[PATCH v3 3/3] fs/proc/task_mmu: execute PROCMAP_QUERY ioctl under per-vma locks

[Suren Baghdasaryan <surenb@google.com>]

Utilize per-vma locks to stabilize vma after lookup without taking
mmap_lock during PROCMAP_QUERY ioctl execution. If vma lock is
contended, we fall back to mmap_lock but take it only momentarily
to lock the vma and release the mmap_lock. In a very unlikely case
of vm_refcnt overflow, this fall back path will fail and ioctl is
done under mmap_lock protection.

This change is designed to reduce mmap_lock contention and prevent
PROCMAP_QUERY ioctl calls from blocking address space updates.

Signed-off-by: Suren Baghdasaryan <surenb@google.com>
---
 fs/proc/task_mmu.c | 84 +++++++++++++++++++++++++++++++++++++---------
 1 file changed, 68 insertions(+), 16 deletions(-)

diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index 45134335e086..0396315dfaee 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -517,28 +517,81 @@ static int pid_maps_open(struct inode *inode, struct file *file)
 		PROCMAP_QUERY_VMA_FLAGS				\
 )
 
-static int query_vma_setup(struct mm_struct *mm)
+#ifdef CONFIG_PER_VMA_LOCK
+
+static int query_vma_setup(struct proc_maps_locking_ctx *lock_ctx)
 {
-	return mmap_read_lock_killable(mm);
+	lock_ctx->locked_vma = NULL;
+	lock_ctx->mmap_locked = false;
+
+	return 0;
 }
 
-static void query_vma_teardown(struct mm_struct *mm, struct vm_area_struct *vma)
+static void query_vma_teardown(struct proc_maps_locking_ctx *lock_ctx)
 {
-	mmap_read_unlock(mm);
+	if (lock_ctx->mmap_locked)
+		mmap_read_unlock(lock_ctx->mm);
+	else
+		unlock_vma(lock_ctx);
+}
+
+static struct vm_area_struct *query_vma_find_by_addr(struct proc_maps_locking_ctx *lock_ctx,
+						     unsigned long addr)
+{
+	struct vm_area_struct *vma;
+	struct vma_iterator vmi;
+
+	if (lock_ctx->mmap_locked)
+		return find_vma(lock_ctx->mm, addr);
+
+	unlock_vma(lock_ctx);
+	rcu_read_lock();
+	vma_iter_init(&vmi, lock_ctx->mm, addr);
+	vma = lock_next_vma(lock_ctx->mm, &vmi, addr);
+	rcu_read_unlock();
+
+	if (!IS_ERR_OR_NULL(vma)) {
+		lock_ctx->locked_vma = vma;
+	} else if (PTR_ERR(vma) == -EAGAIN) {
+		/* Fallback to mmap_lock on vma->vm_refcnt overflow */
+		mmap_read_lock(lock_ctx->mm);
+		vma = find_vma(lock_ctx->mm, addr);
+		lock_ctx->mmap_locked = true;
+	}
+
+	return vma;
+}
+
+#else /* CONFIG_PER_VMA_LOCK */
+
+static int query_vma_setup(struct proc_maps_locking_ctx *lock_ctx)
+{
+	return mmap_read_lock_killable(lock_ctx->mm);
 }
 
-static struct vm_area_struct *query_vma_find_by_addr(struct mm_struct *mm, unsigned long addr)
+static void query_vma_teardown(struct proc_maps_locking_ctx *lock_ctx)
 {
-	return find_vma(mm, addr);
+	mmap_read_unlock(lock_ctx->mm);
 }
 
-static struct vm_area_struct *query_matching_vma(struct mm_struct *mm,
+static struct vm_area_struct *query_vma_find_by_addr(struct proc_maps_locking_ctx *lock_ctx,
+						     unsigned long addr)
+{
+	return find_vma(lock_ctx->mm, addr);
+}
+
+#endif  /* CONFIG_PER_VMA_LOCK */
+
+static struct vm_area_struct *query_matching_vma(struct proc_maps_locking_ctx *lock_ctx,
 						 unsigned long addr, u32 flags)
 {
 	struct vm_area_struct *vma;
 
 next_vma:
-	vma = query_vma_find_by_addr(mm, addr);
+	vma = query_vma_find_by_addr(lock_ctx, addr);
+	if (IS_ERR(vma))
+		return vma;
+
 	if (!vma)
 		goto no_vma;
 
@@ -579,11 +632,11 @@ static struct vm_area_struct *query_matching_vma(struct mm_struct *mm,
 	return ERR_PTR(-ENOENT);
 }
 
-static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg)
+static int do_procmap_query(struct mm_struct *mm, void __user *uarg)
 {
+	struct proc_maps_locking_ctx lock_ctx = { .mm = mm };
 	struct procmap_query karg;
 	struct vm_area_struct *vma;
-	struct mm_struct *mm;
 	const char *name = NULL;
 	char build_id_buf[BUILD_ID_SIZE_MAX], *name_buf = NULL;
 	__u64 usize;
@@ -610,17 +663,16 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg)
 	if (!!karg.build_id_size != !!karg.build_id_addr)
 		return -EINVAL;
 
-	mm = priv->lock_ctx.mm;
 	if (!mm || !mmget_not_zero(mm))
 		return -ESRCH;
 
-	err = query_vma_setup(mm);
+	err = query_vma_setup(&lock_ctx);
 	if (err) {
 		mmput(mm);
 		return err;
 	}
 
-	vma = query_matching_vma(mm, karg.query_addr, karg.query_flags);
+	vma = query_matching_vma(&lock_ctx, karg.query_addr, karg.query_flags);
 	if (IS_ERR(vma)) {
 		err = PTR_ERR(vma);
 		vma = NULL;
@@ -705,7 +757,7 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg)
 	}
 
 	/* unlock vma or mmap_lock, and put mm_struct before copying data to user */
-	query_vma_teardown(mm, vma);
+	query_vma_teardown(&lock_ctx);
 	mmput(mm);
 
 	if (karg.vma_name_size && copy_to_user(u64_to_user_ptr(karg.vma_name_addr),
@@ -725,7 +777,7 @@ static int do_procmap_query(struct proc_maps_private *priv, void __user *uarg)
 	return 0;
 
 out:
-	query_vma_teardown(mm, vma);
+	query_vma_teardown(&lock_ctx);
 	mmput(mm);
 	kfree(name_buf);
 	return err;
@@ -738,7 +790,7 @@ static long procfs_procmap_ioctl(struct file *file, unsigned int cmd, unsigned l
 
 	switch (cmd) {
 	case PROCMAP_QUERY:
-		return do_procmap_query(priv, (void __user *)arg);
+		return do_procmap_query(priv->lock_ctx.mm, (void __user *)arg);
 	default:
 		return -ENOIOCTLCMD;
 	}
-- 
2.50.1.565.gc32cd1483b-goog