Hi Guillaume, On 05/19/2011 08:03 AM, Guillaume Zajac wrote: > --- > src/emulator.c | 4 ++++ > 1 files changed, 4 insertions(+), 0 deletions(-) > > diff --git a/src/emulator.c b/src/emulator.c > index c17b901..251bed8 100644 > --- a/src/emulator.c > +++ b/src/emulator.c > @@ -219,15 +219,19 @@ static void dun_ath_cb(GAtServer *server, GAtServerRequestType type, > if (val != 0) > goto error; > > + g_at_server_suspend(em->server); > g_at_ppp_unref(em->ppp); > em->ppp = NULL; > + g_at_server_resume(em->server); > > g_at_server_send_final(server, G_AT_SERVER_RESULT_OK); > break; > > case G_AT_SERVER_REQUEST_TYPE_COMMAND_ONLY: > + g_at_server_suspend(em->server); > g_at_ppp_unref(em->ppp); > em->ppp = NULL; > + g_at_server_resume(em->server); > > g_at_server_send_final(server, G_AT_SERVER_RESULT_OK); > break; This really sounds wrong. I suggest that we not set the disconnect function in GAtPPP when unrefing if GAtPPP is currently suspended.