From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Mon, 25 Feb 2002 16:06:57 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Mon, 25 Feb 2002 16:06:44 -0500 Received: from gateway-1237.mvista.com ([12.44.186.158]:14845 "EHLO hermes.mvista.com") by vger.kernel.org with ESMTP id ; Mon, 25 Feb 2002 16:06:11 -0500 Message-ID: <3C7AA727.316EC197@mvista.com> Date: Mon, 25 Feb 2002 13:05:43 -0800 From: george anzinger Organization: Monta Vista Software X-Mailer: Mozilla 4.77 [en] (X11; U; Linux 2.2.12-20b i686) X-Accept-Language: en MIME-Version: 1.0 To: "linux-kernel@vger.kernel.org" Subject: Universal Regs address Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org ®s is needed by the deliver signal code and currently is supplied by the system call interface to the system calls that need it. This requires that any new system call to have (at least in some archs) special code in the system call trap area to pass the ®s, or does it? In an arch in which the call stack address decreases as calls are made, isn't: ®s = stack_base+size of(stack) - size of(struct regs); an for stacks that increase: ®s = stack_base; The only time this would not be true, unless I am missing something, is if the system call is made from kernel space. Is this an issue? Do we ever need ®s if called from the kernel? If not, can we tell the call was from the kernel? comments? -- George george@mvista.com High-res-timers: http://sourceforge.net/projects/high-res-timers/ Real time sched: http://sourceforge.net/projects/rtsched/