At 07:41 PM 9/21/2001 +0100, Simon Cozens wrote:
>On Fri, Sep 21, 2001 at 02:24:43PM -0400, Dan Sugalski wrote:
> > Doing this by hand with -O3, you can see a speedup of around a factor
> of 45
> > over an unoptimised runops loop, so it's definitely worth doing in some
> > cases...
>
>Cool! Parrot JIT!
Well, almost. More like a JIC compiler. :-)
Portable JITs are *tough*. We'll get there though, and if we make sure this
continues to work it'll make writing the JIT and/or TIL code easier.
>But that tells us *just* how time-consuming our runops loop is.
No, not quite. Injecting *just* the function bodies also cuts out the
function call overhead--we don't have to execute the preamble and postamble
code, which is a significant fraction of the code each of our (really
small) functions execute. It also gives the optimizer a lot more to chew
on, so the generated code can be more efficient.
It really tells us just how time-consuming uncertainty is.
Dan
--------------------------------------"it's like this"-------------------
Dan Sugalski even samurai
[EMAIL PROTECTED] have teddy bears and even
teddy bears get drunk
