Bug 19122 - keyboard & touchpad io breaks after short uptime
Summary: keyboard & touchpad io breaks after short uptime
Status: REJECTED INSUFFICIENT_DATA
Alias: None
Product: ACPI
Classification: Unclassified
Component: EC (show other bugs)
Hardware: i386 Linux
: P1 high
Assignee: other_modules
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-09-26 20:26 UTC by A. Stammler
Modified: 2011-03-31 08:52 UTC (History)
3 users (show)

See Also:
Kernel Version: 2.6.35
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description A. Stammler 2010-09-26 20:26:19 UTC
When the netbook has been up and running (GDM with a user logged in) for between 10 and 70 minutes, all keyboard and touchpad input (including the separate power button above the keyboard) becomes impossible. Everything else is unaffected, so 
the system can be shut down via ssh, bash and halt. But at the end of the shutdown process it will not power off, the power LED at the power button remains lit and the other LEDs stay as they were. The computer still will not react to any input. The only way to restart it is to remove both power sources: the mains transformer as well as the battery. When power is restored, the computer will boot if the power button is pressed, but the wifi (WLAN) LED will be off even though it usually comes back on if it was on on shutdown. The num lock and caps lock LEDs will light up erratically a few times, then the system will boot normally.
Comment 1 Len Brown 2010-09-30 02:19:14 UTC
is it possible to boot this system with "acpi=off".
if yes, do you still see this issue, or does it go away?

What is the most recent upstream linux kernel that did
not show this problem?

which netbook are we talking about here?
Comment 2 A. Stammler 2010-09-30 15:01:24 UTC
I'll try to test this setting ASAP. What exactly will it do? I suspect it might switch off
something I need.

I switched from 2.6.26 (which did not show this problem) first to 2.6.32 and then to
2.6.35.

It's a Fox Conn netbook.

At the moment I'm trying to narrow down the circumstances under which the problem occurs,
without much luck yet, though.

Thanks for trying to help.

Axel

On Thu, 30 Sep 2010, bugzilla-daemon@bugzilla.kernel.org wrote:

> https://bugzilla.kernel.org/show_bug.cgi?id=19122
> --- Comment #1 from Len Brown <lenb@kernel.org>  2010-09-30 02:19:14 ---
> is it possible to boot this system with "acpi=off".
> if yes, do you still see this issue, or does it go away?
Comment 3 A. Stammler 2010-10-05 07:22:42 UTC
I'm trying the "acpi=off" setting. It's too early to give any results but the system's behaviour has changed: the hard disk is constantly running (it has never done this before), so this would not not really help even if it solved the other problem.
Comment 4 A. Stammler 2010-10-05 21:24:56 UTC
It seems to work so far, i.e. the problem hasn't occurred with this setting. But the power key now causes the computer to be suddenly switched off whereas so far I have been used to the Shutdown box appearing. Again, I think this is not the solution for this problem.
Comment 5 A. Stammler 2010-10-05 21:28:16 UTC
(In reply to comment #1)
> is it possible to boot this system with "acpi=off".

Do you have any further tips for me at this time?
Comment 6 A. Stammler 2010-10-06 20:07:23 UTC
(In reply to comment #1)
> is it possible to boot this system with "acpi=off".
> if yes, do you still see this issue, or does it go away?
> 
> What is the most recent upstream linux kernel that did
> not show this problem?
> 
> which netbook are we talking about here?

In summary: Yes, it is possible, and the issue seems to go away. The setting seems to cause other problems. One problem I haven't mentioned yet: If I shut down the system via the "halt" command or the shutdown menu item, the operating system shuts down and says "System halted." but it doesn't switch off the computer. All in all, this seems not a good way out. Could you please respond?
Comment 7 Zhang Rui 2011-03-21 06:27:37 UTC
can you get the dmesg output via ssh when the system is broken?
Comment 8 Zhang Rui 2011-03-23 03:07:35 UTC
acpi=off is just used to narrow down the problem.
Using it as a workaround is not acceptable.
so please help me to get the dmesg output after system is broken?
And please attach the acpidump output of this laptop as well.
Comment 9 Zhang Rui 2011-03-31 08:52:45 UTC
please feel free to re-open it if the problem still exists in the latest upstream kernel.

Note You need to log in before you can comment on or make changes to this bug.