From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 27280 invoked by alias); 29 Apr 2010 06:56:18 -0000 Received: (qmail 27263 invoked by uid 22791); 29 Apr 2010 06:56:16 -0000 X-SWARE-Spam-Status: No, hits=-1.9 required=5.0 tests=BAYES_00,T_RP_MATCHES_RCVD X-Spam-Check-By: sourceware.org Received: from erik.ddg.lth.se (HELO erik.ddg.lth.se) (130.235.20.69) by sourceware.org (qpsmtpd/0.43rc1) with ESMTP; Thu, 29 Apr 2010 06:56:10 +0000 Received: from filip.ddg.lth.se (root@filip.ddg.lth.se [130.235.20.70]) by erik.ddg.lth.se (8.14.1/8.14.1) with ESMTP id o3T6tuZu003995; Thu, 29 Apr 2010 08:55:56 +0200 Received: from webmail.student.lth.se (daemon@localhost.localdomain [127.0.0.1]) by filip.ddg.lth.se (8.13.4/8.13.4/Debian-3sarge3) with ESMTP id o3T6tuOL011223; Thu, 29 Apr 2010 08:55:56 +0200 Received: from 90.237.163.38 (SquirrelMail authenticated user d03jp) by webmail.student.lth.se with HTTP; Thu, 29 Apr 2010 08:55:56 +0200 (CEST) Message-ID: <32779.90.237.163.38.1272524156.squirrel@webmail.student.lth.se> In-Reply-To: <4BD8991C.7090107@codesourcery.com> References: <201004270401.o3R41xZ8012337@ignucius.se.axis.com> <59949.90.237.230.43.1272393707.squirrel@webmail.student.lth.se> <4BD8991C.7090107@codesourcery.com> Date: Thu, 29 Apr 2010 07:07:00 -0000 Subject: Re: GCC porting tutorials From: "Jonas Paulsson" To: "Bernd Schmidt" Cc: "Hans-Peter Nilsson" , radu.hobincu@arh.pub.ro, gcc@gcc.gnu.org User-Agent: SquirrelMail/1.4.9a MIME-Version: 1.0 Content-Type: text/plain;charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-IsSubscribed: yes Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Id: List-Archive: List-Post: List-Help: Sender: gcc-owner@gcc.gnu.org X-SW-Source: 2010-04/txt/msg00933.txt.bz2 It feels good to know that the widening mults issue has been resolved as it was a bit of a disapointment I noted the erratic behaviour with GCC 4.4.1. Perhaps you would care to comment on what to expect as a user now, then? Another issue that gave me porting problems was the SIMD memory accesses, for e g doing a wide load into two adjacent narrow registers with one instruction. This was resolved earlier on the mailinglist to not be handleable on RTL, so I wonder now if anything has been done for this, as it too seems rather reasonable, just like the widening loads? /Jonas Paulsson > On 04/27/2010 08:41 PM, Jonas Paulsson wrote: > >>> I've recently been bitten by the lost-widening-multiplication- >>> when-in-loop issue myself, and noted it for revisit Some Day. >>> Fixing that by other means made a whopping 27% improvement for >>> the application where I saw it: a hot loop doing a MDCT using >>> Q31 fixed-point, where a common operation is to widen-multiply >>> two 32-bit numbers and pick the high 32 bits of the 64-bit >>> result (being the result divided by 2; i.e. Q30-ish for that >>> operation). >>> >>> brgds, H-P > > Widening multiplies should work again after r158633 and 158643. If > there are cases that still don't work for you, please file bugzilla PRs. > > > Bernd > >