[Bug java/20104] New: gij fails on ppc and not on x86

[Bug java/20104] New: gij fails on ppc and not on x86

[overholt at redhat dot com]

I need to use ecj (the Eclipse java compiler) via gij (long story) so I am
attempting something like this:

CLASSPATH=jdtcore.jar:jdtCompilerAdapter.jar:libgcj-4.0.0.jar gij \
org.eclipse.jdt.internal.compiler.batch.Main HelloWorld.java

On x86 I get no output and a class file, but on ppc I just get "Aborted."  This
happens with CVS head as of yesterday:

$ gij --version
gij (GNU libgcj) version 4.0.0 20050219 (experimental)

gdb doesn't work for me in this case (via its CVS head or via the Fedora rawhide
RPMs).  I'll post the output of strace as an attachment

-- 
           Summary: gij fails on ppc and not on x86
           Product: gcc
           Version: 4.0.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: java
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: overholt at redhat dot com
                CC: gcc-bugs at gcc dot gnu dot org,java-prs at gcc dot gnu
                    dot org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[overholt at redhat dot com]

------- Additional Comments From overholt at redhat dot com  2005-02-20 15:11 -------
Created an attachment (id=8239)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=8239&action=view)
output of strace gij org.eclipse.jdt.internal.compiler.batch.Main
HelloWorld.java

Stripped-down strace output (the full one is 1.1 MB which I'll post if
necessary).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-02-20 15:15 -------
(In reply to comment #0)
> gdb doesn't work for me in this case (via its CVS head or via the Fedora rawhide
> RPMs).  I'll post the output of strace as an attachment

The reason for gdb failing is most likely because of PR 19769, there is a workaround to work around 
the gcc bug.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[aph at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |tromey at redhat dot com


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[aph at gcc dot gnu dot org]

------- Additional Comments From aph at gcc dot gnu dot org  2005-02-22 15:05 -------
The failure is here:

package org.eclipse.jdt.internal.compiler.impl;
public class CompilerOptions ...
{
	void updateSeverity(long irritant, Object severityString) {
		if (ERROR.equals(severityString)) {
			this.errorThreshold |= irritant;

and the corresponding bytecode is:

Method name:"updateSeverity" Signature: (long,java.lang.Object)void
Attribute "Code", length:208, max_stack:7, max_locals:4, code_length:110
  0: ldc <String "error">
  2: aload_3
  3: invokevirtual <Method java.lang.String.equals (java.lang.Object)boolean>

The aload_3 looks wrong to me.

So, there are two separate issues:

1.  Why was aload_3 generated?  Which compiler genrated it?

2.  How did this get past the libgcj verifier?

tromey added to CC list to investigate the verifier.

overholt, please let me know how this jar file was generated.

Bug upgraded to critical.


-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |critical
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
   Last reconfirmed|0000-00-00 00:00:00         |2005-02-22 15:05:57
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[overholt at redhat dot com]

------- Additional Comments From overholt at redhat dot com  2005-02-22 15:09 -------
(In reply to comment #3)

> 1.  Why was aload_3 generated?  Which compiler genrated it?
<snip>
> overholt, please let me know how this jar file was generated.

This jar file was generated by ecj.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[aph at gcc dot gnu dot org]

------- Additional Comments From aph at gcc dot gnu dot org  2005-02-22 15:23 -------
Ok, so the file was generated by ecj, but was ecj precompiled or interpreted?  

What VM was ecj running on?

Does ecj generate this same bogus file for
org.eclipse.jdt.internal.compiler.impl.CompilerOptions when running on other
hardware?


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[overholt at redhat dot com]

------- Additional Comments From overholt at redhat dot com  2005-02-22 15:25 -------
(In reply to comment #5)
> Ok, so the file was generated by ecj, but was ecj precompiled or interpreted?  

ecj was running interpreted at the time
 
> What VM was ecj running on?

gij

> Does ecj generate this same bogus file for
> org.eclipse.jdt.internal.compiler.impl.CompilerOptions when running on other
> hardware?

I will investigate other hardware now.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[aph at gcc dot gnu dot org]

------- Additional Comments From aph at gcc dot gnu dot org  2005-02-22 16:04 -------
Forget that, second arg is a long.  My mistake.

The compiled code is OK.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug java/20104] gij fails on ppc and not on x86

[andreast at gcc dot gnu dot org]

------- Additional Comments From andreast at gcc dot gnu dot org  2005-02-22 20:07 -------
I'll add a tc in libffi soon. It is ready here but I have a bunch of other tests
I'll submit when I finished my other libffi stuff.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug libffi/20104] gij fails on ppc and not on x86

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-23 13:56 -------
Subject: Bug 20104

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	andreast@gcc.gnu.org	2005-02-23 13:56:01

Modified files:
	libffi         : ChangeLog 
Added files:
	libffi/testsuite/libffi.call: return_ll1.c 

Log message:
	2005-02-23  Andreas Tobler  <a.tobler@schweiz.ch>
	
	PR libffi/20104
	* testsuite/libffi.call/return_ll1.c: New test case.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/libffi/ChangeLog.diff?cvsroot=gcc&r1=1.220&r2=1.221
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/libffi/testsuite/libffi.call/return_ll1.c.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug libffi/20104] gij fails on ppc and not on x86

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-02-23 14:05 -------
Subject: Bug 20104

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	aph@gcc.gnu.org	2005-02-23 14:05:15

Modified files:
	libffi/src/powerpc: ffi.c 
	libffi         : ChangeLog 

Log message:
	2005-02-22  Andrew Haley  <aph@redhat.com>
	
	PR libffi/20104
	* src/powerpc/ffi.c (ffi_prep_cif_machdep): Bump alignment to
	odd-numbered register pairs for 64-bit integer types.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/libffi/src/powerpc/ffi.c.diff?cvsroot=gcc&r1=1.11&r2=1.12
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/libffi/ChangeLog.diff?cvsroot=gcc&r1=1.221&r2=1.222



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug libgcj/20104] gij fails on ppc and not on x86

[aph at gcc dot gnu dot org]

------- Additional Comments From aph at gcc dot gnu dot org  2005-02-23 14:06 -------
Try it now.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING
          Component|libffi                      |libgcj


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug libffi/20104] gij fails on ppc and not on x86

[overholt at redhat dot com]

------- Additional Comments From overholt at redhat dot com  2005-02-24 20:40 -------
This is fixed for me as of 20050223.  Changing to FIXED.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104

[Bug libffi/20104] gij fails on ppc and not on x86

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.0.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20104