gc_sync_ffs now uses gc_sync, gc_posedge and gc_negedge. In the case where g_sync_edge = "negative", gc_sync_ffs uses the falling edge of the clock signal, but the underlying blocks use the rising one.