Merge tag 'vfs-6.12-rc5.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs

Pull vfs fixes from Christian Brauner:
 "afs:
   - Fix a lock recursion in afs_wake_up_async_call() on ->notify_lock

 netfs:
   - Drop the references to a folio immediately after the folio has been
     extracted to prevent races with future I/O collection

   - Fix a documenation build error

   - Downgrade the i_rwsem for buffered writes to fix a cifs reported
     performance regression when switching to netfslib

  vfs:
   - Explicitly return -E2BIG from openat2() if the specified size is
     unexpectedly large. This aligns openat2() with other extensible
     struct based system calls

   - When copying a mount namespace ensure that we only try to remove
     the new copy from the mount namespace rbtree if it has already been
     added to it

  nilfs:
   - Clear the buffer delay flag when clearing the buffer state clags
     when a buffer head is discarded to prevent a kernel OOPs

  ocfs2:
   - Fix an unitialized value warning in ocfs2_setattr()

  proc:
   - Fix a kernel doc warning"

* tag 'vfs-6.12-rc5.fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs:
  proc: Fix W=1 build kernel-doc warning
  afs: Fix lock recursion
  fs: Fix uninitialized value issue in from_kuid and from_kgid
  fs: don't try and remove empty rbtree node
  netfs: Downgrade i_rwsem for a buffered write
  nilfs2: fix kernel bug due to missing clearing of buffer delay flag
  openat2: explicitly return -E2BIG for (usize > PAGE_SIZE)
  netfs: fix documentation build error
  netfs: In readahead, put the folio refs as soon extracted

10 files changed, 95 insertions, 65 deletions

diff --git a/fs/afs/internal.h b/fs/afs/internal.h
index 6e1d3c4daf72..52aab09a32a9 100644
--- a/fs/afs/internal.h
+++ b/fs/afs/internal.h
@@ -130,6 +130,7 @@ struct afs_call {
 	wait_queue_head_t	waitq;		/* processes awaiting completion */
 	struct work_struct	async_work;	/* async I/O processor */
 	struct work_struct	work;		/* actual work processor */
+	struct work_struct	free_work;	/* Deferred free processor */
 	struct rxrpc_call	*rxcall;	/* RxRPC call handle */
 	struct rxrpc_peer	*peer;		/* Remote endpoint */
 	struct key		*key;		/* security for this call */
@@ -1331,6 +1332,7 @@ extern int __net_init afs_open_socket(struct afs_net *);
 extern void __net_exit afs_close_socket(struct afs_net *);
 extern void afs_charge_preallocation(struct work_struct *);
 extern void afs_put_call(struct afs_call *);
+void afs_deferred_put_call(struct afs_call *call);
 void afs_make_call(struct afs_call *call, gfp_t gfp);
 void afs_wait_for_call_to_complete(struct afs_call *call);
 extern struct afs_call *afs_alloc_flat_call(struct afs_net *,
diff --git a/fs/afs/rxrpc.c b/fs/afs/rxrpc.c
index c453428f3c8b..9f2a3bb56ec6 100644
--- a/fs/afs/rxrpc.c
+++ b/fs/afs/rxrpc.c
@@ -18,6 +18,7 @@
 
 struct workqueue_struct *afs_async_calls;
 
+static void afs_deferred_free_worker(struct work_struct *work);
 static void afs_wake_up_call_waiter(struct sock *, struct rxrpc_call *, unsigned long);
 static void afs_wake_up_async_call(struct sock *, struct rxrpc_call *, unsigned long);
 static void afs_process_async_call(struct work_struct *);
@@ -149,6 +150,7 @@ static struct afs_call *afs_alloc_call(struct afs_net *net,
 	call->debug_id = atomic_inc_return(&rxrpc_debug_id);
 	refcount_set(&call->ref, 1);
 	INIT_WORK(&call->async_work, afs_process_async_call);
+	INIT_WORK(&call->free_work, afs_deferred_free_worker);
 	init_waitqueue_head(&call->waitq);
 	spin_lock_init(&call->state_lock);
 	call->iter = &call->def_iter;
@@ -159,6 +161,36 @@ static struct afs_call *afs_alloc_call(struct afs_net *net,
 	return call;
 }
 
