1 files changed, 63 insertions, 108 deletions

diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c
index e9f236af1927..79b281e32f4c 100644
--- a/fs/ocfs2/cluster/tcp.c
+++ b/fs/ocfs2/cluster/tcp.c
@@ -1,33 +1,17 @@
-/* -*- mode: c; c-basic-offset: 8; -*-
- *
- * vim: noexpandtab sw=8 ts=8 sts=0:
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
  *
  * Copyright (C) 2004 Oracle.  All rights reserved.
  *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public
- * License as published by the Free Software Foundation; either
- * version 2 of the License, or (at your option) any later version.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
- * General Public License for more details.
- *
- * You should have received a copy of the GNU General Public
- * License along with this program; if not, write to the
- * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
- * Boston, MA 021110-1307, USA.
- *
  * ----
  *
- * Callers for this were originally written against a very simple synchronus
+ * Callers for this were originally written against a very simple synchronous
  * API.  This implementation reflects those simple callers.  Some day I'm sure
  * we'll need to move to a more robust posting/callback mechanism.
  *
  * Transmit calls pass in kernel virtual addresses and block copying this into
  * the socket's tx buffers via a usual blocking sendmsg.  They'll block waiting
- * for a failed socket to timeout.  TX callers can also pass in a poniter to an
+ * for a failed socket to timeout.  TX callers can also pass in a pointer to an
  * 'int' which gets filled with an errno off the wire in response to the
  * message they send.
  *
@@ -62,6 +46,7 @@
 #include <linux/net.h>
 #include <linux/export.h>
 #include <net/tcp.h>
+#include <trace/events/sock.h>
 
 #include <linux/uaccess.h>
 
@@ -116,7 +101,7 @@ static struct socket *o2net_listen_sock;
  * o2net_wq.  teardown detaches the callbacks before destroying the workqueue.
  * quorum work is queued as sock containers are shutdown.. stop_listening
  * tears down all the node's sock containers, preventing future shutdowns
- * and queued quroum work, before canceling delayed quorum work and
+ * and queued quorum work, before canceling delayed quorum work and
  * destroying the work queue.
  */
 static struct workqueue_struct *o2net_wq;
@@ -601,6 +586,8 @@ static void o2net_data_ready(struct sock *sk)
 	void (*ready)(struct sock *sk);
 	struct o2net_sock_container *sc;
 
+	trace_sk_data_ready(sk);
+
 	read_lock_bh(&sk->sk_callback_lock);
 	sc = sk->sk_user_data;
 	if (sc) {
@@ -737,7 +724,7 @@ static void o2net_shutdown_sc(struct work_struct *work)
 	if (o2net_unregister_callbacks(sc->sc_sock->sk, sc)) {
 		/* we shouldn't flush as we're in the thread, the
 		 * races with pending sc work structs are harmless */
-		del_timer_sync(&sc->sc_idle_timeout);
+		timer_delete_sync(&sc->sc_idle_timeout);
 		o2net_sc_cancel_delayed_work(sc, &sc->sc_keepalive_work);
 		sc_put(sc);
 		kernel_sock_shutdown(sc->sc_sock, SHUT_RDWR);
@@ -916,7 +903,7 @@ static int o2net_recv_tcp_msg(struct socket *sock, void *data, size_t len)
 {
 	struct kvec vec = { .iov_len = len, .iov_base = data, };
 	struct msghdr msg = { .msg_flags = MSG_DONTWAIT, };
-	iov_iter_kvec(&msg.msg_iter, READ, &vec, 1, len);
+	iov_iter_kvec(&msg.msg_iter, ITER_DEST, &vec, 1, len);
 	return sock_recvmsg(sock, &msg, MSG_DONTWAIT);
 }
 
@@ -943,19 +930,22 @@ out:
 }
 
 static void o2net_sendpage(struct o2net_sock_container *sc,
-			   void *kmalloced_virt,
-			   size_t size)
+			   void *virt, size_t size)
 {
 	struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
+	struct msghdr msg = {};
+	struct bio_vec bv;
 	ssize_t ret;
 
