Sierra Wireless EM7355

[Bjørn Mork]

Aleksander Morgado <aleksander at aleksander.es> writes:
> On Fri, Jun 20, 2014 at 8:37 PM, Erinn Looney-Triggs
> <erinn.looneytriggs at gmail.com> wrote:
>> [20 Jun 2014, 12:36:33] [Debug] [/dev/cdc-wdm0] Sent message
>> (translated)...
>> <<<<<< Header:
>> <<<<<<   length      = 16
>> <<<<<<   type        = open (0x00000001)
>> <<<<<<   transaction = 3
>> <<<<<< Contents:
>> <<<<<<   max_control_transfer = 4096
>>
>> [20 Jun 2014, 12:36:33] [Debug] [/dev/cdc-wdm0] Received message...
>>>>>>>> RAW: length = 16 data   =
>>>>>>>> 01:00:00:80:10:00:00:00:03:00:00:00:02:00:00:00
>>
>> error: couldn't open the MbimDevice: Failure
>
>
> Oh well... this is the device returning MBIM_STATUS_ERROR_FAILURE
> during the Open sequence... No idea. Maybe the device is already open
> and we shouldn't re-open it?

Hmm, it would not be correct to return failure in that case.  The spec
is quite explicit that the function should try to resyncronize state
again by silently closing the existing session and opening a new one,
returning success.

What the spec says and what the firmware does are of course different
things, but I cannot remember having seen failures in such cases
either.

> Can you try to run it again, but this time like this?
>
> $ sudo mbimcli -d /dev/cdc-wdm0 --query-device-caps --verbose --no-open=10

Yes, that would be useful anyway just to eliminate that cause.

I wonder what else could make the open fail?  I note that the firmware
could have returned MBIM_STATUS_BUSY if that was the problem.  'FAILURE'
seems so much more permanent... 


Bjørn