Re: [PATCHv4 1/8] devfreq: event: Add new devfreq_event class to provide basic data for devfreq governor

[Chanwoo Choi <cw00.choi@samsung.com>]

Dear Myungjoo,

Thanks for your review.

On 12/18/2014 03:24 PM, MyungJoo Ham wrote:
> Hi Chanwoo,
> 
> I love the idea and I now have a little mechanical issues in your code.
> 
>> ---
>>  drivers/devfreq/Kconfig         |   2 +
>>  drivers/devfreq/Makefile        |   5 +-
>>  drivers/devfreq/devfreq-event.c | 449 ++++++++++++++++++++++++++++++++++++++++
>>  drivers/devfreq/event/Makefile  |   1 +
>>  include/linux/devfreq.h         | 160 ++++++++++++++
>>  5 files changed, 616 insertions(+), 1 deletion(-)
>>  create mode 100644 drivers/devfreq/devfreq-event.c
>>  create mode 100644 drivers/devfreq/event/Makefile
>>
>> diff --git a/drivers/devfreq/Kconfig b/drivers/devfreq/Kconfig
>> index faf4e70..4d15b62 100644
>> --- a/drivers/devfreq/Kconfig
>> +++ b/drivers/devfreq/Kconfig
>> @@ -87,4 +87,6 @@ config ARM_EXYNOS5_BUS_DEVFREQ
>>  	  It reads PPMU counters of memory controllers and adjusts the
>>  	  operating frequencies and voltages with OPP support.
>>  
>> +comment "DEVFREQ Event Drivers"
>> +
>>  endif # PM_DEVFREQ
>> diff --git a/drivers/devfreq/Makefile b/drivers/devfreq/Makefile
>> index 16138c9..a1ffabe 100644
>> --- a/drivers/devfreq/Makefile
>> +++ b/drivers/devfreq/Makefile
>> @@ -1,4 +1,4 @@
>> -obj-$(CONFIG_PM_DEVFREQ)	+= devfreq.o
>> +obj-$(CONFIG_PM_DEVFREQ)		+= devfreq.o devfreq-event.o
>>  obj-$(CONFIG_DEVFREQ_GOV_SIMPLE_ONDEMAND)	+= governor_simpleondemand.o
>>  obj-$(CONFIG_DEVFREQ_GOV_PERFORMANCE)	+= governor_performance.o
>>  obj-$(CONFIG_DEVFREQ_GOV_POWERSAVE)	+= governor_powersave.o
>> @@ -7,3 +7,6 @@ obj-$(CONFIG_DEVFREQ_GOV_USERSPACE)	+= governor_userspace.o
>>  # DEVFREQ Drivers
>>  obj-$(CONFIG_ARM_EXYNOS4_BUS_DEVFREQ)	+= exynos/
>>  obj-$(CONFIG_ARM_EXYNOS5_BUS_DEVFREQ)	+= exynos/
>> +
>> +# DEVFREQ Event Drivers
>> +obj-$(CONFIG_PM_DEVFREQ)		+= event/
>>
> 
> It looks getting mature fast.
> However, I would like to suggest you to
> 
> allow not to compile devfreq-event.c and not include its compiled object
>   if devfreq.c is required but devfreq-event.c is not required.
>   (e.g., add CONFIG_PM_DEVFREQ_EVENT and let it be enabled when needed)
>   just a little concern for lightweight devices.
>     (this change might require a bit more work on the header as well)
>   - Or do you think devfreq-event.c will become almost mandatory for
>    most devfreq drivers?

I agree your opinion.
I'll add CONFIG_PM_DEVFREQ_EVENT according to your comment.

> 
> 
> [snip]
> 
>> diff --git a/drivers/devfreq/devfreq-event.c b/drivers/devfreq/devfreq-event.c
>> new file mode 100644
>> index 0000000..0e1948e
>> --- /dev/null
>> +++ b/drivers/devfreq/devfreq-event.c
>> @@ -0,0 +1,449 @@
>> +/*
>> + * devfreq-event: Generic DEVFREQ Event class driver
> 
> DEVFREQ is a generic DVFS mechanism (or subsystem).
> 
> Plus, I thought devfreq-event is considered to be a "framework"
> for devfreq event class drivers. Am I mistaken?

