Fix: sessiond: double socket close on allocation failure

[lttng-tools.git] / src / bin / lttng-sessiond / register.c

/*
 * Copyright (C) 2011 - David Goulet <david.goulet@polymtl.ca>
 *                      Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
 *               2013 - Jérémie Galarneau <jeremie.galarneau@efficios.com>
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License, version 2 only,
 * as published by the Free Software Foundation.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License along
 * with this program; if not, write to the Free Software Foundation, Inc.,
 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
 */

#include <stddef.h>
#include <stdlib.h>
#include <urcu.h>
#include <common/futex.h>
#include <common/macros.h>
#include <common/utils.h>
#include <sys/stat.h>

#include "register.h"
#include "lttng-sessiond.h"
#include "testpoint.h"
#include "health-sessiond.h"
#include "fd-limit.h"
#include "shm.h"
#include "utils.h"
#include "thread.h"

struct thread_state {
	struct lttng_pipe *quit_pipe;
	struct ust_cmd_queue *ust_cmd_queue;
	sem_t ready;
	bool running;
	int application_socket;
};

/*
 * Creates the application socket.
 */
static int create_application_socket(void)
{
	int ret = 0;
	int apps_sock;
	const mode_t old_umask = umask(0);

	/* Create the application unix socket */
	apps_sock = lttcomm_create_unix_sock(config.apps_unix_sock_path.value);
	if (apps_sock < 0) {
		ERR("Create unix sock failed: %s", config.apps_unix_sock_path.value);
		ret = -1;
		goto end;
	}

	/* Set the cloexec flag */
	ret = utils_set_fd_cloexec(apps_sock);
	if (ret < 0) {
		ERR("Unable to set CLOEXEC flag to the app Unix socket (fd: %d). "
				"Continuing but note that the consumer daemon will have a "
				"reference to this socket on exec()", apps_sock);
	}

	/* File permission MUST be 666 */
	ret = chmod(config.apps_unix_sock_path.value,
			S_IRUSR | S_IWUSR | S_IRGRP | S_IWGRP | S_IROTH | S_IWOTH);
	if (ret < 0) {
		PERROR("Set file permissions failed on %s",
				config.apps_unix_sock_path.value);
		goto error_close_socket;
	}

	DBG3("Session daemon application socket created (fd = %d) ", apps_sock);
	ret = apps_sock;
end:
	umask(old_umask);
	return ret;
error_close_socket:
	if (close(apps_sock)) {
		PERROR("Failed to close application socket in error path");
	}
	apps_sock = -1;
	ret = -1;
	goto end;
}

/*
 * Notify UST applications using the shm mmap futex.
 */
static int notify_ust_apps(int active, bool is_root)
{
	char *wait_shm_mmap;

	DBG("Notifying applications of session daemon state: %d", active);

	/* See shm.c for this call implying mmap, shm and futex calls */
	wait_shm_mmap = shm_ust_get_mmap(config.wait_shm_path.value, is_root);
	if (wait_shm_mmap == NULL) {
		goto error;
	}

	/* Wake waiting process */
	futex_wait_update((int32_t *) wait_shm_mmap, active);

	/* Apps notified successfully */
	return 0;

error:
	return -1;
}

static void cleanup_application_registration_thread(void *data)
{
	struct thread_state *thread_state = data;

	if (!data) {
		return;
	}

	lttng_pipe_destroy(thread_state->quit_pipe);
	free(thread_state);
}

static void set_thread_status(struct thread_state *thread_state, bool running)
{
	DBG("Marking application registration thread's state as %s", running ? "running" : "error");
	thread_state->running = running;
	sem_post(&thread_state->ready);
}

static bool wait_thread_status(struct thread_state *thread_state)
{
	DBG("Waiting for application registration thread to be ready");
	sem_wait(&thread_state->ready);
	if (thread_state->running) {
		DBG("Application registration thread is ready");
	} else {
		ERR("Initialization of application registration thread failed");
	}

	return thread_state->running;
}

static void thread_init_cleanup(void *data)
{
	struct thread_state *thread_state = data;

	set_thread_status(thread_state, false);
}

/*
 * This thread manage application registration.
 */
static void *thread_application_registration(void *data)
{
	int sock = -1, i, ret, pollfd, err = -1;
	uint32_t revents, nb_fd;
	struct lttng_poll_event events;
	/*
	 * Gets allocated in this thread, enqueued to a global queue, dequeued
	 * and freed in the manage apps thread.
	 */
	struct ust_command *ust_cmd = NULL;
	const bool is_root = (getuid() == 0);
	struct thread_state *thread_state = data;
	const int application_socket = thread_state->application_socket;
	const int quit_pipe_read_fd = lttng_pipe_get_readfd(
			thread_state->quit_pipe);

