Merge tag 'vfs-6.11.inode' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs inode / dentry updates from Christian Brauner:
 "This contains smaller performance improvements to inodes and dentries:

  inode:

   - Add rcu based inode lookup variants.

     They avoid one inode hash lock acquire in the common case thereby
     significantly reducing contention. We already support RCU-based
     operations but didn't take advantage of them during inode
     insertion.

     Callers of iget_locked() get the improvement without any code
     changes. Callers that need a custom callback can switch to
     iget5_locked_rcu() as e.g., did btrfs.

     With 20 threads each walking a dedicated 1000 dirs * 1000 files
     directory tree to stat(2) on a 32 core + 24GB ram vm:

        before: 3.54s user 892.30s system 1966% cpu 45.549 total
        after:  3.28s user 738.66s system 1955% cpu 37.932 total (-16.7%)

     Long-term we should pick up the effort to introduce more
     fine-grained locking and possibly improve on the currently used
     hash implementation.

   - Start zeroing i_state in inode_init_always() instead of doing it in
     individual filesystems.

     This allows us to remove an unneeded lock acquire in new_inode()
     and not burden individual filesystems with this.

  dcache:

   - Move d_lockref out of the area used by RCU lookup to avoid
     cacheline ping poing because the embedded name is sharing a
     cacheline with d_lockref.

   - Fix dentry size on 32bit with CONFIG_SMP=y so it does actually end
     up with 128 bytes in total"

* tag 'vfs-6.11.inode' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  fs: fix dentry size
  vfs: move d_lockref out of the area used by RCU lookup
  bcachefs: remove now spurious i_state initialization
  xfs: remove now spurious i_state initialization in xfs_inode_alloc
  vfs: partially sanitize i_state zeroing on inode creation
  xfs: preserve i_state around inode_init_always in xfs_reinit_inode
  btrfs: use iget5_locked_rcu
  vfs: add rcu-based find_inode variants for iget ops

4 files changed, 86 insertions, 30 deletions

diff --git a/fs/bcachefs/fs.c b/fs/bcachefs/fs.c
index fa1fee05cf8f..0c7d1bc0548a 100644
--- a/fs/bcachefs/fs.c
+++ b/fs/bcachefs/fs.c
@@ -244,7 +244,6 @@ static struct bch_inode_info *__bch2_new_inode(struct bch_fs *c)
 	inode->ei_flags = 0;
 	mutex_init(&inode->ei_quota_lock);
 	memset(&inode->ei_devs_need_flush, 0, sizeof(inode->ei_devs_need_flush));
-	inode->v.i_state = 0;
 
 	if (unlikely(inode_init_always(c->vfs_sb, &inode->v))) {
 		kmem_cache_free(bch2_inode_cache, inode);
diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c
index 3a2b902b2d1f..d62c96f00ff8 100644
--- a/fs/btrfs/inode.c
+++ b/fs/btrfs/inode.c
@@ -5587,7 +5587,7 @@ static struct inode *btrfs_iget_locked(struct super_block *s, u64 ino,
 	args.ino = ino;
 	args.root = root;
 
-	inode = iget5_locked(s, hashval, btrfs_find_actor,
+	inode = iget5_locked_rcu(s, hashval, btrfs_find_actor,
 			     btrfs_init_locked_inode,
 			     (void *)&args);
 	return inode;
diff --git a/fs/inode.c b/fs/inode.c
index e0815acc5abb..f356fe2ec2b6 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -162,6 +162,7 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
 	inode->i_sb = sb;
 	inode->i_blkbits = sb->s_blocksize_bits;
 	inode->i_flags = 0;
+	inode->i_state = 0;
 	atomic64_set(&inode->i_sequence, 0);
 	atomic_set(&inode->i_count, 1);
 	inode->i_op = &empty_iops;
@@ -231,6 +232,7 @@ int inode_init_always(struct super_block *sb, struct inode *inode)
 
 	if (unlikely(security_inode_alloc(inode)))
 		return -ENOMEM;
+
 	this_cpu_inc(nr_inodes);
 
 	return 0;
@@ -886,36 +888,45 @@ long prune_icache_sb(struct super_block *sb, struct shrink_control *sc)
 	return freed;
 }
 
-static void __wait_on_freeing_inode(struct inode *inode);
+static void __wait_on_freeing_inode(struct inode *inode, bool locked);
 /*
  * Called with the inode lock held.
  */
 static struct inode *find_inode(struct super_block *sb,
 				struct hlist_head *head,
 				int (*test)(struct inode *, void *),
-				void *data)
+				void *data, bool locked)
 {
 	struct inode *inode = NULL;
 
