usb/159919: Patch for HUAWEI E173 (u3g/umodem)

[Hans Petter Selasky]

On Saturday 20 August 2011 16:30:17 Poul-Henning Kamp wrote:
> The following reply was made to PR usb/159919; it has been noted by GNATS.
> 
> From: "Poul-Henning Kamp" <phk at phk.freebsd.dk>
> To: Hans Petter Selasky <hselasky at c2i.net>
> Cc: freebsd-usb at freebsd.org, FreeBSD-gnats-submit at freebsd.org
> Subject: Re: usb/159919: Patch for HUAWEI E173 (u3g/umodem)
> Date: Sat, 20 Aug 2011 14:07:25 +0000
> 
>  In message <201108201555.04355.hselasky at c2i.net>, Hans Petter Selasky 
writes:
>  >On Friday 19 August 2011 18:01:29 Poul-Henning Kamp wrote:
>  >>         With this patch I can get a Huawei E173 3G dongle to work.
>  >>         I'm not sure why I have to teach umodem about it specifically.
>  >>         It comes up as /dev/cuaU[012] rather than /dev/cuaU0.[012]
>  >
>  >Hi,
>  >
>  >What happens if you leave out the u3g.c patch?
>  >
>  >Could you post dmesg and output from:
>  >
>  >usbconfig -d X.Y dump_device_desc dump_curr_config_desc show_ifdrv
> 
>  Dang, I should have thought of including that.
> 
>  It's too late now the E173 just passed through my lab for a few
>  hours and it probably won't be back in the forseeable future.
> 
>  Without any patches it comes up with the E173_INIT product-id and
>  da0/cd0 attaches.
> 
>  The HUAWEI_INIT mode just hangs the device, nothing happens or
>  attaches afterwards.
> 
>  With the HUAWEI_INITSCSI it disaspears, then comes back with the E173
>  product-id, da0/cd0 attaches, but umodem for some reason does not.
> 
>  The patch makes umodem find it, but as three separate devices,
>  not as one device with three subdevices.

I found a descriptor dump here:

http://www.draisberghof.de/usb_modeswitch/bb/viewtopic.php?t=552

I see now why the u3g.c driver is not attaching, because "bDeviceClass = 0" 
and we expect 0xFF.

Nick, do you have any comments on this?

PHK: Are you on #bsdusb on EF-net?

--HPS