[dpdk-dev] [PATCH 3/6] vhost: add reconnect ability

[Yuanhan Liu <yuanhan.liu@linux.intel.com>]

Allow reconnecting on failure when both RTE_VHOST_USER_RECONNECT and
RTE_VHOST_USER_CLIENT flags are set.

Reconnecting means two things here:

- when DPDK app starts first and QEMU (as the server) is not started,
  without reconnecting, DPDK app would simply fail on vhost-user
  registration.

- when QEMU reboots, without reconnecting, you can't re-establish the
  connection without restarting DPDK app.

This patch make it work well for both above cases. It simply creates
a new thread, and keep trying calling "connect()", until it succeeds.

Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
---
 lib/librte_vhost/rte_virtio_net.h            |  1 +
 lib/librte_vhost/vhost_user/vhost-net-user.c | 63 +++++++++++++++++++++++++---
 2 files changed, 59 insertions(+), 5 deletions(-)

diff --git a/lib/librte_vhost/rte_virtio_net.h b/lib/librte_vhost/rte_virtio_net.h
index c84e7ab..f354d52 100644
--- a/lib/librte_vhost/rte_virtio_net.h
+++ b/lib/librte_vhost/rte_virtio_net.h
@@ -52,6 +52,7 @@
 #include <rte_ether.h>
 
 #define RTE_VHOST_USER_CLIENT		(1ULL << 0)
+#define RTE_VHOST_USER_RECONNECT	(1ULL << 1)
 
 struct rte_mbuf;
 
diff --git a/lib/librte_vhost/vhost_user/vhost-net-user.c b/lib/librte_vhost/vhost_user/vhost-net-user.c
index aa98717..07bce6e 100644
--- a/lib/librte_vhost/vhost_user/vhost-net-user.c
+++ b/lib/librte_vhost/vhost_user/vhost-net-user.c
@@ -59,6 +59,7 @@ struct vhost_user_socket {
 	char *path;
 	int listenfd;
 	int is_server;
+	int reconnect;
 };
 
 struct vhost_user_connection {
@@ -78,6 +79,7 @@ struct vhost_user {
 
 static void vhost_user_server_new_connection(int fd, void *data, int *remove);
 static void vhost_user_msg_handler(int fd, void *dat, int *remove);
+static int vhost_user_create_client(struct vhost_user_socket *vsocket);
 
 static struct vhost_user vhost_user = {
 	.fdset = {
@@ -304,6 +306,8 @@ vhost_user_msg_handler(int connfd, void *dat, int *remove)
 	vid = conn->vid;
 	ret = read_vhost_message(connfd, &msg);
 	if (ret <= 0 || msg.request >= VHOST_USER_MAX) {
+		struct vhost_user_socket *vsocket = conn->vsocket;
+
 		if (ret < 0)
 			RTE_LOG(ERR, VHOST_CONFIG,
 				"vhost read message failed\n");
@@ -319,6 +323,9 @@ vhost_user_msg_handler(int connfd, void *dat, int *remove)
 		free(conn);
 		vhost_destroy_device(vid);
 
+		if (vsocket->reconnect)
+			vhost_user_create_client(vsocket);
+
 		return;
 	}
 
@@ -470,6 +477,33 @@ err:
 	return -1;
 }
 
+struct reconnect_info {
+	struct sockaddr_un un;
+	int fd;
+	struct vhost_user_socket *vsocket;
+};
+
+static void *
+vhost_user_client_reconnect(void *arg)
+{
+	struct reconnect_info *reconn = arg;
+	int ret;
+
+	RTE_LOG(ERR, VHOST_CONFIG, "reconnecting...\n");
+	while (1) {
+		ret = connect(reconn->fd, (struct sockaddr *)&reconn->un,
+				sizeof(reconn->un));
+		if (ret == 0)
+			break;
+		sleep(1);
+	}
+
+	vhost_user_add_connection(reconn->fd, reconn->vsocket);
+	free(reconn);
+
+	return NULL;
+}
+
 static int
 vhost_user_create_client(struct vhost_user_socket *vsocket)
 {
@@ -477,22 +511,40 @@ vhost_user_create_client(struct vhost_user_socket *vsocket)
 	int ret;
 	struct sockaddr_un un;
 	const char *path = vsocket->path;
+	struct reconnect_info *reconn;
+	pthread_t tid;
 
 	fd = create_unix_socket(path, &un, vsocket->is_server);
 	if (fd < 0)
 		return -1;
 
 	ret = connect(fd, (struct sockaddr *)&un, sizeof(un));
-	if (ret < 0) {
-		RTE_LOG(ERR, VHOST_CONFIG, "failed to connect to %s: %s\n",
-			path, strerror(errno));
+	if (ret == 0) {
+		vhost_user_add_connection(fd, vsocket);
+		return 0;
+	}
+
+	RTE_LOG(ERR, VHOST_CONFIG,
+		"failed to connect to %s: %s\n",
+		path, strerror(errno));
+
+	if (!vsocket->reconnect) {
 		close(fd);
 		return -1;
 	}
 
-	vhost_user_add_connection(fd, vsocket);
+	/* Create a thread to try reconnecting, to not block the caller. */
+	reconn = malloc(sizeof(*reconn));
+	reconn->un = un;
+	reconn->fd = fd;
+	reconn->vsocket = vsocket;
+	ret = pthread_create(&tid, NULL, vhost_user_client_reconnect, reconn);
+	if (ret < 0) {
+		close(fd);
+		RTE_LOG(ERR, VHOST_CONFIG, "failed to create reconnect thread");
+	}
 
-	return 0;
+	return ret;
 }
 
 /*
@@ -524,6 +576,7 @@ rte_vhost_driver_register(const char *path, uint64_t flags)
 	vsocket->path = strdup(path);
 
 	if ((flags & RTE_VHOST_USER_CLIENT) != 0) {
+		vsocket->reconnect = !!(flags & RTE_VHOST_USER_RECONNECT);
 		ret = vhost_user_create_client(vsocket);
 	} else {
 		vsocket->is_server = 1;
-- 
1.9.0