Re: 5.0/i386 regressions

[Daniel Horecki <morr+lists%morr.pl@localhost>]

Je Tue, 27 Jan 2009 10:11:05 +1100 matthew green <mrg%eterna.com.au@localhost>
scribis:

>    
>    I have encounted this problem too. I've learned that disabling
>    ichlpcib* allows to shutdown cleanly. 
>    
>    Fragments from dmesg:
>    
>    ichlpcib0 at pci0 dev 31 function 0
>    ichlpcib0: vendor 0x8086 product 0x24cc (rev. 0x01)
>    timecounter: Timecounter "ichlpcib0" frequency 3579545 Hz quality
> 1000 ichlpcib0: 24-bit timer
>    ichlpcib0: TCO (watchdog) timer configured.
>    ichlpcib0: SpeedStep enabled
>    
>    But Celeron-M used in my Thinkpad doesn't have SpeedStep.
> 
> 
> could you try some tests for this?
> 
> in sys/arch/x86/pci/ichlpcib.c lpcibattach() function, try 
> commenting out each of the 4 things this driver sets up:
> 
>       pmtimer_configure(self);
> 
>       tcotimer_configure(self);
> 
>       speedstep_configure(self);
> 
>       lpcib_hpet_configure(self);
> 
> 
> if you can test multiple times with only one commented, or
> only one left in, and figure out which part causes the
> reboot to fail, that would be very helpful.
> 
> thanks,
> 

I've just tried that. :-)

Commenting only speedstep_configure seems to be working (I can shutdown
clean). I haven't tried commenting any other functions. So,
speedstep_configure shouldn't be running if CPU is Celeron-M.

Daniel

-- 
Daniel Horecki
http://morr.pl http://linux.pl http://netbsd.pl
HAIL ERIS!