Subject: Re: Support for HArdware?
To: None <>
From: Sam Leffler <>
List: current-users
Date: 04/18/2007 22:43:37
Paul Goyette wrote:
> On Wed, 18 Apr 2007, David Young wrote:
>>> I guess it's not likely to be as easy as just adding another device ID
>>> to the code that decides whether or not to claim the device?  :)
>> Since it is Atheros, I do not see why not.
> Well, now the fun begins, because the routine that needs to be updated
> is ath_hal_probe() which is contained in the blob file at
> src/sys/contrib/dev/ath/public/i386-elf.hal.o.u
> I guess it's time to try to contact the author.  Yet while his name is
> given in the README file, there's no E-mail address.
> * Copyright (c) 2002-2006 Sam Leffler.
> * Copyright (c) 2002-2006 Atheros Communications, Inc.
> Hmmm, maybe I'll just go find some other device to use instead.    :)

Given how hard you looked to find me I'm not sure why I'm responding :)

You've chopped context but I think you've got a card based on the latest
11n chipset.  If so the code in netbsd cvs does not support it.  There
is a test hal at that does have support for
operating the card in legacy mode (only) but to use it with netbsd
requires updates to the ath driver.  Unfortunately I've got no time to
deal with those updates.

Beware that if you make the effort to use the hal it's been reported
that h/w crypto is broken.  I'll deal with it sometime late summer;
probably when I bring in all the 11n work I've got.