Related articles |
---|
Ada95 EPIC/IA-64 compilers? cordieritenono@mediaone.net (Keith) (1999-11-28) |
Re: Ada95 EPIC/IA-64 compilers? guerby@acm.org (Laurent Guerby) (1999-11-28) |
From: | Laurent Guerby <guerby@acm.org> |
Newsgroups: | comp.compilers |
Date: | 28 Nov 1999 09:49:34 -0500 |
Organization: | Club-Internet (France) |
References: | 99-11-168 |
Keywords: | Ada, architecture, code |
"Keith " <cordieritenono@mediaone.net> writes:
> I wonder if anyone is working on incorporating the necessary compiler
> features required to take advantage of the EPIC/IA-64 architecture for
> Ada95?
GNAT, the GNU Ada 95 compiler, will benefit from the work done on the
GCC backend-end to support the IA64. I don't know for other
compilers, but I guess there's some work running too.
> It is not just another target. What are the issues of
> designing an Ada95 EPIC/IA-64 compiler? Real-time/multitasking
> considerations
I don't understand what you're referring to here. May be it's critical
for compilers targetting EPIC chips to have imprecise exceptions
authorized by the language, but that's already the case for Ada 95,
see RM 11.6. As for computing WCET safely without being too
conservative, may be the situation is better with EPIC than RISC
(cache problems which are likely to be the same put aside), I can't
say. For multitasking, may be the OS RW/user RO registers will help
some tasking operations to become very fast (like getting the thread
id, min/max stack adresses and things like that), but that's more an
OS issue.
--LG
Return to the
comp.compilers page.
Search the
comp.compilers archives again.