+static void afs_free_call(struct afs_call *call)
+{
+	struct afs_net *net = call->net;
+	int o;
+
+	ASSERT(!work_pending(&call->async_work));
+
+	rxrpc_kernel_put_peer(call->peer);
+
+	if (call->rxcall) {
+		rxrpc_kernel_shutdown_call(net->socket, call->rxcall);
+		rxrpc_kernel_put_call(net->socket, call->rxcall);
+		call->rxcall = NULL;
+	}
+	if (call->type->destructor)
+		call->type->destructor(call);
+
+	afs_unuse_server_notime(call->net, call->server, afs_server_trace_put_call);
+	kfree(call->request);
+
+	o = atomic_read(&net->nr_outstanding_calls);
+	trace_afs_call(call->debug_id, afs_call_trace_free, 0, o,
+		       __builtin_return_address(0));
+	kfree(call);
+
+	o = atomic_dec_return(&net->nr_outstanding_calls);
+	if (o == 0)
+		wake_up_var(&net->nr_outstanding_calls);
+}
+
 /*
  * Dispose of a reference on a call.
  */
@@ -173,32 +205,34 @@ void afs_put_call(struct afs_call *call)
 	o = atomic_read(&net->nr_outstanding_calls);
 	trace_afs_call(debug_id, afs_call_trace_put, r - 1, o,
 		       __builtin_return_address(0));
+	if (zero)
+		afs_free_call(call);
+}
 
-	if (zero) {
-		ASSERT(!work_pending(&call->async_work));
-		ASSERT(call->type->name != NULL);
-
-		rxrpc_kernel_put_peer(call->peer);
-
-		if (call->rxcall) {
-			rxrpc_kernel_shutdown_call(net->socket, call->rxcall);
-			rxrpc_kernel_put_call(net->socket, call->rxcall);
-			call->rxcall = NULL;
-		}
-		if (call->type->destructor)
-			call->type->destructor(call);
+static void afs_deferred_free_worker(struct work_struct *work)
+{
+	struct afs_call *call = container_of(work, struct afs_call, free_work);
 
-		afs_unuse_server_notime(call->net, call->server, afs_server_trace_put_call);
-		kfree(call->request);
+	afs_free_call(call);
+}
 
-		trace_afs_call(call->debug_id, afs_call_trace_free, 0, o,
-			       __builtin_return_address(0));
-		kfree(call);
+/*
+ * Dispose of a reference on a call, deferring the cleanup to a workqueue
+ * to avoid lock recursion.
+ */
+void afs_deferred_put_call(struct afs_call *call)
+{
+	struct afs_net *net = call->net;
+	unsigned int debug_id = call->debug_id;
+	bool zero;
+	int r, o;
 
-		o = atomic_dec_return(&net->nr_outstanding_calls);
-		if (o == 0)
-			wake_up_var(&net->nr_outstanding_calls);
-	}
+	zero = __refcount_dec_and_test(&call->ref, &r);
+	o = atomic_read(&net->nr_outstanding_calls);
+	trace_afs_call(debug_id, afs_call_trace_put, r - 1, o,
+		       __builtin_return_address(0));
+	if (zero)
+		schedule_work(&call->free_work);
 }
 
 static struct afs_call *afs_get_call(struct afs_call *call,
@@ -640,7 +674,8 @@ static void afs_wake_up_call_waiter(struct sock *sk, struct rxrpc_call *rxcall,
 }
 
 /*
- * wake up an asynchronous call
+ * Wake up an asynchronous call.  The caller is holding the call notify
+ * spinlock around this, so we can't call afs_put_call().
  */
 static void afs_wake_up_async_call(struct sock *sk, struct rxrpc_call *rxcall,
 				   unsigned long call_user_ID)
@@ -657,7 +692,7 @@ static void afs_wake_up_async_call(struct sock *sk, struct rxrpc_call *rxcall,
 			       __builtin_return_address(0));
 
 		if (!queue_work(afs_async_calls, &call->async_work))
-			afs_put_call(call);
+			afs_deferred_put_call(call);
 	}
 }
 
