Intel 6250 and WiMax

[PseudoCylon]

> ------------------------------
>
> Message: 13
> Date: Mon, 9 Apr 2012 12:59:43 +0400
> From: Lev Serebryakov <lev at serebryakov.spb.ru>
> Subject: Re: Intel 6250 and WiMax
> To: Bernhard Schmidt <bschmidt at freebsd.org>
> Cc: freebsd-wireless at freebsd.org
> Message-ID: <992091892.20120409125943 at serebryakov.spb.ru>
> Content-Type: text/plain; charset=utf-8
>
> Hello, Bernhard.
> You wrote 9 ะฐะฟั EURO ะตะปั  2012 ะณ., 12:52:01:
>
>>>  Here are driver for Samsung WiMax USB dongle, and it works as-is,
>>>  without any WiMax stack, as simple ethernet NIC. But, maybe, non-usb
>>>  devices are other story, I don't know.
>> Guess it depends on the device driver, it might contain its own
>> stack.
>  It looks like Samsung USB stick contains stack in firmware, because
>  driver is VERY simple and has size of several KILOBYTES of C code,
>  really, only couple of USB commands like "set SSID", "get signal
>  strength", "send frame" and "poll for received frame", and it's all.
>

It wasn't particularly WiMax, but when I worked on HSPA+ device,
usie(4) it worked like that, set up usb xfers and throw ethernet
packets at the device, that was it. Probably, one simple generic stack
would work for most of them if needed.

The problem I had was there is no mechanism to pass device info (i.e
link state up/down, signal strength, IP addresses) to other part of
the system. I thought about adding it, but I haven't because there
isn't lots of demand.


AK