comparison gtk/gtkmain.c @ 14480:1de5d45426e2

[gaim-migrate @ 17199] This should fix the bug where GStreamer would sometimes fail to initialize correctly. See the insanely long comment I added to the code if you want an explanation. I'm open to suggestions for a better way to fix this. committer: Tailor Script <[e-mail]>

14479:aca9a7b62a23

static SnLauncheeContext *sn_context = NULL;
static SnDisplay *sn_display = NULL;
#endif

#ifdef HAVE_SIGNAL_H
/*
 * Lists of signals we wish to catch and those we wish to ignore.
 * Each list terminated with -1
 */
static int catch_sig_list[] = {

	return ret;
}

#ifdef HAVE_SIGNAL_H
static void
clean_pid(void)
{
	int status;
	pid_t pid;

	do {
		pid = waitpid(-1, &status, WNOHANG);
	} while (pid != 0 && pid != (pid_t)-1);

	if ((pid == (pid_t) - 1) && (errno != ECHILD)) {
		char errmsg[BUFSIZ];
		snprintf(errmsg, BUFSIZ, "Warning: waitpid() returned %d", pid);
		perror(errmsg);
	}
}

char *segfault_message;

static void

	case SIGSEGV:
		fprintf(stderr, "%s", segfault_message);
		abort();
		break;
	case SIGCHLD:
		clean_pid();
		signal(SIGCHLD, sighandler);    /* restore signal catching on this one! */
		break;
	default:
		gaim_debug_warning("sighandler", "Caught signal %d\n", sig);
		gaim_connections_disconnect_all();

#endif

	gtk_main();

#ifdef HAVE_SIGNAL_H
	g_free(segfault_message);
#endif

#ifdef _WIN32
	gtkwgaim_cleanup();

14480:1de5d45426e2

static SnLauncheeContext *sn_context = NULL;
static SnDisplay *sn_display = NULL;
#endif

#ifdef HAVE_SIGNAL_H
static guint clean_pid_timeout = 0;

/*
 * Lists of signals we wish to catch and those we wish to ignore.
 * Each list terminated with -1
 */
static int catch_sig_list[] = {

	return ret;
}

#ifdef HAVE_SIGNAL_H
static void sighandler(int sig);

/**
 * Reap all our dead children.  Sometimes Gaim forks off a separate
 * process to do some stuff.  When that process exits we are
 * informed about it so that we can call waitpid() and let it
 * stop being a zombie.
 *
 * We used to do this immediately when our signal handler was
 * called, but because of GStreamer we now wait one second before
 * reaping anything.  Why?  For some reason GStreamer fork()s
 * during their initialization process.  I don't understand why...
 * but they do it, and there's nothing we can do about it.
 *
 * Anyway, so then GStreamer waits for its child to die and then
 * it continues with the initialization process.  This means that
 * we have a race condition where GStreamer is waitpid()ing for its
 * child to die and we're catching the SIGCHLD signal.  If GStreamer
 * is awarded the zombied process then everything is ok.  But if Gaim
 * reaps the zombie process then the GStreamer initialization sequence
 * fails.
 *
 * So the ugly solution is to wait a second to give GStreamer time to
 * reap that bad boy.
 *
 * GStreamer 0.10.10 and newer have a gst_register_fork_set_enabled()
 * function that can be called by applications to disable forking
 * during initialization.  But it's not in 0.10.0, so we shouldn't
 * use it.
 */
static gboolean
clean_pid(gpointer data)
{
	int status;
	pid_t pid;

	clean_pid_timeout = 0;

	do {
		pid = waitpid(-1, &status, WNOHANG);
	} while (pid != 0 && pid != (pid_t)-1);

	if ((pid == (pid_t) - 1) && (errno != ECHILD)) {
		char errmsg[BUFSIZ];
		snprintf(errmsg, BUFSIZ, "Warning: waitpid() returned %d", pid);
		perror(errmsg);
	}

	/* Restore signal catching */
	signal(SIGCHLD, sighandler);

	/* This timer should not be called again by glib */
	return FALSE;
}

char *segfault_message;

static void

	case SIGSEGV:
		fprintf(stderr, "%s", segfault_message);
		abort();
		break;
	case SIGCHLD:
		if (clean_pid_timeout > 0)
			gaim_timeout_remove(clean_pid_timeout);
		clean_pid_timeout = gaim_timeout_add(1000, clean_pid, NULL);
		break;
	default:
		gaim_debug_warning("sighandler", "Caught signal %d\n", sig);
		gaim_connections_disconnect_all();

#endif

	gtk_main();

#ifdef HAVE_SIGNAL_H
	if (clean_pid_timeout > 0)
		gaim_timeout_remove(clean_pid_timeout);
	g_free(segfault_message);
#endif

#ifdef _WIN32
	gtkwgaim_cleanup();