+	if (locked)
+		lockdep_assert_held(&inode_hash_lock);
+	else
+		lockdep_assert_not_held(&inode_hash_lock);
+
+	rcu_read_lock();
 repeat:
-	hlist_for_each_entry(inode, head, i_hash) {
+	hlist_for_each_entry_rcu(inode, head, i_hash) {
 		if (inode->i_sb != sb)
 			continue;
 		if (!test(inode, data))
 			continue;
 		spin_lock(&inode->i_lock);
 		if (inode->i_state & (I_FREEING|I_WILL_FREE)) {
-			__wait_on_freeing_inode(inode);
+			__wait_on_freeing_inode(inode, locked);
 			goto repeat;
 		}
 		if (unlikely(inode->i_state & I_CREATING)) {
 			spin_unlock(&inode->i_lock);
+			rcu_read_unlock();
 			return ERR_PTR(-ESTALE);
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
+		rcu_read_unlock();
 		return inode;
 	}
+	rcu_read_unlock();
 	return NULL;
 }
 
@@ -924,29 +935,39 @@ repeat:
  * iget_locked for details.
  */
 static struct inode *find_inode_fast(struct super_block *sb,
-				struct hlist_head *head, unsigned long ino)
+				struct hlist_head *head, unsigned long ino,
+				bool locked)
 {
 	struct inode *inode = NULL;
 
+	if (locked)
+		lockdep_assert_held(&inode_hash_lock);
+	else
+		lockdep_assert_not_held(&inode_hash_lock);
+
+	rcu_read_lock();
 repeat:
-	hlist_for_each_entry(inode, head, i_hash) {
+	hlist_for_each_entry_rcu(inode, head, i_hash) {
 		if (inode->i_ino != ino)
 			continue;
 		if (inode->i_sb != sb)
 			continue;
 		spin_lock(&inode->i_lock);
 		if (inode->i_state & (I_FREEING|I_WILL_FREE)) {
-			__wait_on_freeing_inode(inode);
+			__wait_on_freeing_inode(inode, locked);
 			goto repeat;
 		}
 		if (unlikely(inode->i_state & I_CREATING)) {
 			spin_unlock(&inode->i_lock);
+			rcu_read_unlock();
 			return ERR_PTR(-ESTALE);
 		}
 		__iget(inode);
 		spin_unlock(&inode->i_lock);
+		rcu_read_unlock();
 		return inode;
 	}
+	rcu_read_unlock();
 	return NULL;
 }
 
@@ -1004,14 +1025,7 @@ EXPORT_SYMBOL(get_next_ino);
  */
 struct inode *new_inode_pseudo(struct super_block *sb)
 {
-	struct inode *inode = alloc_inode(sb);
-
-	if (inode) {
-		spin_lock(&inode->i_lock);
-		inode->i_state = 0;
-		spin_unlock(&inode->i_lock);
-	}
-	return inode;
+	return alloc_inode(sb);
 }
 
 /**
@@ -1161,7 +1175,7 @@ struct inode *inode_insert5(struct inode *inode, unsigned long hashval,
 
 again:
 	spin_lock(&inode_hash_lock);
-	old = find_inode(inode->i_sb, head, test, data);
+	old = find_inode(inode->i_sb, head, test, data, true);
 	if (unlikely(old)) {
 		/*
 		 * Uhhuh, somebody else created the same inode under us.
@@ -1235,7 +1249,6 @@ struct inode *iget5_locked(struct super_block *sb, unsigned long hashval,
 		struct inode *new = alloc_inode(sb);
 
 		if (new) {
-			new->i_state = 0;
 			inode = inode_insert5(new, hashval, test, set, data);
 			if (unlikely(inode != new))
 				destroy_inode(new);
@@ -1246,6 +1259,47 @@ struct inode *iget5_locked(struct super_block *sb, unsigned long hashval,
 EXPORT_SYMBOL(iget5_locked);
 
