jensting@diku.dk (Jens Tingleff) (06/02/90)
Does anyone have any experience with/information about the
"portable optimising compiler for Modula-2" described in an article
by Michael L. Powell in SIGPLAN Notices 19:6 (the compiler/article
refered to in the Dragon book) ?
jensting@diku.dk is
Jens Tingleff MSc EE, Research Assistent at DIKU
Institute of Computer Science, Copenhagen University
Snail mail: DIKU Universitetsparken 1 DK2100 KBH O
[I heard him give the paper, which was great, but haven't heard if it's
actually in use. -John]
--
Send compilers articles to compilers@esegue.segue.boston.ma.us
{spdcc | ima | lotus}!esegue. Meta-mail to compilers-request@esegue.
Please send responses to the author of the message, not the poster.scott@cs.rochester.edu (Michael Scott) (06/06/90)
In article <1990Jun1.193655.5436@esegue.segue.boston.ma.us> jensting@diku.dk (Jens Tingleff) writes: | Does anyone have any experience with/information about the | "portable optimising compiler for Modula-2" described in an article | by Michael L. Powell in SIGPLAN Notices 19:6 (the compiler/article | refered to in the Dragon book) ? Yes, we used the compiler extensively at Wisconsin back in the early 80's. It was kind of buggy, as one would expect for an experimental non-product, but it produced reasonable code for the VAX, ran reasonably fast, and had an unusually clean interface to C/Unix. We also used it here at Rochester for a number of years, and wrote a back end for it for the 680x0, allowing us to compile for the Sun-2, Sun-3, and BBN Butterfly. We're now a SPARC shop, mostly, and have abandoned the old compiler. Meanwhile, folks at DEC have been busy inventing Modula-3, and have a very nice compiler for it that is available to universities for little or no fee. Contact kalsow@src.dec.com. -- Michael L. Scott Computer Science Dept. (716) 275-7745, 5671 University of Rochester scott@cs.rochester.edu Rochester, NY 14627 ...!rochester!scott -- Send compilers articles to compilers@esegue.segue.boston.ma.us {spdcc | ima | lotus}!esegue. Meta-mail to compilers-request@esegue. Please send responses to the author of the message, not the poster.
grunwald@foobar.colorado.edu (Dirk Grunwald) (06/06/90)
>>>>> On 1 Jun 90 19:36:55 GMT, jensting@diku.dk (Jens Tingleff) said:
JT> Does anyone have any experience with/information about the
JT> "portable optimising compiler for Modula-2" described in an article
JT> by Michael L. Powell in SIGPLAN Notices 19:6 (the compiler/article
JT> refered to in the Dragon book) ?
Whilest employeed at DEC-ERL one summer, Bill Bush & I beat (slightly) on this
compiler. I translated the internal tree representation into common lisp
lists, from which we were doing dataflow analysis for automatic
parallelization.
From my remembering, it was a very clean compiler internally (thank goodness),
much cleaner than e.g., Berkeley 'pc' & you could even understand the
optimization strategies. Amazing. Various small benchmarks showed that the
compiler did perform very well on VAXen.
Sadly, it all came to naught as DEC blew ERL out of the water that summer. You
can get a Modula-3 to C translator from gatekeeper.dec.com. Don't know if
it's the same structure, though.
Dirk Grunwald -- Univ. of Colorado at Boulder (grunwald@foobar.colorado.edu)
(grunwald@boulder.colorado.edu)
--
Send compilers articles to compilers@esegue.segue.boston.ma.us
{spdcc | ima | lotus}!esegue. Meta-mail to compilers-request@esegue.
Please send responses to the author of the message, not the poster.moss@cs.umass.edu (Eliot Moss) (06/08/90)
A couple of notes regarding Dirk's mention of a "Modula-3 to C translator":
1) It is *not* a simple translator, but a true compiler whose target language
happens to be C rather than assembly. (Dirk may have known that; the word
translator is a little ambiguous, though, so I thought I would offer
clarification).
2) The Modula-3 compiler is, I believe, descended from a Modula-2+ compiler at
DEC SRC (with earlier work also done at DEC WRL, I believe). Greg Nelson
(gnelson@src.dec.com) can tell you more about its relationship to other
compilers developed in DEC's research labs.
The Modula-3 compiler is alive and well and almost up to version 1.5. I
suspect it is not highly optimizing, but relies on the C compiler for a lot of
the optimizations. My research group at UMass is working on Gnu Modula-3 using
the Gnu C back end, so it will be a native optimizing compiler (first target
is VAX, but gm3 should be about as easily ported as gcc). Cheers!
--
J. Eliot B. Moss, Assistant Professor
Department of Computer and Information Science
Lederle Graduate Research Center
University of Massachusetts
Amherst, MA 01003
(413) 545-4206; Moss@cs.umass.edu
--
Send compilers articles to compilers@esegue.segue.boston.ma.us
{spdcc | ima | lotus}!esegue. Meta-mail to compilers-request@esegue.
Please send responses to the author of the message, not the poster.