Re: 4p-1.5-wip with just in time compilation

Coos Haak <chforth@xxxxxxxxx> writes Re: 4p-1.5-wip with just in time compilation
[..]
I remember the posting.

Would that mean that my pityful 16 bit CHForth on this AMD Athlon(TM) XP
2000+ would run faster when I used
XCHG BP,SP PUSH BX XCHG BP,SP POP BX
instead of
LEA BP,-2[BP] MOV [BP],BX POP BX
to do >R etc.? Shorter code altogether.

I am not sure I understand what you mean?

Anyway, my bet would be on

bx -> [bp -2 +] mov,
bx pop,
[bp -2 +] -> bp lea,

but I would test all three of them.

-marcel

.