Andrew Whitworth wrote:
> Very interesting, all. However, I have two questions:
> 
> 1) What's the benefit to passing around all these C function pointers
> (which, I notice, aren't properly parsed in ATTR definitions yet),

Yeah, aren't they beautiful?  I can't actually use the typedefs for the
attr types either, because the typedef lines don't show up in the
autogenerated header file.  I created TT #544 for pmc2c naivety and will
get back to that stuff later.


> above creating a subclass of ManagedStruct to use
> executable-code-specific clone() and destroy() methods? Asked a
> different way, how many different configurations do we forsee having,
> and are they enough to justify tacking this kind of ability onto the
> ManagedStruct PMC?

Apparently, yes.  I think chromatic wants to make this stuff generally
available to NCI bindings, and possibly other places.  Please see
http://irclog.perlgeek.de/parrot/2009-04-07#i_1045352 for the (rather
disorganized, spread out and entangled) design discussion.


> 2) Is ManagedStruct really the best thing to be holding JIT-related
> code pages? I can think of other jobs that a real JITBuffer PMC could
> be used for eventually besides simply holding, cloning, and freeing
> code pages.

I've no idea.  At the time, we looked at ManagedStruct (which has the
benefit of being what nci jit was already using), CPointer, subclasses
of those, and entirely new PMCs.  It's also possible to argue against
the complexity of ManagedStruct directly:
http://irclog.perlgeek.de/parrot/2009-04-07#i_1045886

So I dunno.  I've done a fair amount of work on this, but I'm certainly
still open to ideas. :)

Mark
_______________________________________________
http://lists.parrot.org/mailman/listinfo/parrot-dev

Reply via email to