Re: Broadcom 11n implementation

[Michael Buesch <mb@bu3sch.de>]

On Thursday 25 December 2008 01:05:30 Johannes Berg wrote:
> Hey,
> 
> I hope those of you who got their presents today enjoyed that. Here's
> one from me for the poor souls like myself who have a Broadcom 11n chip:
> 
> I've been publishing specifications for things we have reverse
> engineered, is anyone interested in implementing this? If so, go wild,
> but synchronise on the list if there are multiple people. John, I think
> you had somebody who was interested, but I can't find the email right
> now.
> 
> Michael has said that he might be able to help a bit (review maybe), but
> won't implement the bulk of it. And I actually need this code ASAP since
> I failed to replace the wireless card in my new MacBook (I couldn't even
> find it).

Of course, I will always be available to give advise and help on the technical details.
With the recent merge of the PHY abstraction in b43 it shouldn't be that hard
anymore to understand the basics of the code. I think stuff is clearly modulised now
and fairly easy to understand, after you once understood the basics of the device.

So for implementing N-PHY, you'll basically work inside of the phy_n.c and
phy_n.h files only (maybe 98% of the work will be there).

So, you could probably look into phy_g.c to get an idea of how this could probably
look like.

Please contact us before you start writing patches, so we can give some additional
advise on coding conventions used in the driver.

-- 
Greetings, Michael.