4 lines
mm/memory.c +4
-0+* attached, otherwise it waits until refcnt is (VMA_LOCK_OFFSET + 1) +* indicating that vma is attached with no readers. +*/ + locked =__vma_enter_locked (vma,VMA_LOCK_OFFSET +1 );Wonder if it would be slightly better if
tgt_refcountwas just 1 (or 0 below invma_mark_detached()) and theVMA_LOCK_OFFSETadded to it in__vma_enter_locked()itself as it's the one adding it in the first place.
Quoted 6 lines
Well, it won't be called
tgt_refcountthen. Maybe "boolvma_attached" and inside__vma_enter_locked()we do:unsigned int
tgt_refcnt=VMA_LOCK_OFFSET+vma_attached? 1 : 0;Is that better?
Yeah I think so as it centralizes the target refcount logic into a single place __vma_enter_locked().
Hm but then it's weird that __vma_start_write() would set vma_attached to true and yet it handles also a case where it's not attached. Maybe call the parameter "detaching" and switch the 0 and 1?
Suren Baghdasaryan
maintainerRe: [PATCH v8 11/16] mm: replace vm_lock and detached flag with a reference count
Quoted 51 lines · Vlastimil Babka
Quoted 49 lines · Suren Baghdasaryan
Quoted 37 lines
Quoted 37 lines
mm/memory.c +28
-1@@ -6370,9 +6370,41 @@ struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm,#endif #ifdef CONFIG_PER_VMA_LOCK +static inline bool __vma_enter_locked (struct vm_area_struct *vma,unsigned int tgt_refcnt) +{ +/* +* If vma is detached then only vma_mark_attached() can raise the +* vm_refcnt. mmap_write_lock prevents racing with vma_mark_attached(). +*/ +if (!refcount_add_not_zero (VMA_LOCK_OFFSET , &vma->vm_refcnt)) +return false; + +rwsem_acquire (&vma->vmlock_dep_map,0 ,0 , _RET_IP_); +rcuwait_wait_event (&vma->vm_mm->vma_writer_wait, +refcount_read (&vma->vm_refcnt) == tgt_refcnt, +TASK_UNINTERRUPTIBLE ); +lock_acquired (&vma->vmlock_dep_map, _RET_IP_); + +return true; +} + +static inline void __vma_exit_locked (struct vm_area_struct *vma,bool *detached) +{ + *detached =refcount_sub_and_test (VMA_LOCK_OFFSET , &vma->vm_refcnt); +rwsem_release (&vma->vmlock_dep_map, _RET_IP_); +} +void __vma_start_write (struct vm_area_struct *vma,unsigned int mm_lock_seq) { -down_write (&vma->vm_lock.lock); +bool locked; + +/* +* __vma_enter_locked() returns false immediately if the vma is not Quoted 8 lines
mm/memory.c struct vm_area_struct *lock_mm_and_find_vma(struct mm_struct *mm, · line 6406
Quoted 4 lines
mm/memory.c +4
-0+* attached, otherwise it waits until refcnt is (VMA_LOCK_OFFSET + 1) +* indicating that vma is attached with no readers. +*/ + locked =__vma_enter_locked (vma,VMA_LOCK_OFFSET +1 );Wonder if it would be slightly better if
tgt_refcountwas just 1 (or 0 below invma_mark_detached()) and theVMA_LOCK_OFFSETadded to it in__vma_enter_locked()itself as it's the one adding it in the first place.Well, it won't be called
tgt_refcountthen. Maybe "boolvma_attached" and inside__vma_enter_locked()we do:
Quoted 8 lines
Quoted 3 lines
unsigned int
tgt_refcnt=VMA_LOCK_OFFSET+vma_attached? 1 : 0;Is that better?
Yeah I think so as it centralizes the target refcount logic into a single place
__vma_enter_locked().
Hm but then it's weird that__vma_start_write()would setvma_attachedto true and yet it handles also a case where it's not attached.
Ah, good point.
Quoted 1 line
Maybe call the parameter "detaching" and switch the 0 and 1?
Yes, that would be less confusing. Thanks for the suggestion, I'll use it.
Suren Baghdasaryan
maintainerPATCH v8 12/16 mm/debug: print vm_refcnt state when dumping the vma
In reply to Suren Baghdasaryanvm_refcnt encodes a number of useful states:
- whether vma is attached or detached
- the number of current vma readers
- presence of a vma writer
Let's include it in the vma dump.
--- mm/debug.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
mm/debug.c +12 -0
@@ -178,6 +178,17 @@ EXPORT_SYMBOL(dump_page);void dump_vma (const struct vm_area_struct *vma) { +#ifdef CONFIG_PER_VMA_LOCK +pr_emerg ("vma %px start %px end %px mm %px\n" +"prot %lx anon_vma %px vm_ops %px\n" +"pgoff %lx file %px private_data %px\n" +"flags: %#lx(%pGv) refcnt %x\n" , + vma, (void *)vma->vm_start, (void *)vma->vm_end, vma->vm_mm, + (unsigned long )pgprot_val (vma->vm_page_prot), + vma->anon_vma, vma->vm_ops, vma->vm_pgoff, + vma->vm_file, vma->vm_private_data, + vma->vm_flags, &vma->vm_flags,refcount_read (&vma->vm_refcnt)); +#else pr_emerg ("vma %px start %px end %px mm %px\n" "prot %lx anon_vma %px vm_ops %px\n" "pgoff %lx file %px private_data %px\n"
@@ -187,6 +198,7 @@ void dump_vma(const struct vm_area_struct *vma) vma->anon_vma, vma->vm_ops, vma->vm_pgoff, vma->vm_file, vma->vm_private_data, vma->vm_flags, &vma->vm_flags); +#endif }EXPORT_SYMBOL (dump_vma);
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 13/16 mm: remove extra vma_numab_state_init() call
In reply to Suren Baghdasaryanvma_init() already memset's the whole vm_area_struct to 0, so there is no need to an additional vma_numab_state_init().
--- include/linux/mm.h | 1 - 1 file changed, 1 deletion(-)
include/linux/mm.h +0 -1
@@ -948,7 +948,6 @@ static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm) vma->vm_mm = mm; vma->vm_ops = &vma_dummy_vm_ops;INIT_LIST_HEAD (&vma->anon_vma_chain); -vma_numab_state_init (vma);vma_lock_init (vma, false); }
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 14/16 mm: prepare lock_vma_under_rcu() for vma reuse possibility
In reply to Suren BaghdasaryanOnce we make vma cache SLAB_TYPESAFE_BY_RCU, it will be possible for a vma to be reused and attached to another mm after lock_vma_under_rcu() locks the vma. lock_vma_under_rcu() should ensure that vma_start_read() is using the original mm and after locking the vma it should ensure that vma->vm_mm has not changed from under us.
--- include/linux/mm.h | 10 ++++++---- mm/memory.c | 7 ++++--- 2 files changed, 10 insertions(+), 7 deletions(-)
include/linux/mm.h +6 -4
@@ -737,8 +737,10 @@ static inline void vma_refcount_put(struct vm_area_struct *vma)* Try to read-lock a vma. The function is allowed to occasionally yield false * locked result to avoid performance overhead, in which case we fall back to * using mmap_lock. The function should never yield false unlocked result. +* False locked result is possible if mm_lock_seq overflows or if vma gets +* reused and attached to a different mm before we lock it. */ -static inline bool vma_start_read (struct vm_area_struct *vma) +static inline bool vma_start_read (struct mm_struct *mm,struct vm_area_struct *vma) {int oldcnt;
@@ -749,7 +751,7 @@ static inline bool vma_start_read(struct vm_area_struct *vma)* we don't rely on for anything - the mm_lock_seq read against which we * need ordering is below. */ -if (READ_ONCE (vma->vm_lock_seq) ==READ_ONCE (vma->vm_mm->mm_lock_seq.sequence)) +if (READ_ONCE (vma->vm_lock_seq) ==READ_ONCE (mm->mm_lock_seq.sequence))return false;/*
@@ -772,7 +774,7 @@ static inline bool vma_start_read(struct vm_area_struct *vma)* after it has been unlocked. * This pairs with RELEASE semantics in vma_end_write_all(). */ -if (unlikely (vma->vm_lock_seq ==raw_read_seqcount (&vma->vm_mm->mm_lock_seq))) { +if (unlikely (vma->vm_lock_seq ==raw_read_seqcount (&mm->mm_lock_seq))) {vma_refcount_put (vma);return false; }
@@ -906,7 +908,7 @@ struct vm_area_struct *lock_vma_under_rcu(struct mm_struct *mm,#else /* CONFIG_PER_VMA_LOCK */ static inline void vma_lock_init (struct vm_area_struct *vma,bool reset_refcnt) {} -static inline bool vma_start_read (struct vm_area_struct *vma) +static inline bool vma_start_read (struct mm_struct *mm,struct vm_area_struct *vma) {return false; }static inline void vma_end_read (struct vm_area_struct *vma) {}static inline void vma_start_write (struct vm_area_struct *vma) {}
mm/memory.c +4 -3
@@ -6465,7 +6465,7 @@ struct vm_area_struct *lock_vma_under_rcu(struct mm_struct *mm,if (!vma) goto inval; -if (!vma_start_read (vma)) +if (!vma_start_read (mm, vma)) goto inval;/*
@@ -6475,8 +6475,9 @@ struct vm_area_struct *lock_vma_under_rcu(struct mm_struct *mm,* fields are accessible for RCU readers. */ -/* Check since vm_start/vm_end might change before we lock the VMA */ -if (unlikely (address < vma->vm_start || address >= vma->vm_end)) +/* Check if the vma we locked is the right one. */ +if (unlikely (vma->vm_mm != mm || + address < vma->vm_start || address >= vma->vm_end)) goto inval_end_read;rcu_read_unlock ();
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerPATCH v8 15/16 mm: make vma cache SLAB_TYPESAFE_BY_RCU
In reply to Suren BaghdasaryanTo enable SLAB_TYPESAFE_BY_RCU for vma cache we need to ensure that object reuse before RCU grace period is over will be detected by lock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is in exit_mmap(). Add the missing vma_mark_detached() in exit_mmap().
Another issue which might trick lock_vma_under_rcu() during vma reuse is vm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma's vm_refcnt and temporarily making it appear as attached. This might trick a racing lock_vma_under_rcu() to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure that vm_refcnt stays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introduce vma_copy() which preserves new vma's vm_refcnt and use it in vm_area_dup(). Since all vmas are in detached state with no current readers when they are freed, lock_vma_under_rcu() will not be able to take vm_refcnt after vma got detached even if vma is reused.
Finally, make vm_area_cachep SLAB_TYPESAFE_BY_RCU. This will facilitate vm_area_struct reuse and will minimize the number of call_rcu() calls.
