From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 4419 invoked by alias); 16 Aug 2002 21:56:02 -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 4399 invoked by uid 71); 16 Aug 2002 21:56:01 -0000 Date: Fri, 16 Aug 2002 15:03:00 -0000 Message-ID: <20020816215601.4398.qmail@sources.redhat.com> To: nobody@gcc.gnu.org Cc: gcc-prs@gcc.gnu.org, From: Tom Tromey Subject: Re: java/7611: Internal error while compiling libjava with -O Reply-To: Tom Tromey X-SW-Source: 2002-08/txt/msg00340.txt.bz2 List-Id: The following reply was made to PR java/7611; it has been noted by GNATS. From: Tom Tromey To: Thomas Greve Cc: gcc-gnats@gcc.gnu.org Subject: Re: java/7611: Internal error while compiling libjava with -O Date: 16 Aug 2002 15:55:43 -0600 >>>>> "Thomas" == Thomas Greve writes: Thomas> java/io/FilePermission.java:153: internal error in generate_bytecode_insn - tree code not implemented: bit_field_ref The problem here is that gcc will sometimes change a field access into a bit-field access. This happens in the tree-simplification routines (offhand I forget exactly where, but I think there is another PR on this topic). The bytecode back end doesn't understand bit fields, and changing this would be difficult. I'm surprised to see this happen on x86. Previously I believe it was only a problem on the alpha. Try building without -O (the .class files are all that matter). That might help. I once started a patch to add a lang-hook to let `gcj -C' disable this particular optimization. I still have it lying around somewhere but I haven't found the time to test it on my alpha box yet :-( Tom