GCC 3.0/Java and control flow analysis?

GCC 3.0/Java and control flow analysis?

[Kai Hofmann]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 821 bytes --]

Hello,

I am working on JShopz 3.0 (a Java based shop under GPL). Because I changed the
base architecture heavily I am searching for a java compiler that does a control
flow analysis and tells me about unused variables, method parameters etc. like
C/C++ did.
So my question is does support GCC 3.0 a control flow optimization for Java?

Please let me know - thanks in advance


Kai Hofmann
begin:vcard 
n:Hofmann;Kai
tel;pager:+49 1681 1764681
tel;cell:+49 174 1419312
tel;fax:+49 421 22096-55
tel;home:+49 421 4844966
tel;work:+49 421 22096-83
x-mozilla-html:FALSE
url: http://www.isl.org/
org:Institut für Seeverkehrswirtschaft und Logistik (ISL);Telematik
adr:;;Universitätsallee GW1 Block A;Bremen;Bremen;28359;Deutschland
version:2.1
email;internet:hofmann@isl.org
title:Dipl.-Inform.
fn:Kai Hofmann
end:vcard

Re: GCC 3.0/Java and control flow analysis?

[Alexandre Petit-Bianco]

Kai Hofmann <hofmann@isl.org> writes:

> I am working on JShopz 3.0 (a Java based shop under GPL). Because I
> changed the base architecture heavily I am searching for a java
> compiler that does a control flow analysis and tells me about unused
> variables, method parameters etc. like C/C++ did.

Gcj inherits from gcc all the infrastructure required to do that. It
should even work with the current compiler, but it actually
doesn't. I'm looking into this...

./A

Re: GCC 3.0/Java and control flow analysis?

[Alexandre Petit-Bianco]

Alexandre Petit-Bianco <apbianco@cygnus.com> writes:

> Gcj inherits from gcc all the infrastructure required to do that. It
> should even work with the current compiler, but it actually
> doesn't. I'm looking into this...

Like this should help, but it's not entirely tested:

  apbianco@venonat[~/tmp]: cat bar.java
  class bar {
      void foo () {
          int i;
          System.out.println ();
      }
  }
  apbianco@venonat[~/tmp]: gcj bar.java -Wunused -c
  bar.java: In class `bar':
  bar.java: In method `bar.foo()':
  bar.java:3: warning: unused variable `i'

./A

2001-08-09  Alexandre Petit-Bianco  <apbianco@redhat.com

	* expr.c (java_lang_expand_expr): Call `expand_end_bindings' and
	`poplevel' in the right order.

Index: expr.c
===================================================================
RCS file: /cvs/gcc/gcc/gcc/java/expr.c,v
retrieving revision 1.114
diff -u -p -r1.114 expr.c
--- expr.c      2001/08/09 04:19:12     1.114
+++ expr.c      2001/08/09 16:24:36
@@ -2505,8 +2505,8 @@ java_lang_expand_expr (exp, target, tmod
            }
          expand_expr (body, const0_rtx, VOIDmode, 0);
          emit_queue ();
-         poplevel (1, 1, 0);
          expand_end_bindings (getdecls (), 1, 0);
+         poplevel (1, 1, 0);
          return const0_rtx;
        }
       return const0_rtx;

Re: GCC 3.0/Java and control flow analysis?

[Alexandre Petit-Bianco]

Alexandre Petit-Bianco <apbianco@cygnus.com> writes:

> Like this should help, but it's not entirely tested:

Now it is. It'll go in along with this:

  http://gcc.gnu.org/ml/java/2001-08/msg00057.html

./A