--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)
include/linux/mm.h +0 -2
@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;
include/linux/mm_types.h +7 -3
@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that
@@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/*
include/linux/slab.h +0 -6
@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. *
kernel/fork.c +45 -27
108 lines not on this page · raw message
mm/mmap.c +2 -1
@@ -1305,7 +1305,8 @@ void exit_mmap(struct mm_struct *mm) do { if (vma->vm_flags & VM_ACCOUNT) nr_accounted += vma_pages(vma); - remove_vma(vma, /* unreachable = */ true); + vma_mark_detached(vma); + remove_vma(vma); count++; cond_resched(); vma = vma_next(&vmi);
mm/vma.c +3 -8
@@ -406,19 +406,14 @@ static bool can_vma_merge_right(struct vma_merge_struct *vmg,/* * Close a vm structure and free it. */ -void remove_vma (struct vm_area_struct *vma,bool unreachable) +void remove_vma (struct vm_area_struct *vma) {might_sleep ();vma_close (vma);if (vma->vm_file)fput (vma->vm_file);mpol_put (vma_policy (vma)); -if (unreachable) { -vma_mark_detached (vma); -__vm_area_free (vma); - }else { -vm_area_free (vma); - } +vm_area_free (vma); }/*
@@ -1201,7 +1196,7 @@ static void vms_complete_munmap_vmas(struct vma_munmap_struct *vms,/* Remove and clean up vmas */ mas_set (mas_detach,0 );mas_for_each (mas_detach, vma,ULONG_MAX ) -remove_vma (vma,/* unreachable = */ false); +remove_vma (vma);vm_unacct_memory (vms->nr_accounted);validate_mm (mm);
mm/vma.h +1 -1
@@ -170,7 +170,7 @@ int do_vmi_munmap(struct vma_iterator *vmi, struct mm_struct *mm, unsigned long start, size_t len, struct list_head *uf, bool unlock); -void remove_vma (struct vm_area_struct *vma,bool unreachable); +void remove_vma (struct vm_area_struct *vma);void unmap_region (struct ma_state *mas,struct vm_area_struct *vma,struct vm_area_struct *prev,struct vm_area_struct *next);
tools/testing/vma/vma_internal.h +1 -6
@@ -697,14 +697,9 @@ static inline void mpol_put(struct mempolicy *) { } -static inline void __vm_area_free (struct vm_area_struct *vma) -{ -free (vma); -} -static inline void vm_area_free (struct vm_area_struct *vma) { -__vm_area_free (vma); +free (vma); }static inline void lru_add_drain (void )
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Re: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 21 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.
Quoted 1 line
Signed-off-by Suren Baghdasaryan <[email protected]>
You could also drop the reset_refcnt parameter of vma_lock_init() now, as the usage in vm_area_dup() should now be just setting 0 over 0. Maybe a VM_WARN_ON if it's not 0 already?
And a comment in vm_area_struct definition to consider vma_copy() when adding any new field?
kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 469
Quoted 5 lines
kernel/fork.c +5
-0+/* +* src->shared.rb may be modified concurrently, but the clone +* will be reinitialized. +*/ +data_race (memcpy (&dest->shared, &src->shared,sizeof (dest->shared)));
The comment makes it sound as if we didn't need to do it at all? But I didn't verify. If we do need it in some cases (i.e. the just allocated vma might have garbage from previous lifetime, but src is well defined and it's a case where it's not reinitialized afterwards) maybe the comment should say? Or if it's either reinitialized later or zeroes at src, we could memset() the zeroes instead of memcpying them, etc.
Suren Baghdasaryan
maintainerRe: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 23 lines · Vlastimil Babka
Quoted 21 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.
Quoted 7 lines
Quoted 1 line
Signed-off-by Suren Baghdasaryan <[email protected]>Reviewed-by Vlastimil Babka <[email protected]>You could also drop the
reset_refcntparameter ofvma_lock_init()now, as the usage invm_area_dup()should now be just setting 0 over 0. Maybe aVM_WARN_ONif it's not 0 already?
Yeah, that's a good idea. Will do.
Quoted 2 lines
And a comment in
vm_area_structdefinition to considervma_copy()when adding any new field?
Sure, will add.
Quoted 7 lines
kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 469
Quoted 6 lines
kernel/fork.c +5
-0+/* +* src->shared.rb may be modified concurrently, but the clone +* will be reinitialized. +*/ +data_race (memcpy (&dest->shared, &src->shared,sizeof (dest->shared)));
Quoted 6 lines
The comment makes it sound as if we didn't need to do it at all? But I didn't verify. If we do need it in some cases (i.e. the just allocated vma might have garbage from previous lifetime, but src is well defined and it's a case where it's not reinitialized afterwards) maybe the comment should say? Or if it's either reinitialized later or zeroes at src, we could
memset()the zeroes instead of memcpying them, etc.
I see vm_area_dup() being used in dup_mmap() and I think this comment is about this usage in case the src vma changes from under us. However, vm_area_dup() is also used when we simply duplicate an existing vma while holding an mmap_write_lock, like in __split_vma(). In these cases there is no possibility of a race and copied value should hold. Maybe I should amend this comment like this:
/*
* src->shared.rb may be modified concurrently when called from dup_mmap(),
* but the clone will reinitialize it.
*/
WDYT?
Quoted 2 lines
Re: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 9 lines · Suren Baghdasaryan
On Fri, Jan 10, 2025 at 7:31 AM Vlastimil Babka <[email protected]> wrote:
I see
vm_area_dup()being used indup_mmap()and I think this comment is about this usage in case the src vma changes from under us. However,vm_area_dup()is also used when we simply duplicate an existing vma while holding anmmap_write_lock, like in__split_vma(). In these cases there is no possibility of a race and copied value should hold. Maybe I should amend this comment like this:
Quoted 6 lines
/*
*src->shared.rbmay be modified concurrently when called fromdup_mmap(),
* but the clone will reinitialize it.
*/WDYT?
Sounds good, thanks!
Quoted 2 lines
Quoted 2 lines
Suren Baghdasaryan
maintainerRe: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
In reply to Suren BaghdasaryanQuoted 27 lines · Suren Baghdasaryan
Quoted 25 lines · Vlastimil Babka
Quoted 22 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>
Quoted 7 lines
Quoted 5 lines
Reviewed-by Vlastimil Babka <[email protected]>You could also drop the
reset_refcntparameter ofvma_lock_init()now, as the usage invm_area_dup()should now be just setting 0 over 0. Maybe aVM_WARN_ONif it's not 0 already?Yeah, that's a good idea. Will do.
Ugh, once I made this change, the newly added VM_WARN_ON() immediately triggered because vm_area_dup() does not memset(0) the entire vma and kmem_cache_alloc(vm_area_cachep) does not always return a reused vma. I could add a vm_area_cachep constructor to always initialize vm_refcnt to 0 but that would lead to more changes. I think I'll keep reset_refcnt for now and will add vm_area_cachep constructor as a follow-up optimization after this patchset is merged.
Quoted 36 lines
Quoted 3 lines
And a comment in
vm_area_structdefinition to considervma_copy()when adding any new field?Sure, will add.
Quoted 7 lines
kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 469
Quoted 6 lines
kernel/fork.c +5
-0+/* +* src->shared.rb may be modified concurrently, but the clone +* will be reinitialized. +*/ +data_race (memcpy (&dest->shared, &src->shared,sizeof (dest->shared)));Quoted 6 lines
The comment makes it sound as if we didn't need to do it at all? But I didn't verify. If we do need it in some cases (i.e. the just allocated vma might have garbage from previous lifetime, but src is well defined and it's a case where it's not reinitialized afterwards) maybe the comment should say? Or if it's either reinitialized later or zeroes at src, we could
memset()the zeroes instead of memcpying them, etc.I see
vm_area_dup()being used indup_mmap()and I think this comment is about this usage in case the src vma changes from under us. However,vm_area_dup()is also used when we simply duplicate an existing vma while holding anmmap_write_lock, like in__split_vma(). In these cases there is no possibility of a race and copied value should hold. Maybe I should amend this comment like this:/*
*src->shared.rbmay be modified concurrently when called fromdup_mmap(),
* but the clone will reinitialize it.
*/WDYT?
Quoted 2 lines
Re: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
In reply to Suren BaghdasaryanQuoted 96 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)include/linux/mm.h +0
-2@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;include/linux/mm_types.h +7
-3@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that @@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/* include/linux/slab.h +0
-6@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. * kernel/fork.c +0
-0
kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 451
Quoted 3 lines
kernel/fork.c +0
-0@@ -449,6 +449,41 @@ struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return vma; }
There exists a copy_vma() which copies the vma to a new area in the mm in rmap. Naming this vma_copy() is confusing :)
It might be better to just put this code in the vm_area_dup() or call it __vm_area_dup(), or __vma_dup() ?
Quoted 39 lines
kernel/fork.c +35
-0+static void vma_copy (const struct vm_area_struct *src,struct vm_area_struct *dest) +{ + dest->vm_mm = src->vm_mm; + dest->vm_ops = src->vm_ops; + dest->vm_start = src->vm_start; + dest->vm_end = src->vm_end; + dest->anon_vma = src->anon_vma; + dest->vm_pgoff = src->vm_pgoff; + dest->vm_file = src->vm_file; + dest->vm_private_data = src->vm_private_data; +vm_flags_init (dest, src->vm_flags); +memcpy (&dest->vm_page_prot, &src->vm_page_prot, +sizeof (dest->vm_page_prot)); +/* +* src->shared.rb may be modified concurrently, but the clone +* will be reinitialized. +*/ +data_race (memcpy (&dest->shared, &src->shared,sizeof (dest->shared))); +memcpy (&dest->vm_userfaultfd_ctx, &src->vm_userfaultfd_ctx, +sizeof (dest->vm_userfaultfd_ctx)); +#ifdef CONFIG_ANON_VMA_NAME + dest->anon_name = src->anon_name; +#endif +#ifdef CONFIG_SWAP +memcpy (&dest->swap_readahead_info, &src->swap_readahead_info, +sizeof (dest->swap_readahead_info)); +#endif +#ifndef CONFIG_MMU + dest->vm_region = src->vm_region; +#endif +#ifdef CONFIG_NUMA + dest->vm_policy = src->vm_policy; +#endif +} +struct vm_area_struct *vm_area_dup (struct vm_area_struct *orig) {struct vm_area_struct *new =kmem_cache_alloc (vm_area_cachep,GFP_KERNEL );
kernel/fork.c struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig) · line 497
Quoted 10 lines
kernel/fork.c +1
-5@@ -458,11 +493,7 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)ASSERT_EXCLUSIVE_WRITER (orig->vm_flags);ASSERT_EXCLUSIVE_WRITER (orig->vm_file); -/* -* orig->shared.rb may be modified concurrently, but the clone -* will be reinitialized. -*/ -data_race (memcpy (new, orig,sizeof (*new))); +vma_copy (orig, new);vma_lock_init (new, true);
I think this suffers from a race still?
That is, we can still race between vm_lock_seq == mm_lock_seq and the lock acquire, where a free and reuse happens. In the even that the reader is caught between the sequence and lock taking, the vma->vmlock_dep_map may not be replaced and it could see the old lock (or zero?) and things go bad:
It could try to take vmlock_dep_map == 0 in read mode.
It can take the old lock, detect the refcnt is wrong and release the new lock.
Thanks,
Liam
Suren Baghdasaryan
maintainerRe: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 105 lines · Liam R. Howlett
Quoted 96 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)include/linux/mm.h +0
-2@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;include/linux/mm_types.h +7
-3@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that @@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/* include/linux/slab.h +0
-6@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. * kernel/fork.c +0
-0kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 451
Quoted 3 lines
kernel/fork.c +0
-0@@ -449,6 +449,41 @@ struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return vma; }There exists a
copy_vma()which copies the vma to a new area in the mm in rmap. Naming thisvma_copy()is confusing :)
Quoted 2 lines
It might be better to just put this code in the
vm_area_dup()or call it__vm_area_dup(), or__vma_dup()?
Hmm. It's not really duplicating a vma but copying its content (no allocation). How about __vm_area_copy() to indicate it is copying vm_area_struct content?
