E1750 Reloaded - QMI - STATIC - DHCP - IPv4

[Dan Williams]

On Fri, 2014-08-15 at 08:24 +0200, poma wrote:
> On 14.08.2014 18:22, Dan Williams wrote:
> > On Thu, 2014-08-14 at 14:39 +0200, poma wrote:
> >> On 13.08.2014 20:21, Aleksander Morgado wrote:
> >>> On Wed, Aug 13, 2014 at 8:41 AM, poma <pomidorabelisima at gmail.com> wrote:
> >>>>> Yes... :) As I said, if this issue is something that will happen to
> >>>>> all Huawei QMI devices, we should just reflect that in the code,
> >>>>> instead of trying something and falling back to some other method if
> >>>>> it fails.
> >>>>>
> >>>>
> >>>> If you have a preliminary patch don't hesitate to send it for testing purposes. ;)
> >>>
> >>>
> >>> Can you try the branch "aleksander/huawei-qmi-always-dhcp" in the
> >>> upstream MM git repo?
> >>> http://cgit.freedesktop.org/ModemManager/ModemManager/log/?h=aleksander/huawei-qmi-always-dhcp
> >>>
> >>
> >> Yea, it works! :)
> >> http://goo.gl/GQwZzX
> > 
> > I tested today with my E397 and E392 and an Olivetti 200 (not Huawei).
> > All three had no problem with static configuration and that
> > configuration looked similar to poma's results.  However, that doesn't
> > static works everywhere, and there are other reports of static not
> > working with Sierra Qualcomm-based 73xx devices too.
> > 
> > I have just ordered an E1750C but I'm not sure when that'll get here.
> > 
> > Dan
> > 
> 
> And how about to be more plastic reggae fantastic regarding to MMBearerIpMethod, 
> e.g. 
> /etc/udev/rules.d/70-persistent-mmbeareripmethod-static.rules
> ATTRS{idVendor}=="<VID>", ATTRS{idProduct}=="<PID>", ENV{MM_BEARER_QMI_FORCE_DHCP}="0"
> so if necessary, it can be switched to STATIC.

The problem here is that Huawei, ZTE, Sierra, and other vendors use the
same VID/PID for wildly different devices and firmware versions.  So you
simply cannot match on the VID/PID for firmware-specific behavior :(

Dan