Re: [RFC PATCH] cpufreq: cpufreq-cpu0: do not allow transitions with regulators suspended

[Nishanth Menon <nm@ti.com>]

On 11/13/2013 07:25 PM, viresh kumar wrote:
> On Wednesday 13 November 2013 08:46 PM, Nishanth Menon wrote:
>> arrgh, my bad.. Apologies for the bad one.. I missed it :( Does the following
>> look equivalent?
> 
> yes.
> 
>> With this, I now see:
> 
>> [   43.212714] cpufreq: cpufreq_add_policy_cpu: Failed to stop governor
>> ^^^ ??
> 
> Ahh, I missed this part. I thought it will fail at some other place where there
> is no error checking :), but that's not true.
> 
> Following should fix it for you and looks to be the right way as well.
> 
> 
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index dc67fa0..30b09d3 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -1530,6 +1530,14 @@ static void cpufreq_bp_resume(void)
>                 }
>         }
> 
> +       if (has_target()) {
> +               if ((ret = __cpufreq_governor(policy, CPUFREQ_GOV_START)) ||
> +                       (ret = __cpufreq_governor(policy, CPUFREQ_GOV_LIMITS))) {
> +                       pr_err("%s: Failed to start governor\n", __func__);
> +                       goto fail;
> +               }
> +       }
> +
>         schedule_work(&policy->update);
> 
>  fail:
> 
I am guessing this is a little too early for restarting policy here
considering syscore_ops->resume is pretty early..

http://pastebin.mozilla.org/3602746 is the equivalent patch for v3.12
http://pastebin.mozilla.org/3602747 is the result.

-- 
Regards,
Nishanth Menon