Quoted 59 lines
Quoted 40 lines
kernel/fork.c +35
-0+static void vma_copy (const struct vm_area_struct *src,struct vm_area_struct *dest) +{ + dest->vm_mm = src->vm_mm; + dest->vm_ops = src->vm_ops; + dest->vm_start = src->vm_start; + dest->vm_end = src->vm_end; + dest->anon_vma = src->anon_vma; + dest->vm_pgoff = src->vm_pgoff; + dest->vm_file = src->vm_file; + dest->vm_private_data = src->vm_private_data; +vm_flags_init (dest, src->vm_flags); +memcpy (&dest->vm_page_prot, &src->vm_page_prot, +sizeof (dest->vm_page_prot)); +/* +* src->shared.rb may be modified concurrently, but the clone +* will be reinitialized. +*/ +data_race (memcpy (&dest->shared, &src->shared,sizeof (dest->shared))); +memcpy (&dest->vm_userfaultfd_ctx, &src->vm_userfaultfd_ctx, +sizeof (dest->vm_userfaultfd_ctx)); +#ifdef CONFIG_ANON_VMA_NAME + dest->anon_name = src->anon_name; +#endif +#ifdef CONFIG_SWAP +memcpy (&dest->swap_readahead_info, &src->swap_readahead_info, +sizeof (dest->swap_readahead_info)); +#endif +#ifndef CONFIG_MMU + dest->vm_region = src->vm_region; +#endif +#ifdef CONFIG_NUMA + dest->vm_policy = src->vm_policy; +#endif +} +struct vm_area_struct *vm_area_dup (struct vm_area_struct *orig) {struct vm_area_struct *new =kmem_cache_alloc (vm_area_cachep,GFP_KERNEL );kernel/fork.c struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig) · line 497
Quoted 10 lines
kernel/fork.c +1
-5@@ -458,11 +493,7 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)ASSERT_EXCLUSIVE_WRITER (orig->vm_flags);ASSERT_EXCLUSIVE_WRITER (orig->vm_file); -/* -* orig->shared.rb may be modified concurrently, but the clone -* will be reinitialized. -*/ -data_race (memcpy (new, orig,sizeof (*new))); +vma_copy (orig, new);vma_lock_init (new, true);I think this suffers from a race still?
That is, we can still race between
vm_lock_seq==mm_lock_seqand the lock acquire, where a free and reuse happens. In the even that the reader is caught between the sequence and lock taking, thevma->vmlock_dep_mapmay not be replaced and it could see the old lock (or zero?) and things go bad:
Quoted 4 lines
It could try to take
vmlock_dep_map== 0 in read mode.It can take the old lock, detect the refcnt is wrong and release the new lock.
I don't think this race can happen. Notice a call to vma_assert_detached() inside vm_area_free(), so before vma is freed and possibly reused, it has to be detached. vma_mark_detached() ensures that there are no current or future readers by executing the __vma_enter_locked() + __vma_exit_locked() sequence if vm_refcnt is not already at 0. Once __vma_exit_locked() is done, vm_refcnt is at 0 and any new reader will be rejected on __refcount_inc_not_zero_limited(), before even checking vm_lock_seq == mm_lock_seq. Even if a reader tries to sneak in between __vma_enter_locked() and __vma_exit_locked() calls, __refcount_inc_not_zero_limited() will reject it because VMA_LOCK_OFFSET is set and VMA_REF_LIMIT will be violated. IOW, when VMA is freed, it's guaranteed to be detached with no current or future readers, therefore "race between vm_lock_seq == mm_lock_seq and the lock acquire, where a free and reuse happens" should not be possible.
Did I understand your concern correctly and does my explanation make sense to you?
Quoted 3 lines
Thanks,
Liam
Re: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 177 lines · Suren Baghdasaryan
Quoted 105 lines · Liam R. Howlett
Quoted 96 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)include/linux/mm.h +0
-2@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;include/linux/mm_types.h +7
-3@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that @@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/* include/linux/slab.h +0
-6@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. * kernel/fork.c +0
-0kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 451
Quoted 3 lines
kernel/fork.c +0
-0@@ -449,6 +449,41 @@ struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return vma; }There exists a
copy_vma()which copies the vma to a new area in the mm in rmap. Naming thisvma_copy()is confusing :)Quoted 2 lines
It might be better to just put this code in the
vm_area_dup()or call it__vm_area_dup(), or__vma_dup()?Hmm. It's not really duplicating a vma but copying its content (no allocation). How about
__vm_area_copy()to indicate it is copyingvm_area_structcontent?Quoted 59 lines
Quoted 40 lines
kernel/fork.c +35
-0+static void vma_copy (const struct vm_area_struct *src,struct vm_area_struct *dest) +{ + dest->vm_mm = src->vm_mm; + dest->vm_ops = src->vm_ops; + dest->vm_start = src->vm_start; + dest->vm_end = src->vm_end; + dest->anon_vma = src->anon_vma; + dest->vm_pgoff = src->vm_pgoff; + dest->vm_file = src->vm_file; + dest->vm_private_data = src->vm_private_data; +vm_flags_init (dest, src->vm_flags); +memcpy (&dest->vm_page_prot, &src->vm_page_prot, +sizeof (dest->vm_page_prot)); +/* +* src->shared.rb may be modified concurrently, but the clone +* will be reinitialized. +*/ +data_race (memcpy (&dest->shared, &src->shared,sizeof (dest->shared))); +memcpy (&dest->vm_userfaultfd_ctx, &src->vm_userfaultfd_ctx, +sizeof (dest->vm_userfaultfd_ctx)); +#ifdef CONFIG_ANON_VMA_NAME + dest->anon_name = src->anon_name; +#endif +#ifdef CONFIG_SWAP +memcpy (&dest->swap_readahead_info, &src->swap_readahead_info, +sizeof (dest->swap_readahead_info)); +#endif +#ifndef CONFIG_MMU + dest->vm_region = src->vm_region; +#endif +#ifdef CONFIG_NUMA + dest->vm_policy = src->vm_policy; +#endif +} +struct vm_area_struct *vm_area_dup (struct vm_area_struct *orig) {struct vm_area_struct *new =kmem_cache_alloc (vm_area_cachep,GFP_KERNEL );kernel/fork.c struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig) · line 497
Quoted 10 lines
kernel/fork.c +1
-5@@ -458,11 +493,7 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)ASSERT_EXCLUSIVE_WRITER (orig->vm_flags);ASSERT_EXCLUSIVE_WRITER (orig->vm_file); -/* -* orig->shared.rb may be modified concurrently, but the clone -* will be reinitialized. -*/ -data_race (memcpy (new, orig,sizeof (*new))); +vma_copy (orig, new);vma_lock_init (new, true);I think this suffers from a race still?
That is, we can still race between
vm_lock_seq==mm_lock_seqand the lock acquire, where a free and reuse happens. In the even that the reader is caught between the sequence and lock taking, thevma->vmlock_dep_mapmay not be replaced and it could see the old lock (or zero?) and things go bad:Quoted 4 lines
It could try to take
vmlock_dep_map== 0 in read mode.It can take the old lock, detect the refcnt is wrong and release the new lock.
Quoted 9 lines
I don't think this race can happen. Notice a call to
vma_assert_detached()insidevm_area_free(), so before vma is freed and possibly reused, it has to be detached.vma_mark_detached()ensures that there are no current or future readers by executing the__vma_enter_locked()+__vma_exit_locked()sequence ifvm_refcntis not already at 0. Once__vma_exit_locked()is done,vm_refcntis at 0 and any new reader will be rejected on__refcount_inc_not_zero_limited(), before even checkingvm_lock_seq==mm_lock_seq.
Isn't the vm_lock_seq check before the ref count in vma_start_read()?
From patch 11/16:
@@ -720,13 +752,19 @@ static inline bool vma_start_read(struct vm_area_struct *vma)
if (READ_ONCE(vma->vm_lock_seq) == READ_ONCE(vma->vm_mm->mm_lock_seq.sequence))
return false;
- if (unlikely(down_read_trylock(&vma->vm_lock.lock) == 0))
+ /*
+ * If VMA_LOCK_OFFSET is set, __refcount_inc_not_zero_limited() will fail
+ * because VMA_REF_LIMIT is less than VMA_LOCK_OFFSET.
+ */
+ if (unlikely(!__refcount_inc_not_zero_limited(&vma->vm_refcnt, &oldcnt,
+ VMA_REF_LIMIT)))
return false;
+ rwsem_acquire_read(&vma->vmlock_dep_map, 0, 1, _RET_IP_);
Quoted 11 lines
Even if a reader tries to sneak in between
__vma_enter_locked()and__vma_exit_locked()calls,__refcount_inc_not_zero_limited()will reject it becauseVMA_LOCK_OFFSETis set andVMA_REF_LIMITwill be violated. IOW, when VMA is freed, it's guaranteed to be detached with no current or future readers, therefore "race betweenvm_lock_seq==mm_lock_seqand the lock acquire, where a free and reuse happens" should not be possible.Did I understand your concern correctly and does my explanation make sense to you?
It is close to what Vlastimil said before.
Here is the sequence for a NULL dereference, refcnt value is not needed:
A: B: C:lock_vma_under_rcu()
vma = mas_walk()
vma_start_read()
vm_lock_seq == mm->mm_lock_seq.sequence
vma_start_write
vma detached and freed
vm_area_dup()
- vma reallocated
- zero vma
rwsem_acquire_read(NULL)
Here is a sequence for unlocking the new lock while locking the old one. The refcnt failure detects the detached state but does not protect against the wrong lock use:
A: B: C:lock_vma_under_rcu()
vma = mas_walk()
vma_start_read()
vm_lock_seq == mm->mm_lock_seq.sequence
vma_start_write
vma detached and freed
vm_area_dup()
- vma reallocated
rwsem_acquire_read(old lock)
__refcount_inc_not_zero_limited() fails
vma_init_lock();
rwsem_release(new lock)I don't think avoiding the copy of the ref count from the old vma is enough to stop these races?
Thanks,
Liam
Suren Baghdasaryan
maintainerRe: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 226 lines · Liam R. Howlett' via kernel-team
Quoted 177 lines · Suren Baghdasaryan
Quoted 105 lines · Liam R. Howlett
Quoted 96 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)include/linux/mm.h +0
-2@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;include/linux/mm_types.h +7
-3@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that @@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/* include/linux/slab.h +0
-6@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. * kernel/fork.c +0
-0kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 451
Quoted 3 lines
kernel/fork.c +0
-0@@ -449,6 +449,41 @@ struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return vma; }There exists a
copy_vma()which copies the vma to a new area in the mm in rmap. Naming thisvma_copy()is confusing :)Quoted 2 lines
It might be better to just put this code in the
vm_area_dup()or call it__vm_area_dup(), or__vma_dup()?Hmm. It's not really duplicating a vma but copying its content (no allocation). How about
__vm_area_copy()to indicate it is copyingvm_area_structcontent?Quoted 59 lines
Quoted 40 lines
kernel/fork.c +35
-0+static void vma_copy (const struct vm_area_struct *src,struct vm_area_struct *dest) +{ + dest->vm_mm = src->vm_mm; + dest->vm_ops = src->vm_ops; + dest->vm_start = src->vm_start; + dest->vm_end = src->vm_end; + dest->anon_vma = src->anon_vma; + dest->vm_pgoff = src->vm_pgoff; + dest->vm_file = src->vm_file; + dest->vm_private_data = src->vm_private_data; +vm_flags_init (dest, src->vm_flags); +memcpy (&dest->vm_page_prot, &src->vm_page_prot, +sizeof (dest->vm_page_prot)); +/* +* src->shared.rb may be modified concurrently, but the clone +* will be reinitialized. +*/ +data_race (memcpy (&dest->shared, &src->shared,sizeof (dest->shared))); +memcpy (&dest->vm_userfaultfd_ctx, &src->vm_userfaultfd_ctx, +sizeof (dest->vm_userfaultfd_ctx)); +#ifdef CONFIG_ANON_VMA_NAME + dest->anon_name = src->anon_name; +#endif +#ifdef CONFIG_SWAP +memcpy (&dest->swap_readahead_info, &src->swap_readahead_info, +sizeof (dest->swap_readahead_info)); +#endif +#ifndef CONFIG_MMU + dest->vm_region = src->vm_region; +#endif +#ifdef CONFIG_NUMA + dest->vm_policy = src->vm_policy; +#endif +} +struct vm_area_struct *vm_area_dup (struct vm_area_struct *orig) {struct vm_area_struct *new =kmem_cache_alloc (vm_area_cachep,GFP_KERNEL );kernel/fork.c struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig) · line 497
Quoted 10 lines
kernel/fork.c +1
-5@@ -458,11 +493,7 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)ASSERT_EXCLUSIVE_WRITER (orig->vm_flags);ASSERT_EXCLUSIVE_WRITER (orig->vm_file); -/* -* orig->shared.rb may be modified concurrently, but the clone -* will be reinitialized. -*/ -data_race (memcpy (new, orig,sizeof (*new))); +vma_copy (orig, new);vma_lock_init (new, true);I think this suffers from a race still?
