From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 4079 invoked by alias); 8 Nov 2004 16:27:40 -0000 Mailing-List: contact binutils-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: binutils-owner@sources.redhat.com Received: (qmail 3981 invoked from network); 8 Nov 2004 16:27:30 -0000 Received: from unknown (HELO beta.dmz-eu.st.com) (164.129.1.35) by sourceware.org with SMTP; 8 Nov 2004 16:27:30 -0000 Received: from zeta.dmz-eu.st.com (ns2.st.com [164.129.230.9]) by beta.dmz-eu.st.com (STMicroelectronics) with ESMTP id 5834FDD80; Mon, 8 Nov 2004 16:27:28 +0000 (GMT) Received: by zeta.dmz-eu.st.com (STMicroelectronics, from userid 60012) id DB074472E8; Mon, 8 Nov 2004 16:27:39 +0000 (GMT) Received: from zeta.dmz-eu.st.com (localhost [127.0.0.1]) by zeta.dmz-eu.st.com (STMicroelectronics) with ESMTP id 95D8B75999; Mon, 8 Nov 2004 16:27:39 +0000 (UTC) Received: from mail1.bri.st.com (mail1.bri.st.com [164.129.8.218]) by zeta.dmz-eu.st.com (STMicroelectronics) with ESMTP id EF726472E2; Mon, 8 Nov 2004 16:27:37 +0000 (GMT) Received: from st.com (chloe.bri.st.com [164.129.15.48]) by mail1.bri.st.com (MOS 3.4.4-GR) with ESMTP id AND00151 (AUTH "joern rennecke"); Mon, 8 Nov 2004 16:27:23 GMT Message-ID: <418F9E6A.90001@st.com> Date: Mon, 08 Nov 2004 16:27:00 -0000 From: Joern RENNECKE User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030922 MIME-Version: 1.0 To: Andrew STUBBS Cc: "'Nick Clifton'" , "'Alexandre Oliva'" , binutils@sources.redhat.com Subject: Re: Broken SH2a patches References: <008401c4c5a5$8db854c0$180f81a4@uk.w2k.superh.com> In-Reply-To: <008401c4c5a5$8db854c0$180f81a4@uk.w2k.superh.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-SW-Source: 2004-11/txt/msg00138.txt.bz2 > I don't like the name 'fake', but I'm not sure what else to suggest. Perhaps >sh2+, sh2e+, sh2++ and sh2e++ (or sh2+shift, sh2e+shift+fsqrt, >sh2+shift+pref and sh2e+double). Of course '+' won't work in C ... hmmm. > >Unfortunately a lot of libraries etc. will end up labelled with these 'fake' >architectures so it is important that the name means something. > When there are only one ot two simple features added, it seems most future-proof to just enumerate them. I.e. the 'fake1' adds dynamic shift to SH2, 'fake2' adds dynamic shift and fsqrt.s to SH2E, 'fake3' adds dynamic shift and prefetch to SH2. 'fake4', however, adds dynamic shift, fsqrt.s, prefetch, and a plethora of double precision arithmetic instructions and ancillary instructions to that. So I think we better name that one from the intersection that generates it, i.e. SH4_SH2A . You can probably draw this more nicely... SH2 .------------'|`--------------------. / | \ SH2 SH2_dshift SH2E (dsp) / \ \ | | / \ \_____________ | | / \ \ | | / \ \ | | / SH3 \ | | / (nommu) SH2E_dshift_sqrt | SH2_dshift_pref |\ /| | | \ | \ / | | | \ | \ / | | | \ | SH3 / | | | \ | /|\ / | | | \ | / | \ / | | | \ .!./ | \ / | | | . \ | | \ / | | SH2A . SH4 | \ / SH4_SH2A | (nofpu) . (nommu) | SH3E / | | | . (nofpu) | /\ / | | ...+..... | .--' | \ / | | / | | / | \ / | SH3 | SH4 | X | (dsp) | (nofpu) | / \ | | | /\ | / \ | | | / '---. | / \ / | | / \ |/ \/ -+---- . SH4A SH4 SH2A <------.... | (nofpu) | | / \ .-' | .--------' '---. / | / \/ SH4AL SH4A (dsp)