From: Suren Baghdasaryan <surenb@google.com>
To: akpm@linux-foundation.org
Cc: michel@lespinasse.org, jglisse@google.com, mhocko@suse.com,
vbabka@suse.cz, hannes@cmpxchg.org, mgorman@techsingularity.net,
dave@stgolabs.net, willy@infradead.org, liam.howlett@oracle.com,
peterz@infradead.org, ldufour@linux.ibm.com, paulmck@kernel.org,
mingo@redhat.com, will@kernel.org, luto@kernel.org,
songliubraving@fb.com, peterx@redhat.com, david@redhat.com,
dhowells@redhat.com, hughd@google.com, bigeasy@linutronix.de,
kent.overstreet@linux.dev, punit.agrawal@bytedance.com,
lstoakes@gmail.com, peterjung1337@gmail.com, rientjes@google.com,
chriscli@google.com, axelrasmussen@google.com, joelaf@google.com,
minchan@google.com, rppt@kernel.org, jannh@google.com,
shakeelb@google.com, tatashin@google.com, edumazet@google.com,
gthelen@google.com, gurua@google.com, arjunroy@google.com,
soheil@google.com, leewalsh@google.com, posk@google.com,
michalechner92@googlemail.com, linux-mm@kvack.org,
linux-arm-kernel@lists.infradead.org,
linuxppc-dev@lists.ozlabs.org, x86@kernel.org,
linux-kernel@vger.kernel.org, kernel-team@android.com,
Suren Baghdasaryan <surenb@google.com>
Subject: [PATCH v3 12/35] mm: add per-VMA lock and helper functions to control it
Date: Wed, 15 Feb 2023 21:17:27 -0800 [thread overview]
Message-ID: <20230216051750.3125598-13-surenb@google.com> (raw)
In-Reply-To: <20230216051750.3125598-1-surenb@google.com>
Introduce per-VMA locking. The lock implementation relies on a
per-vma and per-mm sequence counters to note exclusive locking:
- read lock - (implemented by vma_start_read) requires the vma
(vm_lock_seq) and mm (mm_lock_seq) sequence counters to differ.
If they match then there must be a vma exclusive lock held somewhere.
- read unlock - (implemented by vma_end_read) is a trivial vma->lock
unlock.
- write lock - (vma_start_write) requires the mmap_lock to be held
exclusively and the current mm counter is assigned to the vma counter.
This will allow multiple vmas to be locked under a single mmap_lock
write lock (e.g. during vma merging). The vma counter is modified
under exclusive vma lock.
- write unlock - (vma_end_write_all) is a batch release of all vma
locks held. It doesn't pair with a specific vma_start_write! It is
done before exclusive mmap_lock is released by incrementing mm
sequence counter (mm_lock_seq).
- write downgrade - if the mmap_lock is downgraded to the read lock, all
vma write locks are released as well (effectivelly same as write
unlock).
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
---
include/linux/mm.h | 82 +++++++++++++++++++++++++++++++++++++++
include/linux/mm_types.h | 8 ++++
include/linux/mmap_lock.h | 13 +++++++
kernel/fork.c | 4 ++
mm/init-mm.c | 3 ++
5 files changed, 110 insertions(+)
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 2992a2d55aee..a056ee170e34 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -623,6 +623,87 @@ struct vm_operations_struct {
unsigned long addr);
};
+#ifdef CONFIG_PER_VMA_LOCK
+static inline void vma_init_lock(struct vm_area_struct *vma)
+{
+ init_rwsem(&vma->lock);
+ vma->vm_lock_seq = -1;
+}
+
+/*
+ * 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.
+ */
+static inline bool vma_start_read(struct vm_area_struct *vma)
+{
+ /* Check before locking. A race might cause false locked result. */
+ if (vma->vm_lock_seq == READ_ONCE(vma->vm_mm->mm_lock_seq))
+ return false;
+
+ if (unlikely(down_read_trylock(&vma->lock) == 0))
+ return false;
+
+ /*
+ * Overflow might produce false locked result.
+ * False unlocked result is impossible because we modify and check
+ * vma->vm_lock_seq under vma->lock protection and mm->mm_lock_seq
+ * modification invalidates all existing locks.
