WLAN: towards a solution for the roaming freeze

[Werner Almesberger]

Andy Green wrote:
> However, if they're willing to update the firmware for free and give us
> a Linux firmware update applet or code we can adapt, it can make sense.
> ~ But I think we should find out about that first since "there's no
> point... if it won't pay off for us".

Well, the first step was to report the bug. Let's see what response
we get before lambasting Atheros for supposed future stubbornness.
Maybe there's even one among the gazillion WMI options that prevents
the bug from happening. Or maybe it's an even larger problem and
this is just one way to trigger it.

Also, having a big bug sitting right in the middle of the firmware
might be just the thing it takes to make Atheros their past policy.
I'm building our case here :-)

Meanwhile, I'm checking what other funnies we have there, because
there are things happening that look odd and don't seem to be
among the usual suspects you find in books or papers.

> But, we have every reason to think it'll be effective
> workaround and we can probably stitch it in by ourselves.

We could actually do better and reset the module automatically when
the crash strikes. But I'd rather avoid complexifying our system by
introducing yet another system daemon that ties in with power
management and so on unless this is truly the best choice.

- Werner