Kernel BUG caused by iwlwifi

[John W. Linville]

On Mon, Mar 19, 2007 at 09:12:07AM +0000, Richard Hughes wrote:
> Not sure what to do here - normally I wouldn't dare report
> out-of-vanilla modules for BUG'ing, but this one is being included in
> rawhide and stops the boot process dead.
> 
> Worth filing a RH bugzilla or a kernel bugzilla? This is with latest
> rawhide.
> 
> Thanks,
> 
> Richard.
> 
> Mar 19 08:53:15 localhost kernel: BUG: workqueue leaked lock or atomic:
> iwlwifi/0/0x00000000/1230
> Mar 19 08:53:15 localhost kernel:     last function:
> ipw_bg_reg_txpower_periodic+0x0/0x33 [iwlwifi]
> Mar 19 08:53:15 localhost kernel: 2 locks held by iwlwifi/0/1230:
> Mar 19 08:53:15 localhost kernel:  #0:  (rtnl_mutex){--..}, at:
> [<c061408a>] mutex_lock+0x21/0x24
> Mar 19 08:53:15 localhost kernel:  #1:  (&sysfs_inode_imutex_key){--..},
> at: [<c061408a>] mutex_lock+0x21/0x24
> Mar 19 08:53:15 localhost kernel:  [<c04061ed>] show_trace_log_lvl
> +0x1a/0x2f
> Mar 19 08:53:15 localhost kernel:  [<c04067b1>] show_trace+0x12/0x14
> Mar 19 08:53:15 localhost kernel:  [<c0406835>] dump_stack+0x16/0x18
> Mar 19 08:53:15 localhost kernel:  [<c04351c8>] run_workqueue+0xfe/0x145
> Mar 19 08:53:15 localhost kernel:  [<c0435b6f>] worker_thread+0xf8/0x124
> Mar 19 08:53:15 localhost kernel:  [<c0438353>] kthread+0xb3/0xdc
> Mar 19 08:53:15 localhost kernel:  [<c0405cd7>] kernel_thread_helper
> +0x7/0x10
> Mar 19 08:53:15 localhost kernel:  =======================

>From the iwlwifi guys:

"If the submitter can load the module with the debug=0x43fff module
parameter it would capture some trace information in the log that might help
shed light on where the lock might be being obtained before the delayed work
item fires."

Give that a try and post the results?

FWIW, go ahead and open a bugzilla with me CC'ed.

Thanks,

John
-- 
John W. Linville
linville at redhat.com