That is, we can still race between
vm_lock_seq==mm_lock_seqand the lock acquire, where a free and reuse happens. In the even that the reader is caught between the sequence and lock taking, thevma->vmlock_dep_mapmay not be replaced and it could see the old lock (or zero?) and things go bad:Quoted 4 lines
It could try to take
vmlock_dep_map== 0 in read mode.It can take the old lock, detect the refcnt is wrong and release the new lock.
Quoted 9 lines
I don't think this race can happen. Notice a call to
vma_assert_detached()insidevm_area_free(), so before vma is freed and possibly reused, it has to be detached.vma_mark_detached()ensures that there are no current or future readers by executing the__vma_enter_locked()+__vma_exit_locked()sequence ifvm_refcntis not already at 0. Once__vma_exit_locked()is done,vm_refcntis at 0 and any new reader will be rejected on__refcount_inc_not_zero_limited(), before even checkingvm_lock_seq==mm_lock_seq.Isn't the
vm_lock_seqcheck before the ref count invma_start_read()?From patch 11/16:
(quoted diff) +7
-1@@ -720,13 +752,19 @@ static inline bool vma_start_read(struct vm_area_struct *vma) if (READ_ONCE(vma->vm_lock_seq) == READ_ONCE(vma->vm_mm->mm_lock_seq.sequence)) return false; - if (unlikely(down_read_trylock(&vma->vm_lock.lock) == 0)) + /* + * If VMA_LOCK_OFFSET is set, __refcount_inc_not_zero_limited() will fail + * because VMA_REF_LIMIT is less than VMA_LOCK_OFFSET. + */ + if (unlikely(!__refcount_inc_not_zero_limited(&vma->vm_refcnt, &oldcnt, + VMA_REF_LIMIT))) return false; + rwsem_acquire_read(&vma->vmlock_dep_map, 0, 1, _RET_IP_);Quoted 11 lines
Even if a reader tries to sneak in between
__vma_enter_locked()and__vma_exit_locked()calls,__refcount_inc_not_zero_limited()will reject it becauseVMA_LOCK_OFFSETis set andVMA_REF_LIMITwill be violated. IOW, when VMA is freed, it's guaranteed to be detached with no current or future readers, therefore "race betweenvm_lock_seq==mm_lock_seqand the lock acquire, where a free and reuse happens" should not be possible.Did I understand your concern correctly and does my explanation make sense to you?
It is close to what Vlastimil said before.
Here is the sequence for a NULL dereference, refcnt value is not needed:
Quoted 8 lines
A: B: C:
lock_vma_under_rcu()
vma =mas_walk()
vma_start_read()
vm_lock_seq==mm->mm_lock_seq.sequence
vma_start_write
vma detached and freed
At this point B makes vm_refcnt==0.
Quoted 5 lines
vm_area_dup()
- vma reallocated
- zero vma
vm_refcnt here is still 0.
Quoted 1 line
Here before calling rwsem_acquire_read(), the reader has to successfully do __refcount_inc_not_zero_limited() and that will fail because vm_refcnt==0. So rwsem_acquire_read(NULL) will not be called. What am I missing?
Quoted 12 lines
rwsem_acquire_read(NULL)Here is a sequence for unlocking the new lock while locking the old one. The refcnt failure detects the detached state but does not protect against the wrong lock use:
A: B: C:lock_vma_under_rcu()
vma =mas_walk()
vma_start_read()
vm_lock_seq==mm->mm_lock_seq.sequence
Quoted 5 lines
vma_start_write
vma detached and freed
vm_area_dup()
- vma reallocated
vm_refcnt is still 0 after reallocation.
Quoted 3 lines
rwsem_acquire_read(old lock)
__refcount_inc_not_zero_limited()fails
Are we looking at the same code
(https://lore.kernel.org/all/[email protected]/)?
The sequence should be reversed here like this:
if (!__refcount_inc_not_zero_limited())
return false;
rwsem_acquire_read(old lock)and because __refcount_inc_not_zero_limited() fails, rwsem_acquire_read(old lock) should never be called.
Quoted 4 lines
vma_init_lock();
rwsem_release(new lock)The reader will not call rwsem_release(new lock) because it failed to acquire the lock.
Quoted 9 lines
I don't think avoiding the copy of the ref count from the old vma is enough to stop these races?
Thanks,
LiamTo unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
Re: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 10 lines · Suren Baghdasaryan
On Fri, Jan 10, 2025 at 11:46 AM 'Liam R. Howlett' via kernel-team
Quoted 3 lines
rwsem_acquire_read(old lock)
__refcount_inc_not_zero_limited()failsAre we looking at the same code
(https://lore.kernel.org/all/[email protected]/)?
The sequence should be reversed here like this:
Quoted 7 lines
if (!__refcount_inc_not_zero_limited()) return false; rwsem_acquire_read(old lock)and because
__refcount_inc_not_zero_limited()fails,rwsem_acquire_read(old lock)should never be called.
We are not, I had an older version of the patch from the 7th.
Sorry for the confusion.
Thanks,
Liam
Suren Baghdasaryan
maintainerRe: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 16 lines · Liam R. Howlett' via kernel-team
Quoted 14 lines · Suren Baghdasaryan
On Fri, Jan 10, 2025 at 11:46 AM 'Liam R. Howlett' via kernel-team
Quoted 3 lines
rwsem_acquire_read(old lock)
__refcount_inc_not_zero_limited()failsAre we looking at the same code
(https://lore.kernel.org/all/[email protected]/)?
The sequence should be reversed here like this:if (!__refcount_inc_not_zero_limited()) return false; rwsem_acquire_read(old lock)
Quoted 7 lines
Quoted 3 lines
and because
__refcount_inc_not_zero_limited()fails,rwsem_acquire_read(old lock)should never be called.We are not, I had an older version of the patch from the 7th.
Sorry for the confusion.
No worries. Thanks for taking a look!
Quoted 6 lines
Thanks,
LiamTo unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
Re: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
In reply to Suren BaghdasaryanQuoted 106 lines · Suren Baghdasaryan
Quoted 105 lines · Liam R. Howlett
Quoted 96 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)include/linux/mm.h +0
-2@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;include/linux/mm_types.h +7
-3@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that @@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/* include/linux/slab.h +0
-6@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. * kernel/fork.c +0
-0kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 451
Quoted 3 lines
kernel/fork.c +0
-0@@ -449,6 +449,41 @@ struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return vma; }There exists a
copy_vma()which copies the vma to a new area in the mm in rmap. Naming thisvma_copy()is confusing :)
Quoted 6 lines
Quoted 2 lines
It might be better to just put this code in the
vm_area_dup()or call it__vm_area_dup(), or__vma_dup()?Hmm. It's not really duplicating a vma but copying its content (no allocation). How about
__vm_area_copy()to indicate it is copyingvm_area_structcontent?
Sorry, I missed this. it's not copying all the content either.
vm_area_init_dup() maybe?
Considering the scope of the series, I'm not sure I want to have a bike shed conversation.. But I also don't want copy_<foo> <foo>_copy confusion in the future.
Suren Baghdasaryan
maintainerRe: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 111 lines · Liam R. Howlett' via kernel-team
Quoted 109 lines · Suren Baghdasaryan
Quoted 107 lines · Liam R. Howlett
Quoted 96 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)include/linux/mm.h +0
-2@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;include/linux/mm_types.h +7
-3@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that @@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/* include/linux/slab.h +0
-6@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. * kernel/fork.c +0
-0kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 451
Quoted 3 lines
kernel/fork.c +0
-0@@ -449,6 +449,41 @@ struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return vma; }There exists a
copy_vma()which copies the vma to a new area in the mm in rmap. Naming thisvma_copy()is confusing :)It might be better to just put this code in the
vm_area_dup()or call it__vm_area_dup(), or__vma_dup()?
Quoted 8 lines
Quoted 3 lines
Hmm. It's not really duplicating a vma but copying its content (no allocation). How about
__vm_area_copy()to indicate it is copyingvm_area_structcontent?Sorry, I missed this. it's not copying all the content either.
vm_area_init_dup()maybe?
Ah, how about vm_area_init_from(src, dest)?
Quoted 7 lines
Considering the scope of the series, I'm not sure I want to have a bike shed conversation.. But I also don't want copy_<foo> <foo>_copy confusion in the future.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
Re: [PATCH v8 15/16] mm: make vma cache SLAB_TYPESAFE_BY_RCU
Quoted 118 lines · Suren Baghdasaryan
Quoted 116 lines · Liam R. Howlett' via kernel-team
Quoted 112 lines · Suren Baghdasaryan
Quoted 105 lines · Liam R. Howlett
Quoted 96 lines · Suren Baghdasaryan
To enable
SLAB_TYPESAFE_BY_RCUfor vma cache we need to ensure that object reuse before RCU grace period is over will be detected bylock_vma_under_rcu().
Current checks are sufficient as long as vma is detached before it is freed. The only place this is not currently happening is inexit_mmap(). Add the missingvma_mark_detached()inexit_mmap().
Another issue which might tricklock_vma_under_rcu()during vma reuse isvm_area_dup(), which copies the entire content of the vma into a new one, overriding new vma'svm_refcntand temporarily making it appear as attached. This might trick a racinglock_vma_under_rcu()to operate on a reused vma if it found the vma before it got reused. To prevent this situation, we should ensure thatvm_refcntstays at detached state (0) when it is copied and advances to attached state only after it is added into the vma tree. Introducevma_copy()which preserves new vma'svm_refcntand use it invm_area_dup(). Since all vmas are in detached state with no current readers when they are freed,lock_vma_under_rcu()will not be able to takevm_refcntafter vma got detached even if vma is reused.
Finally, makevm_area_cachepSLAB_TYPESAFE_BY_RCU. This will facilitatevm_area_structreuse and will minimize the number ofcall_rcu()calls.Signed-off-by Suren Baghdasaryan <[email protected]>--- include/linux/mm.h | 2 - include/linux/mm_types.h | 10 +++-- include/linux/slab.h | 6 --- kernel/fork.c | 72 ++++++++++++++++++++------------ mm/mmap.c | 3 +- mm/vma.c | 11 ++--- mm/vma.h | 2 +- tools/testing/vma/vma_internal.h | 7 +--- 8 files changed, 59 insertions(+), 54 deletions(-)include/linux/mm.h +0
-2@@ -258,8 +258,6 @@ void setup_initial_init_mm(void *start_code, void *end_code,struct vm_area_struct *vm_area_alloc (struct mm_struct *);struct vm_area_struct *vm_area_dup (struct vm_area_struct *);void vm_area_free (struct vm_area_struct *); -/* Use only if VMA has no other users */ -void __vm_area_free (struct vm_area_struct *vma);#ifndef CONFIG_MMU extern struct rb_root nommu_region_tree;include/linux/mm_types.h +7
-3@@ -582,6 +582,12 @@ static inline void *folio_get_private(struct folio *folio)typedef unsigned long vm_flags_t ; +/* +* freeptr_t represents a SLUB freelist pointer, which might be encoded +* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. +*/ +typedef struct {unsigned long v; }freeptr_t ; +/* * A region containing a mapping of a non-memory backed file under NOMMU * conditions. These are held in a global tree and are pinned by the VMAs that @@ -695,9 +701,7 @@ struct vm_area_struct {unsigned long vm_start;unsigned long vm_end; }; -#ifdef CONFIG_PER_VMA_LOCK -struct rcu_head vm_rcu;/* Used for deferred freeing. */ -#endif +freeptr_t vm_freeptr;/* Pointer used by SLAB_TYPESAFE_BY_RCU */ };/* include/linux/slab.h +0
-6@@ -234,12 +234,6 @@ enum _slab_flag_bits {#define SLAB_NO_OBJ_EXT __SLAB_FLAG_UNUSED#endif -/* -* freeptr_t represents a SLUB freelist pointer, which might be encoded -* and not dereferenceable if CONFIG_SLAB_FREELIST_HARDENED is enabled. -*/ -typedef struct {unsigned long v; }freeptr_t ; -/* * ZERO_SIZE_PTR will be returned for zero sized kmalloc requests. * kernel/fork.c +0
-0kernel/fork.c struct vm_area_struct *vm_area_alloc(struct mm_struct *mm) · line 451
Quoted 3 lines
kernel/fork.c +0
-0@@ -449,6 +449,41 @@ struct vm_area_struct *vm_area_alloc(struct mm_struct *mm)return vma; }There exists a
copy_vma()which copies the vma to a new area in the mm in rmap. Naming thisvma_copy()is confusing :)Quoted 2 lines
It might be better to just put this code in the
vm_area_dup()or call it__vm_area_dup(), or__vma_dup()?Hmm. It's not really duplicating a vma but copying its content (no allocation). How about
__vm_area_copy()to indicate it is copyingvm_area_structcontent?
Quoted 6 lines
Quoted 3 lines
Sorry, I missed this. it's not copying all the content either.
vm_area_init_dup()maybe?Ah, how about
vm_area_init_from(src, dest)?
