From mboxrd@z Thu Jan  1 00:00:00 1970
From: will.deacon@arm.com (Will Deacon)
Date: Sun, 6 Jan 2013 16:22:00 +0000
Subject: [PATCH] ARM: call disable_nonboot_cpus() from machine_shutdown()
In-Reply-To: <50E5E967.3020805@wwwdotorg.org>
References: <1357160861-26282-1-git-send-email-swarren@wwwdotorg.org>
 <50E4C9C6.20609@codeaurora.org>
 <20130103120259.GB5193@mudshark.cambridge.arm.com>
 <20130103122100.GN2631@n2100.arm.linux.org.uk>
 <50E5E967.3020805@wwwdotorg.org>
Message-ID: <20130106162200.GC11025@mudshark.cambridge.arm.com>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

On Thu, Jan 03, 2013 at 08:26:15PM +0000, Stephen Warren wrote:
> On 01/03/2013 05:21 AM, Russell King - ARM Linux wrote:
> > On Thu, Jan 03, 2013 at 12:02:59PM +0000, Will Deacon wrote:
> >> You need the smp_send_stop call in order to send the cpu_kill (looks like
> >> tegra needs die and then kill). So you really need hotplug support as well
> >> as suspend for this to do much (if not, the secondaries end up spinning
> >> with interrupts disabled which is probably the best we can do anyway).
> >>
> >> We could add SUSPEND as a KEXEC dependency if SMP (we already have HOTPLUG
> >> there) if you like?
> > 
> > Or we could look into bringing in the code to do this when KEXEC is
> > enabled - which would mean an amount of restructuring of the Kconfig
> > files.
> 
> I'm not sure if any of this thread means we should hold off on this
> patch, or just that the Kconfig could/should be enhanced later?

Well, Russell's suggestion looked easy enough to have a crack out so you
could always post a series implementing it along with this patch.

> One thing I did just notice with my patch: disable_nonboot_cpus() ends
> up being called twice for the poweroff path:
> 
> [   30.461847] Disabling non-boot CPUs ...
> [   30.478797] CPU1: shutdown
> [   30.492104] Power down.
> [   30.494578] Disabling non-boot CPUs ...
> 
> Is this worth worrying about?

It's harmless but it's also pretty horrible. Unfortunately, I don't see
what we can do about it: it's a direct side-effect of generic code calling
disable_nonboot_cpus for poweroff and not for kexec.

Will