 /**
+ * iget5_locked_rcu - obtain an inode from a mounted file system
+ * @sb:		super block of file system
+ * @hashval:	hash value (usually inode number) to get
+ * @test:	callback used for comparisons between inodes
+ * @set:	callback used to initialize a new struct inode
+ * @data:	opaque data pointer to pass to @test and @set
+ *
+ * This is equivalent to iget5_locked, except the @test callback must
+ * tolerate the inode not being stable, including being mid-teardown.
+ */
+struct inode *iget5_locked_rcu(struct super_block *sb, unsigned long hashval,
+		int (*test)(struct inode *, void *),
+		int (*set)(struct inode *, void *), void *data)
+{
+	struct hlist_head *head = inode_hashtable + hash(sb, hashval);
+	struct inode *inode, *new;
+
+again:
+	inode = find_inode(sb, head, test, data, false);
+	if (inode) {
+		if (IS_ERR(inode))
+			return NULL;
+		wait_on_inode(inode);
+		if (unlikely(inode_unhashed(inode))) {
+			iput(inode);
+			goto again;
+		}
+		return inode;
+	}
+
+	new = alloc_inode(sb);
+	if (new) {
+		inode = inode_insert5(new, hashval, test, set, data);
+		if (unlikely(inode != new))
+			destroy_inode(new);
+	}
+	return inode;
+}
+EXPORT_SYMBOL_GPL(iget5_locked_rcu);
+
+/**
  * iget_locked - obtain an inode from a mounted file system
  * @sb:		super block of file system
  * @ino:	inode number to get
@@ -1263,9 +1317,7 @@ struct inode *iget_locked(struct super_block *sb, unsigned long ino)
 	struct hlist_head *head = inode_hashtable + hash(sb, ino);
 	struct inode *inode;
 again:
-	spin_lock(&inode_hash_lock);
-	inode = find_inode_fast(sb, head, ino);
-	spin_unlock(&inode_hash_lock);
+	inode = find_inode_fast(sb, head, ino, false);
 	if (inode) {
 		if (IS_ERR(inode))
 			return NULL;
@@ -1283,7 +1335,7 @@ again:
 
 		spin_lock(&inode_hash_lock);
 		/* We released the lock, so.. */
-		old = find_inode_fast(sb, head, ino);
+		old = find_inode_fast(sb, head, ino, true);
 		if (!old) {
 			inode->i_ino = ino;
 			spin_lock(&inode->i_lock);
@@ -1419,7 +1471,7 @@ struct inode *ilookup5_nowait(struct super_block *sb, unsigned long hashval,
 	struct inode *inode;
 
 	spin_lock(&inode_hash_lock);
-	inode = find_inode(sb, head, test, data);
+	inode = find_inode(sb, head, test, data, true);
 	spin_unlock(&inode_hash_lock);
 
 	return IS_ERR(inode) ? NULL : inode;
@@ -1474,7 +1526,7 @@ struct inode *ilookup(struct super_block *sb, unsigned long ino)
 	struct inode *inode;
 again:
 	spin_lock(&inode_hash_lock);
-	inode = find_inode_fast(sb, head, ino);
+	inode = find_inode_fast(sb, head, ino, true);
 	spin_unlock(&inode_hash_lock);
 
 	if (inode) {
@@ -2235,17 +2287,21 @@ EXPORT_SYMBOL(inode_needs_sync);
  * wake_up_bit(&inode->i_state, __I_NEW) after removing from the hash list
  * will DTRT.
  */
-static void __wait_on_freeing_inode(struct inode *inode)
+static void __wait_on_freeing_inode(struct inode *inode, bool locked)
 {
 	wait_queue_head_t *wq;
 	DEFINE_WAIT_BIT(wait, &inode->i_state, __I_NEW);
 	wq = bit_waitqueue(&inode->i_state, __I_NEW);
 	prepare_to_wait(wq, &wait.wq_entry, TASK_UNINTERRUPTIBLE);
 	spin_unlock(&inode->i_lock);
-	spin_unlock(&inode_hash_lock);
+	rcu_read_unlock();
+	if (locked)
+		spin_unlock(&inode_hash_lock);
 	schedule();
 	finish_wait(wq, &wait.wq_entry);
-	spin_lock(&inode_hash_lock);
+	if (locked)
+		spin_lock(&inode_hash_lock);
+	rcu_read_lock();
 }
 
 static __initdata unsigned long ihash_entries;
diff --git a/fs/xfs/xfs_icache.c b/fs/xfs/xfs_icache.c
index 9967334ea99f..cf629302d48e 100644
--- a/fs/xfs/xfs_icache.c
+++ b/fs/xfs/xfs_icache.c
@@ -86,9 +86,8 @@ xfs_inode_alloc(
 		return NULL;
 	}
 
-	/* VFS doesn't initialise i_mode or i_state! */
+	/* VFS doesn't initialise i_mode! */
 	VFS_I(ip)->i_mode = 0;
-	VFS_I(ip)->i_state = 0;
 	mapping_set_large_folios(VFS_I(ip)->i_mapping);
 
 	XFS_STATS_INC(mp, vn_active);
@@ -314,6 +313,7 @@ xfs_reinit_inode(
 	dev_t			dev = inode->i_rdev;
 	kuid_t			uid = inode->i_uid;
 	kgid_t			gid = inode->i_gid;
+	unsigned long		state = inode->i_state;
 
 	error = inode_init_always(mp->m_super, inode);
 
@@ -324,6 +324,7 @@ xfs_reinit_inode(
 	inode->i_rdev = dev;
 	inode->i_uid = uid;
 	inode->i_gid = gid;
+	inode->i_state = state;
 	mapping_set_large_folios(inode->i_mapping);
 	return error;
 }