From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 6047 invoked by alias); 1 Aug 2003 08:45:36 -0000 Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm Precedence: bulk List-Archive: List-Post: List-Help: Sender: gcc-owner@gcc.gnu.org Received: (qmail 6029 invoked from network); 1 Aug 2003 08:45:35 -0000 Received: from unknown (HELO air.cs.bath.ac.uk) (138.38.108.3) by sources.redhat.com with SMTP; 1 Aug 2003 08:45:35 -0000 Received: from fire.cs.bath.ac.uk ([138.38.108.1] helo=ralph.cs.bath.ac.uk) by air.cs.bath.ac.uk with esmtp (Exim 3.16 #3) id 19iVXb-0007d3-00 for gcc@gcc.gnu.org; Fri, 01 Aug 2003 09:45:35 +0100 Subject: Re: GCC From: Martin To: gcc@gcc.gnu.org In-Reply-To: References: Content-Type: text/plain Content-Transfer-Encoding: 7bit Date: Fri, 01 Aug 2003 09:56:00 -0000 Message-Id: <1059727535.32270.507.camel@ralph.cs.bath.ac.uk> Mime-Version: 1.0 X-SW-Source: 2003-08/txt/msg00015.txt.bz2 > > So why would these people think it's difficult to work with the people > > on this list, and to contribute code (and what can be done about it)? > It's possible to look at it from the other viewpoint, e.g. educate > processor designers to not design bad instruction sets. I'd be willing > to write a document describing "Things not to do when designing a > processor for GCC" since it feels like I've dealt with every single bad > idea ever conceived for a processor... It strikes me this is part of a larger problem. It seems there is (in many companies / institutions there is a large divide between the people who design processors and the people who write compilers. I've heard of a couple of projects where there was essentially one team that did both MIPS and Bulldog/VLIW come to mind. Hennessy & Patterson suggest looking at usage profiles of instructions in their book on computer architecture but don't go so far as to suggest developing the compiler and the processor together. It seems that some of the early RISC work brought the two camps closer together but ATM there is nowhere near enough dialog and co-operation on most projects. Well - that's my impression of the situation. Perhaps it might be an idea to try and set up some sort of collaboration with the Open Cores project ( http://www.opencores.net ). Cheers, - Martin