Hi.  Since you've been talking about math coprocessor emulation,  I 
thought I'd bring up something I sent to the list some time ago.

	I'm running NetBSD 1.0A on a 486-DLC-40 (no FPU).  I get math errors in 
Maple, MIT Scheme, Netscape (running Java applets), and others.

	For example, when I try to plot something in Maple, the plot window goes 
up and goes black and just when it's about to render:

 math_emulate: 0xd9fa not implemented


And under MIT Scheme:

(/ 5 2)
;Value: 
;Hardware trap SIGSEGV
;Type D to debug error, Q to quit back to REP loop: 

So I go to the debugger:

Hardware trap SIGSEGV (code = 0) at offset #x26
within #[compiled-code-block 6] (runtime/dragon4.inf)


	The first time I had this problem, someone told me to rebuild my kernel 
with options_math_emulate active.  I started doing so, only to realize that 
it's already enabled in my kernel and somebody with a 386 (no fpu) gets the 
same Scheme error with the default 1.1 kernel (math_emulate enabled).

	The strange thing I noticed is that in my kernel config (and the default 
1.1 kernel) the call to a physical math coprocessor (npx0, I think) is not 
commented out.  When I asked Charles Hannum whether commenting that should 
fix my problem, he said to leave it in.

	Any ideas?


						-Dan



-- 


****************************************************************************
>From page 5 of Mandelbrot's *The Fractal Geometry of Nature*:

	Scientists will (I am sure) be surprised and delighted that quite a
few shapes they had to call grainy, hydralike, in-between, pimply, POCKY,
ramified, seaweedy, strange, tangled, torturous, wiggly, wispy, wrinkled,
and the like, can henceforth be approached in rigorous and vigorous
quantitative fashion.
****************************************************************************