From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 7409 invoked by alias); 15 Jun 2006 14:58:31 -0000 Received: (qmail 7398 invoked by uid 22791); 15 Jun 2006 14:58:30 -0000 X-Spam-Check-By: sourceware.org Received: from smtp111.sbc.mail.mud.yahoo.com (HELO smtp111.sbc.mail.mud.yahoo.com) (68.142.198.210) by sourceware.org (qpsmtpd/0.31) with SMTP; Thu, 15 Jun 2006 14:58:28 +0000 Received: (qmail 68280 invoked from network); 15 Jun 2006 14:58:26 -0000 Received: from unknown (HELO lucon.org) (hjjean@sbcglobal.net@71.146.67.96 with login) by smtp111.sbc.mail.mud.yahoo.com with SMTP; 15 Jun 2006 14:58:26 -0000 Received: by lucon.org (Postfix, from userid 1000) id 520E663EEC; Thu, 15 Jun 2006 07:58:25 -0700 (PDT) Date: Thu, 15 Jun 2006 15:00:00 -0000 From: "H. J. Lu" To: Nick Clifton Cc: binutils@sources.redhat.com Subject: Re: PATCH: Add --alt-nops=short|long to x86/x86-64 assemblers Message-ID: <20060615145825.GB7470@lucon.org> References: <20060613190254.GB21884@lucon.org> <448FBB66.7040806@redhat.com> <20060614150158.GA30299@lucon.org> <449027CF.3070705@redhat.com> <20060614172000.GA31328@lucon.org> <449111FC.8010909@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <449111FC.8010909@redhat.com> User-Agent: Mutt/1.4.2.1i Mailing-List: contact binutils-help@sourceware.org; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: binutils-owner@sourceware.org X-SW-Source: 2006-06/txt/msg00246.txt.bz2 On Thu, Jun 15, 2006 at 08:53:32AM +0100, Nick Clifton wrote: > Hi H. J. > > >x86/x86-64 assemblers don't support -mtune/-march/-mcpu. > > Ahh. How does the assembler know which instruction set variants are > valid then ? (I am thinking of the all the different SSE, PowerNow, etc > variants). By default, the x86/x86-64 assemblers take everything. You can limit the instruction set with .arch directive. Until now, we never generate instructions beyond i386 ourselves. So it isn't a problem. > > >We need a switch to tell > >assembler if the new nop instructions can be used at all. Secondly, > >not all modern processors prefer "long" versions. We need another > >switch to tell which kind of the new nop instructions should be used, > >short or long. > > In which case the default presumably ought to be the short version and > the long version should only be enabled if explicitly requested via a > command line switch. > > >Maybe I should add -march= and -mtune= to assembler. > > It sounds like it would be a good idea, although if it is only to > support this new feature then you may not want to go that far. I am considering adding -mtune= and updating document to indicate that all instructions are allowed by default. But I will do some experiment first. H.J.