Re: Purify patent (was Re: gawk memory leak)

Hans-Juergen Boehm <boehm@mti.mti.sgi.com>
17 May 1997 00:10:19 -0400

          From comp.compilers
Related articles
[4 earlier articles]
Re: gawk memory leak pfoxSPAMOFF@lehman.com (Paul David Fox) (1997-04-13)
Purify patent (was Re: gawk memory leak) elan@jeeves.net (Elan Feingold) (1997-05-04)
Re: Purify patent (was Re: gawk memory leak) krish@cs.purdue.edu (Sailesh Krishnamurthy) (1997-05-08)
Re: Purify patent (was Re: gawk memory leak) clark@quarry.zk3.dec.com (1997-05-08)
Re: Purify patent (was Re: gawk memory leak) dds@flavors.com (Duncan Smith) (1997-05-09)
Re: Purify patent (was Re: gawk memory leak) pfox@lehman.com (Paul David Fox) (1997-05-13)
Re: Purify patent (was Re: gawk memory leak) boehm@mti.mti.sgi.com (Hans-Juergen Boehm) (1997-05-17)
Partial evaluation vs flow-graph analysis fabre@gr.osf.org (Christian Fabre) (1997-05-22)
Re: Partial evaluation vs flow-graph analysis mossin@diku.dk (1997-05-25)
Re: Partial evaluation vs flow-graph analysis Jacques.Noye@emn.fr (1997-05-25)
Re: Partial evaluation vs flow-graph analysis thorn@spamblock.lalla.irisa.fr (Tommy Thorn) (1997-05-27)
| List of all articles for this month |
From: Hans-Juergen Boehm <boehm@mti.mti.sgi.com>
Newsgroups: comp.compilers
Date: 17 May 1997 00:10:19 -0400
Organization: Silicon Graphics Inc., Mountain View, CA
References: 97-03-165 97-04-020 97-04-022 97-04-037 97-04-070 97-05-019 97-05-090 97-05-162
Keywords: debug, tools, legal

Paul David Fox wrote:


> I dont think it matters whether you instrument object code or .exe
> code for this patent. Its only the two combined acts of
> 'instrumentation' and 'memory leak detection'. (I maybe wrong - I
> cannot remember what I did read).


I don't quite understand this. Their leak detection technique was not
new, at least as far as I know. I believe it is exactly what is
described in Mark Weiser's and my Sept. 88 SP&E paper, combined with
some older techniques.


I'm not sure in what way code instrumentation helps with leak
detection. For leak detection, you just need to intercept malloc, and
add code that does most of what a conservative GC does, together with
some stack backtracing facilities. Our collector has been used for
leak detection. Aside from the mediocre-to-nonexistent stack tracing
code and UI, it does fine.


I do agree that code instrumentation for access checking complements
leak detection.


(I vaguely remember looking at one of the Purify patents a long time
ago. I thought it just dealt with code instrumentation.)


--
Hans-Juergen Boehm
boehm@mti.sgi.com
--

Post a followup to this message

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