Re: ar6000 problems (was: Re: Proposal for further openmoko kernel development)

[Peter Korsgaard]

> "Mark" == Mark Brown  writes:

Hi,

 Mark> Getting the driver into staging is probably still a good idea -
 Mark> even if you can't fix the firmware bugs it'll help everyone
 Mark> with the hardware pool their efforts.

Exactly. I don't have a freerunnner, but will soon start working on
wireless support for another board (s3c6410 based) with an ar6000.

-- 
Bye, Peter Korsgaard

Re: ar6000 problems (was: Re: Proposal for further openmoko kernel development)

[Mark Brown]

On Fri, Aug 21, 2009 at 11:26:36AM +0400, Paul Fertser wrote:
> On Thu, Aug 20, 2009 at 11:48:38PM +0900, Harald Welte wrote:

> > isn't that the ideal candidate for greg-kh's experimental driver tree?  He 
> > says
> > he accepts anything that compiles and is GPL licensed...

> Harald, i hope you know that Werner did a tremendous work trying to improve
> wifi reliability for us, 15000 freerunner users. He came to conclusion that
> even when somebody at least as brave as Hercules (it requires quite some
> courage to deal with large amounts of shit) will "fix" or "rewrite the
> relevant portions of" ar6000 driver we'd be still SOL due to the firmware
> bugs. What bothered him most is the fact firmware halted on any assert()
> because it was compiled with debug options on.

Getting the driver into staging is probably still a good idea - even if
you can't fix the firmware bugs it'll help everyone with the hardware
pool their efforts.