Huawei me906s-158

[Bjørn Mork]

Andreas Fett <andreas.fett at secunet.com> writes:
> On 04/03/16 15:33, Bjørn Mork wrote:
>> Don't know what we can do about that.  If you have Windows, and it uses
>> MBIM, then it would be very useful to get a dump of that (use the open
>> source http://desowin.org/usbpcap/ tool).
>
> Attached you will find a pcap file from the same system as before
> running win 10 (64bit). I did a connect and disconnect of the modem.
> Windows obviously did some network traffic on it's own.
>
> Hope this helps.

Thanks.  At least it shows that Windows has no problems talking MBIM to
the modem.  That's very useful to know.

I was hoping for a capture that would show any interesting differences
in the setup sequence, which could possible explain why the modem
appears dead in Linux.  But I realize that this is difficult to capture
for an internal device on Windows.

Never mind. It's a long shot anyway. It would be the first time such
problems affected the control channel.  Not very likely.

Since the modem definitely works in MBIM modem, it's more likely that
there is something interfering with the MBIM control channel in Linux.
For example some unknown process reading from the character device, or
something sending unknown commands to it which makes it lock up.  We
have seen examples of both those problems in the past.

Use "lsof /dev/cdc-wdm0" to make sure there are no unwanted listeners.

Make sure nothing probes the device for any other protocol than MBIM
(some Sierra Wireless Qualcomm based devices are for example known to
switch control protocol to whatever probes it first, regardless of USB
descriptors saying MBIM).

Yes, I am running out of ideas...



Bjørn