[V3,2/4] cpufreq: add support for intermediate (stable) frequencies

Douglas Anderson, recently pointed out an interesting problem due to which
udelay() was expiring earlier than it should.

While transitioning between frequencies few platforms may temporarily switch to
a stable frequency, waiting for the main PLL to stabilize.

For example: When we transition between very low frequencies on exynos, like
between 200MHz and 300MHz, we may temporarily switch to a PLL running at 800MHz.
No CPUFREQ notification is sent for that. That means there's a period of time
when we're running at 800MHz but loops_per_jiffy is calibrated at between 200MHz
and 300MHz. And so udelay behaves badly.

To get this fixed in a generic way, lets introduce another set of callbacks
get_intermediate() and target_intermediate(), only for drivers with
target_index() and CPUFREQ_ASYNC_NOTIFICATION unset.

get_intermediate should return a stable intermediate frequency platform wants to
switch to, and target_intermediate() should set CPU to to that frequency, before
jumping to the frequency corresponding to 'index'. Core will take care of
sending notifications and driver doesn't have to handle them in
target_intermediate() or target_index().

NOTE: Once set to intermediate frequency, driver isn't expected to fail for the
following ->target_index() call, if it fails core will issue a WARN().

Tested-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
---
 Documentation/cpu-freq/cpu-drivers.txt | 19 ++++++++++++++--
 drivers/cpufreq/cpufreq.c              | 40 +++++++++++++++++++++++++++-------
 include/linux/cpufreq.h                | 15 +++++++++++++
 3 files changed, 64 insertions(+), 10 deletions(-)

Viresh,

On Fri, May 16, 2014 at 9:51 PM, Viresh Kumar <viresh.kumar@linaro.org> wrote:

> +1.7 get_intermediate and target_intermediate
> +--------------------------------------------
> +
> +Only for drivers with target_index() and CPUFREQ_ASYNC_NOTIFICATION unset.
> +
> +get_intermediate should return a stable intermediate frequency platform wants to
> +switch to, and target_intermediate() should set CPU to to that frequency, before
> +jumping to the frequency corresponding to 'index'. Core will take care of
> +sending notifications and driver doesn't have to handle them in
> +target_intermediate() or target_index().

Is it worth documenting that if we implement target_intermediate()
that target_index() must not fail?  That means that any failure-prone
things (like setting a regulator) should happen in target_index().


>  2. Frequency Table Helpers
> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
> index 9bf12a2..f38f2f2 100644
> --- a/drivers/cpufreq/cpufreq.c
> +++ b/drivers/cpufreq/cpufreq.c
> @@ -1819,27 +1819,50 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
>  static int __target_index(struct cpufreq_policy *policy,
>                           struct cpufreq_frequency_table *freq_table, int index)
>  {
> -       struct cpufreq_freqs freqs;
> +       struct cpufreq_freqs freqs = {.old = policy->cur, .flags = 0};
>         int retval = -EINVAL;
>         bool notify;
>
>         notify = !(cpufreq_driver->flags & CPUFREQ_ASYNC_NOTIFICATION);
> +       if (!notify)
> +               goto skip_notify;

I'm personally not a fan of using goto here.  All you're trying to do
is avoiding a level of indentation?  If it really matters that much
then create a sub function.  IMHO goto should generally be reserved
for error handling.


>
> -       if (notify) {
> -               freqs.old = policy->cur;
> -               freqs.new = freq_table[index].frequency;
> -               freqs.flags = 0;
> +       /* Handle switching to intermediate frequency */
> +       if (cpufreq_driver->get_intermediate) {
> +               freqs.new = cpufreq_driver->get_intermediate(policy, index);

Would it be worth it to change this to?

intermediate = 0
if (cpufreq_driver->get_intermediate)
  intermediate = cpufreq_driver->get_intermediate();
if (intermediate)

...the idea being that a driver may use an intermediate frequency for
some transitions but not for all.  For instance: on tegra if you
happen to change to the exact clock frequency of the intermediate PLL
it just stays there.  There's no need for two notifications in that
case.  There may be other systems that can optimize some transitions
to totally skip the intermediate stage (maybe you've got an
non-glitching divider somewhere so you can optimize a transition from
1.4GHz to 700MHz to go w/ no intermediate jump).


> -               pr_debug("%s: cpu: %d, oldfreq: %u, new freq: %u\n",
> +               pr_debug("%s: cpu: %d, switching to intermediate freq: oldfreq: %u, intermediate freq: %u\n",
>                                 __func__, policy->cpu, freqs.old, freqs.new);
>
>                 cpufreq_freq_transition_begin(policy, &freqs);
> +               retval = cpufreq_driver->target_intermediate(policy, freqs.new);

It feels like you want to pass in "index" here too, just in case.  A
driver may need to make decisions about other clocks based on the
eventual final frequency.  They could cache it themselves from the
get_intermediate() call, but that seems ugly.


