On Monday 24 March 2008 02:59:53 kala mazoo wrote: > ....Faith? No, seriously....point taken, but we do know -something- > ...ie; as you indicated earlier, a few people have reported this > problem before..not just me, and so even though we don't know exactly > why it is so or which hardware combination(s) exhibit this behaviour > ....(yet!).....we>do< know that -some- examples of bcm4318 based wireless > PCI LAN adapters...particularly the asus WL-138G V2....suffer from this > problem in some hardware setups.
So, why don't you collect that info, verify it and put it into the wiki? > Do we know if anybody out there has this card working 'as expected' yet? I have plenty of 4318s working. > If so, what hardware setup are they using? ..... i386 and powerpc. > My point is, anyone going to the b43 website would be forgiven > for presuming that this card *does* work with the b43 driver, because > nothing says otherwise. We are not even sure yet that it's the _card_ that makes it fail. It could also be the architecture. > Sure, a quick check of the bcm43xx-dev-mailing-list will have them > discover > otherwise...(and if they knew this aforehand they'd probably AVOID > buying > one of these card types in the first place)....but *most* people don't > do > things like this - they'll just go to the website, see that bcm4318 is > supported and working and expect things to work... > > So, your question is 'how can we add a warning about something we don't > know?' Well...in the supported/unsupported section of the b43 website, > you could include a paragraph like this ; > > Note that some users of the b43 driver > code are reporting issues with some (not > all?) > bcm4318 based hardware implementations. In > particular, > the ASUS branded cards, model WL-138G V2, > are being > reported as not working by a number of > users, but as > yet we are unsure as to exactly what is > causing this > problem, or in what specific hardware > combinations the > problem manifests itself. This issue is > being actively > investigated and hopefully will be > resolved in the > fullness of time, but for now 'your > mileage may vary' > with this particular brand/model of card. > See the > bcm43xx-dev mailing list archives for > more... In general such warning messages tend to live forever. We still have people telling others that we have TX power issues. The actual issue is gone for a long time, but still people think it's there. That's the reason why I do _not_ put such warnings on the page. As removing them again (from peoples heads) is impossible. > > The driver does not work on x86_64 with IOMMU. That's because the IOMMU > > code is broken. (not a b43 bug) > > > > ....Yes, this machine is an AMD64 (AM2 x2) setup --- I'll hunt down > the exact kernel param to boot with and get back to you with any results... Thanks. PS: Could you please hit the return key at about 80col? It's painful to quote your posts. -- Greetings Michael. _______________________________________________ Bcm43xx-dev mailing list [email protected] https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
