Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: Regression in 5.1.0-rc2: PowerBook G4 Aluminum fails to boot -
 bisected to commit 0df977eafc792
To:     Larry Finger <Larry.Finger@lwfinger.net>,
        Michael Ellerman <mpe@ellerman.id.au>,
        Benjamin Herrenschmidt <benh@kernel.crashing.org>,
        Paul Mackerras <paulus@samba.org>,
        ppc-dev <linuxppc-dev@lists.ozlabs.org>
Cc:     LKML <linux-kernel@vger.kernel.org>
References: <acc2d5a5-2fd4-4367-c8ef-497da951c208@lwfinger.net>
From:   Christophe Leroy <christophe.leroy@c-s.fr>
Message-ID: <ee247f37-8889-934c-571c-01c43fc31071@c-s.fr>
Date:   Mon, 25 Mar 2019 07:53:01 +0100
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.6.0
MIME-Version: 1.0
In-Reply-To: <acc2d5a5-2fd4-4367-c8ef-497da951c208@lwfinger.net>
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Language: fr
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk


Le 25/03/2019 à 01:49, Larry Finger a écrit :
> A build of kernel 5.1.0-rc2 resulted in a failure to boot on my 
> PowerBook G4 Aluminum. The bootstrap loads the initial kernel and issues 
> the appropriate start, but the machine hangs at that point.

Can you please be more explicit ? What do you mean by "issues the 
appropriate start" ? What is "that point" ? Any messages on the console ?

Thanks
Christophe


> 
> The problem does not depend on the choice of PPC32 processor type. This 
> machine has a 7447A according to /proc/cpuinfo.
> 
> The problem was bisected to the following:
> 
> commit 0df977eafc792a5365a7f81d8d5920132e03afad
> Author: Christophe Leroy <christophe.leroy@c-s.fr>
> Date:   Thu Feb 21 10:37:54 2019 +0000
> 
>      powerpc/6xx: Don't use SPRN_SPRG2 for storing stack pointer while 
> in RTAS
> 
>      When calling RTAS, the stack pointer is stored in SPRN_SPRG2
>      in order to be able to restore it in case of machine check in RTAS.
> 
>      As machine check is not a perfomance critical path, this patch
>      frees SPRN_SPRG2 by using a field in thread struct instead.
> 
>      Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
>      Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
> 
> I reverted this patch and found that the system began execution, and 
> then failed, likely due to the reassignment of SPRN_SPRG2.
> 
> I had found this problem with 5.1.0-rc1, but -rc2 was out by the time I 
> finished the bisection. Unfortunately, none of the changes in -rc2 fixed 
> the problem.
> 
> Attached is the .config that I used.
> 
> Thanks,
> 
> Larry