Re: [PATCH 2/2] PM / devfreq: exynos-nocp: Remove redundant code to reset load_count/total_count

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

On 2016년 09월 20일 18:38, Axel Lin wrote:
> 2016-09-20 17:34 GMT+08:00 Chanwoo Choi <cw00.choi@samsung.com>:
>> Hi,
>>
>> On 2016년 09월 20일 17:57, Axel Lin wrote:
>>> This is done by devfreq_event_get_event() so remove the redundant code in
>>> exynos_nocp_get_event().
>>>
>>> Signed-off-by: Axel Lin <axel.lin@ingics.com>
>>> ---
>>>  drivers/devfreq/event/exynos-nocp.c | 3 ---
>>>  1 file changed, 3 deletions(-)
>>>
>>> diff --git a/drivers/devfreq/event/exynos-nocp.c b/drivers/devfreq/event/exynos-nocp.c
>>> index a584140..49e712a 100644
>>> --- a/drivers/devfreq/event/exynos-nocp.c
>>> +++ b/drivers/devfreq/event/exynos-nocp.c
>>> @@ -176,9 +176,6 @@ static int exynos_nocp_get_event(struct devfreq_event_dev *edev,
>>>       return 0;
>>>
>>>  out:
>>> -     edata->load_count = 0;
>>> -     edata->total_count = 0;
>>> -
>>>       dev_err(nocp->dev, "Failed to read the counter of NoC probe device\n");
>>>
>>>       return ret;
>>>
>>
>> As you said, devfreq_event_get_event() initializes both load_count and total_count
>> before calling edev->desc->ops->get_event(). Looks good to me.
> 
> Actually, the point is if edev->desc->ops->get_event() fails,
> the code in devfreq_event_get_event() will reset
> edata->total_count and edata->load_count to 0.

You're right. devfreq_event_get_event() initializes the data
when before calling edev->desc->ops->get_event() or fail happen.

-- 
Best Regards,
Chanwoo Choi