Re: Possible to write compiler to Java VM?

ncohen@watson.ibm.com (Norman H. Cohen)
13 Feb 1996 00:20:56 -0500

          From comp.compilers

Related articles
[25 earlier articles]
Re: Possible to write compiler to Java VM? gmckee@cloudnine.com (1996-02-02)
Re: Possible to write compiler to Java VM? darius@phidani.be (Darius Blasband) (1996-02-04)
Re: Possible to write compiler to Java VM? salomon@silver.cs.umanitoba.ca (1996-02-09)
Re: Possible to write compiler to Java VM? macrakis@osf.org (1996-02-09)
Re: Possible to write compiler to Java VM? shellgate!rgh@uu4.psi.com (Richard G. Hash) (1996-02-09)
Re: Possible to write compiler to Java VM? gmckee@cloudnine.com (1996-02-12)
Re: Possible to write compiler to Java VM? ncohen@watson.ibm.com (1996-02-13)
Re: Possible to write compiler to Java VM? wwgrol@pseserv3.fw.hac.com (W. Wesley Groleau (Wes)) (1997-01-02)
| List of all articles for this month |
From: ncohen@watson.ibm.com (Norman H. Cohen)
Newsgroups: comp.lang.java,comp.compilers,comp.lang.ada
Date: 13 Feb 1996 00:20:56 -0500
Organization: IBM T.J. Watson Research Center
References: 96-01-037 96-02-051 96-02-082
Keywords: C, design

macrakis@osf.org (Stavros Macrakis) writes:


|> There are some unsafe features in Ada other than the explicit ones.
|> For instance, aliasing is "erroneous" (thus allowing efficient
|> Fortran-style optimization of array code in subroutines), but no
|> runtime checking is required.


Not quite. Aliasing in itself has never been erroneous. In Ada 83,
forms of aliasing that caused the effect of a call to depend on the
parameter-passing mechanism were erroneous. In Ada 95, such aliasing
is no longer "erroneous"--meaning that the subsequent behavior of the
computation is undefined--but a "bounded error"--meaning the behavior
must be either that resulting from call-by-reference or that resulting
from copy-in-copy-out.


Uninitialized variables are a better example of a safety breach that
can occur even in an Ada program that does not explicitly use unsafe
features. The Normalize_Scalars pragma in the Ada-95 Safety and
Security annex is designed to increase the likelihood that an
uninitialized variable will raise the exception Constraint_Error
rather than corrupting the computation in unpredictable ways, so that
such errors can be smoked out during testing.


--
Norman H. Cohen ncohen@watson.ibm.com
--


Post a followup to this message

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