[Openmcl-devel] A little more LAP stuff...

R. Matthew Emerson rme at clozure.com
Wed Feb 17 19:29:35 UTC 2010


On Feb 16, 2010, at 8:07 PM, Jon S. Anthony wrote:

> On Tue, 2010-02-16 at 20:03 -0500, Jon S. Anthony wrote:
>> 
>> Is the basic difference between ($ n) and ($ 'n), (where n is some
>> positive integer) that the latter is a kind of shorthand indicating that
>> we are talking about fixnums and that this will be turned into the
>> equivalent of ($ (* 4 n)), i.e., fixnum n (where ?? == 32/64)?  While

Yes, ($ 'n) is shorthand for ($ (ash n target::fixnum-shift))

> x8632: ($ (* 4 n))
> x8664: ($ (* 8 n))

That's the effect.

>> the former is literally native value n as represented in full 32 bits?
> 
> And I guess this one should obviously end in "full 32/64 bits?"

Well, a raw immediate operand of the appropriate size.




More information about the Openmcl-devel mailing list