You're right. just "class driver" description is not proper.
I'll modify the description of devfreq-event.c as following:
or If you have other opinion, would you please let me know about it?

	devfreq-event: DEVFREQ-Event Framework to provide raw data of Non-CPU Devices.


> [snip]
> 
>> +struct devfreq_event_dev *devfreq_event_add_edev(struct device *dev,
>> +                                               struct devfreq_event_desc *desc)
>> +{
>> +       struct devfreq_event_dev *edev;
>> +       static atomic_t event_no = ATOMIC_INIT(0);
>> +       int ret;
>> +
>> +       if (!dev || !desc)
>> +               return ERR_PTR(-EINVAL);
>> +
>> +       if (!desc->name || !desc->ops)
>> +               return ERR_PTR(-EINVAL);
>> +
>> +       if (!desc->ops->set_event || !desc->ops->get_event)
>> +               return ERR_PTR(-EINVAL);
>> +
>> +       edev = devm_kzalloc(dev, sizeof(*edev), GFP_KERNEL);
>> +       if (!edev)
>> +               return ERR_PTR(-ENOMEM);
>> +
>> +       mutex_lock(&devfreq_event_list_lock);
> 
> You seem to lock that global lock too long. That lock is only required
> while you operate the list. The data to be protected by this mutex is
> devfreq_event_list. Until the new entry is added to the list, the new
> entry is free from protection. (may be delayed right before list_add)

OK. I'll move global lock right before calling list_add() on below.

> 
>> +       mutex_init(&edev->lock);
>> +       edev->desc = desc;
>> +       edev->dev.parent = dev;
>> +       edev->dev.class = devfreq_event_class;
>> +       edev->dev.release = devfreq_event_release_edev;
>> +
>> +       dev_set_name(&edev->dev, "event.%d", atomic_inc_return(&event_no) - 1);
>> +       ret = device_register(&edev->dev);
>> +       if (ret < 0) {
>> +               put_device(&edev->dev);
>> +               mutex_unlock(&devfreq_event_list_lock);
>> +               return ERR_PTR(ret);
>> +       }
>> +       dev_set_drvdata(&edev->dev, edev);
>> +
>> +       INIT_LIST_HEAD(&edev->node);
>> +       list_add(&edev->node, &devfreq_event_list);
>> +       mutex_unlock(&devfreq_event_list_lock);
>> +
>> +       return edev;
>> +}
> 
> 
> 
> [snip / reversed maybe.. sorry]
> 
>> +/**
>> + * devfreq_event_is_enabled() - Check whether devfreq-event dev is enabled or
>> + *                             not.
>> + * @edev       : the devfreq-event device
>> + *
>> + * Note that this function check whether devfreq-event dev is enabled or not.
>> + * If return true, the devfreq-event dev is enabeld. If return false, the
>> + * devfreq-event dev is disabled.
>> + */
>> +bool devfreq_event_is_enabled(struct devfreq_event_dev *edev)
>> +{
>> +       bool enabled = false;
>> +
>> +       if (!edev || !edev->desc)
>> +               return enabled;
>> +
>> +       mutex_lock(&edev->lock);
>> +
>> +       if (edev->enable_count > 0)
>> +               enabled = true;
>> +
>> +       if (edev->desc->ops && edev->desc->ops->is_enabled)
>> +               enabled |= edev->desc->ops->is_enabled(edev);
> 
> What does it mean when enabled_count > 0 and ops->is_enabled() is false? or..
> What does it mean when enabled_count = 0 and ops->is_enabled() is true?
> 
> If you do enable_count in the subsystem, why would we rely on
> ops->is_enabled()? Are you assuming that a device MAY turn itself off
> without any kernel control (ops->disable()) and it is still a correct
> behabior?

You're right. devfreq_event_is_enabled() has ambiguous operation according to your comment.

I'll only control the enable_count in the subsystem without ops->is_enabled()
and then remove the is_enabled function in the structre devfreq_event_ops.

Best Regards,
Chanwoo Choi