From mboxrd@z Thu Jan 1 00:00:00 1970 From: Michal Simek Subject: Re: [microblaze-uclinux] [RFC PATCH] linker script: unify usage of discard definition Date: Sat, 04 Jul 2009 20:40:18 +0200 Message-ID: <4A4FA212.8030506@monstr.eu> References: <4A4E9638.20304@kernel.org> Reply-To: monstr@monstr.eu Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Return-path: Received: from mail-fx0-f218.google.com ([209.85.220.218]:51944 "EHLO mail-fx0-f218.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752216AbZGDSkX (ORCPT ); Sat, 4 Jul 2009 14:40:23 -0400 In-Reply-To: <4A4E9638.20304@kernel.org> Sender: linux-arch-owner@vger.kernel.org List-ID: To: microblaze-uclinux@itee.uq.edu.au Cc: "linux-arch@vger.kernel.org" , Linux Kernel , Michal Simek , Sam Ravnborg , Paul Mundt , Tony Luck , tj@kernel.org Tejun Heo wrote: > Discarded sections in different archs share some commonality but have > considerable differences. This led to linker script for each arch > implementing its own /DISCARD/ definition, which makes maintaining > tedious and adding new entries error-prone. > > This patch makes all linker scripts to move discard definitions to the > end of the linker script and use the common DISCARDS macro. As ld > uses the first matching section definition, archs can include default > discarded sections by including them earlier in the linker script. > > On x86-64 which explicitly includes EXIT_TEXT, this produces identical > output as before and all but one archs included directives (include or > discard) for all default discarded sections, so this patch shouldn't > cause any behavior difference. > > The only problem case is microblaze, which didn't explicitly included > EXIT_TEXT and DATA but didn't discard them either leaving ld to create > identically names output sections for them. This patch will make the > end result different by throwing away exit text and data. It looks > like the ommission was accidental and throwing away is the right thing > to do, but I won't push it till someone who knows microblaze confirms > it. As you expected noMMU Microblaze kernel has some problems with that patch. (MMU version is fine) I'll take a look at it closer yesterday. Thanks, Michal > ___________________________ > microblaze-uclinux mailing list > microblaze-uclinux@itee.uq.edu.au > Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux > Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/ > > -- Michal Simek, Ing. (M.Eng) w: www.monstr.eu p: +42-0-721842854