On Tue, 30 May 2006 15:28:56 -0400, "Steven M. Bellovin"
<smb@cs.columbia.edu> wrote:

> Thanks to Perry for a clue -- it seems to be a kernel stack overflow.
> Changing UPAGES (in param.h) from 3 to 5 made the problem go away.
> 
I've uploaded some (slightly blurry but -- I think -- readable jpgs of the
backtrace at http://www.machshav.com/~smb/bt1.jpg and
http://www.machshav.com/~smb/bt2.jpg

		--Steven M. Bellovin, http://www.cs.columbia.edu/~smb