linux-mips
[Top] [All Lists]

Re: [PATCH 1/3] Remove '-mno-explicit-relocs' option when CONFIG_BUILD_E

To: Daniel Jacobowitz <dan@debian.org>
Subject: Re: [PATCH 1/3] Remove '-mno-explicit-relocs' option when CONFIG_BUILD_ELF64
From: Franck Bui-Huu <vagabon.xyz@gmail.com>
Date: Tue, 14 Aug 2007 18:22:03 +0200
Cc: linux-mips@linux-mips.org
Dkim-signature: a=rsa-sha1; c=relaxed/relaxed; d=gmail.com; s=beta; h=domainkey-signature:received:received:message-id:date:from:user-agent:mime-version:to:cc:subject:references:in-reply-to:content-type:content-transfer-encoding; b=VRztNMkme4dghJD6eADybM0ZMyqedcYfo0k72CxBt1q6itdixjd4cCUVL48wzRtX6310J1OsBls6KcN/ISaxsaVmw278KVFvA+iKU/6Mvj9RMROhK4ZYmamXbt/07Nd2VMxnsvgX7JRWly4duT3285rlEvQ8xG+vgbWaYWzunOQ=
Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=beta; h=received:message-id:date:from:user-agent:mime-version:to:cc:subject:references:in-reply-to:content-type:content-transfer-encoding; b=qT/oWi9oN0wsW127Wk5JP8THpX1UrNhPuSlILTnsNrnjVv2sfn9corov8MFXZVnLBmr9dweaYubCDLkjcCAA6or9e1r2v7RRsJ1UQgdzK3wcgNG6tWfNXaPdF7yfY3OfBeZSnqN2G13DgVqhzID5dfuH1ZFdDj/KrmkXWK1xRUc=
In-reply-to: <20070813193059.GA17969@caradoc.them.org>
Original-recipient: rfc822;linux-mips@linux-mips.org
References: <11715446603241-git-send-email-fbuihuu@gmail.com> <20070809151812.GA28142@caradoc.them.org> <46BB6D6C.2050601@gmail.com> <20070813193059.GA17969@caradoc.them.org>
Sender: linux-mips-bounce@linux-mips.org
User-agent: Thunderbird 2.0.0.5 (X11/20070719)
Daniel Jacobowitz wrote:
> Yes, I rebased them on top of linux-mips HEAD as of Thursday.
> 

Ok, I rebased them too, and compare them to your diff and got almost
the same result except that you seem to not use the last patch 3/3
version. But it shouldn't hurt.

> My kernel crashed, so I gave your patchset a try, and it got a lot
> further and crashed somewhere different.
> 

It just means that before applying the patchset your kernel was
wrongly configured: you probably set "CONFIG_BUILD_ELF64=y" whereas
your kernel load address is in CKSEG0, is that correct ?

If so, your current issue may be related to a completely different
area and maybe a git-bisect session could help but don't forget to set
"CONFIG_BUILD_ELF64=n" before.

Oh, maybe before doing the git-bisect session, there's one thing you
could try: when building the kernel with the patchset applied do:

                $ make KBUILD_SYM32=no

It's equivalent to "CONFIG_BUILD_ELF64=y" _but_ the kernel will use
the correct definitions for kernel address translations.

In any cases, your dmesg could help.

                Franck

<Prev in Thread] Current Thread [Next in Thread>