panic with today's stable (sort of resolved) PAE related ?

[Mike Tancsa]

One other note about this issue. I have a number of other boxes running 
post PAE commits without problem, but this machine's function is a little 
different in that
a) it does spam scanning via a milter app
b) load averages can spike up dramatically due to bursts of spam.

I occurred to me the other big commits in this period were the libc_r

  MFC: Add wrapper for kqueue() to keep track of the allocated fd and allow
        it to be closed. This fixes a file descriptor leak when closing a
        kqueue() fd.

I have saved all the crash files (7 of them) as well as corresponding debug 
kernels in case other information is required.

         ---Mike

At 12:13 PM 21/08/2003 -0400, Mike Tancsa wrote:

>OK, I cvsup'd back to August 7th (default date=2003.08.07.00.00.00), 
>buildworld / kernel etc and the box has been stable for almost 24hrs 
>now.  Prior to that, it would be rebooting a few times a day.
>
>The panics would be like the one below. I also tried a stable up to the 
>19th, but still panics.
>
>Below are some more panics and some output tegge requested