Sure, thanks.
Suren Baghdasaryan
maintainerPATCH v8 16/16 docs/mm: document latest changes to vm_lock
In reply to Suren BaghdasaryanChange the documentation to reflect that vm_lock is integrated into vma and replaced with vm_refcnt.
Document newly introduced vma_start_read_locked{_nested} functions.
--- Documentation/mm/process_addrs.rst | 44 ++++++++++++++++++------------ 1 file changed, 26 insertions(+), 18 deletions(-)
Documentation/mm/process_addrs.rst +26 -18
@@ -716,9 +716,14 @@ calls :c:func:`!rcu_read_lock` to ensure that the VMA is looked up in an RCU critical section, then attempts to VMA lock it via :c:func:`!vma_start_read`, before releasing the RCU lock via :c:func:`!rcu_read_unlock`. -VMA read locks hold the read lock on the :c:member:`!vma->vm_lock` semaphore for -their duration and the caller of :c:func:`!lock_vma_under_rcu` must release it -via :c:func:`!vma_end_read`. +In cases when the user already holds mmap read lock, :c:func:`!vma_start_read_locked` +and :c:func:`!vma_start_read_locked_nested` can be used. These functions do not +fail due to lock contention but the caller should still check their return values +in case they fail for other reasons. + +VMA read locks increment :c:member:`!vma.vm_refcnt` reference counter for their +duration and the caller of :c:func:`!lock_vma_under_rcu` must drop it via +:c:func:`!vma_end_read`. VMA **write** locks are acquired via :c:func:`!vma_start_write` in instances where a VMA is about to be modified, unlike :c:func:`!vma_start_read` the lock is always
@@ -726,9 +731,9 @@ acquired. An mmap write lock **must** be held for the duration of the VMA write lock, releasing or downgrading the mmap write lock also releases the VMA write lock so there is no :c:func:`!vma_end_write` function. -Note that a semaphore write lock is not held across a VMA lock. Rather, a -sequence number is used for serialisation, and the write semaphore is only -acquired at the point of write lock to update this. +Note that when write-locking a VMA lock, the :c:member:`!vma.vm_refcnt` is temporarily +modified so that readers can detect the presense of a writer. The reference counter is +restored once the vma sequence number used for serialisation is updated. This ensures the semantics we require - VMA write locks provide exclusive write access to the VMA.
@@ -738,7 +743,7 @@ Implementation details The VMA lock mechanism is designed to be a lightweight means of avoiding the use of the heavily contended mmap lock. It is implemented using a combination of a -read/write semaphore and sequence numbers belonging to the containing +reference counter and sequence numbers belonging to the containing :c:struct:`!struct mm_struct` and the VMA. Read locks are acquired via :c:func:`!vma_start_read`, which is an optimistic
@@ -779,28 +784,31 @@ release of any VMA locks on its release makes sense, as you would never want to keep VMAs locked across entirely separate write operations. It also maintains correct lock ordering. -Each time a VMA read lock is acquired, we acquire a read lock on the -:c:member:`!vma->vm_lock` read/write semaphore and hold it, while checking that -the sequence count of the VMA does not match that of the mm. +Each time a VMA read lock is acquired, we increment :c:member:`!vma.vm_refcnt` +reference counter and check that the sequence count of the VMA does not match +that of the mm. -If it does, the read lock fails. If it does not, we hold the lock, excluding -writers, but permitting other readers, who will also obtain this lock under RCU. +If it does, the read lock fails and :c:member:`!vma.vm_refcnt` is dropped. +If it does not, we keep the reference counter raised, excluding writers, but +permitting other readers, who can also obtain this lock under RCU. Importantly, maple tree operations performed in :c:func:`!lock_vma_under_rcu` are also RCU safe, so the whole read lock operation is guaranteed to function correctly. -On the write side, we acquire a write lock on the :c:member:`!vma->vm_lock` -read/write semaphore, before setting the VMA's sequence number under this lock, -also simultaneously holding the mmap write lock. +On the write side, we set a bit in :c:member:`!vma.vm_refcnt` which can't be +modified by readers and wait for all readers to drop their reference count. +Once there are no readers, VMA's sequence number is set to match that of the +mm. During this entire operation mmap write lock is held. This way, if any read locks are in effect, :c:func:`!vma_start_write` will sleep until these are finished and mutual exclusion is achieved. -After setting the VMA's sequence number, the lock is released, avoiding -complexity with a long-term held write lock. +After setting the VMA's sequence number, the bit in :c:member:`!vma.vm_refcnt` +indicating a writer is cleared. From this point on, VMA's sequence number will +indicate VMA's write-locked state until mmap write lock is dropped or downgraded. -This clever combination of a read/write semaphore and sequence count allows for +This clever combination of a reference counter and sequence count allows for fast RCU-based per-VMA lock acquisition (especially on page fault, though utilised elsewhere) with minimal complexity around lock ordering.
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
Suren Baghdasaryan
maintainerQuoted 79 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
1. movesvm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.Pagefault microbenchmarks show performance improvement:
Hmean faults/cpu-1 507926.5547 ( 0.00%) 506519.3692 * -0.28%* Hmean faults/cpu-4 479119.7051 ( 0.00%) 481333.6802 * 0.46%* Hmean faults/cpu-7 452880.2961 ( 0.00%) 455845.6211 * 0.65%* Hmean faults/cpu-12 347639.1021 ( 0.00%) 352004.2254 * 1.26%* Hmean faults/cpu-21 200061.2238 ( 0.00%) 229597.0317 * 14.76%* Hmean faults/cpu-30 145251.2001 ( 0.00%) 164202.5067 * 13.05%* Hmean faults/cpu-48 106848.4434 ( 0.00%) 120641.5504 * 12.91%* Hmean faults/cpu-56 92472.3835 ( 0.00%) 103464.7916 * 11.89%* Hmean faults/sec-1 507566.1468 ( 0.00%) 506139.0811 * -0.28%* Hmean faults/sec-4 1880478.2402 ( 0.00%) 1886795.6329 * 0.34%* Hmean faults/sec-7 3106394.3438 ( 0.00%) 3140550.7485 * 1.10%* Hmean faults/sec-12 4061358.4795 ( 0.00%) 4112477.0206 * 1.26%* Hmean faults/sec-21 3988619.1169 ( 0.00%) 4577747.1436 * 14.77%* Hmean faults/sec-30 3909839.5449 ( 0.00%) 4311052.2787 * 10.26%* Hmean faults/sec-48 4761108.4691 ( 0.00%) 5283790.5026 * 10.98%* Hmean faults/sec-56 4885561.4590 ( 0.00%) 5415839.4045 * 10.85%*Changes since v7 [4]:
- Removed additional parameter forvma_iter_store()and introducedvma_iter_store_attached()instead, per Vlastimil Babka and Liam R. Howlett
- Fixed coding style nits, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Liam R. Howlett
- Added Acked-by, per Davidlohr Bueso
- Removed unnecessary patch changeingnommu.c
- Folded a fixup patch [5] into the patch it was fixing
- Changed calculation in__refcount_add_not_zero_limited()to avoid overflow, to change the limit to be inclusive and to useINT_MAXto indicate no limits, per Vlastimil Babka and Matthew Wilcox
- Folded a fixup patch [6] into the patch it was fixing
- Addedvm_refcntrules summary in the changelog, per Liam R. Howlett
- Changed writers to not incrementvm_refcntand adjustedVMA_REF_LIMITto not reserve one count for a writer, per Liam R. Howlett
- Changedvma_refcount_put()to wake up writers only when the last reader is leaving, per Liam R. Howlett
- Fixedrwsem_acquire_read()parameters when read-locking a vma to match the waydown_read_trylock()does lockdep, per Vlastimil Babka
- Foldedvma_lockdep_init()intovma_lock_init()for simplicity
- Brought backvma_copy()to keepvm_refcountat 0 during reuse, per Vlastimil BabkaWhat I did not include in this patchset:
- Liam's suggestion to changedump_vma()output since it's unclear to me how it should look like. The patch is for debug only and not critical for the rest of the series, we can change the output later or even drop it if necessary.[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
[4] https://lore.kernel.org/all/[email protected]/
[5] https://lore.kernel.org/all/[email protected]/
[6] https://lore.kernel.org/all/[email protected]/
Quoted 2 lines
Patchset applies over mm-unstable after reverting v7 (current SHA range:
588f0086398e-fb2270654630)
^^^ Please note that to apply this patchset over mm-unstable you should revert the previous version. Thanks!
Quoted 42 lines
Suren Baghdasaryan (16):
mm: introduce vma_start_read_locked{_nested} helpers
mm: move per-vma lock intovm_area_struct
mm: mark vma as detached until it's added into vma tree
mm: introducevma_iter_store_attached()to use with attached vmas
mm: mark vmas detached upon exit
types: move struct rcuwait intotypes.h
mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
mm: movemmap_init_lock()out of the header file
mm: uninline the main body ofvma_start_write()
refcount: introduce __refcount_{add|inc}_not_zero_limited
mm: replacevm_lockand detached flag with a reference count mm/debug: printvm_refcntstate when dumping the vma
mm: remove extravma_numab_state_init()call
mm: preparelock_vma_under_rcu()for vma reuse possibility
mm: make vma cacheSLAB_TYPESAFE_BY_RCU
docs/mm: document latest changes tovm_lockDocumentation/mm/process_addrs.rst | 44 +++++---- include/linux/mm.h | 152 ++++++++++++++++++++++------- include/linux/mm_types.h | 36 ++++--- include/linux/mmap_lock.h | 6 -- include/linux/rcuwait.h | 13 +-- include/linux/refcount.h | 20 +++- include/linux/slab.h | 6 -- include/linux/types.h | 12 +++ kernel/fork.c | 128 +++++++++++------------- mm/debug.c | 12 +++ mm/init-mm.c | 1 + mm/memory.c | 94 +++++++++++++++--- mm/mmap.c | 3 +- mm/userfaultfd.c | 32 +++--- mm/vma.c | 23 ++--- mm/vma.h | 15 ++- tools/testing/vma/linux/atomic.h | 5 + tools/testing/vma/vma_internal.h | 93 ++++++++---------- 18 files changed, 435 insertions(+), 260 deletions(-)--
2.47.1.613.gc27f4b7a9f-goog
Peter Zijlstra
maintainerQuoted 13 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
Quoted 8 lines
1. moves
vm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.
Does not clean up that reattach nonsense :-(
Suren Baghdasaryan
maintainerQuoted 24 lines · Peter Zijlstra
Quoted 13 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:Quoted 8 lines
1. moves
vm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.
