Re: Value range tracing (Michel Schinz)
9 Dec 1995 19:24:50 -0500

          From comp.compilers

Related articles
Value range tracing (1995-12-01)
Re: value range tracing (David Whalley) (1995-12-09)
Re: Value range tracing (1995-12-09)
Re: Value range tracing (1995-12-09)
Re: Value range tracing (1995-12-09)
Value range tracing (Dave Lloyd) (1995-12-09)
Re: Value range tracing (1995-12-09)
Re: Value range tracing (1995-12-09)
Value range tracing (1995-12-09)
Re: Value range tracing (John Gough) (1995-12-09)
Re: Value range tracing (Jason Patterson) (1995-12-09)
[3 later articles]
| List of all articles for this month |

From: (Michel Schinz)
Newsgroups: comp.compilers
Date: 9 Dec 1995 19:24:50 -0500
Organization: Compilers Central
References: 95-12-014
Keywords: analysis

Jeremy Fitzhardinge wrote:
: I'm wondering if people bother with value range tracing. That is,
: keeping track of the possible range of variables at each point in the
: program.

The first (?) version of the Self compiler, written by Craig Chambers,
did so (I don't know if the current Self compiler also does).

This information was used to eliminate array range checks, overflow
tests, and even "constant fold" integer comparaisons. For example, if
"i" is known to be in range 1-10, and "j" in range 20-30, the compiler
could replace an expression like "i < j" by the constant "true".

These techniques are described in Craig's excellent thesis:

Craig Chambers "The Design and Implementation of the SELF Compiler, an
Optimizing Compiler for Object-Oriented Programming Languages". PhD
Thesis, Stanford University, March 1992.


Post a followup to this message

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