#include "fd-limit.h"
#include "jul-thread.h"
#include "lttng-sessiond.h"
+#include "session.h"
+#include "utils.h"
/*
* Note that there is not port here. It's set after this URI is parsed so we
* can let the user define a custom one. However, localhost is ALWAYS the
* default listening address.
*/
-static const char *default_reg_uri = "tcp://localhost";
+static const char *default_reg_uri =
+ "tcp://" DEFAULT_NETWORK_VIEWER_BIND_ADDRESS;
+
+/*
+ * Update JUL application using the given socket. This is done just after
+ * registration was successful.
+ *
+ * This is a quite heavy call in terms of locking since the session list lock
+ * AND session lock are acquired.
+ */
+static void update_jul_app(int sock)
+{
+ struct ltt_session *session, *stmp;
+ struct ltt_session_list *list;
+
+ list = session_get_list();
+ assert(list);
+
+ session_lock_list();
+ cds_list_for_each_entry_safe(session, stmp, &list->head, list) {
+ session_lock(session);
+ if (session->ust_session) {
+ jul_update(&session->ust_session->domain_jul, sock);
+ }
+ session_unlock(session);
+ }
+ session_unlock_list();
+}
+
+/*
+ * Destroy a JUL application by socket.
+ */
+static void destroy_jul_app(int sock)
+{
+ struct jul_app *app;
+
+ assert(sock >= 0);
+
+ /*
+ * Not finding an application is a very important error that should NEVER
+ * happen. The hash table deletion is ONLY done through this call even on
+ * thread cleanup.
+ */
+ rcu_read_lock();
+ app = jul_find_app_by_sock(sock);
+ assert(app);
+ rcu_read_unlock();
+
+ /* RCU read side lock is taken in this function call. */
+ jul_delete_app(app);
+
+ /* The application is freed in a RCU call but the socket is closed here. */
+ jul_destroy_app(app);
+}
+
+/*
+ * Cleanup remaining JUL apps in the hash table. This should only be called in
+ * the exit path of the thread.
+ */
+static void clean_jul_apps_ht(void)
+{
+ struct lttng_ht_node_ulong *node;
+ struct lttng_ht_iter iter;
+
+ DBG3("[jul-thread] Cleaning JUL apps ht");
+
+ rcu_read_lock();
+ cds_lfht_for_each_entry(jul_apps_ht_by_sock->ht, &iter.iter, node, node) {
+ struct jul_app *app;
+
+ app = caa_container_of(node, struct jul_app, node);
+ destroy_jul_app(app->sock->fd);
+ }
+ rcu_read_unlock();
+}
/*
* Create and init socket from uri.
ret = sock->ops->bind(sock);
if (ret < 0) {
+ WARN("Another session daemon is using this JUL port. JUL support "
+ "will be deactivated to prevent interfering with the tracing.");
goto error;
}
lttcomm_destroy_sock(sock);
}
+/*
+ * Handle a new JUL registration using the reg socket. After that, a new JUL
+ * application is added to the global hash table and attach to an UST app
+ * object. If r_app is not NULL, the created app is set to the pointer.
+ *
+ * Return the new FD created upon accept() on success or else a negative errno
+ * value.
+ */
+static int handle_registration(struct lttcomm_sock *reg_sock,
+ struct jul_app **r_app)
+{
+ int ret;
+ pid_t pid;
+ ssize_t size;
+ struct jul_app *app;
+ struct jul_register_msg msg;
+ struct lttcomm_sock *new_sock;
+
+ assert(reg_sock);
+
+ new_sock = reg_sock->ops->accept(reg_sock);
+ if (!new_sock) {
+ ret = -ENOTCONN;
+ goto error;
+ }
+
+ size = new_sock->ops->recvmsg(new_sock, &msg, sizeof(msg), 0);
+ if (size < sizeof(msg)) {
+ ret = -errno;
+ goto error_socket;
+ }
+ pid = be32toh(msg.pid);
+
+ DBG2("[jul-thread] New registration for pid %d on socket %d", pid,
+ new_sock->fd);
+
+ app = jul_create_app(pid, new_sock);
+ if (!app) {
+ ret = -ENOMEM;
+ goto error_socket;
+ }
+
+ /*
+ * Add before assigning the socket value to the UST app so it can be found
+ * concurrently.
+ */
+ jul_add_app(app);
+
+ /*
+ * We don't need to attach the JUL app to the app. If we ever do
+ * so, we should consider both registration order of JUL before
+ * app and app before JUL.
+ */
+
+ if (r_app) {
+ *r_app = app;
+ }
+
+ return new_sock->fd;
+
+error_socket:
+ new_sock->ops->close(new_sock);
+ lttcomm_destroy_sock(new_sock);
+error:
+ return ret;
+}
+
/*
* This thread manage application notify communication.
*/
rcu_register_thread();
rcu_thread_online();
+ /* JUL initialization call MUST be called before starting the thread. */
+ assert(jul_apps_ht_by_sock);
+
/* Create pollset with size 2, quit pipe and socket. */
ret = sessiond_set_thread_pollset(&events, 2);
if (ret < 0) {
goto error;
}
nb_fd = ret;
+ DBG3("[jul-thread] %d fd ready", nb_fd);
for (i = 0; i < nb_fd; i++) {
/* Fetch once the poll data */
goto error;
}
- /* FIXME: Nullify the JUL socket for the associated ust app. */
- } else if (revents & (LPOLLIN | LPOLLPRI)) {
- /*
- * FIXME: Handle JUL registration which must link an UST-app
- * and this JUL socket.
- */
+ destroy_jul_app(pollfd);
+ } else if (revents & (LPOLLIN)) {
+ int new_fd;
+ struct jul_app *app = NULL;
+
+ /* Pollin event of JUL app socket should NEVER happen. */
+ assert(pollfd == reg_sock->fd);
+
+ new_fd = handle_registration(reg_sock, &app);
+ if (new_fd < 0) {
+ WARN("[jul-thread] JUL registration failed. Ignoring.");
+ /* Somehow the communication failed. Just continue. */
+ continue;
+ }
+ /* Should not have a NULL app on success. */
+ assert(app);
+
+ /* Only add poll error event to only detect shutdown. */
+ ret = lttng_poll_add(&events, new_fd,
+ LPOLLERR | LPOLLHUP | LPOLLRDHUP);
+ if (ret < 0) {
+ destroy_jul_app(new_fd);
+ continue;
+ }
+
+ /* Update newly registered app. */
+ update_jul_app(new_fd);
+
+ /* On failure, the poll will detect it and clean it up. */
+ (void) jul_send_registration_done(app);
} else {
ERR("Unknown poll events %u for sock %d", revents, pollfd);
continue;
}
exit:
+ /* Whatever happens, try to delete it and exit. */
+ (void) lttng_poll_del(&events, reg_sock->fd);
error:
destroy_tcp_socket(reg_sock);
error_tcp_socket:
error_poll_create:
DBG("[jul-thread] is cleaning up and stopping.");
+ if (jul_apps_ht_by_sock) {
+ clean_jul_apps_ht();
+ lttng_ht_destroy(jul_apps_ht_by_sock);
+ }
+
rcu_thread_offline();
rcu_unregister_thread();
return NULL;
projects / lttng-tools.git / blobdiff