Quoted 1 line
Does not clean up that reattach nonsense :-(
Oh, no. I think it does. That's why in [1] I introduce vma_iter_store_attached() to be used on already attached vmas and to avoid marking them attached again. Also I added assertions in vma_mark_attached()/vma_mark_detached() to avoid re-attaching or re-detaching. Unless I misunderstood your comment?
Peter Zijlstra
maintainerQuoted 25 lines · Suren Baghdasaryan
Quoted 24 lines · Peter Zijlstra
Quoted 21 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
1. movesvm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.
Quoted 7 lines
Quoted 1 line
Does not clean up that reattach nonsense :-(
Oh, no. I think it does. That's why in [1] I introduce
vma_iter_store_attached()to be used on already attached vmas and to avoid marking them attached again. Also I added assertions in vma_mark_attached()/vma_mark_detached() to avoid re-attaching or re-detaching. Unless I misunderstood your comment?
Hmm, I'll go read the thing again, maybe I missed it.
Peter Zijlstra
maintainerQuoted 28 lines · Peter Zijlstra
Quoted 27 lines · Suren Baghdasaryan
Quoted 25 lines · Peter Zijlstra
Quoted 13 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:Quoted 8 lines
1. moves
vm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.Does not clean up that reattach nonsense :-(
Quoted 7 lines
Quoted 5 lines
Oh, no. I think it does. That's why in [1] I introduce
vma_iter_store_attached()to be used on already attached vmas and to avoid marking them attached again. Also I added assertions in vma_mark_attached()/vma_mark_detached() to avoid re-attaching or re-detaching. Unless I misunderstood your comment?Hmm, I'll go read the thing again, maybe I missed it.
You're right. I was looking for the approach that changed the need to reattach, by moving the point of no return.
This should do for now.
Let me see if I can find time today to finally do a proper reading.
Thanks!
Btw the subject became rather incomplete given all the series does :)
Quoted 23 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
1. movesvm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.Pagefault microbenchmarks show performance improvement:
Quoted 16 lines
Hmean faults/cpu-1 507926.5547 ( 0.00%) 506519.3692 * -0.28%* Hmean faults/cpu-4 479119.7051 ( 0.00%) 481333.6802 * 0.46%* Hmean faults/cpu-7 452880.2961 ( 0.00%) 455845.6211 * 0.65%* Hmean faults/cpu-12 347639.1021 ( 0.00%) 352004.2254 * 1.26%* Hmean faults/cpu-21 200061.2238 ( 0.00%) 229597.0317 * 14.76%* Hmean faults/cpu-30 145251.2001 ( 0.00%) 164202.5067 * 13.05%* Hmean faults/cpu-48 106848.4434 ( 0.00%) 120641.5504 * 12.91%* Hmean faults/cpu-56 92472.3835 ( 0.00%) 103464.7916 * 11.89%* Hmean faults/sec-1 507566.1468 ( 0.00%) 506139.0811 * -0.28%* Hmean faults/sec-4 1880478.2402 ( 0.00%) 1886795.6329 * 0.34%* Hmean faults/sec-7 3106394.3438 ( 0.00%) 3140550.7485 * 1.10%* Hmean faults/sec-12 4061358.4795 ( 0.00%) 4112477.0206 * 1.26%* Hmean faults/sec-21 3988619.1169 ( 0.00%) 4577747.1436 * 14.77%* Hmean faults/sec-30 3909839.5449 ( 0.00%) 4311052.2787 * 10.26%* Hmean faults/sec-48 4761108.4691 ( 0.00%) 5283790.5026 * 10.98%* Hmean faults/sec-56 4885561.4590 ( 0.00%) 5415839.4045 * 10.85%*
Given how patch 2 discusses memory growth due to moving the lock, should also patch 11 discuss how the replacement with refcount reduces the memory footprint? And/or the cover letter could summarize the impact of the whole series in that aspect? Perhaps the refcount doesn't reduce anything as it's smaller but sits alone in the cacheline? Could it be grouped with some non-hot fields instead as a followup, so could we get to <=192 (non-debug) size without impacting performance?
Quoted 79 lines
Changes since v7 [4]:
- Removed additional parameter forvma_iter_store()and introducedvma_iter_store_attached()instead, per Vlastimil Babka and Liam R. Howlett
- Fixed coding style nits, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Liam R. Howlett
- Added Acked-by, per Davidlohr Bueso
- Removed unnecessary patch changeingnommu.c
- Folded a fixup patch [5] into the patch it was fixing
- Changed calculation in__refcount_add_not_zero_limited()to avoid overflow, to change the limit to be inclusive and to useINT_MAXto indicate no limits, per Vlastimil Babka and Matthew Wilcox
- Folded a fixup patch [6] into the patch it was fixing
- Addedvm_refcntrules summary in the changelog, per Liam R. Howlett
- Changed writers to not incrementvm_refcntand adjustedVMA_REF_LIMITto not reserve one count for a writer, per Liam R. Howlett
- Changedvma_refcount_put()to wake up writers only when the last reader is leaving, per Liam R. Howlett
- Fixedrwsem_acquire_read()parameters when read-locking a vma to match the waydown_read_trylock()does lockdep, per Vlastimil Babka
- Foldedvma_lockdep_init()intovma_lock_init()for simplicity
- Brought backvma_copy()to keepvm_refcountat 0 during reuse, per Vlastimil BabkaWhat I did not include in this patchset:
- Liam's suggestion to changedump_vma()output since it's unclear to me how it should look like. The patch is for debug only and not critical for the rest of the series, we can change the output later or even drop it if necessary.[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
[4] https://lore.kernel.org/all/[email protected]/
[5] https://lore.kernel.org/all/[email protected]/
[6] https://lore.kernel.org/all/[email protected]/Patchset applies over mm-unstable after reverting v7 (current SHA range:
588f0086398e-fb2270654630)Suren Baghdasaryan (16):
mm: introduce vma_start_read_locked{_nested} helpers
mm: move per-vma lock intovm_area_struct
mm: mark vma as detached until it's added into vma tree
mm: introducevma_iter_store_attached()to use with attached vmas
mm: mark vmas detached upon exit
types: move struct rcuwait intotypes.h
mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
mm: movemmap_init_lock()out of the header file
mm: uninline the main body ofvma_start_write()
refcount: introduce __refcount_{add|inc}_not_zero_limited
mm: replacevm_lockand detached flag with a reference count mm/debug: printvm_refcntstate when dumping the vma
mm: remove extravma_numab_state_init()call
mm: preparelock_vma_under_rcu()for vma reuse possibility
mm: make vma cacheSLAB_TYPESAFE_BY_RCU
docs/mm: document latest changes tovm_lockDocumentation/mm/process_addrs.rst | 44 +++++---- include/linux/mm.h | 152 ++++++++++++++++++++++------- include/linux/mm_types.h | 36 ++++--- include/linux/mmap_lock.h | 6 -- include/linux/rcuwait.h | 13 +-- include/linux/refcount.h | 20 +++- include/linux/slab.h | 6 -- include/linux/types.h | 12 +++ kernel/fork.c | 128 +++++++++++------------- mm/debug.c | 12 +++ mm/init-mm.c | 1 + mm/memory.c | 94 +++++++++++++++--- mm/mmap.c | 3 +- mm/userfaultfd.c | 32 +++--- mm/vma.c | 23 ++--- mm/vma.h | 15 ++- tools/testing/vma/linux/atomic.h | 5 + tools/testing/vma/vma_internal.h | 93 ++++++++---------- 18 files changed, 435 insertions(+), 260 deletions(-)
Suren Baghdasaryan
maintainerQuoted 44 lines · Vlastimil Babka
Btw the subject became rather incomplete given all the series does :)
Quoted 23 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
1. movesvm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.Pagefault microbenchmarks show performance improvement:
Quoted 16 lines
Hmean faults/cpu-1 507926.5547 ( 0.00%) 506519.3692 * -0.28%* Hmean faults/cpu-4 479119.7051 ( 0.00%) 481333.6802 * 0.46%* Hmean faults/cpu-7 452880.2961 ( 0.00%) 455845.6211 * 0.65%* Hmean faults/cpu-12 347639.1021 ( 0.00%) 352004.2254 * 1.26%* Hmean faults/cpu-21 200061.2238 ( 0.00%) 229597.0317 * 14.76%* Hmean faults/cpu-30 145251.2001 ( 0.00%) 164202.5067 * 13.05%* Hmean faults/cpu-48 106848.4434 ( 0.00%) 120641.5504 * 12.91%* Hmean faults/cpu-56 92472.3835 ( 0.00%) 103464.7916 * 11.89%* Hmean faults/sec-1 507566.1468 ( 0.00%) 506139.0811 * -0.28%* Hmean faults/sec-4 1880478.2402 ( 0.00%) 1886795.6329 * 0.34%* Hmean faults/sec-7 3106394.3438 ( 0.00%) 3140550.7485 * 1.10%* Hmean faults/sec-12 4061358.4795 ( 0.00%) 4112477.0206 * 1.26%* Hmean faults/sec-21 3988619.1169 ( 0.00%) 4577747.1436 * 14.77%* Hmean faults/sec-30 3909839.5449 ( 0.00%) 4311052.2787 * 10.26%* Hmean faults/sec-48 4761108.4691 ( 0.00%) 5283790.5026 * 10.98%* Hmean faults/sec-56 4885561.4590 ( 0.00%) 5415839.4045 * 10.85%*
Quoted 4 lines
Given how patch 2 discusses memory growth due to moving the lock, should also patch 11 discuss how the replacement with refcount reduces the memory footprint? And/or the cover letter could summarize the impact of the whole series in that aspect?
That's a good idea. I can amend the cover letter and the description of patch 11 to include size information.
Quoted 4 lines
Perhaps the refcount doesn't reduce
anything as it's smaller but sits alone in the cacheline? Could it be grouped with some non-hot fields instead as a followup, so could we get to <=192 (non-debug) size without impacting performance?
Yes, absolutely. Before this series, vm_area_struct was roughly 168 bytes and vm_lock was 40 bytes. After the changes vm_area_struct becomes 256 bytes. I was planning to pack the fields as a follow-up patch similar to an earlier one [1] and bring the size of vm_area_struct to < 192. I felt this patchset already does many things and did not include it here but I can add it at the end of this patchset if you think it's essential.
