[ltp] FIXED (but still need debug help): tpb/nvram bewilderment

Richard Neill linux-thinkpad@linux-thinkpad.org
Mon, 17 Jul 2006 04:24:25 +0100


This is a *really* odd problem.
I found the workaround quite by accident:

Discovery #1)
   Cold-Boot -> KDE   = fail
   But Cold Boot -> ICEWM -> Logout -> KDE = OK

Leading....e.v.e.n.t.u.a.l.l.y...to the discovery that
it is the KDE Battery meter which is to blame!

This is the klaptop batt meter - without any of the ACPI extras (power 
profiles, cpu throttling etc) enabled.

In the broken state, `cat /dev/nvram` does not change in response to any
of the button presses. Furthermore, echo -n "on" > /proc/acpi/ibm/light
also fails. A cold boot is required to re-initialise it.

The important test case is this:

a)Cold boot (kernel 2.6.17); Start KDE; Start KDE batt monitor
     => nvram is unusable, writes to /proc/acpi/ibm fail

b)Cold boot; start kde; ACPI read; ACPI write;
THEN start KDE batt monitor
     => all OK.


By ACPI Read; ACPI write, I use "Press the Fn-f12 button, in response to 
which acpid flashes the thinklight". The physical button press is required!

Any ideas why this might be?

Also, is there another battery meter which isn't the KDE one, but which 
will go in the systray?  Or how to start KDE's batt monitor from a 
shell-script rather than kcontrol?

Thanks for your help,

Richard



Richard Neill wrote:
> Dear All,
> 
> This seems rather odd. Somehow, tpb (on an X20) has stopped working.
> (I've already tried rebooting) - although it worked fine a few days ago. 
> Yes, I have been reconfiguring quite a bit, but nothing as dramatic as a 
> new kernel. I wonder whether anyone has any suggestions.
> 
>   Observations:
> 
> 1)/dev/nvram is still there, with correct permissions (same as on a 
> working A22p)
> 
> 2)tpb is totally dead, although things like Fn-PgUp for the thinklight 
> still work fine.
> 
> 3)Volume up/down don't work any more for volume level - although they do 
> work  for unmute. Volume mute works fine. In either case, there is a 1 
> second delay.
> 
> 4) watch --differences=cumulative -n 0 xxd /dev/nvram
> does not show anything (on the X20, although it does on the A22p).
> 
> Has anyone seen this before?  I'm sure it's my fault for breaking it - 
> but I can't see what I did.
> 
> Thanks,
> 
> Richard
> 
> 
> 

-- 
rn214@hermes.cam.ac.uk  **  http://www.richardneill.org
Richard Neill, Trinity College, Cambridge, CB21TQ, U.K.