nietzsche at lysator.liu.se nietzsche at lysator.liu.se
Tue Mar 1 12:57:46 UTC 2016

OK, this is the pcap dump of the start and connect sequence on Windows 10
with Telenor SE.


> Ive added the NDP fixup to the driver to no avail.
> I'll do some USB-dumping on Windows. Good news is that it support the
> huawei embedded AT commands.
> //M
> Den 1 mar 2016 11:57 skrev Bjørn Mork <bjorn at mork.no>:
>> Markus Gothe <nietzsche at lysator.liu.se> writes:
>> > Bjørn, what's the status of this device? Did people get it working
>> with NDP fixes???
>> No, not as far as I know.  You have probably seen this thread, which is
>> without success so far:
>> http://www.draisberghof.de/usb_modeswitch/bb/viewtopic.php?f=4&t=2485
>> There are similar (or the same?) issues with the E3372:
>> https://bugs.freedesktop.org/show_bug.cgi?id=94340
>> I don't really know what is going on here.  Looks like "NDP to end" is
>> not the (only) problem.  The symptoms are similar to what we saw when I
>> messed up the setup sequence: The modem silently ignores our
>> SET_NTB_FORMAT reuquest and continues to operate in 32bit mode.  So I
>> am
>> wondering if it might be a problem with our setup.  Maybe the modem
>> doesn't like being told to do 16bit mode while the buffers are > 2^16?
>> If you have the hardware and can reproduce the problem, then it might be
>> worth trying to move the initial buffer setup before the SET_NTB_FORMAT.
>> Maybe cdc_ncm_setup(dev) can be moved into cdc_ncm_init(dev)?  I don't
>> remember why I split it like that...  Might have been a good reason.
>> Or not.
>> Bjørn
> _______________________________________________
> libmbim-devel mailing list
> libmbim-devel at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/libmbim-devel
-------------- next part --------------
A non-text attachment was scrubbed...
Name: K5160.pcap
Type: application/vnd.tcpdump.pcap
Size: 35024 bytes
Desc: not available
URL: <https://lists.freedesktop.org/archives/libmbim-devel/attachments/20160301/cd0a927e/attachment-0001.pcap>

More information about the libmbim-devel mailing list