No, its no good to write extra machine descriptions for these, as they are
basically x86 machines. The x86 backend, though, will be (and already is
being) reworked.
egcs already has, and future gcc versions will, too, have -mk6 and -march=k6
switches.
> - - have the compiler do more efficient multipipelining, this
> will be especially useful for the Rise chip
work in progres, look at latest egcs.
> - - replace one expensive opcode by multiple (parallelizible)
> cheap ones when useful (-frisc for pgcc???)
yes, being worked on ;)
> - - make egcs do better instruction scheduling by avoiding
> unneeded branches and/or doing predicative branching
> (a'la Merced) in software (if it's worth it at all:-)
egcs already can take limited advantage of profile data, for example, and it
is thought about giving the compiler enough intelligence (and assistance) to
optimize fast paths without goto.
-----==- |
----==-- _ |
---==---(_)__ __ ____ __ Marc Lehmann +--
--==---/ / _ \/ // /\ \/ / pcg@goof.com |e|
-=====/_/_//_/\_,_/ /_/\_\ --+
The choice of a GNU generation |
|
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/