[PRL] Keith Cooper @ Harvard 10/16

Mitchell Wand wand at ccs.neu.edu
Thu Sep 25 13:25:30 EDT 2003


Title:     Keith Cooper, Rice University

Date:      October 16, 2003

Old Location: 

New Location: Maxwell Dworkin G125, (Ice Cream at 3:30PM - Maxwell
Dworkin 2nd Floor Lounge Area)

Time:      4 p.m. - 5 p.m.

Harvard University
33 Oxford St., 
Cambridge, MA 02138
tel: (617) 496-1440  fax: (617) 495-9837

Colloquium

Rethinking Compiler Structure


  	  
Abstract

For more than forty years, we have built compilers that follow a
simple structural model: they apply a fixed set of passes to every
program.  A typical compiler has a handful of optimization levels that
add extra passes into the process. This approach has allowed us to
make progress on compilation and optimization.  It has allowed us to
build and debug compilers.  However, it has not placed us in a
position where we can deliver high-quality compilers in a timely
fashion. Despite forty years of research on optimization and code
quality, we still have trouble producing compilers that generate
excellent code for new architectures and new applications.

One area that research has largely avoided is the structure of the
compilers that we build.  Modern compilers are organized along the
same basic lines that were used in the first Fortran compiler, in the
late 1950s.  The time has come to fundamentally rethink the way that
we organize and execute optimizing compilers.  In our research, we are
building and evaluating an adaptive compiler.  This compiler changes
its behavior in response to both the application and the target
machine. It uses a simple feedback mechanism in an attempt to minimize
an explicit objective function, such as execution time or code size.

This talk will present preliminary results from some large scale
experiments with our prototype system. It will discuss some of the
engineering challenges that adaptivity presents. It will discuss the
impact that our results should have on the way that we build and
structure compilers.  It will suggest directions for future research
in adaptive compilation.

This talk should be accessible to a broad audience.  

Host:  Professor Michael Smith


Calendar URL
- ------------
http://www.fas.harvard.edu/

***********************************************************



More information about the PRL mailing list