Bug 34272

Summary: Kernel panic when issuing "echo disabled > /proc/acpi/ibm/bluetooth" during startup
Product: Drivers Reporter: Stefan Bauer (stefan.andreas.bauer)
Component: BluetoothAssignee: drivers_bluetooth (drivers_bluetooth)
Status: CLOSED UNREPRODUCIBLE    
Severity: normal CC: florian, maciej.rutecki, rjw
Priority: P1    
Hardware: All   
OS: Linux   
Kernel Version: 2.6.39-rc5 Tree: Mainline
Regression: Yes
Bug Depends on:    
Bug Blocks: 32012    
Attachments: lsusb -v
The promised photo showing the panic message
Kernel .config

Description Stefan Bauer 2011-05-03 05:29:55 UTC
Created attachment 56292 [details]
lsusb -v

I've configured acpid to issue the above given command when I press Fn+F6 to disable the BT module built into my laptop. Running 2.6.39-rc5, the kernel panics when I press Fn+F6 during startup, preferably when X/gdm is starting. I wasn't able to reproduce the problem once the system was "settled".

I never faced this problem in 2.6.38 or any prior version.

Attached you find a photo (sorry for the bad quality, but it should be readable) of the messages and the lsusb output.

Steps to reproduce:
- Turn on the laptop. The bluetooth module is activated by default.
- Press Fn+F6 after acpid is started, i.e. while gdm is starting
- watch the kernel panic

Hardware:
- Thinkpad Z60m

Software:
- Gentoo Linux
- if you need version numbers of some particular userspace programs, let me know
Comment 1 Stefan Bauer 2011-05-03 05:31:00 UTC
Created attachment 56302 [details]
The promised photo showing the panic message
Comment 2 Stefan Bauer 2011-05-03 05:33:40 UTC
Created attachment 56312 [details]
Kernel .config
Comment 3 Stefan Bauer 2011-05-03 05:58:10 UTC
OK, now I was able to trigger the bug during a running KDE session. So I withdraw my statement that the problem isn't reproducible once the boot process is finished.

I also suspect this bug being the reason that entering Suspend to RAM fails with a flashing Caps Lock LED since 2.6.39-rc5. Suspend to RAM worked flawlessly with .38.
Comment 4 Rafael J. Wysocki 2011-05-04 16:38:47 UTC
Does 2.6.39-rc6 fix it?
Comment 5 Stefan Bauer 2011-05-05 01:26:02 UTC
(In reply to comment #4)
> Does 2.6.39-rc6 fix it?

Apparently, yes. I can't reproduce the panic with 2.6.39-rc6, and suspend works again. Thank you!
Comment 6 Florian Mickler 2011-05-09 09:32:55 UTC
Thanks for the update.