From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 25298 invoked by alias); 8 Apr 2002 17:36:02 -0000 Mailing-List: contact gcc-prs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-prs-owner@gcc.gnu.org Received: (qmail 25280 invoked by uid 71); 8 Apr 2002 17:36:01 -0000 Date: Mon, 08 Apr 2002 10:36:00 -0000 Message-ID: <20020408173601.25279.qmail@sources.redhat.com> To: nobody@gcc.gnu.org Cc: gcc-prs@gcc.gnu.org, From: Tom Tromey Subject: Re: java/6163: anonymous class verifier error Reply-To: Tom Tromey X-SW-Source: 2002-04/txt/msg00495.txt.bz2 List-Id: The following reply was made to PR java/6163; it has been noted by GNATS. From: Tom Tromey To: nferrier@tapsellferrier.co.uk Cc: gcc-gnats@gcc.gnu.org Subject: Re: java/6163: anonymous class verifier error Date: 08 Apr 2002 11:35:54 -0600 >>>>> "Nic" == Nic Ferrier writes: Nic> When compiling to byte-code source with anonymous classes GCJ Nic> produces code which causes verifier errors with Sun's Nic> interpreter. I looked at this. Are you using the 1.2 JDK? Sun changed the verifier a bit between 1.2 and 1.4. In particular they added a special exception for field initialization in methods. This change is the root of the problem you are seeing. When I run the `gcj -C' output against the 1.2 java interpreter, I get a verification error. When I use the 1.4 interpreter, it works. It also works with gij (gij specifically implements this particular verifier change). Inspecting the method in test$1.class, I see that the code in question is definitely one of these field assignments. So my current thinking is that this is not a bug. What do you think? Tom