+ */
+ if (unlikely(vma->vm_lock_seq == READ_ONCE(vma->vm_mm->mm_lock_seq))) {
+ up_read(&vma->lock);
+ return false;
+ }
+ return true;
+}
+
+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->lock);
+ rcu_read_unlock();
+}
+
+static inline void vma_start_write(struct vm_area_struct *vma)
+{
+ int mm_lock_seq;
+
+ mmap_assert_write_locked(vma->vm_mm);
+
+ /*
+ * current task is holding mmap_write_lock, both vma->vm_lock_seq and
+ * mm->mm_lock_seq can't be concurrently modified.
+ */
+ mm_lock_seq = READ_ONCE(vma->vm_mm->mm_lock_seq);
+ if (vma->vm_lock_seq == mm_lock_seq)
+ return;
+
+ down_write(&vma->lock);
+ vma->vm_lock_seq = mm_lock_seq;
+ up_write(&vma->lock);
+}
+
+static inline void vma_assert_write_locked(struct vm_area_struct *vma)
+{
+ mmap_assert_write_locked(vma->vm_mm);
+ /*
+ * current task is holding mmap_write_lock, both vma->vm_lock_seq and
+ * mm->mm_lock_seq can't be concurrently modified.
+ */
+ VM_BUG_ON_VMA(vma->vm_lock_seq != READ_ONCE(vma->vm_mm->mm_lock_seq), vma);
+}
+
+#else /* CONFIG_PER_VMA_LOCK */
+
+static inline void vma_init_lock(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) {}
+static inline void vma_start_write(struct vm_area_struct *vma) {}
+static inline void vma_assert_write_locked(struct vm_area_struct *vma) {}
+
+#endif /* CONFIG_PER_VMA_LOCK */
+
static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm)
{
static const struct vm_operations_struct dummy_vm_ops = {};
@@ -631,6 +712,7 @@ static inline void vma_init(struct vm_area_struct *vma, struct mm_struct *mm)
vma->vm_mm = mm;
vma->vm_ops = &dummy_vm_ops;
INIT_LIST_HEAD(&vma->anon_vma_chain);
+ vma_init_lock(vma);
}
/* Use when VMA is not part of the VMA tree and needs no locking */
diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
index fb4e2afad787..e268723eaf44 100644
--- a/include/linux/mm_types.h
+++ b/include/linux/mm_types.h
@@ -508,6 +508,11 @@ struct vm_area_struct {
vm_flags_t __private __vm_flags;
};
+#ifdef CONFIG_PER_VMA_LOCK
+ int vm_lock_seq;
+ struct rw_semaphore lock;
+#endif
+
/*
* For areas with an address space and backing store,
* linkage into the address_space->i_mmap interval tree.
@@ -633,6 +638,9 @@ struct mm_struct {
* init_mm.mmlist, and are protected
* by mmlist_lock
*/
+#ifdef CONFIG_PER_VMA_LOCK
+ int mm_lock_seq;
+#endif
unsigned long hiwater_rss; /* High-watermark of RSS usage */
diff --git a/include/linux/mmap_lock.h b/include/linux/mmap_lock.h
index e49ba91bb1f0..aab8f1b28d26 100644
--- a/include/linux/mmap_lock.h
+++ b/include/linux/mmap_lock.h
@@ -72,6 +72,17 @@ static inline void mmap_assert_write_locked(struct mm_struct *mm)
VM_BUG_ON_MM(!rwsem_is_locked(&mm->mmap_lock), mm);
}
+#ifdef CONFIG_PER_VMA_LOCK
+static inline void vma_end_write_all(struct mm_struct *mm)
+{
+ mmap_assert_write_locked(mm);
+ /* No races during update due to exclusive mmap_lock being held */
+ WRITE_ONCE(mm->mm_lock_seq, mm->mm_lock_seq + 1);
+}
+#else
+static inline void vma_end_write_all(struct mm_struct *mm) {}
+#endif
+
static inline void mmap_init_lock(struct mm_struct *mm)
{
init_rwsem(&mm->mmap_lock);
@@ -114,12 +125,14 @@ static inline bool mmap_write_trylock(struct mm_struct *mm)
static inline void mmap_write_unlock(struct mm_struct *mm)
{
__mmap_lock_trace_released(mm, true);
+ vma_end_write_all(mm);
up_write(&mm->mmap_lock);
}
static inline void mmap_write_downgrade(struct mm_struct *mm)
{
__mmap_lock_trace_acquire_returned(mm, false, true);
+ vma_end_write_all(mm);
downgrade_write(&mm->mmap_lock);
}
diff --git a/kernel/fork.c b/kernel/fork.c
index 314d51eb91da..9141427a98b2 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -474,6 +474,7 @@ struct vm_area_struct *vm_area_dup(struct vm_area_struct *orig)
*/
