Re: spec for VM or language

From: Bill Pugh (
Date: Fri Jun 25 1999 - 11:47:37 EDT

At 7:50 PM +0000 6/24/99, David G. Wonnacott wrote:
>Is there some reason to consider (a) two provably equivalent models,
>(b) one unified model (the current state of affairs?), and (c) a JVM
>model and a statement that the JLS model is whatever translates into
>the JVM model, but not (d) a JLS model and a rule of the form "any
>compiler/JVM pair must preserve this model".
> >From (d) we would then derive a rule about the standard JVM that
>ensures any bytecodes are compilant, but we also have the option of
>producing more sophisticated compiler/JVM variants, based on future
>experience (these could even become parts of future versions of the
>JVM standard).

The problem is that compilers and VM's don't come in matched pairs.
Bytecode generated by any compiler should run on any JVM.

> Variants would have to tag bytecodes to make sure they
>weren't executed in a mismatched JVM, but I assume there's already a
>mechanism for this.

It wouldn't be valid Java is tags changed the semantics of the bytecodes.

        Bill Pugh

This is the JavaMemoryModel mailing list, managed by Majordomo 1.94.4.

To send a message to the list, email
To send a request to the list, email and put
your request in the body of the message (use the request "help" for help).
For more information, visit

This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:12 EDT