lkml.org 
[lkml]   [2014]   [Dec]   [23]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH] cpufreq: suspend cpufreq governors on shutdown
From
On 24 December 2014 at 05:41, Doug Anderson <dianders@chromium.org> wrote:
> We should stop cpufreq governors when we shut down the system. If we
> don't do this, we can end up with this deadlock:

Can you also add what exactly happens in such deadlock? Some lockdeps?
Or just a hang ?

> 1. cpufreq governor may be running on a CPU other than CPU0.
> 2. In machine_restart() we call smp_send_stop() which stops CPUs.
> If one of these CPUs was actively running a cpufreq governor
> then it may have the mutex / spinlock needed to access the main
> PMIC in the system (perhaps over I2C)
> 3. If a machine needs access to the main PMIC in order to shutdown
> then it will never get it since the mutex was lost when the other
> CPU stopped.
>
> Let's avoid the race by stopping the cpufreq governor at shutdown,
> which is a sensible thing to do anyway.
>
> Signed-off-by: Doug Anderson <dianders@chromium.org>
> ---
> NOTE: This was developed / tested / on a 3.14 kernel with backports.
> I have confirmed that it compiles on a mainline kernel and doesn't
> crash, but I haven't verified that there isn't some other fix in
> mainline that also fixes this problem. If you are aware of such a fix
> then please drop this patch.

No, its a new problem.

> drivers/cpufreq/cpufreq.c | 12 ++++++++++++
> 1 file changed, 12 insertions(+)
>
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index a09a29c..bd89721 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -27,6 +27,7 @@
> #include <linux/mutex.h>
> #include <linux/slab.h>
> #include <linux/suspend.h>
> +#include <linux/syscore_ops.h>
> #include <linux/tick.h>
> #include <trace/events/power.h>
>
> @@ -2550,6 +2551,15 @@ int cpufreq_unregister_driver(struct cpufreq_driver *driver)
> }
> EXPORT_SYMBOL_GPL(cpufreq_unregister_driver);
>
> +static void cpufreq_shutdown(void)
> +{
> + cpufreq_suspend();
> +}
> +
> +static struct syscore_ops cpufreq_syscore_ops = {
> + .shutdown = cpufreq_shutdown,

directly pass cpufreq_suspend() here and add a note over the
cpufreq_syscore_ops on why it exists here.

> +};
> +
> static int __init cpufreq_core_init(void)
> {
> if (cpufreq_disabled())
> @@ -2558,6 +2568,8 @@ static int __init cpufreq_core_init(void)
> cpufreq_global_kobject = kobject_create();
> BUG_ON(!cpufreq_global_kobject);
>
> + register_syscore_ops(&cpufreq_syscore_ops);
> +
> return 0;
> }
> core_initcall(cpufreq_core_init);
> --
> 2.2.0.rc0.207.ga3a616c
>


\
 
 \ /
  Last update: 2014-12-24 03:21    [W:0.030 / U:0.284 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site