From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id 380623858CDB; Mon, 6 Feb 2023 07:42:52 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 380623858CDB DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gcc.gnu.org; s=default; t=1675669372; bh=IlOLYo7S9zfMNYERA/gZ9GLlfj+hIqyCjUlzcZGPfRk=; h=From:To:Subject:Date:In-Reply-To:References:From; b=my6cwxQJd4aKcWweL3YOUvf//VxahPtRTv1YVHxKbuKBeK5MvUOxBnDt/1Y2Jn9yX 6FHUKNLrF4XMhecA79wrK3zUvHv443YqpQACSyGw8ddnugCAuDQPaAGPhe2TEYzL+b ua0/gkMTsEG36KIQN63L1wUqFGjD9tzVEA8Cq2ZQ= From: "rguenth at gcc dot gnu.org" To: gcc-bugs@gcc.gnu.org Subject: [Bug debug/108656] [12/13 Regression] '-fcompare-debug' failure (length) w/ -O2 -fno-ipa-pure-const -fno-tree-dce --param early-inlining-insns=0 since r12-5236 Date: Mon, 06 Feb 2023 07:42:51 +0000 X-Bugzilla-Reason: CC X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: gcc X-Bugzilla-Component: debug X-Bugzilla-Version: 13.0 X-Bugzilla-Keywords: compare-debug-failure X-Bugzilla-Severity: normal X-Bugzilla-Who: rguenth at gcc dot gnu.org X-Bugzilla-Status: NEW X-Bugzilla-Resolution: X-Bugzilla-Priority: P2 X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org X-Bugzilla-Target-Milestone: 12.3 X-Bugzilla-Flags: X-Bugzilla-Changed-Fields: Message-ID: In-Reply-To: References: Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/ Auto-Submitted: auto-generated MIME-Version: 1.0 List-Id: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D108656 --- Comment #4 from Richard Biener --- In general we avoid disallowing attribute combinations that at least in the= ory make sense. pure/const are about memory side-effects while returns_twice is about control flow, so in this regard I don't see how they should conflict. That we exclude !gimple_has_side_effects from call_can_make_abnormal_goto is somewhat of a chicken-and-egg thing - "side effect" is something not explicitely encoded in the IL but a returns_twice results in explicit encoding via abnormal edges. But then we cannot use that for CFG construction.=