Yeah, the rfkill literally makes the chip go away from the PCI lists, it
seems. But your problems can certainly be somehow driver-related, I
haven't actually _tested_ that thing at all, I got the machine just as the
merge window started, so I've had no time to even play with it.
Well, it's also trying to support both 2.4.x and 2.6 etc. Yeah, that is
horrid. At the same time, I can't say that the "rewrite it entirely"
approach of the previous-gen cards has worked very well either, since that
seems to have just perpetuated the problems. It would be great to try to
educate them, but I'm not finding even an email address in the sources.
Oh well. I can't really complain, since just the fact that they bothered
to even make sources available still makes them pretty responsible people,
even if the sources are pretty dang ugly.
All the capitalization etc makes me think they are all old DOS and Windows
programmers. UCHAR and USHORT indeed, along with CamelCase functions.
I dunno.
Linus
--