From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 2757 invoked by alias); 18 Nov 2003 21:12:29 -0000 Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-bugs-owner@gcc.gnu.org Received: (qmail 2750 invoked by uid 48); 18 Nov 2003 21:12:28 -0000 Date: Tue, 18 Nov 2003 21:12:00 -0000 Message-ID: <20031118211228.2749.qmail@sources.redhat.com> From: "tromey at gcc dot gnu dot org" To: gcc-bugs@gcc.gnu.org In-Reply-To: <20031118161405.13107.bonniot@users.sf.net> References: <20031118161405.13107.bonniot@users.sf.net> Reply-To: gcc-bugzilla@gcc.gnu.org Subject: [Bug java/13107] Wrong verification error in gij: recursive subroutine call X-Bugzilla-Reason: CC X-SW-Source: 2003-11/txt/msg01688.txt.bz2 List-Id: ------- Additional Comments From tromey at gcc dot gnu dot org 2003-11-18 21:12 ------- The bug here is that we propagate subroutine information into an exception handler. However, clearing subroutine state in state::set_exception results in verifier regressions elsewhere (see the mauve verifier tests). The underlying problem seems to be that our representation of subroutines is incorrect. Perhaps we need to not keep track of the "current subroutine" but instead just keep the PC info in the return_address_type and look at that. More investigation is required, alas. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=13107