From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 25288 invoked by alias); 1 Feb 2006 03:14:46 -0000 Received: (qmail 25262 invoked by uid 48); 1 Feb 2006 03:14:43 -0000 Date: Wed, 01 Feb 2006 03:14:00 -0000 Message-ID: <20060201031443.25261.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug c/26004] [4.1/4.2 Regression] gcc errors on valid code [SVO] In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "mmitchel at gcc dot gnu dot org" 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 X-SW-Source: 2006-02/txt/msg00031.txt.bz2 List-Id: ------- Comment #7 from mmitchel at gcc dot gnu dot org 2006-02-01 03:14 ------- This is a design bug. The middle-end should not be calling mark_addressable hooks in the front ends. The fix may be to factor the front end mark_addressable functions into a part that just does the appropriate marking/recursion for language-dependent tree codes, and a separate part that enforces semantic constraints. The former function would be the one available to the middle end; the latter (which would presumably call the former) would be used by the front end itself. This should be a relatively straightforward fix. -- mmitchel at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P3 |P2 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=26004