[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Re: Re: [f-cpu] Memory add/sub



hi there,

This instruction (pointer add, could be "padd", or "madd" but
it sounds too crayzy, moomoo...) is annoying. it is doing
what two other instructions already do : add/sub and load->R0.

The ONLY reason i see to include it is that it allows
a "3-address" instruction, allowing some other sofware
constructs that are less constraining. That's all.
Scheduling, latency, etc. : this will change in the future.
Maybe a future architecture will be smart enough to recognize
that an arbitrary operation will create a pointer. Who knows.

I don't see this as important : the whole core is not finished
yet and everybody wants his instruction. Yet, the manual
is very hard to maintain. Then guess why i prefer to concentrate
on the code.

Include this instruction in the set if you want but take your
responsibilities then. If you ask for something that can't
be implemented, i won't do any magic.

'night,
YG (tired).

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