> @@ -2361,7 +2384,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data)
>             !(driver_data->setpolicy || driver_data->target_index ||
>                     driver_data->target) ||
>              (driver_data->setpolicy && (driver_data->target_index ||
> -                   driver_data->target)))
> +                   driver_data->target)) ||
> +            (!!driver_data->get_intermediate ^ !!driver_data->target_intermediate))

I'm OK with the !! trick, but using ^ here seems more confusing.  Why
not use "!="?
  (!!driver_data->get_intermediate != !!driver_data->target_intermediate))

-Doug
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Doug,

On 20 May 2014 22:18, Doug Anderson <dianders@chromium.org> wrote:

> Is it worth documenting that if we implement target_intermediate()
> that target_index() must not fail?  That means that any failure-prone
> things (like setting a regulator) should happen in target_index().

You meant target_intermediate() is the last line, right ?

Yeah, we can add that..

>>  2. Frequency Table Helpers
>> diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
>> index 9bf12a2..f38f2f2 100644
>> --- a/drivers/cpufreq/cpufreq.c
>> +++ b/drivers/cpufreq/cpufreq.c
>> @@ -1819,27 +1819,50 @@ EXPORT_SYMBOL(cpufreq_unregister_notifier);
>>  static int __target_index(struct cpufreq_policy *policy,
>>                           struct cpufreq_frequency_table *freq_table, int index)
>>  {
>> -       struct cpufreq_freqs freqs;
>> +       struct cpufreq_freqs freqs = {.old = policy->cur, .flags = 0};
>>         int retval = -EINVAL;
>>         bool notify;
>>
>>         notify = !(cpufreq_driver->flags & CPUFREQ_ASYNC_NOTIFICATION);
>> +       if (!notify)
>> +               goto skip_notify;
>
> I'm personally not a fan of using goto here.  All you're trying to do
> is avoiding a level of indentation?  If it really matters that much
> then create a sub function.  IMHO goto should generally be reserved
> for error handling.

Yeah, I was trying that :) .. Another routine wouldn't be right here
as the POST_NOTIFICATION will be handled in this routine only.

I will try again to see if I can write some better code here, but wouldn't
promise that :)

>> -       if (notify) {
>> -               freqs.old = policy->cur;
>> -               freqs.new = freq_table[index].frequency;
>> -               freqs.flags = 0;
>> +       /* Handle switching to intermediate frequency */
>> +       if (cpufreq_driver->get_intermediate) {
>> +               freqs.new = cpufreq_driver->get_intermediate(policy, index);
>
> Would it be worth it to change this to?
>
> intermediate = 0
> if (cpufreq_driver->get_intermediate)
>   intermediate = cpufreq_driver->get_intermediate();
> if (intermediate)
>
> ...the idea being that a driver may use an intermediate frequency for
> some transitions but not for all.  For instance: on tegra if you
> happen to change to the exact clock frequency of the intermediate PLL
> it just stays there.  There's no need for two notifications in that
> case.  There may be other systems that can optimize some transitions
> to totally skip the intermediate stage (maybe you've got an
> non-glitching divider somewhere so you can optimize a transition from
> 1.4GHz to 700MHz to go w/ no intermediate jump).

Hmm, will try to fix that as well. Looks like a valid point.

>> -               pr_debug("%s: cpu: %d, oldfreq: %u, new freq: %u\n",
>> +               pr_debug("%s: cpu: %d, switching to intermediate freq: oldfreq: %u, intermediate freq: %u\n",
>>                                 __func__, policy->cpu, freqs.old, freqs.new);
>>
>>                 cpufreq_freq_transition_begin(policy, &freqs);
>> +               retval = cpufreq_driver->target_intermediate(policy, freqs.new);
>
> It feels like you want to pass in "index" here too, just in case.  A
> driver may need to make decisions about other clocks based on the
> eventual final frequency.  They could cache it themselves from the
> get_intermediate() call, but that seems ugly.

I had index here initially, but then it looked like they may need to perform
get_intermediate() again from this routine and so sending the intermediate
freq is probably better. So, probably just wait for some drivers which may
need index here ?

>> @@ -2361,7 +2384,8 @@ int cpufreq_register_driver(struct cpufreq_driver *driver_data)
>>             !(driver_data->setpolicy || driver_data->target_index ||
>>                     driver_data->target) ||
>>              (driver_data->setpolicy && (driver_data->target_index ||
>> -                   driver_data->target)))
>> +                   driver_data->target)) ||
>> +            (!!driver_data->get_intermediate ^ !!driver_data->target_intermediate))
>
> I'm OK with the !! trick, but using ^ here seems more confusing.  Why
> not use "!="?
>   (!!driver_data->get_intermediate != !!driver_data->target_intermediate))

Will work as well :)
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/