From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 21755 invoked by alias); 14 Feb 2006 15:41:02 -0000 Received: (qmail 21726 invoked by uid 48); 14 Feb 2006 15:41:00 -0000 Date: Tue, 14 Feb 2006 15:41:00 -0000 Message-ID: <20060214154100.21725.qmail@sourceware.org> X-Bugzilla-Reason: CC References: Subject: [Bug c++/23372] [4.0/4.1/4.2 Regression] Temporary aggregate copy not elided when passing parameters by value In-Reply-To: Reply-To: gcc-bugzilla@gcc.gnu.org To: gcc-bugs@gcc.gnu.org From: "rguenth 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/msg01543.txt.bz2 List-Id: ------- Comment #33 from rguenth at gcc dot gnu dot org 2006-02-14 15:40 ------- The attached patch is bogus, and a correct one doesn't fix the first testcase (as the attached one didn't, too). As analyzed previously, expand does not deal with void g(A*) (a) { struct A D.2007; : D.2007 = *a_1; f (D.2007) [tail call]; return; } and TER doesn't produce (non-gimple) f (*a). Still TER looks like the only place where we could get this fixed, because we still have dataflow information left. Also a real struct copyprop pass will not help here. As TER / outof-ssa is not something I want to look into, unassigning this. -- rguenth at gcc dot gnu dot org changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|rguenth at gcc dot gnu dot |unassigned at gcc dot gnu |org |dot org Status|ASSIGNED |NEW http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23372