Version 2.2.3

[lttng-ust.git] / liblttng-ust / lttng-ust-comm.c

diff --git a/liblttng-ust/lttng-ust-comm.c b/liblttng-ust/lttng-ust-comm.c
index a06489fbef8a32bc5d29a12d844278575881a548..e8c85304e798b446ecf17110cc9e3fd70464527f 100644 (file)
--- a/liblttng-ust/lttng-ust-comm.c
+++ b/liblttng-ust/lttng-ust-comm.c
@@ -689,6 +689,30 @@ int get_wait_shm(struct sock_info *sock_info, size_t mmap_size)
         */
        wait_shm_fd = shm_open(sock_info->wait_shm_path, O_RDONLY, 0);
        if (wait_shm_fd >= 0) {
+               int32_t tmp_read;
+               ssize_t len;
+               size_t bytes_read = 0;
+
+               /*
+                * Try to read the fd. If unable to do so, try opening
+                * it in write mode.
+                */
+               do {
+                       len = read(wait_shm_fd,
+                               &((char *) &tmp_read)[bytes_read],
+                               sizeof(tmp_read) - bytes_read);
+                       if (len > 0) {
+                               bytes_read += len;
+                       }
+               } while ((len < 0 && errno == EINTR)
+                       || (len > 0 && bytes_read < sizeof(tmp_read)));
+               if (bytes_read != sizeof(tmp_read)) {
+                       ret = close(wait_shm_fd);
+                       if (ret) {
+                               ERR("close wait_shm_fd");
+                       }
+                       goto open_write;
+               }
                goto end;
        } else if (wait_shm_fd < 0 && errno != ENOENT) {
                /*
@@ -699,9 +723,11 @@ int get_wait_shm(struct sock_info *sock_info, size_t mmap_size)
                ERR("Error opening shm %s", sock_info->wait_shm_path);
                goto end;
        }
+
+open_write:
        /*
-        * If the open failed because the file did not exist, try
-        * creating it ourself.
+        * If the open failed because the file did not exist, or because
+        * the file was not truncated yet, try creating it ourself.
         */
        URCU_TLS(lttng_ust_nest_count)++;
        pid = fork();
@@ -896,8 +922,6 @@ void *ust_listener_thread(void *arg)
 {
        struct sock_info *sock_info = arg;
        int sock, ret, prev_connect_failed = 0, has_waited = 0;
-       int open_sock[2];
-       int i;
        long timeout;
 
        /* Restart trying to connect to the session daemon */
@@ -935,55 +959,35 @@ restart:
                sock_info->notify_socket = -1;
        }
 
-       /* Register */
-       for (i = 0; i < 2; i++) {
-               ret = ustcomm_connect_unix_sock(sock_info->sock_path);
-               if (ret < 0) {
-                       DBG("Info: sessiond not accepting connections to %s apps socket", sock_info->name);
-                       prev_connect_failed = 1;
-
-                       ust_lock();
+       /*
+        * Register. We need to perform both connect and sending
+        * registration message before doing the next connect otherwise
+        * we may reach unix socket connect queue max limits and block
+        * on the 2nd connect while the session daemon is awaiting the
+        * first connect registration message.
+        */
+       /* Connect cmd socket */
+       ret = ustcomm_connect_unix_sock(sock_info->sock_path);
+       if (ret < 0) {
+               DBG("Info: sessiond not accepting connections to %s apps socket", sock_info->name);
+               prev_connect_failed = 1;
 
-                       if (lttng_ust_comm_should_quit) {
-                               goto quit;
-                       }
+               ust_lock();
 
-                       /*
-                        * If we cannot find the sessiond daemon, don't delay
-                        * constructor execution.
-                        */
-                       ret = handle_register_done(sock_info);
-                       assert(!ret);
-                       ust_unlock();
-                       goto restart;
+               if (lttng_ust_comm_should_quit) {
+                       goto quit;
                }
-               open_sock[i] = ret;
-       }
 
-       sock_info->socket = open_sock[0];
-       sock_info->notify_socket = open_sock[1];
-
-       timeout = get_notify_sock_timeout();
-       if (timeout >= 0) {
                /*
-                * Give at least 10ms to sessiond to reply to
-                * notifications.
+                * If we cannot find the sessiond daemon, don't delay
+                * constructor execution.
                 */
-               if (timeout < 10)
-                       timeout = 10;
-               ret = ustcomm_setsockopt_rcv_timeout(sock_info->notify_socket,
-                               timeout);
-               if (ret < 0) {
-                       WARN("Error setting socket receive timeout");
-               }
-               ret = ustcomm_setsockopt_snd_timeout(sock_info->notify_socket,
-                               timeout);
-               if (ret < 0) {
-                       WARN("Error setting socket send timeout");
-               }
-       } else if (timeout < -1) {
-               WARN("Unsupported timeout value %ld", timeout);
+               ret = handle_register_done(sock_info);
+               assert(!ret);
+               ust_unlock();
+               goto restart;
        }
+       sock_info->socket = ret;
 
        ust_lock();
 
@@ -1019,6 +1023,60 @@ restart:
                ust_unlock();
                goto restart;
        }
+
+       ust_unlock();
+
+       /* Connect notify socket */
+       ret = ustcomm_connect_unix_sock(sock_info->sock_path);
+       if (ret < 0) {
+               DBG("Info: sessiond not accepting connections to %s apps socket", sock_info->name);
+               prev_connect_failed = 1;
+
+               ust_lock();
+
+               if (lttng_ust_comm_should_quit) {
+                       goto quit;
+               }
+
+               /*
+                * If we cannot find the sessiond daemon, don't delay
+                * constructor execution.
+                */
+               ret = handle_register_done(sock_info);
+               assert(!ret);
+               ust_unlock();
+               goto restart;
+       }
+       sock_info->notify_socket = ret;
+
+       timeout = get_notify_sock_timeout();
+       if (timeout >= 0) {
+               /*
+                * Give at least 10ms to sessiond to reply to
+                * notifications.
+                */
+               if (timeout < 10)
+                       timeout = 10;
+               ret = ustcomm_setsockopt_rcv_timeout(sock_info->notify_socket,
+                               timeout);
+               if (ret < 0) {
+                       WARN("Error setting socket receive timeout");
+               }
+               ret = ustcomm_setsockopt_snd_timeout(sock_info->notify_socket,
+                               timeout);
+               if (ret < 0) {
+                       WARN("Error setting socket send timeout");
+               }
+       } else if (timeout < -1) {
+               WARN("Unsupported timeout value %ld", timeout);
+       }
+
+       ust_lock();
+
+       if (lttng_ust_comm_should_quit) {
+               goto quit;
+       }
+
        ret = register_to_sessiond(sock_info->notify_socket,
                        USTCTL_SOCKET_NOTIFY);
        if (ret < 0) {