cpuidle: allow to disable C states of the ladder governor
Subsystems (2)
cover cpuidle: allow to disable C states of the ladder governor
Hi,
when trying to disable a specific C state in order to improve the idle responsiveness of a system but still enjoy some power saving, we noted that writing to the related sysfs variable had no effect.
This patch allows to disable C states of the ladder governor.
-Carsten.
PATCH 1/1 Honor state disabling in the cpuidle ladder governor
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if CONFIG_CPU_IDLE is selected, the menu governor additionally requires CONFIG_NO_HZ.
A particular C state can be disabled by writing to the sysfs file /sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system where CONFIG_NO_HZ is not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to /dev/cpu_dma_latency and keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
--- drivers/cpuidle/governors/ladder.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/cpuidle/governors/ladder.c +3 -1
@@ -88,6 +88,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !drv->states[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;
@@ -100,7 +101,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (drv->states[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
Re: [PATCH 1/1] Honor state disabling in the cpuidle ladder governor
Quoted 14 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.
Quoted 3 lines
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Yes, I agree that currently that disabling a particular C-state is not reflected in working of ladder governor. This patch is needed to fix it on ladder too.
Also wanted to clarify on the intended implementation here, if there are say 5 C-states on a system, disabling 2nd
state would also end by disabling all the remaining 3 deeper states too as ladder governor enters the lightest state first, and will only move on to the next deeper state if a idle period was long enough as per the implementation.
If one is disabling only the deepest state, then it would work as intended.
Cheers,
Deepthi
Quoted 30 lines
Signed-off-by Carsten Emde <[email protected]>--- drivers/cpuidle/governors/ladder.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)Index:
linux-3.4.4-rt14-rc2-64/drivers/cpuidle/governors/ladder.c
===================================================================drivers/cpuidle/governors/ladder.c +3
-1@@ -88,6 +88,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !drv->states[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;@@ -100,7 +101,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (drv->states[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
Re: [PATCH 1/1] Honor state disabling in the cpuidle ladder governor
Quoted 24 lines · Deepthi Dharwar
Quoted 14 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.Quoted 3 lines
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Yes, I agree that currently that disabling a particular C-state is not reflected in working of ladder governor. This patch is needed to fix it on ladder too.
Quoted 8 lines
Also wanted to clarify on the intended implementation here, if there are say 5 C-states on a system, disabling 2nd
state would also end by disabling all the remaining 3 deeper states too as ladder governor enters the lightest state first, and will only move on to the next deeper state if a idle period was long enough as per the implementation.
If one is disabling only the deepest state, then it would work as intended.
Yes, the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect.
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?
-Carsten.
Re: [PATCH 1/1] Honor state disabling in the cpuidle ladder governor
Quoted 33 lines · Carsten Emde
Quoted 24 lines · Deepthi Dharwar
Quoted 14 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.Quoted 3 lines
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Yes, I agree that currently that disabling a particular C-state is not reflected in working of ladder governor. This patch is needed to fix it on ladder too.
Quoted 8 lines
Also wanted to clarify on the intended implementation here, if there are say 5 C-states on a system, disabling 2nd
state would also end by disabling all the remaining 3 deeper states too as ladder governor enters the lightest state first, and will only move on to the next deeper state if a idle period was long enough as per the implementation.
If one is disabling only the deepest state, then it would work as intended.
Quoted 5 lines
Yes, the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect.
Agree, as per the ladder design.
Quoted 5 lines
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?
No, I dont think thats necessary, current code suffices it. The disable flag is knob we are giving to the user . So may be just document the intended use of disable flag working alongside design of ladder governor.
Cheers
Deepthi
Quoted 2 lines
-Carsten.
PATCH v2 1/1 Honor state disabling in the cpuidle ladder governor - documented
Quoted 46 lines · Deepthi Dharwar
Quoted 32 lines · Carsten Emde
Quoted 23 lines · Deepthi Dharwar
Quoted 14 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.Quoted 3 lines
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Yes, I agree that currently that disabling a particular C-state is not reflected in working of ladder governor. This patch is needed to fix it on ladder too.
Quoted 8 lines
Also wanted to clarify on the intended implementation here, if there are say 5 C-states on a system, disabling 2nd
state would also end by disabling all the remaining 3 deeper states too as ladder governor enters the lightest state first, and will only move on to the next deeper state if a idle period was long enough as per the implementation.
If one is disabling only the deepest state, then it would work as intended.Quoted 5 lines
Yes, the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect.
Agree, as per the ladder design.
Quoted 4 lines
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?
Quoted 4 lines
No, I dont think thats necessary, current code suffices it. The disable flag is knob we are giving to the user . So may be just document the intended use of disable flag working alongside design of ladder governor.
Here comes v2 with a related section added to the documentation.
-Carsten.
Subject: Honor state disabling in the cpuidle ladder governor
From: Carsten Emde <[email protected]>
Date: Tue, 17 Jul 2012 19:50:13 +0100
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if CONFIG_CPU_IDLE is selected, the menu governor additionally requires CONFIG_NO_HZ.
A particular C state can be disabled by writing to the sysfs file /sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system where CONFIG_NO_HZ is not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to /dev/cpu_dma_latency and keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Note that the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect. A related section has been addded to the documentation.
