Bug 808237

Summary: System suspends when powering off bluetooth keyboard with kernel 3.3
Product: [Fedora] Fedora Reporter: Roy <nouveau>
Component: gnome-shellAssignee: Owen Taylor <otaylor>
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: high Docs Contact:
Priority: unspecified    
Version: 17CC: beland, maxamillion, otaylor, samkraju, walters
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-02-20 20:08:35 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Roy 2012-03-29 21:54:50 UTC
Description of problem:
A couple of days ago I updated the kernel of my HTPC (nVidia ION) to the 3.3.0-4 release. This kernel comes with amongst many things one particular new feature: the capability of reading the battery of my Logitech MediaBoard Pro bluetooth keyboard (Whoohoow!)
Now I am genuinely happy with this new feature, as I always had to guess the battery level of this device. However it has a serious side-effect.
The battery is registered, as per standard, in /sys/class/battery/hid-MAC_ADDR-... Gnome picks up this battery and shows it in the shell. When I then turn off the keyboard, it either reads 0 or disappears from the system interface (didn't verify), making Gnome think it has reached a critical level. The actions available for Gnome to execute when reaching such a level are either suspend or power off (why is there no "do nothing" option in the menu?! It's Linux!). Result: as soon as I turn off the external keyboard, Gnome suspends.

Version-Release number of selected component (if applicable):
kernel-3.3.0-4.fc16.x86_64
gnome-shell-3.2.2.1-1.fc16.x86_64

How reproducible:
Log in to Gnome using your bluetooth keyboard with battery reporting. Then turn it off (and perhaps on again).

Steps to Reproduce:
1. Boot 3.3 kernel
2. Turn on Logitech MediaBoard Pro
3. log in to the system in Gnome
4. Turn keyboard off
  
Actual results:
A system going into suspend or shutting down, depending on the settings, unable to get it out as the bluetooth keyboard is not working when suspended

Expected results:
A system that lives happily ever after

Additional info:
If the problem seems difficult to fix, I would be satisfied with having the "no action" option in the dropdown list in power settings -> when battery level critical. This at least works around the major issue while you have time to come up with a more sustainable solution. :-)

Comment 1 Sergio Basto 2012-05-03 00:55:29 UTC
Hi, from bug #806548, kernel 3.3.x: terrible slow boot / shutdown in VMWare

kernel-3.3.4-3.fc17 has been submitted as an update for Fedora 17.
https://admin.fedoraproject.org/updates/kernel-3.3.4-3.fc17

Comment 2 Roy 2012-05-03 09:09:39 UTC
Sergio, did you even read the bug description? As much as I appreciate the attention, this problem is in no way related to performance issues in virtual machines.

Comment 3 Sergio Basto 2012-05-03 23:44:40 UTC
(In reply to comment #2)
> Sergio, did you even read the bug description? As much as I appreciate the
> attention, this problem is in no way related to performance issues in virtual
> machines.

ok I read now in diagonal ,in first place, I interpreter this title with some commas like : System stuck when powering off, on bluetooth keyboard, with kernel 3.3.

Sorry !

Comment 4 Roy 2012-05-04 08:33:03 UTC
On a more productive note. On kernel 3.3.4-1 I cannot reproduce this bug any more. Seems the battery of the keyboard is no longer registered in /sys/class/battery . I'm actually guessing this bug is now masked by another one that prevents the battery from being read out. Either that or support is deliberately not compiled along.

Comment 5 Roy 2012-05-31 08:52:08 UTC
Kernel 3.3.7-1.fc17 once again exposes this bug.

Comment 6 Christopher Beland 2013-02-20 20:08:35 UTC
Thanks for this report; looks like a lot of folks are having the same problem, so consolidating bugs.

*** This bug has been marked as a duplicate of bug 816775 ***