Re: Ada vs. C performance, was Possible to write compiler to Java VM?

Laurent Guerby wrote:
[...]
> I agree that optimization is a commercial issue, so pinpointing
> early Ada compilers is not really relevant (except for creating myth
> of inefficiency ;-).


Agreed; the quality of optimisation is an economic function, as well
as a technical one. However, doesn't work like ``Frameworks for
Intra- and Interprocedural Dataflow Analysis''
(http://www.cs.washington.edu/research/projects/cecil/www/Papers/engines.html)
foreshadow a day when various optimisation techniques can be easily
used in many compilers? That is, won't the incremental cost of
implementing an optimiser go down radically once frameworks (or
someother reuse mechanism) makes them relatively simple to apply to
many different compilers? Will optimisations that exploit
language-specific features then become too expensive (relatively) to
implement?


> A minor point, it's easy, given a factor k, to write an Ada program
> that will outperform the "same" C program, just put a large local
> array in a procedure and call it a lot, in C it must be initialized to
> zero, and this is not the case in Ada so it will be faster (how
> semantic can affect performance ;-). (I don't think there's a way to
> have non zero'ed data on the stack in C.)


void foo() {
      int a[100];
      /* ... */
}


It may be a minor point, but I object to misinformation being spread
like this. Perhaps this is how myths get started? C does *not*
initialise automatic arrays. Perhaps you are confusing this with
static value initialisation.
--

Return to the comp.compilers page.
Search the comp.compilers archives again.