	DBG("[thread] Manage application registration started");

	pthread_cleanup_push(thread_init_cleanup, NULL);
	health_register(health_sessiond, HEALTH_SESSIOND_TYPE_APP_REG);

	ret = lttcomm_listen_unix_sock(application_socket);
	if (ret < 0) {
		goto error_listen;
	}

	if (testpoint(sessiond_thread_registration_apps)) {
		goto error_create_poll;
	}

	/*
	 * Pass 2 as size here for the thread quit pipe and apps_sock. Nothing
	 * more will be added to this poll set.
	 */
	ret = lttng_poll_create(&events, 2, LTTNG_CLOEXEC);
	if (ret < 0) {
		goto error_create_poll;
	}

	/* Add the application registration socket */
	ret = lttng_poll_add(&events, application_socket, LPOLLIN | LPOLLRDHUP);
	if (ret < 0) {
		goto error_poll_add;
	}

	/* Add the application registration socket */
	ret = lttng_poll_add(&events, quit_pipe_read_fd, LPOLLIN | LPOLLRDHUP);
	if (ret < 0) {
		goto error_poll_add;
	}

	set_thread_status(thread_state, true);
	pthread_cleanup_pop(0);

	while (1) {
		DBG("Accepting application registration");

		/* Inifinite blocking call, waiting for transmission */
	restart:
		health_poll_entry();
		ret = lttng_poll_wait(&events, -1);
		health_poll_exit();
		if (ret < 0) {
			/*
			 * Restart interrupted system call.
			 */
			if (errno == EINTR) {
				goto restart;
			}
			goto error;
		}

		nb_fd = ret;

		for (i = 0; i < nb_fd; i++) {
			health_code_update();

			/* Fetch once the poll data */
			revents = LTTNG_POLL_GETEV(&events, i);
			pollfd = LTTNG_POLL_GETFD(&events, i);

			/* Thread quit pipe has been closed. Killing thread. */
			if (pollfd == quit_pipe_read_fd) {
				err = 0;
				goto exit;
			} else {
				/* Event on the registration socket */
				if (revents & LPOLLIN) {
					sock = lttcomm_accept_unix_sock(application_socket);
					if (sock < 0) {
						goto error;
					}

					/*
					 * Set socket timeout for both receiving and ending.
					 * app_socket_timeout is in seconds, whereas
					 * lttcomm_setsockopt_rcv_timeout and
					 * lttcomm_setsockopt_snd_timeout expect msec as
					 * parameter.
					 */
					if (config.app_socket_timeout >= 0) {
						(void) lttcomm_setsockopt_rcv_timeout(sock,
								config.app_socket_timeout * 1000);
						(void) lttcomm_setsockopt_snd_timeout(sock,
								config.app_socket_timeout * 1000);
					}

					/*
					 * Set the CLOEXEC flag. Return code is useless because
					 * either way, the show must go on.
					 */
					(void) utils_set_fd_cloexec(sock);

					/* Create UST registration command for enqueuing */
					ust_cmd = zmalloc(sizeof(struct ust_command));
					if (ust_cmd == NULL) {
						PERROR("ust command zmalloc");
						ret = close(sock);
						if (ret) {
							PERROR("close");
						}
						sock = -1;
						goto error;
					}

					/*
					 * Using message-based transmissions to ensure we don't
					 * have to deal with partially received messages.
					 */
					ret = lttng_fd_get(LTTNG_FD_APPS, 1);
					if (ret < 0) {
						ERR("Exhausted file descriptors allowed for applications.");
						free(ust_cmd);
						ret = close(sock);
						if (ret) {
							PERROR("close");
						}
						sock = -1;
						continue;
					}

					health_code_update();
					ret = ust_app_recv_registration(sock, &ust_cmd->reg_msg);
					if (ret < 0) {
						free(ust_cmd);
						/* Close socket of the application. */
						ret = close(sock);
						if (ret) {
							PERROR("close");
						}
						lttng_fd_put(LTTNG_FD_APPS, 1);
						sock = -1;
						continue;
					}
					health_code_update();

					ust_cmd->sock = sock;
					sock = -1;

