Re: global optimizing compilers vs. unix/make/rcs development methodology

cliffc@rice.edu (Cliff Click)
Mon, 2 Aug 1993 14:57:33 GMT

          From comp.compilers

Related articles
global optimizing compilers vs. unix/make/rcs development methodology merlin@neuro.usc.edu (1993-08-01)
Re: global optimizing compilers vs. unix/make/rcs development methodol cliffc@rice.edu (1993-08-02)
Re: global optimizing compilers vs. unix/make/rcs development methodol preston@dawn.cs.rice.edu (1993-08-02)
global optimizing compilers vs. unix/make/rcs development methodology metzger@bach.convex.com (1993-08-02)
| List of all articles for this month |

Newsgroups: comp.compilers
From: cliffc@rice.edu (Cliff Click)
Keywords: optimize
Organization: Center for Research on Parallel Computations
References: 93-08-007
Date: Mon, 2 Aug 1993 14:57:33 GMT

merlin@neuro.usc.edu (merlin) writes:


      Current unix/make/rcs software developent methodology generally encourages
      a style of coding in which huge computer programs and systems are broken
      up into a large number of very small source modules which may be
      incrementally compiled following source code updated.
      ...
      Is anyone working on updating make or developing compilers which are able
      do global optimization of procedures spanning huge source code trees? How
      is such optimization accomplished by existing and/or planned compilers?


      Alexander-James Annala


At Rice University we have a large, long-standing ongoing research
compiler which does global analysis with incremental compilation.


The "danger" is that you make use of facts from module X to optimize
module Y; later you modify module X and those facts no longer hold.
Rice's solution is to note when the facts change in the "bad" direction
(i.e. they lead to incorrect, instead of just pessimal, code) and force
re-compilation of module Y with the correct facts.


Here is one reference. There are several other papers published
somewhere, but I don't have access to my bibliography right now.


@unpublished{BCKT:Inter,
author={M. Burke and K. Cooper and Ken Kennedy and L. Torczon},
title={Interprocedural Optimization: Eliminating Unnecessary
Recompilation},
note={submitted to TOPLAS},
location=28}




Cliff
--
cliffc@cs.rice.edu -- Massively Scalar Compiler Group
--


Post a followup to this message

Return to the comp.compilers page.
Search the comp.compilers archives again.