diff mbox

[2/4] cpu: expose pm_qos_resume_latency for each cpu

Message ID 1472114562-2736-2-git-send-email-alex.shi@linaro.org (mailing list archive)
State Not Applicable, archived
Headers show

Commit Message

Alex Shi Aug. 25, 2016, 8:42 a.m. UTC
The cpu-dma PM QoS constraint impacts all the cpus in the system. There
is no way to let the user to choose a PM QoS constraint per cpu.

The following patch exposes to the userspace a per cpu based sysfs file
in order to let the userspace to change the value of the PM QoS latency
constraint.

This change is inoperative in its form and the cpuidle governors have to
take into account the per cpu latency constraint in addition to the
global cpu-dma latency constraint in order to operate properly.

BTW
The pm_qos_resume_latency usage defined in
Documentation/ABI/testing/sysfs-devices-power
The /sys/devices/.../power/pm_qos_resume_latency_us attribute
contains the PM QoS resume latency limit for the given device,
which is the maximum allowed time it can take to resume the
device, after it has been suspended at run time, from a resume
request to the moment the device will be ready to process I/O,
in microseconds.  If it is equal to 0, however, this means that
the PM QoS resume latency may be arbitrary.

Signed-off-by: Alex Shi <alex.shi@linaro.org>
To: linux-kernel@vger.kernel.org
To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: linux-pm@vger.kernel.org
Cc: Ulf Hansson <ulf.hansson@linaro.org>
Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
---
 drivers/base/cpu.c | 3 +++
 1 file changed, 3 insertions(+)

Comments

Alex Shi Aug. 31, 2016, 10:45 a.m. UTC | #1
Hi,

Is there any concern on this patch?

Regards
Alex


On 08/25/2016 04:42 PM, Alex Shi wrote:
> The cpu-dma PM QoS constraint impacts all the cpus in the system. There
> is no way to let the user to choose a PM QoS constraint per cpu.
> 
> The following patch exposes to the userspace a per cpu based sysfs file
> in order to let the userspace to change the value of the PM QoS latency
> constraint.
> 
> This change is inoperative in its form and the cpuidle governors have to
> take into account the per cpu latency constraint in addition to the
> global cpu-dma latency constraint in order to operate properly.
> 
> BTW
> The pm_qos_resume_latency usage defined in
> Documentation/ABI/testing/sysfs-devices-power
> The /sys/devices/.../power/pm_qos_resume_latency_us attribute
> contains the PM QoS resume latency limit for the given device,
> which is the maximum allowed time it can take to resume the
> device, after it has been suspended at run time, from a resume
> request to the moment the device will be ready to process I/O,
> in microseconds.  If it is equal to 0, however, this means that
> the PM QoS resume latency may be arbitrary.
> 
> Signed-off-by: Alex Shi <alex.shi@linaro.org>
> To: linux-kernel@vger.kernel.org
> To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: linux-pm@vger.kernel.org
> Cc: Ulf Hansson <ulf.hansson@linaro.org>
> Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
> ---
>  drivers/base/cpu.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
> index 4c28e1a..ba11e23 100644
> --- a/drivers/base/cpu.c
> +++ b/drivers/base/cpu.c
> @@ -17,6 +17,7 @@
>  #include <linux/of.h>
>  #include <linux/cpufeature.h>
>  #include <linux/tick.h>
> +#include <linux/pm_qos.h>
>  
>  #include "base.h"
>  
> @@ -376,6 +377,8 @@ int register_cpu(struct cpu *cpu, int num)
>  
>  	per_cpu(cpu_sys_devices, num) = &cpu->dev;
>  	register_cpu_under_node(num, cpu_to_node(num));
> +	if (dev_pm_qos_expose_latency_limit(&cpu->dev, 0))
> +		pr_debug("CPU%d: add resume latency failed\n", num);
>  
>  	return 0;
>  }
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Greg Kroah-Hartman Aug. 31, 2016, 1:18 p.m. UTC | #2
On Thu, Aug 25, 2016 at 04:42:40PM +0800, Alex Shi wrote:
> The cpu-dma PM QoS constraint impacts all the cpus in the system. There
> is no way to let the user to choose a PM QoS constraint per cpu.
> 
> The following patch exposes to the userspace a per cpu based sysfs file
> in order to let the userspace to change the value of the PM QoS latency
> constraint.
> 
> This change is inoperative in its form and the cpuidle governors have to
> take into account the per cpu latency constraint in addition to the
> global cpu-dma latency constraint in order to operate properly.
> 
> BTW
> The pm_qos_resume_latency usage defined in
> Documentation/ABI/testing/sysfs-devices-power
> The /sys/devices/.../power/pm_qos_resume_latency_us attribute
> contains the PM QoS resume latency limit for the given device,
> which is the maximum allowed time it can take to resume the
> device, after it has been suspended at run time, from a resume
> request to the moment the device will be ready to process I/O,
> in microseconds.  If it is equal to 0, however, this means that
> the PM QoS resume latency may be arbitrary.
> 
> Signed-off-by: Alex Shi <alex.shi@linaro.org>
> To: linux-kernel@vger.kernel.org
> To: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: linux-pm@vger.kernel.org
> Cc: Ulf Hansson <ulf.hansson@linaro.org>
> Cc: Daniel Lezcano <daniel.lezcano@linaro.org>
> ---
>  drivers/base/cpu.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
> index 4c28e1a..ba11e23 100644
> --- a/drivers/base/cpu.c
> +++ b/drivers/base/cpu.c
> @@ -17,6 +17,7 @@
>  #include <linux/of.h>
>  #include <linux/cpufeature.h>
>  #include <linux/tick.h>
> +#include <linux/pm_qos.h>
>  
>  #include "base.h"
>  
> @@ -376,6 +377,8 @@ int register_cpu(struct cpu *cpu, int num)
>  
>  	per_cpu(cpu_sys_devices, num) = &cpu->dev;
>  	register_cpu_under_node(num, cpu_to_node(num));
> +	if (dev_pm_qos_expose_latency_limit(&cpu->dev, 0))
> +		pr_debug("CPU%d: add resume latency failed\n", num);

