Fix: get_wait_shm() ust mutex deadlock (add 2 missing exit calls)

[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 12385dfca95762d32839b4e5917a92f8b82b4c58..16a634069a0beb1d33616860128449c95e0d1923 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.
  */
@@ -130,6 +136,15 @@ extern void ltt_ring_buffer_client_overwrite_exit(void);
 extern void ltt_ring_buffer_client_discard_exit(void);
 extern void ltt_ring_buffer_metadata_client_exit(void);
 
+/*
+ * Force a read (imply TLS fixup for dlopen) of TLS variables.
+ */
+static
+void lttng_fixup_nest_count_tls(void)
+{
+       asm volatile ("" : : "m" (lttng_ust_nest_count));
+}
+
 static
 int setup_local_apps(void)
 {
@@ -469,7 +484,9 @@ 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;
 
@@ -518,9 +535,9 @@ int get_wait_shm(struct sock_info *sock_info, size_t mmap_size)
                        ret = ftruncate(wait_shm_fd, mmap_size);
                        if (ret) {
                                PERROR("ftruncate");
-                               exit(EXIT_FAILURE);
+                               _exit(EXIT_FAILURE);
                        }
-                       exit(EXIT_SUCCESS);
+                       _exit(EXIT_SUCCESS);
                }
                /*
                 * For local shm, we need to have rw access to accept
@@ -532,13 +549,13 @@ int get_wait_shm(struct sock_info *sock_info, size_t mmap_size)
                 */
                if (!sock_info->global && errno != EACCES) {
                        ERR("Error opening shm %s", sock_info->wait_shm_path);
-                       exit(EXIT_FAILURE);
+                       _exit(EXIT_FAILURE);
                }
                /*
                 * The shm exists, but we cannot open it RW. Report
                 * success.
                 */
-               exit(EXIT_SUCCESS);
+               _exit(EXIT_SUCCESS);
        } else {
                return -1;
        }
@@ -838,6 +855,7 @@ void __attribute__((constructor)) lttng_ust_init(void)
 {
        struct timespec constructor_timeout;
        sigset_t sig_all_blocked, orig_parent_mask;
+       pthread_attr_t thread_attr;
        int timeout_mode;
        int ret;
 
@@ -852,6 +870,7 @@ void __attribute__((constructor)) lttng_ust_init(void)
        lttng_fixup_event_tls();
        lttng_fixup_ringbuffer_tls();
        lttng_fixup_vtid_tls();
+       lttng_fixup_nest_count_tls();
 
        /*
         * We want precise control over the order in which we construct
@@ -883,23 +902,41 @@ 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(&global_apps.ust_listener, NULL,
-                       ust_listener_thread, &global_apps);
+       ret = pthread_attr_init(&thread_attr);
+       if (ret) {
+               ERR("pthread_attr_init: %s", strerror(ret));
+       }
+       ret = pthread_attr_setdetachstate(&thread_attr, PTHREAD_CREATE_DETACHED);
+       if (ret) {
+               ERR("pthread_attr_setdetachstate: %s", strerror(ret));
+       }
 
+       ret = pthread_create(&global_apps.ust_listener, &thread_attr,
+                       ust_listener_thread, &global_apps);
+       if (ret) {
+               ERR("pthread_create global: %s", strerror(ret));
+       }
        if (local_apps.allowed) {
-               ret = pthread_create(&local_apps.ust_listener, NULL,
+               ret = pthread_create(&local_apps.ust_listener, &thread_attr,
                                ust_listener_thread, &local_apps);
+               if (ret) {
+                       ERR("pthread_create local: %s", strerror(ret));
+               }
        } else {
                handle_register_done(&local_apps);
        }
+       ret = pthread_attr_destroy(&thread_attr);
+       if (ret) {
+               ERR("pthread_attr_destroy: %s", strerror(ret));
+       }
 
        /* 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) {
@@ -975,12 +1012,14 @@ 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));
                }
        }
        /*
@@ -1012,6 +1051,8 @@ 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);
@@ -1037,6 +1078,8 @@ 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 */
@@ -1054,6 +1097,8 @@ 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();