Fix: liblttng-ust-fork deadlock

[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 76c8ee3da149ff49ecfe4633d7904927f1683de9..7a9097014d26e4d483f4d4a9f7d98cbd581807d9 100644 (file)
--- a/liblttng-ust/lttng-ust-comm.c
+++ b/liblttng-ust/lttng-ust-comm.c
@@ -78,6 +78,12 @@ static sem_t constructor_wait;
  */
 static int sem_count = { 2 };
 
+/*
+ * Counting nesting within lttng-ust. Used to ensure that calling fork()
+ * from liblttng-ust does not execute the pre/post fork handlers.
+ */
+static int __thread lttng_ust_nest_count;
+
 /*
  * Info about socket and associated listener thread.
  */
@@ -236,6 +242,7 @@ int handle_message(struct sock_info *sock_info,
        struct ustcomm_ust_reply lur;
        int shm_fd, wait_fd;
        union ust_args args;
+       ssize_t len;
 
        ust_lock();
 
@@ -346,6 +353,22 @@ end:
                        goto error;
                }
        }
+       /*
+        * LTTNG_UST_TRACEPOINT_FIELD_LIST_GET needs to send the field
+        * after the reply.
+        */
+       if (lur.ret_code == USTCOMM_OK) {
+               switch (lum->cmd) {
+               case LTTNG_UST_TRACEPOINT_FIELD_LIST_GET:
+                       len = ustcomm_send_unix_sock(sock,
+                               &args.field_list.entry,
+                               sizeof(args.field_list.entry));
+                       if (len != sizeof(args.field_list.entry)) {
+                               ret = -1;
+                               goto error;
+                       }
+               }
+       }
        /*
         * We still have the memory map reference, and the fds have been
         * sent to the sessiond. We can therefore close those fds. Note
@@ -464,7 +487,9 @@ int get_wait_shm(struct sock_info *sock_info, size_t mmap_size)
         * If the open failed because the file did not exist, try
         * creating it ourself.
         */
+       lttng_ust_nest_count++;
        pid = fork();
+       lttng_ust_nest_count--;
        if (pid > 0) {
                int status;
 
@@ -878,15 +903,20 @@ void __attribute__((constructor)) lttng_ust_init(void)
        sigfillset(&sig_all_blocked);
        ret = pthread_sigmask(SIG_SETMASK, &sig_all_blocked, &orig_parent_mask);
        if (ret) {
-               PERROR("pthread_sigmask: %s", strerror(ret));
+               ERR("pthread_sigmask: %s", strerror(ret));
        }
 
-       ret = pthread_create(&local_apps.ust_listener, NULL,
-                       ust_listener_thread, &local_apps);
-
+       ret = pthread_create(&global_apps.ust_listener, NULL,
+                       ust_listener_thread, &global_apps);
+       if (ret) {
+               ERR("pthread_create global: %s", strerror(ret));
+       }
        if (local_apps.allowed) {
-               ret = pthread_create(&global_apps.ust_listener, NULL,
-                               ust_listener_thread, &global_apps);
+               ret = pthread_create(&local_apps.ust_listener, NULL,
+                               ust_listener_thread, &local_apps);
+               if (ret) {
+                       ERR("pthread_create local: %s", strerror(ret));
+               }
        } else {
                handle_register_done(&local_apps);
        }
@@ -894,7 +924,7 @@ void __attribute__((constructor)) lttng_ust_init(void)
        /* Restore original signal mask in parent */
        ret = pthread_sigmask(SIG_SETMASK, &orig_parent_mask, NULL);
        if (ret) {
-               PERROR("pthread_sigmask: %s", strerror(ret));
+               ERR("pthread_sigmask: %s", strerror(ret));
        }
 
        switch (timeout_mode) {
@@ -970,12 +1000,14 @@ void __attribute__((destructor)) lttng_ust_exit(void)
        /* cancel threads */
        ret = pthread_cancel(global_apps.ust_listener);
        if (ret) {
-               ERR("Error cancelling global ust listener thread");
+               ERR("Error cancelling global ust listener thread: %s",
+                       strerror(ret));
        }
        if (local_apps.allowed) {
                ret = pthread_cancel(local_apps.ust_listener);
                if (ret) {
-                       ERR("Error cancelling local ust listener thread");
+                       ERR("Error cancelling local ust listener thread: %s",
+                               strerror(ret));
                }
        }
        /*
@@ -1007,6 +1039,8 @@ void ust_before_fork(sigset_t *save_sigset)
        sigset_t all_sigs;
        int ret;
 
+       if (lttng_ust_nest_count)
+               return;
        /* Disable signals */
        sigfillset(&all_sigs);
        ret = sigprocmask(SIG_BLOCK, &all_sigs, save_sigset);
@@ -1032,6 +1066,8 @@ static void ust_after_fork_common(sigset_t *restore_sigset)
 
 void ust_after_fork_parent(sigset_t *restore_sigset)
 {
+       if (lttng_ust_nest_count)
+               return;
        DBG("process %d", getpid());
        rcu_bp_after_fork_parent();
        /* Release mutexes and reenable signals */
@@ -1049,6 +1085,8 @@ void ust_after_fork_parent(sigset_t *restore_sigset)
  */
 void ust_after_fork_child(sigset_t *restore_sigset)
 {
+       if (lttng_ust_nest_count)
+               return;
        DBG("process %d", getpid());
        /* Release urcu mutexes */
        rcu_bp_after_fork_child();