[Bug java/16188] New: [3.5 Regression] wrong code generation

["bonniot at users dot sf dot net" <gcc-bugzilla at gcc dot gnu dot org>]

I attach a compiled class, for which gcj 3.5.0 20040613 generates wrong code (at
least, the execution gives a wrong result). There is no call to any external
library, so it must be the code generation itself.

Looking at the bytecode for the main method, it is easy to see that to see that
the local variable 2 should be assigned a newly created array. So the println
call should print an array reference. Instead it prints null. Interpretation
with gij (same version) works fine (prints non-null), as well as compilation
with gcj 3.3.4 (hence the regression).

(I have no Java source to show for this, which was generated by the Nice compiler)

-- 
           Summary: [3.5 Regression] wrong code generation
           Product: gcc
           Version: 3.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: java
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: bonniot at users dot sf dot net
                CC: gcc-bugs at gcc dot gnu dot org,java-prs at gcc dot gnu
                    dot org
  GCC host triplet: i386-debian-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16188