Huawei modem (12d1:1573) resets with AT^GETPORTMODE

[David McCullough]

Aleksander Morgado wrote the following:
> Hey,
> 
> >>
> >> Just tried out a new modem module from Huawei (12d1:1573) and it seems get
> >> quite upset with the AT^GETPORTMODE command.  Without modem manager running
> >> I can issue this command on any of the suitable tty ports and the modem will
> >> HW reset after 15 seconds or so.
> >>
> >> Here is a patch that disables the command on this modem,  not sure if this
> >> is the best way to do it so happy to be pointed in the right direction.
> >
> > So looks like the correct method for doing this is needed :-)
> >
> > The patch below is causing me crashes on a different Huawei modem.  It dies
> > in "mm_device_get_product" because either self is invalid or self->priv is
> > corrupted.
> >
> > Sorry for the bad patch, any help to create a correct fix appreciated :-)
> >
> > Thanks,
> > Davidm
> >
> >> diff --git a/plugins/huawei/mm-plugin-huawei.c b/plugins/huawei/mm-plugin-huawei.c
> >> index 89fe0fc..23cd4eb 100644
> >> --- a/plugins/huawei/mm-plugin-huawei.c
> >> +++ b/plugins/huawei/mm-plugin-huawei.c
> >> @@ -285,6 +285,12 @@ huawei_custom_init_step (HuaweiCustomInitContext *ctx)
> >>          return;
> >>      }
> >>
> >> +     /* this particular modem crashes/resets if you run AT^GETPORTMODE */
> >> +     if (mm_device_get_product (mm_port_get_device (MM_PORT (ctx->port))) == 0x1573) {
> >> +             ctx->getportmode_done = TRUE;
> >> +             ctx->getportmode_retries = 0;
> >> +     }
> >> +
> >>      /* Try to get a port map from the modem */
> >>      if (!ctx->getportmode_done) {
> >>          if (ctx->getportmode_retries == 0) {
> >>
> 
> What's the model name? I've got a Huawei ME906 which shows up with
> that VID/PID pair, firmware revision 11.234.09.05.00, and I haven't
> seen that behaviour myself. Have you tried to install the latest
> firmware?

This is running as delivered in the mail.  Not even sure where I go to
get firmware.

	# mmcli -m 0 --command=i
	response: 'Manufacturer: Huawei Technologies Co., Ltd.
	Model: MU609
	Revision: 12.107.07.05.457
	IMEI: 357784044010427
	+GCAP: +CGSM,+DS,+ES'

some bits from 'mmcli -m 0':

	drivers: 'option1, cdc_ether'
    ports:   'eth1 (net), ttyUSB1 (qcdm), ttyUSB2 (at), ttyUSB0 (at), ttyUSB4 (at)'

Let me know if you want any other info.

> How many ttys does this modem export? Does it also export a wwan iface?

5 ttys and a cdc_ether (eth1) interface,  all up and working fine with
the above patch.

> If we really have to handle this case, we should setup a new udev tag;
> e.g. ID_MM_HUAWEI_NO_PORTMODE, and just initialize
> ctx->getportmode_retries=0 if the tag is found. This would allow us to
> just add a new udev rule to handle new devices showing the same
> symptom, instead of requiring to rebuild the Huawei plugin. But first,
> I would like to confirm that the issue is not a firmware error which
> is already fixed in newer firmware.

Ok,  I had considered that option, seems a bit more involved ;-)

Cheers,
Davidm

-- 
David McCullough,  david.mccullough at accelecon.com,   Ph: 0410 560 763