+	bvec_set_virt(&bv, virt, size);
+	iov_iter_bvec(&msg.msg_iter, ITER_SOURCE, &bv, 1, size);
+
 	while (1) {
+		msg.msg_flags = MSG_DONTWAIT | MSG_SPLICE_PAGES;
 		mutex_lock(&sc->sc_send_lock);
-		ret = sc->sc_sock->ops->sendpage(sc->sc_sock,
-						 virt_to_page(kmalloced_virt),
-						 offset_in_page(kmalloced_virt),
-						 size, MSG_DONTWAIT);
+		ret = sock_sendmsg(sc->sc_sock, &msg);
 		mutex_unlock(&sc->sc_send_lock);
+
 		if (ret == size)
 			break;
 		if (ret == (ssize_t)-EAGAIN) {
@@ -1006,14 +996,12 @@ static int o2net_tx_can_proceed(struct o2net_node *nn,
 }
 
 /* Get a map of all nodes to which this node is currently connected to */
-void o2net_fill_node_map(unsigned long *map, unsigned bytes)
+void o2net_fill_node_map(unsigned long *map, unsigned int bits)
 {
 	struct o2net_sock_container *sc;
 	int node, ret;
 
-	BUG_ON(bytes < (BITS_TO_LONGS(O2NM_MAX_NODES) * sizeof(unsigned long)));
-
-	memset(map, 0, bytes);
+	bitmap_zero(map, bits);
 	for (node = 0; node < O2NM_MAX_NODES; ++node) {
 		if (!o2net_tx_can_proceed(o2net_nn_from_num(node), &sc, &ret))
 			continue;
@@ -1212,7 +1200,6 @@ static int o2net_process_message(struct o2net_sock_container *sc,
 			msglog(hdr, "bad magic\n");
 			ret = -EINVAL;
 			goto out;
-			break;
 	}
 
 	/* find a handler for it */
@@ -1432,7 +1419,7 @@ out:
 	return ret;
 }
 
-/* this work func is triggerd by data ready.  it reads until it can read no
+/* this work func is triggered by data ready.  it reads until it can read no
  * more.  it interprets 0, eof, as fatal.  if data_ready hits while we're doing
  * our work the work struct will be marked and we'll be called again. */
 static void o2net_rx_until_empty(struct work_struct *work)
@@ -1455,22 +1442,6 @@ static void o2net_rx_until_empty(struct work_struct *work)
 	sc_put(sc);
 }
 
-static int o2net_set_nodelay(struct socket *sock)
-{
-	int val = 1;
-
-	return kernel_setsockopt(sock, SOL_TCP, TCP_NODELAY,
-				    (void *)&val, sizeof(val));
-}
-
-static int o2net_set_usertimeout(struct socket *sock)
-{
-	int user_timeout = O2NET_TCP_USER_TIMEOUT;
-
-	return kernel_setsockopt(sock, SOL_TCP, TCP_USER_TIMEOUT,
-				(void *)&user_timeout, sizeof(user_timeout));
-}
-
 static void o2net_initialize_handshake(void)
 {
 	o2net_hand->o2hb_heartbeat_timeout_ms = cpu_to_be32(
@@ -1512,12 +1483,13 @@ static void o2net_sc_send_keep_req(struct work_struct *work)
 	sc_put(sc);
 }
 
-/* socket shutdown does a del_timer_sync against this as it tears down.
+/* socket shutdown does a timer_delete_sync against this as it tears down.
  * we can't start this timer until we've got to the point in sc buildup
  * where shutdown is going to be involved */
 static void o2net_idle_timer(struct timer_list *t)
 {
-	struct o2net_sock_container *sc = from_timer(sc, t, sc_idle_timeout);
+	struct o2net_sock_container *sc = timer_container_of(sc, t,
+							     sc_idle_timeout);
 	struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
 #ifdef CONFIG_DEBUG_FS
 	unsigned long msecs = ktime_to_ms(ktime_get()) -
@@ -1584,15 +1556,13 @@ static void o2net_start_connect(struct work_struct *work)
 	struct sockaddr_in myaddr = {0, }, remoteaddr = {0, };
 	int ret = 0, stop;
 	unsigned int timeout;
-	unsigned int noio_flag;
+	unsigned int nofs_flag;
 
 	/*
-	 * sock_create allocates the sock with GFP_KERNEL. We must set
-	 * per-process flag PF_MEMALLOC_NOIO so that all allocations done
-	 * by this process are done as if GFP_NOIO was specified. So we
-	 * are not reentering filesystem while doing memory reclaim.
+	 * sock_create allocates the sock with GFP_KERNEL. We must
+	 * prevent the filesystem from being reentered by memory reclaim.
 	 */
-	noio_flag = memalloc_noio_save();
+	nofs_flag = memalloc_nofs_save();
 	/* if we're greater we initiate tx, otherwise we accept */
 	if (o2nm_this_node() <= o2net_num_from_nn(nn))
 		goto out;
@@ -1639,12 +1609,13 @@ static void o2net_start_connect(struct work_struct *work)
 	sc->sc_sock = sock; /* freed by sc_kref_release */
 
 	sock->sk->sk_allocation = GFP_ATOMIC;
+	sock->sk->sk_use_task_frag = false;
 
 	myaddr.sin_family = AF_INET;
 	myaddr.sin_addr.s_addr = mynode->nd_ipv4_address;
 	myaddr.sin_port = htons(0); /* any port */
 