[1] https://lore.kernel.org/all/[email protected]/
Quoted 81 lines
Quoted 80 lines
Changes since v7 [4]:
- Removed additional parameter forvma_iter_store()and introducedvma_iter_store_attached()instead, per Vlastimil Babka and Liam R. Howlett
- Fixed coding style nits, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Liam R. Howlett
- Added Acked-by, per Davidlohr Bueso
- Removed unnecessary patch changeingnommu.c
- Folded a fixup patch [5] into the patch it was fixing
- Changed calculation in__refcount_add_not_zero_limited()to avoid overflow, to change the limit to be inclusive and to useINT_MAXto indicate no limits, per Vlastimil Babka and Matthew Wilcox
- Folded a fixup patch [6] into the patch it was fixing
- Addedvm_refcntrules summary in the changelog, per Liam R. Howlett
- Changed writers to not incrementvm_refcntand adjustedVMA_REF_LIMITto not reserve one count for a writer, per Liam R. Howlett
- Changedvma_refcount_put()to wake up writers only when the last reader is leaving, per Liam R. Howlett
- Fixedrwsem_acquire_read()parameters when read-locking a vma to match the waydown_read_trylock()does lockdep, per Vlastimil Babka
- Foldedvma_lockdep_init()intovma_lock_init()for simplicity
- Brought backvma_copy()to keepvm_refcountat 0 during reuse, per Vlastimil BabkaWhat I did not include in this patchset:
- Liam's suggestion to changedump_vma()output since it's unclear to me how it should look like. The patch is for debug only and not critical for the rest of the series, we can change the output later or even drop it if necessary.[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
[4] https://lore.kernel.org/all/[email protected]/
[5] https://lore.kernel.org/all/[email protected]/
[6] https://lore.kernel.org/all/[email protected]/Patchset applies over mm-unstable after reverting v7 (current SHA range:
588f0086398e-fb2270654630)Suren Baghdasaryan (16):
mm: introduce vma_start_read_locked{_nested} helpers
mm: move per-vma lock intovm_area_struct
mm: mark vma as detached until it's added into vma tree
mm: introducevma_iter_store_attached()to use with attached vmas
mm: mark vmas detached upon exit
types: move struct rcuwait intotypes.h
mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
mm: movemmap_init_lock()out of the header file
mm: uninline the main body ofvma_start_write()
refcount: introduce __refcount_{add|inc}_not_zero_limited
mm: replacevm_lockand detached flag with a reference count mm/debug: printvm_refcntstate when dumping the vma
mm: remove extravma_numab_state_init()call
mm: preparelock_vma_under_rcu()for vma reuse possibility
mm: make vma cacheSLAB_TYPESAFE_BY_RCU
docs/mm: document latest changes tovm_lockDocumentation/mm/process_addrs.rst | 44 +++++---- include/linux/mm.h | 152 ++++++++++++++++++++++------- include/linux/mm_types.h | 36 ++++--- include/linux/mmap_lock.h | 6 -- include/linux/rcuwait.h | 13 +-- include/linux/refcount.h | 20 +++- include/linux/slab.h | 6 -- include/linux/types.h | 12 +++ kernel/fork.c | 128 +++++++++++------------- mm/debug.c | 12 +++ mm/init-mm.c | 1 + mm/memory.c | 94 +++++++++++++++--- mm/mmap.c | 3 +- mm/userfaultfd.c | 32 +++--- mm/vma.c | 23 ++--- mm/vma.h | 15 ++- tools/testing/vma/linux/atomic.h | 5 + tools/testing/vma/vma_internal.h | 93 ++++++++---------- 18 files changed, 435 insertions(+), 260 deletions(-)
Suren Baghdasaryan
maintainerQuoted 67 lines · Suren Baghdasaryan
Quoted 44 lines · Vlastimil Babka
Btw the subject became rather incomplete given all the series does :)
Quoted 23 lines · Suren Baghdasaryan
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
1. movesvm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.Pagefault microbenchmarks show performance improvement:
Quoted 16 lines
Hmean faults/cpu-1 507926.5547 ( 0.00%) 506519.3692 * -0.28%* Hmean faults/cpu-4 479119.7051 ( 0.00%) 481333.6802 * 0.46%* Hmean faults/cpu-7 452880.2961 ( 0.00%) 455845.6211 * 0.65%* Hmean faults/cpu-12 347639.1021 ( 0.00%) 352004.2254 * 1.26%* Hmean faults/cpu-21 200061.2238 ( 0.00%) 229597.0317 * 14.76%* Hmean faults/cpu-30 145251.2001 ( 0.00%) 164202.5067 * 13.05%* Hmean faults/cpu-48 106848.4434 ( 0.00%) 120641.5504 * 12.91%* Hmean faults/cpu-56 92472.3835 ( 0.00%) 103464.7916 * 11.89%* Hmean faults/sec-1 507566.1468 ( 0.00%) 506139.0811 * -0.28%* Hmean faults/sec-4 1880478.2402 ( 0.00%) 1886795.6329 * 0.34%* Hmean faults/sec-7 3106394.3438 ( 0.00%) 3140550.7485 * 1.10%* Hmean faults/sec-12 4061358.4795 ( 0.00%) 4112477.0206 * 1.26%* Hmean faults/sec-21 3988619.1169 ( 0.00%) 4577747.1436 * 14.77%* Hmean faults/sec-30 3909839.5449 ( 0.00%) 4311052.2787 * 10.26%* Hmean faults/sec-48 4761108.4691 ( 0.00%) 5283790.5026 * 10.98%* Hmean faults/sec-56 4885561.4590 ( 0.00%) 5415839.4045 * 10.85%*Quoted 4 lines
Given how patch 2 discusses memory growth due to moving the lock, should also patch 11 discuss how the replacement with refcount reduces the memory footprint? And/or the cover letter could summarize the impact of the whole series in that aspect?
That's a good idea. I can amend the cover letter and the description of patch 11 to include size information.
Quoted 4 lines
Perhaps the refcount doesn't reduce
anything as it's smaller but sits alone in the cacheline? Could it be grouped with some non-hot fields instead as a followup, so could we get to <=192 (non-debug) size without impacting performance?Yes, absolutely. Before this series,
vm_area_structwas roughly 168 bytes andvm_lockwas 40 bytes. After the changesvm_area_structbecomes 256 bytes. I was planning to pack the fields as a follow-up patch similar to an earlier one [1] and bring the size ofvm_area_structto < 192. I felt this patchset already does many things and did not include it here but I can add it at the end of this patchset if you think it's essential.
Quoted 1 line
Actually I tried to rewrite the above patch [1] over the latest patchset and it's pretty much the same. I think I should include it at the end of this patchset as it's pretty simple. Planning to post v9 tomorrow morning, so if you don't want it in this patchset, please let me know.
Thanks!
Quoted 82 lines
Quoted 82 lines
Quoted 81 lines
Changes since v7 [4]:
- Removed additional parameter forvma_iter_store()and introducedvma_iter_store_attached()instead, per Vlastimil Babka and Liam R. Howlett
- Fixed coding style nits, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Liam R. Howlett
- Added Acked-by, per Davidlohr Bueso
- Removed unnecessary patch changeingnommu.c
- Folded a fixup patch [5] into the patch it was fixing
- Changed calculation in__refcount_add_not_zero_limited()to avoid overflow, to change the limit to be inclusive and to useINT_MAXto indicate no limits, per Vlastimil Babka and Matthew Wilcox
- Folded a fixup patch [6] into the patch it was fixing
- Addedvm_refcntrules summary in the changelog, per Liam R. Howlett
- Changed writers to not incrementvm_refcntand adjustedVMA_REF_LIMITto not reserve one count for a writer, per Liam R. Howlett
- Changedvma_refcount_put()to wake up writers only when the last reader is leaving, per Liam R. Howlett
- Fixedrwsem_acquire_read()parameters when read-locking a vma to match the waydown_read_trylock()does lockdep, per Vlastimil Babka
- Foldedvma_lockdep_init()intovma_lock_init()for simplicity
- Brought backvma_copy()to keepvm_refcountat 0 during reuse, per Vlastimil BabkaWhat I did not include in this patchset:
- Liam's suggestion to changedump_vma()output since it's unclear to me how it should look like. The patch is for debug only and not critical for the rest of the series, we can change the output later or even drop it if necessary.[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
[4] https://lore.kernel.org/all/[email protected]/
[5] https://lore.kernel.org/all/[email protected]/
[6] https://lore.kernel.org/all/[email protected]/Patchset applies over mm-unstable after reverting v7 (current SHA range:
588f0086398e-fb2270654630)Suren Baghdasaryan (16):
mm: introduce vma_start_read_locked{_nested} helpers
mm: move per-vma lock intovm_area_struct
mm: mark vma as detached until it's added into vma tree
mm: introducevma_iter_store_attached()to use with attached vmas
mm: mark vmas detached upon exit
types: move struct rcuwait intotypes.h
mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
mm: movemmap_init_lock()out of the header file
mm: uninline the main body ofvma_start_write()
refcount: introduce __refcount_{add|inc}_not_zero_limited
mm: replacevm_lockand detached flag with a reference count mm/debug: printvm_refcntstate when dumping the vma
mm: remove extravma_numab_state_init()call
mm: preparelock_vma_under_rcu()for vma reuse possibility
mm: make vma cacheSLAB_TYPESAFE_BY_RCU
docs/mm: document latest changes tovm_lockDocumentation/mm/process_addrs.rst | 44 +++++---- include/linux/mm.h | 152 ++++++++++++++++++++++------- include/linux/mm_types.h | 36 ++++--- include/linux/mmap_lock.h | 6 -- include/linux/rcuwait.h | 13 +-- include/linux/refcount.h | 20 +++- include/linux/slab.h | 6 -- include/linux/types.h | 12 +++ kernel/fork.c | 128 +++++++++++------------- mm/debug.c | 12 +++ mm/init-mm.c | 1 + mm/memory.c | 94 +++++++++++++++--- mm/mmap.c | 3 +- mm/userfaultfd.c | 32 +++--- mm/vma.c | 23 ++--- mm/vma.h | 15 ++- tools/testing/vma/linux/atomic.h | 5 + tools/testing/vma/vma_internal.h | 93 ++++++++---------- 18 files changed, 435 insertions(+), 260 deletions(-)
Suren Baghdasaryan
maintainerQuoted 2 lines · Vlastimil Babka
Btw the subject became rather incomplete given all the series does :)
Missed this one. What do you think is worth mentioning here? It's true, the patchset does many small things but I wanted to outline the main conceptual changes. Please LMK if you think there are more changes big enough to be mentioned here.
Quoted 130 lines · Suren Baghdasaryan
Quoted 23 lines
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
1. movesvm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.Pagefault microbenchmarks show performance improvement:
Quoted 16 lines
Hmean faults/cpu-1 507926.5547 ( 0.00%) 506519.3692 * -0.28%* Hmean faults/cpu-4 479119.7051 ( 0.00%) 481333.6802 * 0.46%* Hmean faults/cpu-7 452880.2961 ( 0.00%) 455845.6211 * 0.65%* Hmean faults/cpu-12 347639.1021 ( 0.00%) 352004.2254 * 1.26%* Hmean faults/cpu-21 200061.2238 ( 0.00%) 229597.0317 * 14.76%* Hmean faults/cpu-30 145251.2001 ( 0.00%) 164202.5067 * 13.05%* Hmean faults/cpu-48 106848.4434 ( 0.00%) 120641.5504 * 12.91%* Hmean faults/cpu-56 92472.3835 ( 0.00%) 103464.7916 * 11.89%* Hmean faults/sec-1 507566.1468 ( 0.00%) 506139.0811 * -0.28%* Hmean faults/sec-4 1880478.2402 ( 0.00%) 1886795.6329 * 0.34%* Hmean faults/sec-7 3106394.3438 ( 0.00%) 3140550.7485 * 1.10%* Hmean faults/sec-12 4061358.4795 ( 0.00%) 4112477.0206 * 1.26%* Hmean faults/sec-21 3988619.1169 ( 0.00%) 4577747.1436 * 14.77%* Hmean faults/sec-30 3909839.5449 ( 0.00%) 4311052.2787 * 10.26%* Hmean faults/sec-48 4761108.4691 ( 0.00%) 5283790.5026 * 10.98%* Hmean faults/sec-56 4885561.4590 ( 0.00%) 5415839.4045 * 10.85%*Given how patch 2 discusses memory growth due to moving the lock, should also patch 11 discuss how the replacement with refcount reduces the memory footprint? And/or the cover letter could summarize the impact of the whole series in that aspect? Perhaps the refcount doesn't reduce anything as it's smaller but sits alone in the cacheline? Could it be grouped with some non-hot fields instead as a followup, so could we get to <=192 (non-debug) size without impacting performance?
