Hi,
I own a SIS630 based notebook from Baycom (model name "Worldbook II"), which
indeed is a rebrand from a Clevo/Kapok model (normal model name 2700C).
I'm experiencing occasional, unreproducable keyboard lockups. No keyboard
input at all is accepted while the machine itself continues to work
properly (processes continue to write to stdout and I can login and
continue to work through SSH). The freezes occur approx. at least once
a week and sometimes three times a day, so there's no real pattern behind
it.
A friend of mine owns the same model and the same problems occur; I also
found a website stating the same problems on it; so it's of general nature
and not specific to my hardware. I didn't find any information about these
freezes running Windows, so I guess it's Linux-kernel specific and not a sole
hardware issue.
The system freezed on all kinds of kernel, either 2.4 and 2.5, either
vanilla and heavily patched.
So, the big question: Which data would I need to collect on the next freeze
that would allow an experienced kernel hacker to track down the problem?
This nasty bug is the only thing in the way to a perfect Linux notebook,
everything else works just perfect.
TIA,
Moritz
On Thu, Aug 29, 2002 at 09:05:33PM +0200, Moritz Muehlenhoff wrote:
> Hi,
> I own a SIS630 based notebook from Baycom (model name "Worldbook II"), which
> indeed is a rebrand from a Clevo/Kapok model (normal model name 2700C).
>
> I'm experiencing occasional, unreproducable keyboard lockups. No keyboard
> input at all is accepted while the machine itself continues to work
> properly (processes continue to write to stdout and I can login and
> continue to work through SSH). The freezes occur approx. at least once
> a week and sometimes three times a day, so there's no real pattern behind
> it.
>
> A friend of mine owns the same model and the same problems occur; I also
> found a website stating the same problems on it; so it's of general nature
> and not specific to my hardware. I didn't find any information about these
> freezes running Windows, so I guess it's Linux-kernel specific and not a sole
> hardware issue.
>
> The system freezed on all kinds of kernel, either 2.4 and 2.5, either
> vanilla and heavily patched.
>
> So, the big question: Which data would I need to collect on the next freeze
> that would allow an experienced kernel hacker to track down the problem?
>
> This nasty bug is the only thing in the way to a perfect Linux notebook,
> everything else works just perfect.
If it freezes with 2.5.32 (or better the latest BK tree), then #define
I8042_DEBUG_DATA in drivers/input/serio/i8042.h and send me the log of
the last i8042 data before the freeze.
--
Vojtech Pavlik
SuSE Labs
Moritz Muehlenhoff wrote:
> Hi,
> I own a SIS630 based notebook from Baycom (model name "Worldbook II"), which
> indeed is a rebrand from a Clevo/Kapok model (normal model name 2700C).
>
> I'm experiencing occasional, unreproducable keyboard lockups.
Me too (2700T). I was meaning to look into it when I had time,
but since I never do now is as good a time as any :-)
If noticed the lockups occur when the "battery low"
light starts flashing. I.E. it's realted to some
APM event. Also the mouse is unresponsive for me
when this happens.
I upgraded to BIOS 1.10 2 days ago actually, just
in case this would magically fix things so I'll
keep you posted.
P?draig.
In stuga.ml.linux.kernel, you wrote:
>> I own a SIS630 based notebook from Baycom (model name "Worldbook II"), which
>> indeed is a rebrand from a Clevo/Kapok model (normal model name 2700C).
>>
>> I'm experiencing occasional, unreproducable keyboard lockups.
>
> Me too (2700).
>
> If noticed the lockups occur when the "battery low"
> light starts flashing. I.E. it's realted to some
> APM event. Also the mouse is unresponsive for me
> when this happens.
There were some lockups when I connected the AC plug, but these
were rare and totally irreproducable as well.
Alan had the same idea as you and gave me the hint to rebuild
the kernel without APM and APIC. I'm running this configuration
since two hours and I'll see if it works a whole week without
freezes. Maybe you can do the same to eliminate statistical
errors.
Moritz Muehlenhoff writes:
> In stuga.ml.linux.kernel, you wrote:
> >> I own a SIS630 based notebook from Baycom (model name "Worldbook II"), which
> >> indeed is a rebrand from a Clevo/Kapok model (normal model name 2700C).
> >>
> >> I'm experiencing occasional, unreproducable keyboard lockups.
> >
> > Me too (2700).
> >
> > If noticed the lockups occur when the "battery low"
> > light starts flashing. I.E. it's realted to some
> > APM event. Also the mouse is unresponsive for me
> > when this happens.
>
> There were some lockups when I connected the AC plug, but these
> were rare and totally irreproducable as well.
> Alan had the same idea as you and gave me the hint to rebuild
> the kernel without APM and APIC.
Some laptops, including the Dell Inspiron ones, (a) don't tell the
kernel APM when they're about to do APM stuff, and (b) have a buggy
APM and/or SMM implementation that breaks if the kernel has enabled
the local APIC. The workaround is to not enable the local APIC. If
this fixes your machine, please post your machine's DMI strings so
we can update the local APIC blacklist rule set in dmi_scan.
Events that would break the Dells include pulling or inserting the
AC plug, and hitting the magic key combo to enter the BIOS screens.
/Mikael