Re: [f-cpu] Manual 0.2.6

[Cedric BAIL <cedric.bail@free.fr>]

> > An idea is to propose a loadchunk that work like loadcons, but with every
> > chunk size. With this we are scalable as much as we want and we can do
> > what we want with any registers size. And if we add this instruction, the
> > loadconsp will be seen like a immediate form of loadchunk.

> > So what about this idea ?
> 
> There is no room in the instruction word. Where should a 64-bit
> constant come from - a register? How does it get there?

Hum, in fact it wasn't clear enough. I mean something like :

loadchunk.b r2, r1

in c that's somtehing like :
	r1 = r1 << 8 | r2.b

The idea is to have this capability for every register size (8, 16, 32, 64).
We need this instructions to load register that are bigger than 64 bits (the
shifter only work on 64 bits, and it isn't its job to work in a inter chunk
capability).

In fact we need the opposite operation :
unloadchunk.b r2, r1

(equal to :
	r2 = r1 & 8
	r1 = r1 >> 8)

Personnaly I didn't like the name because we are thinking it's something
like a memory operation, but I have no better idea.

Cedric
*************************************************************
To unsubscribe, send an e-mail to majordomo@seul.org with
unsubscribe f-cpu       in the body. http://f-cpu.seul.org/