Re: Re: Parallelizing (WAS: Death by pointers.)

"John Carter" <ECE@dwaf-hri.pwv.gov.za> writes:


> And this where this diatribe affects this forum... Simple processors
> are MUCH easier to write good compilers for. Far too much effort and
> ingenuity of this community is going into rearranging things to
> fit smoothly into weird pipelines. As a programmer, I don't really
> give a brass fart about pipeline optimization.


> .... I really care about
> reliable compilers. I really care about language design. I really
> care about such things as exception handling, garbage collection etc.
> etc. I heartily wish pipeline optimization would go away and let the
> compiler writers head for the real issues.


[I hope this is not regarded as a polemic -- I kept it as short as
  I could, though I have strong opinions on this subject.]


Vote with your dollars, and I can think of several people who'd be more than
happy to work on the real issues. Till then, the unreal issues (and I do
agree with your judgement here) pay the mortgage. Anyhow, exception handling
is a solved problem (theory and practice), even in the presence of an
optimizer, and garbage collection is near enough (for instance, I know a
good solution to the GC-safety problem, but it needs to be published
first).


And, if you really mean what you say, there are several alternatives
that address the "real" issues, but don't yet have the all-singing-all-dancing
pipeline-tweaking-potato-slicing optimizers that Fortran compilers do.
Off the top of my head, I can think of (my two favorites) Modula-3 and
Scheme (both available for free), Eiffel, Python, Sather, other Lisps, ML,
and Java. If enough people used one of these languages, you'd also get
pipeline optimization, but you wouldn't have to suffer for it.


speaking for myself
(and other frustrated would-be implementors of "real" languages),


David Chase
--

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