Re: Generating Java Bytecode

"Bill Mangione-Smith" <billms@ee.ucla.edu>
21 Nov 1996 23:14:18 -0500

          From comp.compilers

Related articles
[9 earlier articles]
Re: Generating Java Bytecode john@dwaf-hri.pwv.gov.za (John Carter) (1996-11-21)
Re: Generating Java Bytecode jhummel@crispix.ICS.UCI.EDU (Joe Hummel) (1996-11-21)
Re: Generating Java Bytecode bmd@cs.kuleuven.ac.be (Bart Demoen) (1996-11-21)
Re: Generating Java Bytecode stephens@math.ruu.nl (Bruce Stephens) (1996-11-21)
Re: Generating Java Bytecode torhr@storm.stud.ntnu.no (1996-11-21)
Re: Generating Java Bytecode kuznetso@MIT.EDU (1996-11-21)
Re: Generating Java Bytecode billms@ee.ucla.edu (Bill Mangione-Smith) (1996-11-21)
Re: Generating Java Bytecode pardo@cs.washington.edu (1996-11-21)
Re: Generating Java Bytecode lynch@frigg.cci.de (Andrew Lynch) (1996-11-24)
Re: Generating Java Bytecode am56@dial.pipex.com (Stefan Heinzmann) (1996-11-24)
Re: Generating Java Bytecode guerby@gnat.com (1996-11-26)
Re: Generating Java Bytecode gvreugde@uwaterloo.ca (1996-11-26)
Re: Generating Java Bytecode jaidi@ubd.edu.bn (Nor Jaidi) (1996-11-26)
[6 later articles]
| List of all articles for this month |

From: "Bill Mangione-Smith" <billms@ee.ucla.edu>
Newsgroups: comp.compilers
Date: 21 Nov 1996 23:14:18 -0500
Organization: UCLA Electrical Engineering
References: 96-11-108 96-11-125
Keywords: Java

Hugo wrote:
> Does anyone have an opinion on generating bytecode for languages other
> than java ?
and John interjected:
> [ ....
> In particular, it specifically doesn't include C's wild 'n' crazy
> pointer rules. -John]


Bill Purvis <w.purvis@daresbury.ac.uk> wrote
> The Java VM does support the basic things you need for C's pointer
> rules - it's the Java Compiler that excludes them.


Bill, are you assuming that the JVM does the type checking specified at
runtime, or leaves it to a verifier at "load" time" (which then could be
bypassed) as is often done. The JVM spec says that you can't add a
constant to a handle. This makes it tough to implement many of the tricks
C programmers do. Not impossible, but much more difficult.


If you assume that the type checks are implemented in the verifier, and
legal JVM code can violate these checks, then things probably get much
easier. But based on the draft spec from last year that I've got it isn't
clear if you would really have a valid JVM program or not. My view, after
reviewing the issue for a recent paper, is not.


Bill
--


Post a followup to this message

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