lttng-relayd: use TCP keep-alive mechanism to detect dead-peer

[lttng-tools.git] / src / bin / lttng-relayd / main.c

diff --git a/src/bin/lttng-relayd/main.c b/src/bin/lttng-relayd/main.c
index 7b313818e5c09a5719339740f47b2787304d461b..b70026242abadc5c2ba54913be6f8de6e743b921 100644 (file)
--- a/src/bin/lttng-relayd/main.c
+++ b/src/bin/lttng-relayd/main.c
@@ -70,6 +70,7 @@
 #include "stream.h"
 #include "connection.h"
 #include "tracefile-array.h"
+#include "tcp_keep_alive.h"
 
 /* command line options */
 char *opt_output_path;
@@ -891,6 +892,15 @@ restart:
                                        lttcomm_destroy_sock(newsock);
                                        goto error;
                                }
+
+                               ret = socket_apply_keep_alive_config(newsock->fd);
+                               if (ret < 0) {
+                                       ERR("Failed to apply TCP keep-alive configuration on socket (%i)",
+                                                       newsock->fd);
+                                       lttcomm_destroy_sock(newsock);
+                                       goto error;
+                               }
+
                                new_conn = connection_create(newsock, type);
                                if (!new_conn) {
                                        lttcomm_destroy_sock(newsock);
@@ -969,12 +979,16 @@ static void *relay_thread_dispatcher(void *data)
 
        health_code_update();
 
-       while (!CMM_LOAD_SHARED(dispatch_thread_exit)) {
+       for (;;) {
                health_code_update();
 
                /* Atomically prepare the queue futex */
                futex_nto1_prepare(&relay_conn_queue.futex);
 
+               if (CMM_LOAD_SHARED(dispatch_thread_exit)) {
+                       break;
+               }
+
                do {
                        health_code_update();
 
@@ -2937,6 +2951,12 @@ exit_init_data:
        health_app_destroy(health_relayd);
 exit_health_app_create:
 exit_options:
+       /*
+        * Wait for all pending call_rcu work to complete before tearing
+        * down data structures. call_rcu worker may be trying to
+        * perform lookups in those structures.
+        */
+       rcu_barrier();
        relayd_cleanup();
 
        /* Ensure all prior call_rcu are done. */