JavaMemoryModel: Language semantics cannot depend upon optimization flags

From: Bill Pugh (pugh@cs.umd.edu)
Date: Fri Aug 01 2003 - 21:07:54 EDT


At 3:41 PM -0700 8/1/03, Joseph Bowbeer wrote:
>
>Concerning case 1, this is the kind of analysis that I would expect to be
>associated with some switch (like -o) or some high-end product (like DashO),
>but that I may not want applied to every app that I run in a JVM.
>

Compiler optimization flags should not change the semantics of the
programming language. Otherwise, Java is not a single portable
programming language.

test case 1 has to be either allowed or disallowed. Making it depend
upon a compiler switch is absolutely unacceptable.

        Bill

-------------------------------
JavaMemoryModel mailing list - http://www.cs.umd.edu/~pugh/java/memoryModel



This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:49 EDT