Merge branch 'devmem-tcp-minor-cleanups-and-ksft-improvements'

Mina Almasry says:

====================
Devmem TCP minor cleanups and ksft improvements

v2: https://lore.kernel.org/20250519023517.4062941-1-almasrymina@google.com
====================

Link: https://patch.msgid.link/20250523230524.1107879-1-almasrymina@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net/core/devmem.c

@@ -178,7 +178,8 @@ err_close_rxq:
 struct net_devmem_dmabuf_binding *
 net_devmem_bind_dmabuf(struct net_device *dev,
 		       enum dma_data_direction direction,
-		       unsigned int dmabuf_fd, struct netlink_ext_ack *extack)
+		       unsigned int dmabuf_fd, struct netdev_nl_sock *priv,
+		       struct netlink_ext_ack *extack)
 {
 	struct net_devmem_dmabuf_binding *binding;
 	static u32 id_alloc_next;
@@ -299,6 +300,8 @@ net_devmem_bind_dmabuf(struct net_device *dev,
 	if (err < 0)
 		goto err_free_chunks;
 
+	list_add(&binding->list, &priv->bindings);
+
 	return binding;
 
 err_free_chunks:

net/core/devmem.h

@@ -11,6 +11,7 @@
 #define _NET_DEVMEM_H
 
 #include <net/netmem.h>
+#include <net/netdev_netlink.h>
 
 struct netlink_ext_ack;
 
@@ -82,7 +83,8 @@ void __net_devmem_dmabuf_binding_free(struct work_struct *wq);
 struct net_devmem_dmabuf_binding *
 net_devmem_bind_dmabuf(struct net_device *dev,
 		       enum dma_data_direction direction,
-		       unsigned int dmabuf_fd, struct netlink_ext_ack *extack);
+		       unsigned int dmabuf_fd, struct netdev_nl_sock *priv,
+		       struct netlink_ext_ack *extack);
 struct net_devmem_dmabuf_binding *net_devmem_lookup_dmabuf(u32 id);
 void net_devmem_unbind_dmabuf(struct net_devmem_dmabuf_binding *binding);
 int net_devmem_bind_dmabuf_to_queue(struct net_device *dev, u32 rxq_idx,
@@ -170,6 +172,7 @@ static inline void __net_devmem_dmabuf_binding_free(struct work_struct *wq)
 static inline struct net_devmem_dmabuf_binding *
 net_devmem_bind_dmabuf(struct net_device *dev, unsigned int dmabuf_fd,
 		       enum dma_data_direction direction,
+		       struct netdev_nl_sock *priv,
 		       struct netlink_ext_ack *extack)
 {
 	return ERR_PTR(-EOPNOTSUPP);

net/core/netdev-genl.c

@@ -908,7 +908,7 @@ int netdev_nl_bind_rx_doit(struct sk_buff *skb, struct genl_info *info)
 	}
 
 	binding = net_devmem_bind_dmabuf(netdev, DMA_FROM_DEVICE, dmabuf_fd,
-					 info->extack);
+					 priv, info->extack);
 	if (IS_ERR(binding)) {
 		err = PTR_ERR(binding);
 		goto err_unlock;
@@ -943,8 +943,6 @@ int netdev_nl_bind_rx_doit(struct sk_buff *skb, struct genl_info *info)
 			goto err_unbind;
 	}
 
-	list_add(&binding->list, &priv->bindings);
-
 	nla_put_u32(rsp, NETDEV_A_DMABUF_ID, binding->id);
 	genlmsg_end(rsp, hdr);
 
@@ -1020,15 +1018,13 @@ int netdev_nl_bind_tx_doit(struct sk_buff *skb, struct genl_info *info)
 		goto err_unlock_netdev;
 	}
 
-	binding = net_devmem_bind_dmabuf(netdev, DMA_TO_DEVICE, dmabuf_fd,
+	binding = net_devmem_bind_dmabuf(netdev, DMA_TO_DEVICE, dmabuf_fd, priv,
 					 info->extack);
 	if (IS_ERR(binding)) {
 		err = PTR_ERR(binding);
 		goto err_unlock_netdev;
 	}
 