-	ret = sock->ops->bind(sock, (struct sockaddr *)&myaddr,
+	ret = sock->ops->bind(sock, (struct sockaddr_unsized *)&myaddr,
 			      sizeof(myaddr));
 	if (ret) {
 		mlog(ML_ERROR, "bind failed with %d at address %pI4\n",
@@ -1652,17 +1623,8 @@ static void o2net_start_connect(struct work_struct *work)
 		goto out;
 	}
 
-	ret = o2net_set_nodelay(sc->sc_sock);
-	if (ret) {
-		mlog(ML_ERROR, "setting TCP_NODELAY failed with %d\n", ret);
-		goto out;
-	}
-
-	ret = o2net_set_usertimeout(sock);
-	if (ret) {
-		mlog(ML_ERROR, "set TCP_USER_TIMEOUT failed with %d\n", ret);
-		goto out;
-	}
+	tcp_sock_set_nodelay(sc->sc_sock->sk);
+	tcp_sock_set_user_timeout(sock->sk, O2NET_TCP_USER_TIMEOUT);
 
 	o2net_register_callbacks(sc->sc_sock->sk, sc);
 
@@ -1676,7 +1638,7 @@ static void o2net_start_connect(struct work_struct *work)
 	remoteaddr.sin_port = node->nd_ipv4_port;
 
 	ret = sc->sc_sock->ops->connect(sc->sc_sock,
-					(struct sockaddr *)&remoteaddr,
+					(struct sockaddr_unsized *)&remoteaddr,
 					sizeof(remoteaddr),
 					O_NONBLOCK);
 	if (ret == -EINPROGRESS)
@@ -1697,7 +1659,7 @@ out:
 	if (mynode)
 		o2nm_node_put(mynode);
 
-	memalloc_noio_restore(noio_flag);
+	memalloc_nofs_restore(nofs_flag);
 	return;
 }
 
@@ -1776,7 +1738,7 @@ static void o2net_hb_node_up_cb(struct o2nm_node *node, int node_num,
 		(msecs_to_jiffies(o2net_reconnect_delay()) + 1);
 
 	if (node_num != o2nm_this_node()) {
-		/* believe it or not, accept and node hearbeating testing
+		/* believe it or not, accept and node heartbeating testing
 		 * can succeed for this node before we got here.. so
 		 * only use set_nn_state to clear the persistent error
 		 * if that hasn't already happened */
@@ -1823,16 +1785,17 @@ static int o2net_accept_one(struct socket *sock, int *more)
 	struct o2nm_node *node = NULL;
 	struct o2nm_node *local_node = NULL;
 	struct o2net_sock_container *sc = NULL;
+	struct proto_accept_arg arg = {
+		.flags = O_NONBLOCK,
+	};
 	struct o2net_node *nn;
-	unsigned int noio_flag;
+	unsigned int nofs_flag;
 
 	/*
-	 * sock_create_lite allocates the sock with GFP_KERNEL. We must set
-	 * per-process flag PF_MEMALLOC_NOIO so that all allocations done
-	 * by this process are done as if GFP_NOIO was specified. So we
-	 * are not reentering filesystem while doing memory reclaim.
+	 * sock_create_lite allocates the sock with GFP_KERNEL. We must
+	 * prevent the filesystem from being reentered by memory reclaim.
 	 */
-	noio_flag = memalloc_noio_save();
+	nofs_flag = memalloc_nofs_save();
 
 	BUG_ON(sock == NULL);
 	*more = 0;
@@ -1843,24 +1806,15 @@ static int o2net_accept_one(struct socket *sock, int *more)
 
 	new_sock->type = sock->type;
 	new_sock->ops = sock->ops;
-	ret = sock->ops->accept(sock, new_sock, O_NONBLOCK, false);
+	ret = sock->ops->accept(sock, new_sock, &arg);
 	if (ret < 0)
 		goto out;
 
