Huawei E3276 problem

[Olof Ermis]

2012/11/9 Bjørn Mork <bjorn at mork.no>:
> Dan Williams <dcbw at redhat.com> writes:
>
>> I always wondered what the "Jungo" devices that were mentioned in that
>> driver were.  Now I know :)  I got my driver out of the bits from
>> HuaweiE173Eu-1_Linux.7z which I forget where I found it.
>>
>> It was somewhat interesting that it does QMI in the driver
>> (GetConnectionStatus or something like that) to set the carrier.  Also,
>> any idea what the "VDF" thing means?
>
> According to the Windows driver inf: Vodafone. I guess it's just a
> cosmetical thingy, for branding.
>
>> The driver I'm looking at has
>> matches for:
>>
>> mbb_NDIS_OPTIMIZED_INTERFACE
>> mbb_NDIS_OPTIMIZED_INTERFACE_VDF
>> mbb_NDIS_OPTIMIZED_INTERFACE_JUNGO
>> mbb_NDIS_OPTIMIZED_INTERFACE_VDF_JUNGO
>>
>> but doesn't lock to USB product IDs, just 12d1 + class/subclass/proto.
>
> Yup, that's the same I've seen.
>
> While looking at the "Jungo" Windows driver, I note that there are 5
> protocol variants supported:
>
> %NCMDeviceDesc% = ew_jucdcncm.ndi, USBCDCNCM\VID_12D1&Subclass_02&Prot_16
> %VDFNCMDesc% = ew_jucdcncm.ndi, USBCDCNCM\VID_12D1&Subclass_02&Prot_46
> %ECMDeviceDesc% = ew_jucdcecm.ndi, USBCDCECM\VID_12D1&Subclass_02&Prot_07
> %ECMDeviceDesc% = ew_jucdcecm.ndi, USBCDCECM\VID_12D1&Subclass_02&Prot_11
> %VDFECMDesc% = ew_jucdcecm.ndi, USBCDCECM\VID_12D1&Subclass_02&Prot_37
>
> The ECM variants could maybe be supported by qmi_wwan, if these
> firmwares export QMI (and these devices are Qualcomm based, which I
> don't know if we have any info on either?).

According to some posts/press-release (ie
http://www.mobil.se/nyheter/nytt-snabbare-4g-modem-till-sverige-1.512705.html,
in Swedish) Huawei claims that the chipset is homemade so it could be
so that it is not Qualcomm as was the case with E392/E398 (I believe).
So most likely the issues we have here with E3276 might happen with
many future Huawei modems.

/Olof

>
>
>
> Bjørn