data_race(memcpy(new, orig, sizeof(*new)));
INIT_LIST_HEAD(&new->anon_vma_chain);
+ vma_init_lock(new);
dup_anon_vma_name(orig, new);
}
return new;
@@ -1147,6 +1148,9 @@ static struct mm_struct *mm_init(struct mm_struct *mm, struct task_struct *p,
seqcount_init(&mm->write_protect_seq);
mmap_init_lock(mm);
INIT_LIST_HEAD(&mm->mmlist);
+#ifdef CONFIG_PER_VMA_LOCK
+ mm->mm_lock_seq = 0;
+#endif
mm_pgtables_bytes_init(mm);
mm->map_count = 0;
mm->locked_vm = 0;
diff --git a/mm/init-mm.c b/mm/init-mm.c
index c9327abb771c..33269314e060 100644
--- a/mm/init-mm.c
+++ b/mm/init-mm.c
@@ -37,6 +37,9 @@ struct mm_struct init_mm = {
.page_table_lock = __SPIN_LOCK_UNLOCKED(init_mm.page_table_lock),
.arg_lock = __SPIN_LOCK_UNLOCKED(init_mm.arg_lock),
.mmlist = LIST_HEAD_INIT(init_mm.mmlist),
+#ifdef CONFIG_PER_VMA_LOCK
+ .mm_lock_seq = 0,
+#endif
.user_ns = &init_user_ns,
.cpu_bitmap = CPU_BITS_NONE,
#ifdef CONFIG_IOMMU_SVA
--
2.39.1
next prev parent reply other threads:[~2023-02-16 5:19 UTC|newest]
Thread overview: 67+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-16 5:17 [PATCH v3 00/35] Per-VMA locks Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 01/35] maple_tree: Be more cautious about dead nodes Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 02/35] maple_tree: Detect dead nodes in mas_start() Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 03/35] maple_tree: Fix freeing of nodes in rcu mode Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 04/35] maple_tree: remove extra smp_wmb() from mas_dead_leaves() Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 05/35] maple_tree: Fix write memory barrier of nodes once dead for RCU mode Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 06/35] maple_tree: Add smp_rmb() to dead node detection Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 07/35] maple_tree: Add RCU lock checking to rcu callback functions Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 08/35] mm: Enable maple tree RCU mode by default Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 09/35] mm: introduce CONFIG_PER_VMA_LOCK Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 10/35] mm: rcu safe VMA freeing Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 11/35] mm: move mmap_lock assert function definitions Suren Baghdasaryan
2023-02-16 5:17 ` Suren Baghdasaryan [this message]
2023-02-16 5:17 ` [PATCH v3 13/35] mm: mark VMA as being written when changing vm_flags Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 14/35] mm/mmap: move VMA locking before vma_adjust_trans_huge call Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 15/35] mm/khugepaged: write-lock VMA while collapsing a huge page Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 16/35] mm/mmap: write-lock VMAs before merging, splitting or expanding them Suren Baghdasaryan
2023-02-23 14:51 ` Hyeonggon Yoo
2023-02-23 14:59 ` Hyeonggon Yoo
2023-02-23 17:46 ` Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 17/35] mm/mmap: write-lock VMA before shrinking or expanding it Suren Baghdasaryan
2023-02-23 20:20 ` Liam R. Howlett
2023-02-23 20:28 ` Liam R. Howlett
2023-02-23 21:16 ` Suren Baghdasaryan
2023-02-24 1:46 ` Liam R. Howlett
2023-02-24 2:06 ` Suren Baghdasaryan
2023-02-24 16:14 ` Liam R. Howlett
2023-02-24 16:19 ` Suren Baghdasaryan
2023-02-27 17:33 ` Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 18/35] mm/mremap: write-lock VMA while remapping it to a new address range Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 19/35] mm: write-lock VMAs before removing them from VMA tree Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 20/35] mm: conditionally write-lock VMA in free_pgtables Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 21/35] mm/mmap: write-lock adjacent VMAs if they can grow into unmapped area Suren Baghdasaryan
