Huawei E3276 problem

[Dan Williams]

On Mon, 2012-11-12 at 11:00 -0600, Dan Williams wrote:
> On Mon, 2012-11-12 at 11:42 +0100, Olof Ermis wrote:
> > (sry for spamming)
> > 
> > Now it actually connected (from minicom):
> > 
> > AT^NDISDUP=1,1,"online.telia.se"
> > OK
> 
> While this device might be different, I believe the format of the
> NDISDUP command for other Huawei device is:
> 
> AT^NDISDUP=<cid>,<1=start, 0=stop>
> 
> So if you use AT+CGDCONT to define an IPV6 context, and then use
> AT^NDISDUP=<cid of IPv6 context>,1 that'll connect using that context.
> 
> I have no idea of all the google hits for AT^NDISDUP=1,1,"myapn" are
> valid or not, the modem might simply ignore anything after the "1,1",
> but only Huawei can tell us for sure.

Ok, so Franko replied on the NM lists about the correct form of the
AT^NDISDUP command for these devices, and it does appear that all
necessary parameters can be set with NDISDUP.  However, that still
doesn't answer the question of whether the first number is in fact a
CID, and if so, how it plays with AT+CGDCONT.  Previous Huawei modems,
some based on Qualcomm chipsets, appear to operate as I'd described
above, with a simple NDISDUP=<cid>,<start/stop>.

Dan

> Dan
> 
> > ^NDISSTAT:1,,,"IPV4"
> > AT^DHCP
> > OK
> > 
> > The blue led is now constantly on so I guess the only thing that is
> > needed now is the network layer right?
> > 
> > Still no wwan network device.
> > 
> > Cheers
> > Olof
> > 
> > 
> > 2012/11/12 Olof Ermis <olof.ermis at gmail.com>:
> > > Hi!
> > >
> > > Have added the patch now. Attached is a new syslog.
> > >
> > > The only difference now is that the /dev/ttyUSB1 is not created, so
> > > according to what you said Björn that looks like a good thing. I can
> > > send
> > >
> > >  echo -e AT^NDISDUP=1,1,"telia.online.se" > /dev/ttyUSB0
> > >
> > > and /dev/ttyUSB0 I get the command echo'ed back but no other feedback.
> > >
> > > lsmod still lists
> > >
> > > option                 29746  0
> > > usb_wwan               14830  1 option
> > > usbserial              27212  2 option,usb_wwan
> > >
> > > so I can't see any cdc-ncm driver loaded.
> > >
> > > No wwanX device (when I do ifconfig).
> > >
> > > Any additional debug or configuration that I should do here?
> > >
> > > Cheers
> > > Olof
> > >
> > > 2012/11/12 Bjørn Mork <bjorn at mork.no>:
> > >> Olof Ermis <olof.ermis at gmail.com> writes:
> > >>
> > >>> Hi!
> > >>>
> > >>> Okay, back to trying now. Attached is the syslog when I connect the
> > >>> dongle. I have added the
> > >>>
> > >>> echo "12d1 1506" > /sys/bus/usb-serial/drivers/option1/new_id
> > >>>
> > >>> lsmod shows:
> > >>>
> > >>> option                 29746  0
> > >>> usb_wwan               14830  1 option
> > >>> usbserial              27212  2 option,usb_wwan
> > >>>
> > >>> I have not yet applied to cdc_ncm patch. What is the minium kernel
> > >>> version for the patch and get this to work? Right now I run on
> > >>> 3.7.0-rc4, but 3.6 (or even 3.2.x) would be nice due to my Raspbian
> > >>> dist need.
> > >>
> > >> Should work fine on both 3.7.0-rc4 and 3.6.x
> > >>
> > >>> I can only connect / send-commands to /dev/ttyUSB0. /dev/ttyUSB1 is
> > >>> completely dead.
> > >>
> > >> /dev/ttyUSB1 should not be there.  It is really the network interface,
> > >> which the option just steals.  This is expected when you use the new_id
> > >> method and no driver has bound to this interface before.
> > >>
> > >> This will be fixed when the drivers handle this device.
> > >>
> > >>
> > >>> On /dev/ttyUSB0 (using cat /dev/ttyUSB0) I get these status reports
> > >>> continuously, maybe this is the PCUI connection strength status:
> > >>>
> > >>> ^RSSI:29
> > >>>
> > >>> ^RSSI:29
> > >>>
> > >>> ^RSSI:29
> > >>>
> > >>> I can send
> > >>>
> > >>> echo -e ATI > /dev/ttyUSB0 and then I get
> > >>>
> > >>> Manufacturer: huawei
> > >>> Model: E3276
> > >>> Revision: 21.192.03.01.07
> > >>> IMEI: 863781010146687
> > >>> +GCAP: +CGSM,+DS,+ES
> > >>>
> > >>> When i send
> > >>>
> > >>>  echo -e AT^NDISDUP=1,1,"telia.online.se" > /dev/ttyUSB0
> > >>>
> > >>> I have got various responses, (ERROR, +CME ERROR: 50) but sometimes
> > >>> (?) I get an echo back
> > >>>
> > >>> AT^NDISDUP=1,1,telia.online.se
> > >>>
> > >>> Led is flashing blue (I have removed PIN-verification).
> > >>
> > >> Looks good.
> > >>
> > >>> So:
> > >>>
> > >>> - Is the AT-command working as it should and the only thing that
> > >>> remains now is the cdc_ncm-patch (see question on version)?
> > >>
> > >> Looks like it.
> > >>
> > >>> - Is /dev/ttyUSB0 then right port you think?
> > >>
> > >> Yes.  That is the only serial port exposed by the modem, and it is
> > >> intended for device management.
> > >>
> > >>> - What device will be exposed when this work? As a device and/or
> > >>> network interface (I only have eth0 and lo right now and no
> > >>> /dev/cdc-something)? Any additional configuration needed to get that
> > >>> to work?
> > >>
> > >> You will have a wwanX network interface.  The cdc_ncm driver will handle
> > >> that.
> > >>
> > >>
> > >> Bjørn
> > _______________________________________________
> > libqmi-devel mailing list
> > libqmi-devel at lists.freedesktop.org
> > http://lists.freedesktop.org/mailman/listinfo/libqmi-devel
> 
> 
> _______________________________________________
> libqmi-devel mailing list
> libqmi-devel at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/libqmi-devel