Hi Denis,
> > ---
> > plugins/ril.c | 19 +++++++++++++++++--
> > 1 file changed, 17 insertions(+), 2 deletions(-)
> >
>
> I went ahead and applied this patch, however, see below:
>
> > diff --git a/plugins/ril.c b/plugins/ril.c index 2298bb6..2ea3fb5
> > 100644
> > --- a/plugins/ril.c
> > +++ b/plugins/ril.c
> > @@ -409,15 +409,30 @@ int ril_enable(struct ofono_modem *modem)
> > return -EINPROGRESS;
> > }
> >
> > +static void power_off_cb(struct ril_msg *message, gpointer user_data)
> > +{
> > + struct cb_data *cbd = user_data;
> > + struct ril_data *rd = cbd->user;
> > + struct ofono_modem *modem = cbd->data;
> > +
> > + if (rd) {
> > + g_ril_unref(rd->ril);
>
> It seems GRil doesn't gracefully handle g_ril_unrefs from within a result
> callback. Have you tried running this with valgrind?
>
> We might need to come up with a fix for GRil...
>
The g_ril_unref should be fine. There are two glib runtime warning messages, it
is because:
In handle_response/gril.c: handle_response, after " req->callback" is called,
"gril" has been cleaned up,
both 'out_queue' and 'command_queue' are NULL.
It seems it's harmless. I will submit a patch to avoid the warnings.
> > + rd->ril = NULL;
> > + }
> > +
> > + ofono_modem_set_powered(modem, FALSE); }
> > +
> > int ril_disable(struct ofono_modem *modem)
> > {
> > struct ril_data *rd = ofono_modem_get_data(modem);
> > + struct cb_data *cbd = cb_data_new(NULL, modem, rd);
> >
> > DBG("%p", modem);
> >
> > - ril_send_power(rd, FALSE, NULL, NULL);
> > + ril_send_power(rd, FALSE, power_off_cb, cbd);
Regards,
Caiwen
_______________________________________________
ofono mailing list
[email protected]
https://lists.ofono.org/mailman/listinfo/ofono
