Seems having a read lock still prevents a write lock to be taken, even if the thread...

diff --git a/src/mod/endpoints/mod_rtmp/mod_rtmp.c b/src/mod/endpoints/mod_rtmp/mod_rtmp.c
index 0d83a41f37a65591c4a3a8aca5a67944299b2e0e..1e772866d01cfc0557ae68b2d01878796b39eef5 100644 (file)
--- a/src/mod/endpoints/mod_rtmp/mod_rtmp.c
+++ b/src/mod/endpoints/mod_rtmp/mod_rtmp.c
@@ -838,6 +838,7 @@ switch_status_t rtmp_session_destroy(rtmp_session_t **rsession)
                
                /* At this point we don't know if the session still exists, so request a fresh pointer to it from the core. */
                if ( (session = switch_core_session_locate((char *)key)) != NULL ) {
+                       switch_core_session_rwunlock(session);
                        switch_core_session_write_lock(session);
                        channel = switch_core_session_get_channel(session);
                        tech_pvt = switch_core_session_get_private(session);