On Thu, 2 Aug 2007, John W. Linville wrote: > On Thu, Aug 02, 2007 at 11:52:57AM +0100, Andy Green wrote: > > Somebody in the thread at some point said: > > > and, moving on, here's what appears to be the final(?) steps in the > > > recipe (from > > > http://forums.fedoraforum.org/forum/showthread.php?t=156282): > > > > > > > > > # bcm43xx-fwcutter -w /lib/firmware ~fedora/wl_apsta.o > > > # echo 'blacklist bcm43xx' >> /etc/modprobe.d/blacklist > > > # modprobe -r bcm43xx > > > # modprobe -r ieee80211softmac > > > # modprobe -r ieee80211_crypt > > > # modprobe -r ieee80211 > > > # modprobe -r bcm43xx_mac80211 > > > # modprobe bcm43xx_mac80211 > > > > > > "Then I plugged along and enabled NetworkManager and > > > NetworkManagerDispatcher" > > > > > > does all of that sound sane? if so, i'll give it a shot. > > > > Since both the old ieee80211 stack driver and the new mac80211-based one > > both want to own the same device ID at the moment, your workaround > > sounds very needed. > > The Fedora kernels carry a patch to geld the PCI ID table of the > bcm43xx driver. So, all that modprobing and blacklisting should > be unnecessary. but i don't always use fedora kernels, i frequently use the latest source and roll my own, so i'm assuming i should still use the above in those cases, right? and will that fedora patch eventually be pushed upstream? rday
