[feisty] bcm43xx driver causes hard-lock on association

[Scott Robinson]

The bcm43xx driver causes a hard-lock upon association.

No OOPS is logged, though I haven't tested in a stripped down configuration yet.

[Scott Robinson (scott-ubuntu)]

Multiple OOPses are thrown. At least one inside of an interrupt so I cannot easily log it on this laptop.

[Scott Robinson (scott-ubuntu)]

Linux geneva 2.6.20-8-generic #2 SMP Tue Feb 13 05:18:42 UTC 2007 i686 GNU/Linux

This is with the new dscape bcm43xx driver.

[Scott Robinson (scott-ubuntu)]

Immediately before the OOPs, the following syslog reliably repeats (different specific values, obviously):

Feb 15 03:30:19 localhost kernel: [ 2817.984000] ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready
Feb 15 03:30:19 localhost kernel: [ 2818.020000] ssb: Switching to core 0
Feb 15 03:30:19 localhost kernel: [ 2818.020000] ssb: Switching to core 1
Feb 15 03:30:19 localhost kernel: [ 2818.020000] KERNEL: assertion (!atomic_read(&sk->sk_wmem_alloc)) failed at net/unix/af_unix.c (344)
Feb 15 03:30:19 localhost kernel: [ 2818.020000] KERNEL: assertion (sk_unhashed(sk)) failed at net/unix/af_unix.c (345)
Feb 15 03:30:19 localhost kernel: [ 2818.020000] KERNEL: assertion (!sk->sk_socket) failed at net/unix/af_unix.c (346)
Feb 15 03:30:19 localhost kernel: [ 2818.020000] Attempt to release alive unix socket: c8cffac0
Feb 15 03:30:19 localhost kernel: [ 2818.024000] KERNEL: assertion (!atomic_read(&sk->sk_wmem_alloc)) failed at net/unix/af_unix.c (344)
Feb 15 03:30:19 localhost kernel: [ 2818.024000] KERNEL: assertion (sk_unhashed(sk)) failed at net/unix/af_unix.c (345)
Feb 15 03:30:19 localhost kernel: [ 2818.024000] KERNEL: assertion (!sk->sk_socket) failed at net/unix/af_unix.c (346)
Feb 15 03:30:19 localhost kernel: [ 2818.024000] Attempt to release alive unix socket: c8cffac0
Feb 15 03:30:19 localhost kernel: [ 2818.024000] KERNEL: assertion (!atomic_read(&sk->sk_wmem_alloc)) failed at net/unix/af_unix.c (344)

[Scott Robinson (scott-ubuntu)]

It is similar to the first OOPs in this post:

http://<email address hidden>/msg02799.html

[Michael Buesch (mb-bu3sch)]

If you are using latest feisty shipped kernel, please wait until ubuntu kernel maintainers updated the driver.
It was figured out that there were merge mistakes and current feisty bcm43xx is completely broken.

[Scott Robinson (scott-ubuntu)]

Have the kernel maintainers been notified of the merge mistakes? I don't know the kernel release schedule, but it has been several days.

[Chris Wagner (chris-wagner)]

(poke) - Any word on this? Is bcm43xx still broken in Feisty? Although I'm not having any hard-locks, my bcm43xx-based wireless device does not seem to be usable (with the bcm43xx driver; I've not yet tried ndiswrapper). Thanks.

[Chris Wagner (chris-wagner)]

This may be (is likely to be) the same as bug 85404,

[Launchpad Janitor (janitor)]

Beginning with the Hardy Heron 8.04 development cycle, all open Ubuntu kernel bugs need to be reported against the "linux" kernel package. We are automatically migrating this bug to the new "linux" package. However, development has already began for the upcoming Intrepid Ibex 8.10 release. It would be helpful if you could test the upcoming release and verify if this is still an issue - http://www.ubuntu.com/testing . If the issue still exists, please update this report by changing the Status of the "linux" task from "Incomplete" to "New". We appreciate your patience and understanding as we make this transition. Thanks!