Le mercredi 18 octobre 2006 à 23:27 +0200, Didier LINK a écrit : > Le mercredi 18 octobre 2006 à 16:08 -0500, Larry Finger a écrit : > > > > > > > Maybe a race condition during firmware initiation (I've udev-100) > > > > There certainly could be a race condition, but I have no idea how it could > > be triggered by the > > signal-quality patch. Have you seen the firmware load failure before? > > Yes, I've begin to use the bcm43xx driver with a wireless-2.6 from git > repository that correspond to a kernel 2.6.18-rc6. Since the begin I've > always this failures. > > I need to : > * stop the interface (kill wpa_supplicant and ifconfig down) > * rmmod bcm43xx > * modprobe 43xx > * start the interface > > for 4-5 times to have a connection without failure. In fact I need to do > this "procedure" between 0 and 10 times, completely random !
After 2 days of test I can confirm that's this race condition is due to conflict between hotplug and udev firmware loading. I've commented the udev's one and now the interface come up instantly and each time :) Thanks for your support ! Didier Link
signature.asc
Description: Ceci est une partie de message numériquement signée
_______________________________________________ Bcm43xx-dev mailing list [email protected] https://lists.berlios.de/mailman/listinfo/bcm43xx-dev
