Re: C++ threads vs. Ada tasks - surprised

["Ira Baxter" <idbaxter@semdesigns.com>]

"Maciej Sobczak" <see.my.homepage@gmail.com> wrote in message 
news:f02a0d10-acaa-47d3-ade8-2e7bd4030bd8@d4g2000yqa.googlegroups.com...

> An important point that was discussed there was the potential gain
> from having tasking built-in the language. I have argued that there is
> none *functional* gain (there is an obvious gain in readability,
> though), or rather that all compiler tricks can be performed with
> standard API as well.

I can't speak for Ada, but it is clear that a compiler that understands
the existence of tasks can generate better code than one can get
from a pure library implementation.  The compiler can preallocate
space for the task stack and task control blocks, can prefill the
areas as needed, can generate specialzed calls on the scheduler,
can optimize away task interactions that it can prove are unnecessary, etc.
Finally, having the compiler know about the tasks can enable easier
expression of the task-based problem, easing coding effort,
debugging and maintenance.

We designed a parallel langauge, PARLANSE, to take advantage of
this for SMP computing.   See 
http://www.semanticdesigns.com/Products/PARLANSE
(It isn't intended to be competition for Ada :)


-- 
Ira Baxter, CTO
www.semanticdesigns.com