diff --git a/fs/namespace.c b/fs/namespace.c
index 93c377816d75..d26f5e6d2ca3 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -3944,7 +3944,9 @@ struct mnt_namespace *copy_mnt_ns(unsigned long flags, struct mnt_namespace *ns,
 	new = copy_tree(old, old->mnt.mnt_root, copy_flags);
 	if (IS_ERR(new)) {
 		namespace_unlock();
-		free_mnt_ns(new_ns);
+		ns_free_inum(&new_ns->ns);
+		dec_mnt_namespaces(new_ns->ucounts);
+		mnt_ns_release(new_ns);
 		return ERR_CAST(new);
 	}
 	if (user_ns != ns->user_ns) {
diff --git a/fs/netfs/buffered_read.c b/fs/netfs/buffered_read.c
index c40e226053cc..af46a598f4d7 100644
--- a/fs/netfs/buffered_read.c
+++ b/fs/netfs/buffered_read.c
@@ -67,7 +67,8 @@ static int netfs_begin_cache_read(struct netfs_io_request *rreq, struct netfs_in
  * Decant the list of folios to read into a rolling buffer.
  */
 static size_t netfs_load_buffer_from_ra(struct netfs_io_request *rreq,
-					struct folio_queue *folioq)
+					struct folio_queue *folioq,
+					struct folio_batch *put_batch)
 {
 	unsigned int order, nr;
 	size_t size = 0;
@@ -82,6 +83,9 @@ static size_t netfs_load_buffer_from_ra(struct netfs_io_request *rreq,
 		order = folio_order(folio);
 		folioq->orders[i] = order;
 		size += PAGE_SIZE << order;
+
+		if (!folio_batch_add(put_batch, folio))
+			folio_batch_release(put_batch);
 	}
 
 	for (int i = nr; i < folioq_nr_slots(folioq); i++)
@@ -120,6 +124,9 @@ static ssize_t netfs_prepare_read_iterator(struct netfs_io_subrequest *subreq)
 		 * that we will need to release later - but we don't want to do
 		 * that until after we've started the I/O.
 		 */
+		struct folio_batch put_batch;
+
+		folio_batch_init(&put_batch);
 		while (rreq->submitted < subreq->start + rsize) {
 			struct folio_queue *tail = rreq->buffer_tail, *new;
 			size_t added;
@@ -132,10 +139,11 @@ static ssize_t netfs_prepare_read_iterator(struct netfs_io_subrequest *subreq)
 			new->prev = tail;
 			tail->next = new;
 			rreq->buffer_tail = new;
-			added = netfs_load_buffer_from_ra(rreq, new);
+			added = netfs_load_buffer_from_ra(rreq, new, &put_batch);
 			rreq->iter.count += added;
 			rreq->submitted += added;
 		}
+		folio_batch_release(&put_batch);
 	}
 
 	subreq->len = rsize;
@@ -348,6 +356,7 @@ static int netfs_wait_for_read(struct netfs_io_request *rreq)
 static int netfs_prime_buffer(struct netfs_io_request *rreq)
 {
 	struct folio_queue *folioq;
+	struct folio_batch put_batch;
 	size_t added;
 
 	folioq = kmalloc(sizeof(*folioq), GFP_KERNEL);
@@ -360,39 +369,14 @@ static int netfs_prime_buffer(struct netfs_io_request *rreq)
 	rreq->submitted = rreq->start;
 	iov_iter_folio_queue(&rreq->iter, ITER_DEST, folioq, 0, 0, 0);
 
-	added = netfs_load_buffer_from_ra(rreq, folioq);
+	folio_batch_init(&put_batch);
+	added = netfs_load_buffer_from_ra(rreq, folioq, &put_batch);
+	folio_batch_release(&put_batch);
 	rreq->iter.count += added;
 	rreq->submitted += added;
 	return 0;
 }
 
-/*
- * Drop the ref on each folio that we inherited from the VM readahead code.  We
- * still have the folio locks to pin the page until we complete the I/O.
- *
- * Note that we can't just release the batch in each queue struct as we use the
- * occupancy count in other places.
- */
-static void netfs_put_ra_refs(struct folio_queue *folioq)
-{
-	struct folio_batch fbatch;
-
-	folio_batch_init(&fbatch);
-	while (folioq) {
-		for (unsigned int slot = 0; slot < folioq_count(folioq); slot++) {
-			struct folio *folio = folioq_folio(folioq, slot);
-			if (!folio)
-				continue;
-			trace_netfs_folio(folio, netfs_folio_trace_read_put);
-			if (!folio_batch_add(&fbatch, folio))
-				folio_batch_release(&fbatch);
-		}
-		folioq = folioq->next;
-	}
-
-	folio_batch_release(&fbatch);
-}
-
 /**
  * netfs_readahead - Helper to manage a read request
  * @ractl: The description of the readahead request
@@ -436,9 +420,6 @@ void netfs_readahead(struct readahead_control *ractl)
 		goto cleanup_free;
 	netfs_read_to_pagecache(rreq);
 
