On Thu, Apr 01, 1999 at 01:53:31PM +0200, Manuel Bouyer wrote:
> Well, I looked more at this yesterday evening. I ran several hardware
> tests programs which didn't find anything wrong on my box. Then I opened
> it and removed and put back the SIMMS.
> I couldn't reproduce the pmap_zero_page panic after that, but still got a core
> dumps from cc1.
> 
> Unless someone else get this, I think this can be classified as
> "hardware problem". Sorry for the noise, and thank to the people (Chuck
> and Jason) who spent some time on it.

No, it's not.  I can't reproduce it any more (dammit) but I was getting it
regularly about a week ago.  I have a machine with a large FFS filesystem I
burn ISO9660 disks from.  Running mkhybrid would cause a pmap_zero_page panic
when the mkhybrid program exited, unless I typed "sync" after it did its last
disk write but before it had printed its final messages and exited.

As soon as I tried to track it down the problem went away -- but I definitely
saw it.  This was on hardware which had been 100% reliable for several months.

Thor