Why is this a debug message?  And you have a struct device, why not use
it?

Also, what userspace changes does this require, or affect?  Any new
sysfs files?

thanks,

greg k-h
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Alex Shi Sept. 1, 2016, 3:39 a.m. UTC | #3
>> @@ -376,6 +377,8 @@ int register_cpu(struct cpu *cpu, int num)
>>  
>>  	per_cpu(cpu_sys_devices, num) = &cpu->dev;
>>  	register_cpu_under_node(num, cpu_to_node(num));
>> +	if (dev_pm_qos_expose_latency_limit(&cpu->dev, 0))
>> +		pr_debug("CPU%d: add resume latency failed\n", num);
> 
> Why is this a debug message?  And you have a struct device, why not use
> it?
> 

Looks no one care about the dev_pm_qos_expose_latency_limit's return
value, so this debug message like gilding the lily. will remove it.

> Also, what userspace changes does this require, or affect?  Any new
> sysfs files?

User can set values on each of cpu, like limit 100ms on cpu0, that means
the cpu0 response time should be in 100ms in possible idle. It similar
with DMA_LATENCY, but that request is for all cpu. This is just for
particular cpu, like a interrupt pined CPU.

echo 100 > /sys/devices/system/cpu/cpu0/power/pm_qos_resume_latency_us
> 
> thanks,
> 
> greg k-h
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Alex Shi Sept. 1, 2016, 3:45 a.m. UTC | #4
On 09/01/2016 11:39 AM, Alex Shi wrote:
> User can set values on each of cpu, like limit 100ms on cpu0, that means
> the cpu0 response time should be in 100ms in possible idle. It similar
> with DMA_LATENCY, but that request is for all cpu. This is just for
> particular cpu, like a interrupt pined CPU.

Sorry for typo!
s/100ms/100us/

> 
> echo 100 > /sys/devices/system/cpu/cpu0/power/pm_qos_resume_latency_us
>> > 
--
To unsubscribe from this list: send the line "unsubscribe linux-pm" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
diff mbox

Patch

diff --git a/drivers/base/cpu.c b/drivers/base/cpu.c
index 4c28e1a..ba11e23 100644
--- a/drivers/base/cpu.c
+++ b/drivers/base/cpu.c
@@ -17,6 +17,7 @@ 
 #include <linux/of.h>
 #include <linux/cpufeature.h>
 #include <linux/tick.h>
+#include <linux/pm_qos.h>
 
 #include "base.h"
 
@@ -376,6 +377,8 @@  int register_cpu(struct cpu *cpu, int num)
 
 	per_cpu(cpu_sys_devices, num) = &cpu->dev;
 	register_cpu_under_node(num, cpu_to_node(num));
+	if (dev_pm_qos_expose_latency_limit(&cpu->dev, 0))
+		pr_debug("CPU%d: add resume latency failed\n", num);
 
 	return 0;
 }