Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1763470AbZD3Ovo (ORCPT ); Thu, 30 Apr 2009 10:51:44 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1761813AbZD3Ove (ORCPT ); Thu, 30 Apr 2009 10:51:34 -0400 Received: from mail-fx0-f158.google.com ([209.85.220.158]:41186 "EHLO mail-fx0-f158.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1759173AbZD3Ove (ORCPT ); Thu, 30 Apr 2009 10:51:34 -0400 Message-ID: <49F9BAF1.9060805@monstr.eu> Date: Thu, 30 Apr 2009 16:51:29 +0200 From: Michal Simek Reply-To: monstr@monstr.eu User-Agent: Thunderbird 2.0.0.18 (X11/20081120) MIME-Version: 1.0 To: Arnd Bergmann CC: Christoph Hellwig , linux-kernel@vger.kernel.org, john.williams@petalogix.com Subject: Re: [PATCH 29/30] microblaze_mmu_v1: stat.h MMU update References: <1240821139-7247-23-git-send-email-monstr@monstr.eu> <200904301553.47556.arnd@arndb.de> <49F9B13C.3060008@monstr.eu> <200904301640.32769.arnd@arndb.de> In-Reply-To: <200904301640.32769.arnd@arndb.de> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 2348 Lines: 54 Arnd Bergmann wrote: > On Thursday 30 April 2009, Michal Simek wrote: > >> 2. header cleanup -> some of them is pain because I have to recompile toolchain >> that's should be a big deal but after recompilation kernel, user apps I am getting some problems. > > Yes, I understand that you have a circular dependency there, but fixing > the ABI can only become harder with time, as more people start using > the current kernel/toolchain combination. Here is the one advantage that in mainline are only some drivers for Microblaze and the rest is in petalinux. And people can use whole distribution. > >> 3. syscall cleanup -> we still have old libc and I can't simple remove ancient syscalls > > I had hoped that the way I set __ARCH_WANT_SYSCALL_* in the microblaze > unistd.h would provide an easy way to migrate, because this still defines > almost all traditional system calls (I haven't double-checked if some more > are currently used on microblaze, but they would be easy to add). > This means that as a first step, rebuilding the tool chain (as above) should > be sufficient and not require source-level changes to libc. From that point, > you can remove the remaining __ARCH_WANT_SYSCALL_* defintions one by one, > and change the libc accordingly. > > I could also try building uclibc with the generic headers if I find the > time. Which version do you use, just upstream or do you have a private > libc tree with more changes? That's good question but harder answer. Glibc is any ancient 3-4 years non mainline version. There are some backports too. I have some comments to it but it is better not to write them. uclibc version is a little bit better. I had functional port half year mainline version + some private fixes. Currently John W is working on it to get working version with gcc 4.1.2. He has some troubles with it but I hope he solve it soon. We have troubles because none have taken care about. I hope that we are able to update all the things and repair everything what wrong it. Michal > > Arnd <>< -- Michal Simek, Ing. (M.Eng) w: www.monstr.eu p: +42-0-721842854 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/