There is a race condition between the writer thread and stop_threads (via deinit): if deinit sends the signal just at the wrong time, then the worker thread willl never wake up, as the main thread is already waiting for a join to finish. The fix is to lock the queue_mutex before signalling, that eliminates the race. This patch does just that. Signed-off-by: Gergely Nagy <algernon@balabit.hu> --- modules/afmongodb/afmongodb.c | 6 ++++-- 1 files changed, 4 insertions(+), 2 deletions(-) diff --git a/modules/afmongodb/afmongodb.c b/modules/afmongodb/afmongodb.c index 42c140e..dade669 100644 --- a/modules/afmongodb/afmongodb.c +++ b/modules/afmongodb/afmongodb.c @@ -1,6 +1,6 @@ /* - * Copyright (c) 2002-2011 BalaBit IT Ltd, Budapest, Hungary - * Copyright (c) 2010-2011 Gergely Nagy <algernon@balabit.hu> + * Copyright (c) 2010-2012 BalaBit IT Ltd, Budapest, Hungary + * Copyright (c) 2010-2012 Gergely Nagy <algernon@balabit.hu> * * This program is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License version 2 as published @@ -403,7 +403,9 @@ static void afmongodb_dd_stop_thread (MongoDBDestDriver *self) { self->writer_thread_terminate = TRUE; + g_mutex_lock(self->queue_mutex); g_cond_signal(self->writer_thread_wakeup_cond); + g_mutex_unlock(self->queue_mutex); g_thread_join(self->writer_thread); } -- 1.7.8.3