From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262423AbVFXUbI (ORCPT ); Fri, 24 Jun 2005 16:31:08 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261630AbVFXUbG (ORCPT ); Fri, 24 Jun 2005 16:31:06 -0400 Received: from smtp.osdl.org ([65.172.181.4]:11721 "EHLO smtp.osdl.org") by vger.kernel.org with ESMTP id S262423AbVFXUan (ORCPT ); Fri, 24 Jun 2005 16:30:43 -0400 Date: Fri, 24 Jun 2005 13:28:26 -0700 From: Andrew Morton To: Adrian Bunk Cc: linux-kernel@vger.kernel.org, fastboot@lists.osdl.org Subject: Re: [-mm patch] i386: enable REGPARM by default Message-Id: <20050624132826.4cdfb63c.akpm@osdl.org> In-Reply-To: <20050624200916.GJ6656@stusta.de> References: <20050624200916.GJ6656@stusta.de> X-Mailer: Sylpheed version 1.0.0 (GTK+ 1.2.10; i386-vine-linux-gnu) Mime-Version: 1.0 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 Adrian Bunk wrote: > > This patch: > - removes the dependency of REGPARM on EXPERIMENTAL > - let REGPARM default to y hm, a compromise. One other concern I have with this is that I expect -mregparm will make kgdb (and now crashdump) less useful. When incoming args are on the stack you have a good chance of being able to see what their value is by walking the stack slots. When the incoming args are in registers I'd expect that it would be a lot harder (or impossible) to work out their value. Have the kdump guys thought about (or encountered) this?