Dan Sugalski <[EMAIL PROTECTED]> wrote:
Copying 640 bytes once, or 640 bytes * 2 * nr of calls? What is inefficient?
This *only* makes a difference for vtable functions written in bytecode. For normal code we're already copying the frames in and out when we make a call and there's no way around that.
Did you even read my proposal? It works[1] for *all* subroutines. I'm not talking about calls from within C.
Yeah, I read the proposal. It's desperately un-thread-safe, which is one of the things that didn't make it out in my last reply. You're moving state data out of the interpreter structure, which is threadsafe and can't be shareable between threads, into the sub pmc, which is shareable. You can't cache state data like this, it means that you can't have two or more threads in the same sub PMC at once. That just won't work.
--
Dan
--------------------------------------it's like this-------------------
Dan Sugalski even samurai
[EMAIL PROTECTED] have teddy bears and even
teddy bears get drunk
