From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kbuild-owner@vger.kernel.org>
Received: from mout.kundenserver.de ([212.227.17.10]:52165 "EHLO
        mout.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
        with ESMTP id S1751449AbcHZOJf (ORCPT
        <rfc822;linux-kbuild@vger.kernel.org>);
        Fri, 26 Aug 2016 10:09:35 -0400
From: Arnd Bergmann <arnd@arndb.de>
Subject: Re: [PATCH] arm: add an option for erratum 657417
Date: Fri, 26 Aug 2016 16:07:49 +0200
References: <1470989957-23671-1-git-send-email-npiggin@gmail.com> <2147968.tzghefjZMo@wuerfel> <20160826211729.592ff5cd@roar.ozlabs.ibm.com>
In-Reply-To: <20160826211729.592ff5cd@roar.ozlabs.ibm.com>
MIME-Version: 1.0
Content-Type: Text/Plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Message-Id: <201608261607.49610.arnd@arndb.de>
Sender: linux-kbuild-owner@vger.kernel.org
List-ID: <linux-kbuild.vger.kernel.org>
To: linux-arm-kernel@lists.infradead.org
Cc: Nicholas Piggin <npiggin@gmail.com>, Nicolas Pitre <nicolas.pitre@linaro.org>, linux-arch@vger.kernel.org, Segher Boessenkool <segher@kernel.crashing.org>, Alan Modra <amodra@gmail.com>, linux-kbuild@vger.kernel.org, Russell King <linux@armlinux.org.uk>, Julian Brown <julian@codesourcery.com>, Michal Marek <mmarek@suse.com>, Stephen Rothwell <sfr@canb.auug.org.au>, Sam Ravnborg <sam@ravnborg.org>

On Friday 26 August 2016, Nicholas Piggin wrote:
> > I still see a problem with big-endian builds failing with
> > thinarc/gc-sections, I'll investigate that some other day,
> > or you could have a look at that if you want to make sure
> > it's an ARM specific problem, not something with your
> > patches in general.
> > 
> > The patch that I sent for enabling the two on ARM blocks
> > out CONFIG_CPU_BIG_ENDIAN, so just revert that hunk to see
> > the problem. It's possible that it only breaks when doing
> > a big-endian build after a little-endian build without
> > a "make clean" inbetween.
> 
> I'm able to build big endian ARM allyesconfig with thin
> archives and gc-sections, worked fine.
> 
> There have been a few bugs in powerpc when failing to notice
> the change when endian of builds was swapped, so maybe you
> got bitten by something similar.

I tracked this down as well now, and it's also a problem on my
local machine, your patches are fine. In order to debug
the other problem, I was building with
"make LD=/home/arnd/..../ld" to try out different
versions of the linker, and that caused the "LD += -EB"
line from arch/arm/Makefile to be ignored.

We should probably override LDFLAGS rather than LD
(and AFLAGS instead of AS) for big-endian builds, but
that is unrelated to your work.

	Arnd