[Openmcl-devel] Apple to switch to Intel - what does this portend?

Hamilton Link helink at sandia.gov
Tue Jun 7 12:39:08 PDT 2005


Iirc that's something I wrote based on conversations with GB. Not to 
put too much in his mouth, but I think the assumption at the time was 
that "porting to x86/P*" meant having a wintel or linux86 version 
maintained concurrently, and there weren't pressing reasons to create 
them.

Now the issue may be that it's still a big pain and will require 
Apple/PPC (jaguar+) and Apple/P* (tiger+) versions to be concurrently 
maintained for a while, but compiler improvements to abstract the lisp 
environment from its compiler back end are desirable anyway, and might 
ease the way to Apple/P* and other targets.

I would be interested in seeing how Rosetta does with openmcl out of 
the box... I wonder if GCs would bollox whatever caching it does.

h

On Jun 6, 2005, at 3:12 PM, David Steuber wrote:

> On Jun 6, 2005, at 4:49 PM, David Steuber wrote:
>
>> Anyway, I'm sorry for spreading any misinformation.
>
> Just as a sort of excuse, here is a direct quote from a knowledgeable 
> person that gave me my impression that OpenMCL would not go to x86:
>
> "Oh, and another good bootstrapping question is "What would be 
> involved in porting openmcl to another platform?" Meaning another risc 
> architecture with plenty of general-purpose registers, like to Sun or 
> SGI boxen? (Porting to x86 or P* probably constitutes yet another 
> question, and a less welcome one, because register allocation is so 
> different. The answer to this one, I know: use some other lisp and 
> write some libraries to ensure you can write to consistent lisp system 
> APIs.)"
>
> It is possible I misread this.
>
> _______________________________________________
> Openmcl-devel mailing list
> Openmcl-devel at clozure.com
> http://clozure.com/mailman/listinfo/openmcl-devel
>





More information about the Openmcl-devel mailing list