 	*more = 1;
 	new_sock->sk->sk_allocation = GFP_ATOMIC;
 
-	ret = o2net_set_nodelay(new_sock);
-	if (ret) {
-		mlog(ML_ERROR, "setting TCP_NODELAY failed with %d\n", ret);
-		goto out;
-	}
-
-	ret = o2net_set_usertimeout(new_sock);
-	if (ret) {
-		mlog(ML_ERROR, "set TCP_USER_TIMEOUT failed with %d\n", ret);
-		goto out;
-	}
+	tcp_sock_set_nodelay(new_sock->sk);
+	tcp_sock_set_user_timeout(new_sock->sk, O2NET_TCP_USER_TIMEOUT);
 
 	ret = new_sock->ops->getname(new_sock, (struct sockaddr *) &sin, 1);
 	if (ret < 0)
@@ -1948,7 +1902,7 @@ out:
 	if (sc)
 		sc_put(sc);
 
-	memalloc_noio_restore(noio_flag);
+	memalloc_nofs_restore(nofs_flag);
 	return ret;
 }
 
@@ -1962,7 +1916,6 @@ static void o2net_accept_many(struct work_struct *work)
 {
 	struct socket *sock = o2net_listen_sock;
 	int	more;
-	int	err;
 
 	/*
 	 * It is critical to note that due to interrupt moderation
@@ -1977,7 +1930,7 @@ static void o2net_accept_many(struct work_struct *work)
 	 */
 
 	for (;;) {
-		err = o2net_accept_one(sock, &more);
+		o2net_accept_one(sock, &more);
 		if (!more)
 			break;
 		cond_resched();
@@ -1988,6 +1941,8 @@ static void o2net_listen_data_ready(struct sock *sk)
 {
 	void (*ready)(struct sock *sk);
 
+	trace_sk_data_ready(sk);
+
 	read_lock_bh(&sk->sk_callback_lock);
 	ready = sk->sk_user_data;
 	if (ready == NULL) { /* check for teardown race */
@@ -2047,7 +2002,7 @@ static int o2net_open_listening_sock(__be32 addr, __be16 port)
 	INIT_WORK(&o2net_listen_work, o2net_accept_many);
 
 	sock->sk->sk_reuse = SK_CAN_REUSE;
-	ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin));
+	ret = sock->ops->bind(sock, (struct sockaddr_unsized *)&sin, sizeof(sin));
 	if (ret < 0) {
 		printk(KERN_ERR "o2net: Error %d while binding socket at "
 		       "%pI4:%u\n", ret, &addr, ntohs(port)); 
@@ -2139,18 +2094,23 @@ void o2net_stop_listening(struct o2nm_node *node)
 
 int o2net_init(void)
 {
+	struct folio *folio;
+	void *p;
 	unsigned long i;
 
 	o2quo_init();
+	o2net_debugfs_init();
 
-	if (o2net_debugfs_init())
+	folio = folio_alloc(GFP_KERNEL | __GFP_ZERO, 0);
+	if (!folio)
 		goto out;
 
-	o2net_hand = kzalloc(sizeof(struct o2net_handshake), GFP_KERNEL);
-	o2net_keep_req = kzalloc(sizeof(struct o2net_msg), GFP_KERNEL);
-	o2net_keep_resp = kzalloc(sizeof(struct o2net_msg), GFP_KERNEL);
-	if (!o2net_hand || !o2net_keep_req || !o2net_keep_resp)
-		goto out;
+	p = folio_address(folio);
+	o2net_hand = p;
+	p += sizeof(struct o2net_handshake);
+	o2net_keep_req = p;
+	p += sizeof(struct o2net_msg);
+	o2net_keep_resp = p;
 
 	o2net_hand->protocol_version = cpu_to_be64(O2NET_PROTOCOL_VERSION);
 	o2net_hand->connector_id = cpu_to_be64(1);
@@ -2177,9 +2137,6 @@ int o2net_init(void)
 	return 0;
 
 out:
-	kfree(o2net_hand);
-	kfree(o2net_keep_req);
-	kfree(o2net_keep_resp);
 	o2net_debugfs_exit();
 	o2quo_exit();
 	return -ENOMEM;
@@ -2188,8 +2145,6 @@ out:
 void o2net_exit(void)
 {
 	o2quo_exit();
-	kfree(o2net_hand);
-	kfree(o2net_keep_req);
-	kfree(o2net_keep_resp);
 	o2net_debugfs_exit();
+	folio_put(virt_to_folio(o2net_hand));
 }