					DBG("UST registration received with pid:%d ppid:%d uid:%d"
							" gid:%d sock:%d name:%s (version %d.%d)",
							ust_cmd->reg_msg.pid, ust_cmd->reg_msg.ppid,
							ust_cmd->reg_msg.uid, ust_cmd->reg_msg.gid,
							ust_cmd->sock, ust_cmd->reg_msg.name,
							ust_cmd->reg_msg.major, ust_cmd->reg_msg.minor);

					/*
					 * Lock free enqueue the registration request. The red pill
					 * has been taken! This apps will be part of the *system*.
					 */
					cds_wfcq_enqueue(&thread_state->ust_cmd_queue->head,
							&thread_state->ust_cmd_queue->tail,
							&ust_cmd->node);

					/*
					 * Wake the registration queue futex. Implicit memory
					 * barrier with the exchange in cds_wfcq_enqueue.
					 */
					futex_nto1_wake(&thread_state->ust_cmd_queue->futex);
				} else if (revents & (LPOLLERR | LPOLLHUP | LPOLLRDHUP)) {
					ERR("Register apps socket poll error");
					goto error;
				} else {
					ERR("Unexpected poll events %u for sock %d", revents, pollfd);
					goto error;
				}
			}
		}
	}

exit:
error:
	/* Notify that the registration thread is gone */
	notify_ust_apps(0, is_root);

	ret = close(application_socket);
	if (ret) {
		PERROR("Failed to close application registration socket");
	}
	if (sock >= 0) {
		ret = close(sock);
		if (ret) {
			PERROR("Failed to close application socket");
		}
		lttng_fd_put(LTTNG_FD_APPS, 1);
	}
	unlink(config.apps_unix_sock_path.value);

error_poll_add:
	lttng_poll_clean(&events);
error_listen:
error_create_poll:
	DBG("UST Registration thread cleanup complete");
	if (err) {
		health_error();
		ERR("Health error occurred in %s", __func__);
	}
	health_unregister(health_sessiond);
	return NULL;
}

static bool shutdown_application_registration_thread(void *data)
{
	struct thread_state *thread_state = data;
	const int write_fd = lttng_pipe_get_writefd(thread_state->quit_pipe);

	return notify_thread_pipe(write_fd) == 1;
}

struct lttng_thread *launch_application_registration_thread(
		struct ust_cmd_queue *cmd_queue)
{
	int ret;
	struct lttng_pipe *quit_pipe;
	struct thread_state *thread_state = NULL;
	struct lttng_thread *thread = NULL;
	const bool is_root = (getuid() == 0);
	int application_socket = -1;

	thread_state = zmalloc(sizeof(*thread_state));
	if (!thread_state) {
		goto error_alloc;
	}
	quit_pipe = lttng_pipe_open(FD_CLOEXEC);
	if (!quit_pipe) {
		goto error;
	}
	thread_state->quit_pipe = quit_pipe;
	thread_state->ust_cmd_queue = cmd_queue;
	application_socket = create_application_socket();
	if (application_socket < 0) {
		goto error;
	}
	thread_state->application_socket = application_socket;
	sem_init(&thread_state->ready, 0, 0);

	thread = lttng_thread_create("UST application registration",
			thread_application_registration,
			shutdown_application_registration_thread,
			cleanup_application_registration_thread,
			thread_state);
	if (!thread) {
		goto error;
	}
	/*
	 * The application registration thread now owns the application socket
	 * and the global thread state. The thread state is used to wait for
	 * the thread's status, but its ownership now belongs to the thread.
	 */
	application_socket = -1;
	if (!wait_thread_status(thread_state)) {
		thread_state = NULL;
		goto error;
	}

	/* Notify all applications to register. */
	ret = notify_ust_apps(1, is_root);
	if (ret < 0) {
		ERR("Failed to notify applications or create the wait shared memory.\n"
			"Execution continues but there might be problems for already\n"
			"running applications that wishes to register.");
	}

	return thread;
error:
	lttng_thread_put(thread);
	cleanup_application_registration_thread(thread_state);
	if (application_socket >= 0) {
		if (close(application_socket)) {
			PERROR("Failed to close application registration socket");
		}
	}
error_alloc:
	return NULL;
}