On Wednesday, November 20, 2002, at 10:08  Uhr, John M McIntosh wrote:

> Ah, so what were the numbers before on your machine, and now. Also 
> compared to the 3.4.0b2 VM that's up there so I can understand
> were 80-95% faster comes from?

Sorry, it was a false-alarm.  It seems that an intermediate version I 
had was exceptionally slow (maybe compiled with gcc 3.1 by mistake or 
something).

Now it's back to 72-73 million bytecodes/s for a 1 GHz G4.

By the way, couldn't passing the pointer to the interpreter struct to 
each function be faster still?  After all, registers are probably 
plentyful, and that would eliminate access to the global structure 
pointer.  Of course, that would mean altering the signature of the C 
functions.

Marcel

-- 
Marcel Weiher				Metaobject Software Technologies
marcel at metaobject.com		www.metaobject.com
Metaprogramming for the Graphic Arts.   HOM, IDEAs, MetaAd etc.