-	list_add(&binding->list, &priv->bindings);
-
 	nla_put_u32(rsp, NETDEV_A_DMABUF_ID, binding->id);
 	genlmsg_end(rsp, hdr);
 

net/core/page_pool.c

@@ -867,8 +867,8 @@ void page_pool_put_unrefed_netmem(struct page_pool *pool, netmem_ref netmem,
 	if (!allow_direct)
 		allow_direct = page_pool_napi_local(pool);
 
-	netmem =
-		__page_pool_put_page(pool, netmem, dma_sync_size, allow_direct);
+	netmem = __page_pool_put_page(pool, netmem, dma_sync_size,
+				      allow_direct);
 	if (netmem && !page_pool_recycle_in_ring(pool, netmem)) {
 		/* Cache full, fallback to free pages */
 		recycle_stat_inc(pool, ring_full);

net/ipv4/tcp.c

@@ -1067,7 +1067,7 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 	int flags, err, copied = 0;
 	int mss_now = 0, size_goal, copied_syn = 0;
 	int process_backlog = 0;
-	bool sockc_valid = true;
+	int sockc_err = 0;
 	int zc = 0;
 	long timeo;
 
@@ -1075,13 +1075,10 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 
 	sockc = (struct sockcm_cookie){ .tsflags = READ_ONCE(sk->sk_tsflags) };
 	if (msg->msg_controllen) {
-		err = sock_cmsg_send(sk, msg, &sockc);
-		if (unlikely(err))
-			/* Don't return error until MSG_FASTOPEN has been
-			 * processed; that may succeed even if the cmsg is
-			 * invalid.
-			 */
-			sockc_valid = false;
+		sockc_err = sock_cmsg_send(sk, msg, &sockc);
+		/* Don't return error until MSG_FASTOPEN has been processed;
+		 * that may succeed even if the cmsg is invalid.
+		 */
 	}
 
 	if ((flags & MSG_ZEROCOPY) && size) {
@@ -1092,7 +1089,7 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 		} else if (sock_flag(sk, SOCK_ZEROCOPY)) {
 			skb = tcp_write_queue_tail(sk);
 			uarg = msg_zerocopy_realloc(sk, size, skb_zcopy(skb),
-						    sockc_valid && !!sockc.dmabuf_id);
+						    !sockc_err && sockc.dmabuf_id);
 			if (!uarg) {
 				err = -ENOBUFS;
 				goto out_err;
@@ -1102,7 +1099,7 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 			else
 				uarg_to_msgzc(uarg)->zerocopy = 0;
 
-			if (sockc_valid && sockc.dmabuf_id) {
+			if (!sockc_err && sockc.dmabuf_id) {
 				binding = net_devmem_get_binding(sk, sockc.dmabuf_id);
 				if (IS_ERR(binding)) {
 					err = PTR_ERR(binding);
@@ -1116,7 +1113,7 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 			zc = MSG_SPLICE_PAGES;
 	}
 
-	if (sockc_valid && sockc.dmabuf_id &&
+	if (!sockc_err && sockc.dmabuf_id &&
 	    (!(flags & MSG_ZEROCOPY) || !sock_flag(sk, SOCK_ZEROCOPY))) {
 		err = -EINVAL;
 		goto out_err;
@@ -1160,9 +1157,8 @@ int tcp_sendmsg_locked(struct sock *sk, struct msghdr *msg, size_t size)
 		/* 'common' sending to sendq */
 	}
 
-	if (!sockc_valid) {
-		if (!err)
-			err = -EINVAL;
+	if (sockc_err) {
+		err = sockc_err;
 		goto out_err;
 	}
 

tools/testing/selftests/drivers/net/hw/devmem.py

@@ -21,30 +21,30 @@ def require_devmem(cfg):
 
