From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 48) id 13CB53858C2C; Thu, 23 Dec 2021 09:15:40 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 13CB53858C2C From: "jakub at gcc dot gnu.org" To: gcc-bugs@gcc.gnu.org Subject: [Bug tree-optimization/103797] Clang vectorized LightPixel while GCC does not Date: Thu, 23 Dec 2021 09:15:40 +0000 X-Bugzilla-Reason: CC X-Bugzilla-Type: changed X-Bugzilla-Watch-Reason: None X-Bugzilla-Product: gcc X-Bugzilla-Component: tree-optimization X-Bugzilla-Version: unknown X-Bugzilla-Keywords: missed-optimization X-Bugzilla-Severity: enhancement X-Bugzilla-Who: jakub at gcc dot gnu.org X-Bugzilla-Status: NEW X-Bugzilla-Resolution: X-Bugzilla-Priority: P3 X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org X-Bugzilla-Target-Milestone: --- 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 X-BeenThere: gcc-bugs@gcc.gnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Gcc-bugs mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Thu, 23 Dec 2021 09:15:41 -0000 https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D103797 --- Comment #15 from Jakub Jelinek --- (In reply to Uro=C5=A1 Bizjak from comment #12) > (In reply to Jakub Jelinek from comment #10) > > At least on your short testcase clang doesn't use divps either. > > We do support mulv2sf3, addv2sf3 etc. but not divv2sf3 I bet because wi= th > > TARGET_MMX_WITH_SSE it would divide by zero in the 3rd and 4th elts, > > but perhaps we could insert 1.0f, 1.0f into those elements of the divis= or > > before using divps? >=20 > It could be done, but I was under impression that the sequence to load 1.= 0f > into topmost elements nullifies the benefit of operation to divide two Sure, so perhaps we should somewhat increase the vectorization cost of V2SF= mode division so that we would use it only if it is part of longer sequences?=