From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262937AbUBZTIO (ORCPT ); Thu, 26 Feb 2004 14:08:14 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262940AbUBZTIO (ORCPT ); Thu, 26 Feb 2004 14:08:14 -0500 Received: from kinesis.swishmail.com ([209.10.110.86]:59141 "EHLO kinesis.swishmail.com") by vger.kernel.org with ESMTP id S262937AbUBZTII (ORCPT ); Thu, 26 Feb 2004 14:08:08 -0500 Message-ID: <403E4681.20603@techsource.com> Date: Thu, 26 Feb 2004 14:18:25 -0500 From: Timothy Miller MIME-Version: 1.0 To: "Nakajima, Jun" CC: richard.brunner@amd.com, linux-kernel@vger.kernel.org Subject: Re: Intel vs AMD64 References: <7F740D512C7C1046AB53446D37200173EA28A5@scsmsx402.sc.intel.com> In-Reply-To: <7F740D512C7C1046AB53446D37200173EA28A5@scsmsx402.sc.intel.com> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Nakajima, Jun wrote: > Thanks for the clarification. > > Yes, "implementation specific" is one of the differences between IA-32e > and AMD64, i.e. that behavior is architecturally defined on AMD64, but > on IA-32e (as I posted): > Near branch with 66H prefix: > As documented in PRM the behavior is implementation specific and > should > avoid using 66H prefix on near branches. In other words, Intel's implementation deviates from the architecture as defined by AMD. So it's not 100% compatible. I just want this point to be clear. If these sorts of branches are common enough (and I suspect they are), then this sort of deviation could have a notable code-size (and L1) impact on code which is compiled to be compatible with both implementations. Why did Intel decide to do that?