Skip to contentSkip to conversation

Problems with b43 using BCM4306 r3

DISCUSSION Doug McLain 2 participants 4 messages Dec 10, 2007 – Dec 12, 2007

lore
Share & tools
lore
#0 lore raw

I have an emachines M6805 with the BCM4306 r3 wifi chip. Using the ndiswrapper driver, all is ok, but using the b43 driver results in constant flow of dmesg output (see link), eventual lockup of wifi under heavy network traffic, and constant xruns when using JACK audio server.
Again, when I use the ndiswrapper driver, everything works ok.

http://nostar.net/ardour/dmesg.txt

#1 lore raw
Quoted 7 lines · On Monday 10 December 2007 21:33:03 Doug McLain wrote:

I have an emachines M6805 with the BCM4306 r3 wifi chip. Using the ndiswrapper driver, all is ok, but using the b43 driver results in constant flow of dmesg output (see link), eventual lockup of wifi under heavy network traffic, and constant xruns when using JACK audio server.
Again, when I use the ndiswrapper driver, everything works ok.

http://nostar.net/ardour/dmesg.txt

That's weird. These messages say that interrupts are not disabled in the IRQ handler. Which driver are you using? Please try vanilla wireless-2.6.

Signature · 1 line
Greetings Michael.
#2 lore raw

The problems only exist on the planet CCRMA

http://ccrma.stanford.edu/planetccrma/software/

low-lat patched kernel that I was using. I'm now running a stock Fedora 8 2.6.23.8-63.fc8 kernel and there dont appear to be anymore wifi problems. My reasons for using the ccrma kernel werent low latency performance so much as a working firewire stack (fedora 7 and above uses a new, broken stack) but I found another method.

Quoted 2 lines · Michael Buesch
Quoted 1 line · On Monday 10 December 2007 21:33:03 Doug McLain wrote:

I have an emachines M6805 with the BCM4306 r3 wifi chip. Using the

ndiswrapper driver, all is ok, but using the b43 driver results in constant flow of dmesg output (see link), eventual lockup of wifi under heavy network traffic, and constant xruns when using JACK audio server.

Quoted 4 lines
Quoted 2 lines

Again, when I use the ndiswrapper driver, everything works ok.
http://nostar.net/ardour/dmesg.txt

That's weird. These messages say that interrupts are not disabled in the

IRQ handler. Which driver are you using? Please try vanilla

Quoted 2 lines

wireless-2.6.

#3 lore raw
Quoted 9 lines · On Wednesday 12 December 2007 01:08:19 [email protected] wrote:

The problems only exist on the planet CCRMA

http://ccrma.stanford.edu/planetccrma/software/

low-lat patched kernel that I was using. I'm now running a stock Fedora 8 2.6.23.8-63.fc8 kernel and there dont appear to be anymore wifi problems. My reasons for using the ccrma kernel werent low latency performance so much as a working firewire stack (fedora 7 and above uses a new, broken stack) but I found another method.

Ah, well. The driver is not supposed to work with this kind of stuff.

Signature · 1 line
Greetings Michael.