From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753960AbaGVL0L (ORCPT ); Tue, 22 Jul 2014 07:26:11 -0400 Received: from terminus.zytor.com ([198.137.202.10]:53528 "EHLO mail.zytor.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753667AbaGVL0J (ORCPT ); Tue, 22 Jul 2014 07:26:09 -0400 Message-ID: <53CE4A41.4020605@zytor.com> Date: Tue, 22 Jul 2014 04:25:53 -0700 From: "H. Peter Anvin" User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.6.0 MIME-Version: 1.0 To: Klemen Jan Enova , linux-kernel@vger.kernel.org CC: mingo@redhat.com, x86@kernel.org Subject: Re: [PATCH] arch: x86: boot: bioscall.S: simplify self-modification of the INT instruction References: <53CE449C.2090001@gmail.com> In-Reply-To: <53CE449C.2090001@gmail.com> X-Enigmail-Version: 1.6 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 07/22/2014 04:01 AM, Klemen Jan Enova wrote: > At the start of the intcall function, the %al register is compared to > 0x3f. If it does not equal 0x3f it is equated to it. Instruction flow > continues in both cases from the label 1. The comparison is therefore > unneeded. > > Testing data is appended. > > Signed-off-and-tested-by: Klemen Jan Enova NAK. This is a performance optimization - the SMC takes a substantial penalty so the write is avoided if necessary. Also, 3f is a label, not 0x3f. -hpa