Quoted 79 lines
Changes since v7 [4]:
- Removed additional parameter forvma_iter_store()and introducedvma_iter_store_attached()instead, per Vlastimil Babka and Liam R. Howlett
- Fixed coding style nits, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Liam R. Howlett
- Added Acked-by, per Davidlohr Bueso
- Removed unnecessary patch changeingnommu.c
- Folded a fixup patch [5] into the patch it was fixing
- Changed calculation in__refcount_add_not_zero_limited()to avoid overflow, to change the limit to be inclusive and to useINT_MAXto indicate no limits, per Vlastimil Babka and Matthew Wilcox
- Folded a fixup patch [6] into the patch it was fixing
- Addedvm_refcntrules summary in the changelog, per Liam R. Howlett
- Changed writers to not incrementvm_refcntand adjustedVMA_REF_LIMITto not reserve one count for a writer, per Liam R. Howlett
- Changedvma_refcount_put()to wake up writers only when the last reader is leaving, per Liam R. Howlett
- Fixedrwsem_acquire_read()parameters when read-locking a vma to match the waydown_read_trylock()does lockdep, per Vlastimil Babka
- Foldedvma_lockdep_init()intovma_lock_init()for simplicity
- Brought backvma_copy()to keepvm_refcountat 0 during reuse, per Vlastimil BabkaWhat I did not include in this patchset:
- Liam's suggestion to changedump_vma()output since it's unclear to me how it should look like. The patch is for debug only and not critical for the rest of the series, we can change the output later or even drop it if necessary.[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
[4] https://lore.kernel.org/all/[email protected]/
[5] https://lore.kernel.org/all/[email protected]/
[6] https://lore.kernel.org/all/[email protected]/Patchset applies over mm-unstable after reverting v7 (current SHA range:
588f0086398e-fb2270654630)Suren Baghdasaryan (16):
mm: introduce vma_start_read_locked{_nested} helpers
mm: move per-vma lock intovm_area_struct
mm: mark vma as detached until it's added into vma tree
mm: introducevma_iter_store_attached()to use with attached vmas
mm: mark vmas detached upon exit
types: move struct rcuwait intotypes.h
mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
mm: movemmap_init_lock()out of the header file
mm: uninline the main body ofvma_start_write()
refcount: introduce __refcount_{add|inc}_not_zero_limited
mm: replacevm_lockand detached flag with a reference count mm/debug: printvm_refcntstate when dumping the vma
mm: remove extravma_numab_state_init()call
mm: preparelock_vma_under_rcu()for vma reuse possibility
mm: make vma cacheSLAB_TYPESAFE_BY_RCU
docs/mm: document latest changes tovm_lockDocumentation/mm/process_addrs.rst | 44 +++++---- include/linux/mm.h | 152 ++++++++++++++++++++++------- include/linux/mm_types.h | 36 ++++--- include/linux/mmap_lock.h | 6 -- include/linux/rcuwait.h | 13 +-- include/linux/refcount.h | 20 +++- include/linux/slab.h | 6 -- include/linux/types.h | 12 +++ kernel/fork.c | 128 +++++++++++------------- mm/debug.c | 12 +++ mm/init-mm.c | 1 + mm/memory.c | 94 +++++++++++++++--- mm/mmap.c | 3 +- mm/userfaultfd.c | 32 +++--- mm/vma.c | 23 ++--- mm/vma.h | 15 ++- tools/testing/vma/linux/atomic.h | 5 + tools/testing/vma/vma_internal.h | 93 ++++++++---------- 18 files changed, 435 insertions(+), 260 deletions(-)
Suren Baghdasaryan
maintainerQuoted 9 lines · Suren Baghdasaryan
Quoted 2 lines · Vlastimil Babka
Btw the subject became rather incomplete given all the series does :)
Missed this one. What do you think is worth mentioning here? It's true, the patchset does many small things but I wanted to outline the main conceptual changes. Please LMK if you think there are more changes big enough to be mentioned here.
I just realized that your comment was only about the subject of this cover letter. Maybe something like this:
per-vma lock and vm_area_struct cache optimizations
Would that be better?
Quoted 131 lines
Quoted 131 lines · Suren Baghdasaryan
Quoted 23 lines
Back when per-vma locks were introduces,
vm_lockwas moved out ofvm_area_structin [1] because of the performance regression caused by false cacheline sharing. Recent investigation [2] revealed that the regressions is limited to a rather old Broadwell microarchitecture and even there it can be mitigated by disabling adjacent cacheline prefetching, see [3].
Splitting single logical structure into multiple ones leads to more complicated management, extra pointer dereferences and overall less maintainable code. When that split-away part is a lock, it complicates things even further. With no performance benefits, there are no reasons for this split. Merging thevm_lockback intovm_area_structalso allowsvm_area_structto useSLAB_TYPESAFE_BY_RCUlater in this patchset. This patchset:
1. movesvm_lockback intovm_area_struct, aligning it at the cacheline boundary and changing the cache to be cacheline-aligned to minimize cacheline sharing;
2. changesvm_area_structinitialization to mark new vma as detached until it is inserted into vma tree;
3. replacesvm_lockandvma->detachedflag with a reference counter;
4. changesvm_area_structcache toSLAB_TYPESAFE_BY_RCUto allow for their reuse and to minimizecall_rcu()calls.Pagefault microbenchmarks show performance improvement:
Quoted 16 lines
Hmean faults/cpu-1 507926.5547 ( 0.00%) 506519.3692 * -0.28%* Hmean faults/cpu-4 479119.7051 ( 0.00%) 481333.6802 * 0.46%* Hmean faults/cpu-7 452880.2961 ( 0.00%) 455845.6211 * 0.65%* Hmean faults/cpu-12 347639.1021 ( 0.00%) 352004.2254 * 1.26%* Hmean faults/cpu-21 200061.2238 ( 0.00%) 229597.0317 * 14.76%* Hmean faults/cpu-30 145251.2001 ( 0.00%) 164202.5067 * 13.05%* Hmean faults/cpu-48 106848.4434 ( 0.00%) 120641.5504 * 12.91%* Hmean faults/cpu-56 92472.3835 ( 0.00%) 103464.7916 * 11.89%* Hmean faults/sec-1 507566.1468 ( 0.00%) 506139.0811 * -0.28%* Hmean faults/sec-4 1880478.2402 ( 0.00%) 1886795.6329 * 0.34%* Hmean faults/sec-7 3106394.3438 ( 0.00%) 3140550.7485 * 1.10%* Hmean faults/sec-12 4061358.4795 ( 0.00%) 4112477.0206 * 1.26%* Hmean faults/sec-21 3988619.1169 ( 0.00%) 4577747.1436 * 14.77%* Hmean faults/sec-30 3909839.5449 ( 0.00%) 4311052.2787 * 10.26%* Hmean faults/sec-48 4761108.4691 ( 0.00%) 5283790.5026 * 10.98%* Hmean faults/sec-56 4885561.4590 ( 0.00%) 5415839.4045 * 10.85%*Given how patch 2 discusses memory growth due to moving the lock, should also patch 11 discuss how the replacement with refcount reduces the memory footprint? And/or the cover letter could summarize the impact of the whole series in that aspect? Perhaps the refcount doesn't reduce anything as it's smaller but sits alone in the cacheline? Could it be grouped with some non-hot fields instead as a followup, so could we get to <=192 (non-debug) size without impacting performance?
Quoted 79 lines
Changes since v7 [4]:
- Removed additional parameter forvma_iter_store()and introducedvma_iter_store_attached()instead, per Vlastimil Babka and Liam R. Howlett
- Fixed coding style nits, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Vlastimil Babka
- Added Reviewed-bys and Acked-bys, per Liam R. Howlett
- Added Acked-by, per Davidlohr Bueso
- Removed unnecessary patch changeingnommu.c
- Folded a fixup patch [5] into the patch it was fixing
- Changed calculation in__refcount_add_not_zero_limited()to avoid overflow, to change the limit to be inclusive and to useINT_MAXto indicate no limits, per Vlastimil Babka and Matthew Wilcox
- Folded a fixup patch [6] into the patch it was fixing
- Addedvm_refcntrules summary in the changelog, per Liam R. Howlett
- Changed writers to not incrementvm_refcntand adjustedVMA_REF_LIMITto not reserve one count for a writer, per Liam R. Howlett
- Changedvma_refcount_put()to wake up writers only when the last reader is leaving, per Liam R. Howlett
- Fixedrwsem_acquire_read()parameters when read-locking a vma to match the waydown_read_trylock()does lockdep, per Vlastimil Babka
- Foldedvma_lockdep_init()intovma_lock_init()for simplicity
- Brought backvma_copy()to keepvm_refcountat 0 during reuse, per Vlastimil BabkaWhat I did not include in this patchset:
- Liam's suggestion to changedump_vma()output since it's unclear to me how it should look like. The patch is for debug only and not critical for the rest of the series, we can change the output later or even drop it if necessary.[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/ZsQyI%2F087V34JoIt@xsang-OptiPlex-9020/
[3] https://lore.kernel.org/all/CAJuCfpEisU8Lfe96AYJDZ+OM4NoPmnw9bP53cT_kbfP_pR+-2g@mail.gmail.com/
[4] https://lore.kernel.org/all/[email protected]/
[5] https://lore.kernel.org/all/[email protected]/
[6] https://lore.kernel.org/all/[email protected]/Patchset applies over mm-unstable after reverting v7 (current SHA range:
588f0086398e-fb2270654630)Suren Baghdasaryan (16):
mm: introduce vma_start_read_locked{_nested} helpers
mm: move per-vma lock intovm_area_struct
mm: mark vma as detached until it's added into vma tree
mm: introducevma_iter_store_attached()to use with attached vmas
mm: mark vmas detached upon exit
types: move struct rcuwait intotypes.h
mm: allow vma_start_read_locked/vma_start_read_locked_nested to fail
mm: movemmap_init_lock()out of the header file
mm: uninline the main body ofvma_start_write()
refcount: introduce __refcount_{add|inc}_not_zero_limited
mm: replacevm_lockand detached flag with a reference count mm/debug: printvm_refcntstate when dumping the vma
mm: remove extravma_numab_state_init()call
mm: preparelock_vma_under_rcu()for vma reuse possibility
mm: make vma cacheSLAB_TYPESAFE_BY_RCU
docs/mm: document latest changes tovm_lockDocumentation/mm/process_addrs.rst | 44 +++++---- include/linux/mm.h | 152 ++++++++++++++++++++++------- include/linux/mm_types.h | 36 ++++--- include/linux/mmap_lock.h | 6 -- include/linux/rcuwait.h | 13 +-- include/linux/refcount.h | 20 +++- include/linux/slab.h | 6 -- include/linux/types.h | 12 +++ kernel/fork.c | 128 +++++++++++------------- mm/debug.c | 12 +++ mm/init-mm.c | 1 + mm/memory.c | 94 +++++++++++++++--- mm/mmap.c | 3 +- mm/userfaultfd.c | 32 +++--- mm/vma.c | 23 ++--- mm/vma.h | 15 ++- tools/testing/vma/linux/atomic.h | 5 + tools/testing/vma/vma_internal.h | 93 ++++++++---------- 18 files changed, 435 insertions(+), 260 deletions(-)
Quoted 11 lines · Suren Baghdasaryan
Quoted 9 lines · Suren Baghdasaryan
Quoted 2 lines · Vlastimil Babka
Btw the subject became rather incomplete given all the series does :)
Missed this one. What do you think is worth mentioning here? It's true, the patchset does many small things but I wanted to outline the main conceptual changes. Please LMK if you think there are more changes big enough to be mentioned here.
Quoted 4 lines
I just realized that your comment was only about the subject of this cover letter. Maybe something like this:
per-vma lock and
vm_area_structcache optimizations
arguably the biggest change here is:
reimplement per-vma lock as a refcount
but yours is ok to, don't want to bikeshed
Suren Baghdasaryan
maintainerQuoted 15 lines · Vlastimil Babka
Quoted 11 lines · Suren Baghdasaryan
Quoted 9 lines · Suren Baghdasaryan
Quoted 2 lines · Vlastimil Babka
Btw the subject became rather incomplete given all the series does :)
Missed this one. What do you think is worth mentioning here? It's true, the patchset does many small things but I wanted to outline the main conceptual changes. Please LMK if you think there are more changes big enough to be mentioned here.
Quoted 8 lines
Quoted 4 lines
I just realized that your comment was only about the subject of this cover letter. Maybe something like this:
per-vma lock and
vm_area_structcache optimizationsarguably the biggest change here is:
reimplement per-vma lock as a refcount
Ok, I'll use that. Thanks!
Quoted 3 lines
but yours is ok to, don't want to bikeshed