2023-02-16 15:34 ` Liam R. Howlett
2023-02-16 19:36 ` Suren Baghdasaryan
2023-02-17 14:50 ` Liam R. Howlett
2023-02-17 15:54 ` Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 22/35] kernel/fork: assert no VMA readers during its destruction Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 23/35] mm/mmap: prevent pagefault handler from racing with mmu_notifier registration Suren Baghdasaryan
2023-02-23 20:06 ` Liam R. Howlett
2023-02-23 20:29 ` Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 24/35] mm: introduce vma detached flag Suren Baghdasaryan
2023-02-23 20:08 ` Liam R. Howlett
2023-02-23 20:34 ` Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 25/35] mm: introduce lock_vma_under_rcu to be used from arch-specific code Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 26/35] mm: fall back to mmap_lock if vma->anon_vma is not yet set Suren Baghdasaryan
2023-02-16 15:44 ` Matthew Wilcox
2023-02-16 19:43 ` Suren Baghdasaryan
2023-02-17 2:14 ` Suren Baghdasaryan
2023-02-17 10:21 ` Hyeonggon Yoo
2023-02-17 16:13 ` Suren Baghdasaryan
2023-02-17 18:49 ` Hyeonggon Yoo
2023-02-17 16:05 ` Matthew Wilcox
2023-02-17 16:10 ` Suren Baghdasaryan
2023-04-03 19:49 ` Matthew Wilcox
2023-02-16 5:17 ` [PATCH v3 27/35] mm: add FAULT_FLAG_VMA_LOCK flag Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 28/35] mm: prevent do_swap_page from handling page faults under VMA lock Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 29/35] mm: prevent userfaults to be handled under per-vma lock Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 30/35] mm: introduce per-VMA lock statistics Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 31/35] x86/mm: try VMA lock-based page fault handling first Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 32/35] arm64/mm: " Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 33/35] powerc/mm: " Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 34/35] mm/mmap: free vm_area_struct without call_rcu in exit_mmap Suren Baghdasaryan
2023-02-16 5:17 ` [PATCH v3 35/35] mm: separate vma->lock from vm_area_struct Suren Baghdasaryan
2023-02-24 9:21 ` [PATCH v3 00/35] Per-VMA locks freak07
2023-02-27 16:50 ` Davidlohr Bueso
2023-02-27 17:22 ` Suren Baghdasaryan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20230216051750.3125598-13-surenb@google.com \
--to=surenb@google.com \
--cc=akpm@linux-foundation.org \
--cc=arjunroy@google.com \
--cc=axelrasmussen@google.com \
--cc=bigeasy@linutronix.de \
--cc=chriscli@google.com \
--cc=dave@stgolabs.net \
--cc=david@redhat.com \
--cc=dhowells@redhat.com \
--cc=edumazet@google.com \
--cc=gthelen@google.com \
--cc=gurua@google.com \
--cc=hannes@cmpxchg.org \
--cc=hughd@google.com \
--cc=jannh@google.com \
--cc=jglisse@google.com \
--cc=joelaf@google.com \
--cc=kent.overstreet@linux.dev \
--cc=kernel-team@android.com \
--cc=ldufour@linux.ibm.com \
--cc=leewalsh@google.com \
--cc=liam.howlett@oracle.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=lstoakes@gmail.com \
--cc=luto@kernel.org \
--cc=mgorman@techsingularity.net \
--cc=mhocko@suse.com \
--cc=michalechner92@googlemail.com \
--cc=michel@lespinasse.org \
--cc=minchan@google.com \
--cc=mingo@redhat.com \
--cc=paulmck@kernel.org \
--cc=peterjung1337@gmail.com \
--cc=peterx@redhat.com \
--cc=peterz@infradead.org \
--cc=posk@google.com \
--cc=punit.agrawal@bytedance.com \
--cc=rientjes@google.com \
--cc=rppt@kernel.org \
--cc=shakeelb@google.com \
--cc=soheil@google.com \
--cc=songliubraving@fb.com \
--cc=tatashin@google.com \
--cc=vbabka@suse.cz \
--cc=will@kernel.org \
--cc=willy@infradead.org \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).