From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267127AbUBSJgh (ORCPT ); Thu, 19 Feb 2004 04:36:37 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267134AbUBSJgh (ORCPT ); Thu, 19 Feb 2004 04:36:37 -0500 Received: from natsmtp01.rzone.de ([81.169.145.166]:7921 "EHLO natsmtp01.rzone.de") by vger.kernel.org with ESMTP id S267127AbUBSJge (ORCPT ); Thu, 19 Feb 2004 04:36:34 -0500 From: Arnd Bergmann To: Pete Zaitcev Subject: Re: sys_tux stolen @s390 in 2.6 Date: Thu, 19 Feb 2004 10:30:59 +0100 User-Agent: KMail/1.5.4 Cc: Martin Schwidefsky , linux-kernel@vger.kernel.org MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200402191030.59229.arnd@arndb.de> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Pete wrote: > in 2.4, syscall #222 was reserved for tux on s390, but now it is used > by sys_readahead. What do we do now? In my copy of tux-3.2.13, the number 242 is used correctly. That number is the one that is reserved in the official linux sources. Martin allocated it exactly one year ago when I sent the patch enabling s390 in tux to Florian La Roche . If you have a really old version of the tux sources, there might be the fallback to number 222 still there (which is a pretty dumb idea, btw). Arnd <>< from tux.c: #if defined(__powerpc__) #define __NR_tux 225 #elif defined(__x86_64__) #define __NR_tux 184 #elif defined(__alpha__) #define __NR_tux 397 #elif defined(__s390__) #define __NR_tux 242 #elif (defined (__i386__) || defined (__arm__)) #define __NR_tux 222 #else #warning unsupported architecture, guessing __NR_tux=222 like x86... #define __NR_tux 222 #endif