Xref: utzoo comp.arch:3886 comp.misc:2080
Path: utzoo!mnetor!uunet!husc6!tut.cis.ohio-state.edu!mailrus!ames!sgi!bron
From: bron@olympus.SGI.COM (Bron C. Nelson)
Newsgroups: comp.arch,comp.misc
Subject: Instruction Scheduling
Message-ID: <12512@sgi.SGI.COM>
Date: 10 Mar 88 17:19:37 GMT
References: <477@imagine.PAWL.RPI.EDU>
Sender: daemon@sgi.SGI.COM
Organization: Silicon Graphics Inc, Mountain View, CA
Lines: 31
Keywords: optimization, pipe-line interlocks, code re-organization
Summary: Starting a new topic

In article <477@imagine.PAWL.RPI.EDU>, jesup@pawl23.pawl.rpi.edu (Randell E. Jesup) writes:
> I think these issues have been beaten to death.  Anyone want a new subject,
> like reorganizing algorithms? (my specialty :-)  Is there anyone else out
> there that does linking before reorganization?  Or given thought to doing
> optimal (instead of near-optimal) reorganization in small code blocks?
> 

I think this is a fine idea.  Instruction scheduling/reorganization
is a neat field that still has lots of room for good ideas.  Also,
it is destined to become an increasingly "hot" topic as memory
latencies increase and asynchronous funtional units become more
and more common even in the micro world (the MIPS R2000 for example
has autonomous integer multiply and divide hardware that can
run overlapped with other "standard" ALU ops, and of course the
floating point co-processor runs autonomously as well).

This isn't really an architectural issue, so probably doesn't
really belong in comp.arch.  However, it does have interesting
architectural impact.  (For example, a machine similar to the
Multi-FLow VLIW (tm I think) machine could probably have been
BUILT years ago, but without something like a trace-scheduler,
the machine would be worthless.)  Also, there doesn't really
seem to be a good place to discuss this except perhaps comp.misc
(or comp.compilers, but I'd rather avoid moderation :-) at least
to start).  Ergo, I will post my first few articles on
this topic to comp.arch, cross-posted to comp.misc.  If enough
comp.arch people complain, we'll use comp.misc exclusively.

Bron Nelson  bron@sgi.com
Silicon Graphics Inc.
Don't blame my employers for my opinions