PVS studio check

[shlyakpavel]

Hello. I have just examined the code with PVS studio static analyzer. The code is pretty clean. However, suspicious place was found:

mate-notification-daemon/src/daemon/daemon.c

Lines 1565 to 1590 in c586fe5

	if (!nt->has_timeout || (!screensaver_active (GTK_WIDGET (nw)) && !fullscreen_window_exists (GTK_WIDGET (nw))))
	{
	theme_show_notification (nw);
	if (sound_file != NULL)
	{
	sound_play_file (GTK_WIDGET (nw), sound_file);
	}
	}
	else
	{
	_NotifyPendingClose* data;
	/* The notification was not shown, so queue up a close
	* for it */
	data = g_new0 (_NotifyPendingClose, 1);
	data->id = id;
	data->daemon = g_object_ref (daemon);
	g_idle_add ((GSourceFunc) _close_notification_not_shown, data);
	}
	g_free (sound_file);
	g_object_set_data (G_OBJECT (nw), "_notify_id", GUINT_TO_POINTER (return_id));
	if (nt)

Pay attention to the line 1565 where nt pointer is first used and the line 1590 where it is checked against nullptr. Null pointer dereference is possible.
Thanks for providing us with stable and useful software!
Have a nice day :)

---

Want to back this issue? Post a bounty on it! We accept bounties via Bountysource.

[raveit65]

Pull request please.

[shlyakpavel]

@raveit65 it may require extra investigation after 5 years
I'll check it later