Re: Can anyone explain this?

["David P. Reese, Jr." <daver@xxxxxxxxxxxx>]

On Sun, Nov 16, 2003 at 03:14:09PM -0800, Matthew Dillon wrote:
> :I have a dual P3 VIA motherboard (from Gigabyte) that I'm currently
> :running multiple operating systems from: Slackware 9.1, NetBSD -CURRENT,
> :DragonFlyBSD, BeOS, Windows XP, and Syllable.
> :
> :Of those six, any time I reboot from DragonFlyBSD, I get the following
> :message from the computer when it POSTs.
> :
> :"CMOS/GPNV Checksum Bad"
> :
> :When I first noticed this pattern, I chalked it up to coincidence, and
> :just reset the CMOS to it's default values with F2 and kept going.
> :However, I'm beginning to doubt this is a coicidence any more.  I have
> :spent all weekend rebooting between operating systems, with different
> :hardware configuration nearly every time, and this *only* happens when I
> :reboot from DragonFly. 
> :
> :Does this make any sense?  Could the OS be screwing up my CMOS?
> :
> :Adam
> 
>     Yes, it is screwing up the CMOS.  Several people have reported this,
>     but so far nobody has been able to track it down.  I don't have a
>     system that does this so I haven't been able to track it down either.
> 
>     Would you like to try to track it down?  It involves throwing an endless
>     loop into the low level boot code to lock the machine up and then hitting
>     reset, then moving the loop to zero in on which part of our codebase is
>     corrupting the CMOS.
> 
> 					-Matt

I'll go ahead and do this after I find the bug that killed linux-sun-jdk14.
Consider it penance for the recent bugs introduced in my syscall split
patches.

-- 
   David P. Reese, Jr.                                     daver@xxxxxxxxxxxx
                                               http://www.gomerbud.com/daver/