From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 10820 invoked by alias); 9 Nov 2005 00:02:58 -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 10791 invoked by uid 48); 9 Nov 2005 00:02:56 -0000 Date: Wed, 09 Nov 2005 00:02:00 -0000 Message-ID: <20051109000256.10790.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug middle-end/24639] [meta-bug] bug to track all uninit variable issues In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "law at redhat dot com" X-SW-Source: 2005-11/txt/msg01212.txt.bz2 List-Id: ------- Comment #4 from law at redhat dot com 2005-11-09 00:02 ------- Created an attachment (id=10181) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=10181&action=view) Proposed patch and new testcases I'm not comfortable self-approving this patch; the option processing code is totally new to me and I have no idea if I got it right. And while I believe the sense of -Wuninitialized is correct (no change in current behavior), I don't think we have reached a consensus on that issue. I will note that after reviewing all the PRs, by far the most common complaint is false positives, #2 is missing warnings, particularly for objects which have had their address taken, #3 is missing warnings in unreachable/dead code (which we can now detect with -Wuninitialized=2 using this patch). -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=24639