-	/* Release the folio refs whilst we're waiting for the I/O. */
-	netfs_put_ra_refs(rreq->buffer);
-
 	netfs_put_request(rreq, true, netfs_rreq_trace_put_return);
 	return;
 
diff --git a/fs/netfs/locking.c b/fs/netfs/locking.c
index 21eab56ee2f9..2249ecd09d0a 100644
--- a/fs/netfs/locking.c
+++ b/fs/netfs/locking.c
@@ -109,6 +109,7 @@ int netfs_start_io_write(struct inode *inode)
 		up_write(&inode->i_rwsem);
 		return -ERESTARTSYS;
 	}
+	downgrade_write(&inode->i_rwsem);
 	return 0;
 }
 EXPORT_SYMBOL(netfs_start_io_write);
@@ -123,7 +124,7 @@ EXPORT_SYMBOL(netfs_start_io_write);
 void netfs_end_io_write(struct inode *inode)
 	__releases(inode->i_rwsem)
 {
-	up_write(&inode->i_rwsem);
+	up_read(&inode->i_rwsem);
 }
 EXPORT_SYMBOL(netfs_end_io_write);
 
diff --git a/fs/netfs/read_collect.c b/fs/netfs/read_collect.c
index b18c65ba5580..3cbb289535a8 100644
--- a/fs/netfs/read_collect.c
+++ b/fs/netfs/read_collect.c
@@ -77,6 +77,8 @@ static void netfs_unlock_read_folio(struct netfs_io_subrequest *subreq,
 			folio_unlock(folio);
 		}
 	}
+
+	folioq_clear(folioq, slot);
 }
 
 /*
diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c
index 9c0b7cddeaae..5436eb0424bd 100644
--- a/fs/nilfs2/page.c
+++ b/fs/nilfs2/page.c
@@ -77,7 +77,8 @@ void nilfs_forget_buffer(struct buffer_head *bh)
 	const unsigned long clear_bits =
 		(BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) |
 		 BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) |
-		 BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected));
+		 BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected) |
+		 BIT(BH_Delay));
 
 	lock_buffer(bh);
 	set_mask_bits(&bh->b_state, clear_bits, 0);
@@ -406,7 +407,8 @@ void nilfs_clear_folio_dirty(struct folio *folio)
 		const unsigned long clear_bits =
 			(BIT(BH_Uptodate) | BIT(BH_Dirty) | BIT(BH_Mapped) |
 			 BIT(BH_Async_Write) | BIT(BH_NILFS_Volatile) |
-			 BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected));
+			 BIT(BH_NILFS_Checked) | BIT(BH_NILFS_Redirected) |
+			 BIT(BH_Delay));
 
 		bh = head;
 		do {
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index ad131a2fc58e..58887456e3c5 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1129,9 +1129,12 @@ int ocfs2_setattr(struct mnt_idmap *idmap, struct dentry *dentry,
 	trace_ocfs2_setattr(inode, dentry,
 			    (unsigned long long)OCFS2_I(inode)->ip_blkno,
 			    dentry->d_name.len, dentry->d_name.name,
-			    attr->ia_valid, attr->ia_mode,
-			    from_kuid(&init_user_ns, attr->ia_uid),
-			    from_kgid(&init_user_ns, attr->ia_gid));
+			    attr->ia_valid,
+				attr->ia_valid & ATTR_MODE ? attr->ia_mode : 0,
+				attr->ia_valid & ATTR_UID ?
+					from_kuid(&init_user_ns, attr->ia_uid) : 0,
+				attr->ia_valid & ATTR_GID ?
+					from_kgid(&init_user_ns, attr->ia_gid) : 0);
 
 	/* ensuring we don't even attempt to truncate a symlink */
 	if (S_ISLNK(inode->i_mode))
diff --git a/fs/open.c b/fs/open.c
index acaeb3e25c88..5da4df2f9b18 100644
--- a/fs/open.c
+++ b/fs/open.c
@@ -1457,6 +1457,8 @@ SYSCALL_DEFINE4(openat2, int, dfd, const char __user *, filename,
 
 	if (unlikely(usize < OPEN_HOW_SIZE_VER0))
 		return -EINVAL;
+	if (unlikely(usize > PAGE_SIZE))
+		return -E2BIG;
 
 	err = copy_struct_from_user(&tmp, sizeof(tmp), how, usize);
 	if (err)
diff --git a/fs/proc/fd.c b/fs/proc/fd.c
index 1f54a54bfb91..5e391cbca7a3 100644
--- a/fs/proc/fd.c
+++ b/fs/proc/fd.c
@@ -77,7 +77,7 @@ static int seq_fdinfo_open(struct inode *inode, struct file *file)
 	return single_open(file, seq_show, inode);
 }
 
-/**
+/*
  * Shared /proc/pid/fdinfo and /proc/pid/fdinfo/fd permission helper to ensure
  * that the current task has PTRACE_MODE_READ in addition to the normal
  * POSIX-like checks.