From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261174AbUKDDkd (ORCPT ); Wed, 3 Nov 2004 22:40:33 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261542AbUKDDkd (ORCPT ); Wed, 3 Nov 2004 22:40:33 -0500 Received: from fw.osdl.org ([65.172.181.6]:56536 "EHLO mail.osdl.org") by vger.kernel.org with ESMTP id S261174AbUKDDka (ORCPT ); Wed, 3 Nov 2004 22:40:30 -0500 Date: Wed, 3 Nov 2004 19:40:27 -0800 (PST) From: Linus Torvalds To: Andrea Arcangeli cc: Kernel Mailing List Subject: Re: RFC: avoid asmlinkage on x86 traps/interrupts In-Reply-To: <20041104020835.GL3571@dualathlon.random> Message-ID: References: <20041103090710.GV3571@dualathlon.random> <20041104020835.GL3571@dualathlon.random> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Thu, 4 Nov 2004, Andrea Arcangeli wrote: > > I wonder why they don't forbid it completely then. I mean, what's magic > about an input parmeter here? Clobbers are about the internals of > the asm Actually, they are mostly about register allocation. And that's apparently why the gcc rules ended up changing: by setting a clobber, you basically allocate that register for the instruction, and tell gcc that it's dead. That's also likely why such a register cannot be an input (or an output): it cannot be allocated. Linus