 @ksft_disruptive
 def check_rx(cfg) -> None:
-    cfg.require_ipver("6")
     require_devmem(cfg)
 
     port = rand_port()
-    listen_cmd = f"{cfg.bin_local} -l -f {cfg.ifname} -s {cfg.addr_v['6']} -p {port}"
+    socat = f"socat -u - TCP{cfg.addr_ipver}:{cfg.addr}:{port},bind={cfg.remote_addr}:{port}"
+    listen_cmd = f"{cfg.bin_local} -l -f {cfg.ifname} -s {cfg.addr} -p {port} -c {cfg.remote_addr} -v 7"
 
-    with bkg(listen_cmd) as socat:
+    with bkg(listen_cmd, exit_wait=True) as ncdevmem:
         wait_port_listen(port)
-        cmd(f"echo -e \"hello\\nworld\"| socat -u - TCP6:[{cfg.addr_v['6']}]:{port}", host=cfg.remote, shell=True)
+        cmd(f"yes $(echo -e \x01\x02\x03\x04\x05\x06) | \
+            head -c 1K | {socat}", host=cfg.remote, shell=True)
 
-    ksft_eq(socat.stdout.strip(), "hello\nworld")
+    ksft_eq(ncdevmem.ret, 0)
 
 
 @ksft_disruptive
 def check_tx(cfg) -> None:
-    cfg.require_ipver("6")
     require_devmem(cfg)
 
     port = rand_port()
-    listen_cmd = f"socat -U - TCP6-LISTEN:{port}"
+    listen_cmd = f"socat -U - TCP{cfg.addr_ipver}-LISTEN:{port}"
 
-    with bkg(listen_cmd, exit_wait=True) as socat:
+    with bkg(listen_cmd) as socat:
         wait_port_listen(port)
-        cmd(f"echo -e \"hello\\nworld\"| {cfg.bin_remote} -f {cfg.ifname} -s {cfg.addr_v['6']} -p {port}", host=cfg.remote, shell=True)
+        cmd(f"echo -e \"hello\\nworld\"| {cfg.bin_remote} -f {cfg.ifname} -s {cfg.addr} -p {port}", host=cfg.remote, shell=True)
 
     ksft_eq(socat.stdout.strip(), "hello\nworld")
 

tools/testing/selftests/drivers/net/hw/ncdevmem.c

@@ -373,7 +373,8 @@ static int configure_flow_steering(struct sockaddr_in6 *server_sin)
 		server_addr = strrchr(server_addr, ':') + 1;
 	}
 
-	return run_command("sudo ethtool -N %s flow-type %s %s %s dst-ip %s %s %s dst-port %s queue %d >&2",
+	/* Try configure 5-tuple */
+	if (run_command("sudo ethtool -N %s flow-type %s %s %s dst-ip %s %s %s dst-port %s queue %d >&2",
 			   ifname,
 			   type,
 			   client_ip ? "src-ip" : "",
@@ -381,7 +382,17 @@ static int configure_flow_steering(struct sockaddr_in6 *server_sin)
 			   server_addr,
 			   client_ip ? "src-port" : "",
 			   client_ip ? port : "",
-			   port, start_queue);
+			   port, start_queue))
+		/* If that fails, try configure 3-tuple */
+		if (run_command("sudo ethtool -N %s flow-type %s dst-ip %s dst-port %s queue %d >&2",
+				ifname,
+				type,
+				server_addr,
+				port, start_queue))
+			/* If that fails, return error */
+			return -1;
+
+	return 0;
 }
 
 static int bind_rx_queue(unsigned int ifindex, unsigned int dmabuf_fd,
@@ -529,7 +540,6 @@ static struct netdev_queue_id *create_queues(void)
 static int do_server(struct memory_buffer *mem)
 {
 	char ctrl_data[sizeof(int) * 20000];
-	struct netdev_queue_id *queues;
 	size_t non_page_aligned_frags = 0;
 	struct sockaddr_in6 client_addr;
 	struct sockaddr_in6 server_sin;