From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 18138 invoked by alias); 17 Apr 2002 20:12:14 -0000 Mailing-List: contact cgen-help@sources.redhat.com; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: cgen-owner@sources.redhat.com Received: (qmail 18105 invoked from network); 17 Apr 2002 20:12:11 -0000 Received: from unknown (HELO albatross.prod.itd.earthlink.net) (207.217.120.120) by sources.redhat.com with SMTP; 17 Apr 2002 20:12:11 -0000 Received: from dialup-63.214.109.98.dial1.boston1.level3.net ([63.214.109.98]) by albatross.prod.itd.earthlink.net with esmtp (Exim 3.33 #1) id 16xvmk-0004O0-00; Wed, 17 Apr 2002 13:12:11 -0700 Mime-Version: 1.0 X-Sender: qsmgmt@pop.earthlink.net Message-Id: Date: Wed, 17 Apr 2002 13:12:00 -0000 To: cgen@sources.redhat.com From: Alan Lehotsky Subject: CGEN vs. VLIW architectures... Cc: apl@alum.mit.edu Content-Type: text/plain; charset="us-ascii" ; format="flowed" X-SW-Source: 2002-q2/txt/msg00000.txt.bz2 With a couple of small patches, I got CGEN to accept the basic specification of a VLIW machine with 256 bit instructions. [Of course, at this point, I only have defined - TWO instructions (one of them's NOP, and the other is bogus :-) - ONE operand (CGEN generates bad C code without any operand definitions.) Of course, since this is a VLIW machine, there IS only one instruction, with about 40 subfields. Is there anyone else working on a VLIW port? Want to coordinate figuring out what needs fixing for machines where the instruction is larger than "long long"? -- Quality Software Management http://home.earthlink.net/~qsmgmt apl@alum.mit.edu (978)287-0435 Voice (978)808-6836 Cell Software Process Improvement / Management Consulting Language Design / Compiler Implementation