Re: [RFC/PATCH 9/9] OMAP2+: cpuidle only influences the MPU state

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Santosh,

On Mon, Jun 27, 2011 at 4:11 PM, Santosh Shilimkar
<santosh.shilimkar@xxxxxx> wrote:
> On 6/26/2011 11:53 PM, Jean Pihet wrote:
>>
>> Hi Santosh,
>>
>> On Sat, Jun 25, 2011 at 3:23 PM, Santosh Shilimkar
>> <santosh.shilimkar@xxxxxx>  wrote:
>>>
>>> On 6/24/2011 7:38 AM, jean.pihet@xxxxxxxxxxxxxx wrote:
>>>>
>>>> From: Jean Pihet<j-pihet@xxxxxx>
>>>>
>>>> Since cpuidle is a CPU centric framework it decides the MPU
>>>> next power state based on the MPU exit_latency and target_residency
>>>> figures.
>>>>
>>>> The rest of the power domains get their next power state programmed
>>>> from the PM_QOS_DEV_WAKEUP_LATENCY class of the PM QoS framework,
>>>> via the device wake-up latency constraints.
>>>>
>>>> Note: the exit_latency and target_residency figures of the MPU
>>>> include the MPU itself and the peripherals needed for the MPU to
>>>> execute instructions (e.g. main memory, caches, IRQ controller,
>>>> MMU etc). Some of those peripherals can belong to other power domains
>>>> than the MPU subsystem and so the corresponding latencies must be
>>>> included in this figure.
>>>>
>>> With above comment, I was expecting that the latency numbers
>>> in the table will change.
>>
>> Not necessarily. I just wanted to have it clearly stated in the commit
>> description.
>> In any case I will review the figures and update them if needed.
>> ...
>>
>>>>  static struct cpuidle_params cpuidle_params_table[] = {
>>>> -       /* C1 */
>>>> +       /* C1 . MPU WFI + Core active */
>>>>        {2 + 2, 5, 1},
>>>> -       /* C2 */
>>>> +       /* C2 . MPU WFI + Core inactive */
>>>>        {10 + 10, 30, 1},
>>>> -       /* C3 */
>>>> +       /* C3 . MPU CSWR + Core inactive */
>>>>        {50 + 50, 300, 1},
>>>> -       /* C4 */
>>>> +       /* C4 . MPU OFF + Core inactive */
>>>>        {1500 + 1800, 4000, 1},
>>>> -       /* C5 */
>>>> +       /* C5 . MPU RET + Core RET */
>>>>        {2500 + 7500, 12000, 1},
>>>> -       /* C6 */
>>>> +       /* C6 . MPU OFF + Core RET */
>>>>        {3000 + 8500, 15000, 1},
>>>> -       /* C7 */
>>>> +       /* C7 . MPU OFF + Core OFF */
>>>>        {10000 + 30000, 300000, 1},
>>>
>>> Latency numbers still seems to include CORE PD latency as
>>> well. Am I missing something Jean?
>>
>> The figures are looking OK.
>>
> Hmmm...
> Well they represent mostly MPU + CORE PD sleep + wakeup latencies
> I suppose so not sure what you mean by OK here.
I mean the figures are representing what they are supposed to. As said
above they might need refining.

Thanks,
Jean

>
> Regards
> Santosh
>
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Index of Archives]     [Linux Arm (vger)]     [ARM Kernel]     [ARM MSM]     [Linux Tegra]     [Linux WPAN Networking]     [Linux Wireless Networking]     [Maemo Users]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite Trails]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux