From mboxrd@z Thu Jan  1 00:00:00 1970
From: hechtb@gmail.com (Bastian Hecht)
Date: Mon, 18 Feb 2013 08:56:46 -0600
Subject: [PATCH 05/06] ARM: shmobile: Common shmobile_scu_base in
 headsmp-scu.S
In-Reply-To: <CANqRtoRtPOjqy+uhWOxpbQMdp=sDaRMAEqZqjMUBcruLB0NuwQ@mail.gmail.com>
References: <20130213134628.12153.47120.sendpatchset@w520>
 <20130213134717.12153.6162.sendpatchset@w520>
 <CABYn4syysDVAKRpjDUc9j26SgmUcjxV3rNf3u2kZMZWurhrjiw@mail.gmail.com>
 <CANqRtoRtPOjqy+uhWOxpbQMdp=sDaRMAEqZqjMUBcruLB0NuwQ@mail.gmail.com>
Message-ID: <CABYn4sxoFJMsFnrf9bNCAGfram0rADv2TSbp_90BJ31DRT2W3w@mail.gmail.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

Hi,

2013/2/17 Magnus Damm <magnus.damm@gmail.com>:
> Hello Bastian,
>
> On Mon, Feb 18, 2013 at 7:20 AM, Bastian Hecht <hechtb@gmail.com> wrote:
>> Hi,
>>
>> this breaks the secondary CPU startup when I test it. I think you
>> can't access shmobile_scu_address this way with the MMU still turned
>> off.
>>
>> But instead of
>>> +       ldr     r1, =shmobile_scu_base
>>> +       ldr     r1, [r1]                @ SCU base address
>>
>> try
>>
>> +        mrc     p15, 4, r1, c15, c0, 0 @ get SCU base address
>>
>> to get it from the coprocessor.
>
> Thanks. Can you please let us know which platform you see this?
>
> I've executed this on sh73a0 and KZM9G myself without any issues.
>
> Could it be that you are powering down the power domain in your own code?

I saw it on the current HEAD of the next branch using my KZM9G. I
bisected it down to your commit. The problem appears right at boot
time when the 2nd CPU should be brought up.

Thanks,

 Bastian


> Thanks,
>
> / magnus