--- Documentation/cpuidle/sysfs.txt | 10 +++++++++- drivers/cpuidle/governors/ladder.c | 4 +++- 2 files changed, 12 insertions(+), 2 deletions(-)
Documentation/cpuidle/sysfs.txt +9 -1
@@ -76,9 +76,17 @@ total 0 * desc : Small description about the idle state (string) -* disable : Option to disable this idle state (bool) +* disable : Option to disable this idle state (bool) -> see note below * latency : Latency to exit out of this idle state (in microseconds) * name : Name of the idle state (string) * power : Power consumed while in this idle state (in milliwatts) * time : Total time spent in this idle state (in microseconds) * usage : Number of times this state was entered (count) + +Note: +The behavior and the effect of the disable variable depends on the +implementation of a particular governor. In the ladder governor, for +example, it is not coherent, i.e. if one is disabling a light state, +then all deeper states are disabled as well, but the disable variable +does not reflect it. Likewise, if one enables a deep state but a lighter +state still is disabled, then this has no effect.
drivers/cpuidle/governors/ladder.c +3 -1
@@ -88,6 +88,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !drv->states[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;
@@ -100,7 +101,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (drv->states[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
PATCH v3 1/1 Honor state disabling in the cpuidle ladder governor - with sanitizer
In reply to Deepthi DharwarQuoted 46 lines · Deepthi Dharwar
Quoted 32 lines · Carsten Emde
Quoted 23 lines · Deepthi Dharwar
Quoted 14 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.Quoted 3 lines
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Yes, I agree that currently that disabling a particular C-state is not reflected in working of ladder governor. This patch is needed to fix it on ladder too.
Quoted 8 lines
Also wanted to clarify on the intended implementation here, if there are say 5 C-states on a system, disabling 2nd
state would also end by disabling all the remaining 3 deeper states too as ladder governor enters the lightest state first, and will only move on to the next deeper state if a idle period was long enough as per the implementation.
If one is disabling only the deepest state, then it would work as intended.Quoted 5 lines
Yes, the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect.
Agree, as per the ladder design.
Quoted 4 lines
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?
Quoted 4 lines
No, I dont think thats necessary, current code suffices it. The disable flag is knob we are giving to the user . So may be just document the intended use of disable flag working alongside design of ladder governor.
It's not necessary - but maybe better. Here comes v3 with a sanitizer. Is this too ugly?
-Carsten.
From: Carsten Emde <[email protected]>
Subject: Honor state disabling in the cpuidle ladder governor
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if CONFIG_CPU_IDLE is selected, the menu governor additionally requires CONFIG_NO_HZ.
A particular C state can be disabled by writing to the sysfs file /sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system where CONFIG_NO_HZ is not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to /dev/cpu_dma_latency and keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
A sanitize mechanism takes care the disable variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled.
--- drivers/cpuidle/governors/ladder.c | 25 ++++++++++++++++++++++++- drivers/cpuidle/sysfs.c | 6 ++++++ include/linux/cpuidle.h | 4 ++++ 3 files changed, 34 insertions(+), 1 deletion(-)
drivers/cpuidle/governors/ladder.c +24 -1
@@ -58,6 +58,25 @@ static inline void ladder_do_selection(s ldev->last_state_idx = new_idx; } +void notify_state_modified (struct cpuidle_state *state) +{ +if (state->disable) { +/* synchronize all deeper states, if any */ +int states = state->state_count -1 - state->state_no; +struct cpuidle_state *last = state + states; + +while (++state <= last) + state->disable =1 ; + }else { +/* synchronize all lighter states, if any */ +int states = state->state_no; +struct cpuidle_state *first = state - states; + +while (--state >= first) + state->disable =0 ; + } +} +/** * ladder_select_state - selects the next state to enter * @drv: cpuidle driver
@@ -88,6 +107,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !drv->states[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;
@@ -100,7 +120,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (drv->states[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
@@ -154,6 +175,8 @@ static int ladder_enable_device(struct c lstate->threshold.promotion_time = state->exit_latency;if (i >0 ) lstate->threshold.demotion_time = state->exit_latency; + + state->notify = notify_state_modified; }return 0 ;
drivers/cpuidle/sysfs.c +6 -0
@@ -239,15 +239,19 @@ static ssize_t store_state_##_name(struc { \long value; \int err; \ +unsigned int old; \if (!capable (CAP_SYS_ADMIN )) \return -EPERM ; \ err =kstrtol (buf,0 , &value); \if (err) \return err; \ + old = state->disable; \if (value) \ state->disable =1 ; \else \ state->disable =0 ; \ +if (state->notify && state->disable != old) \ + state->notify (state); \return size; \ }
@@ -377,6 +381,8 @@ int cpuidle_add_state_sysfs(struct cpuidkfree (kobj); goto error_state; } + drv->states[i].state_no = i; + drv->states[i].state_count = device->state_count;kobject_uevent (&kobj->kobj,KOBJ_ADD ); device->kobjs[i] = kobj; }
include/linux/cpuidle.h +4 -0
@@ -47,6 +47,10 @@ struct cpuidle_state {int power_usage;/* in mW */ unsigned int target_residency;/* in US */ unsigned int disable; +unsigned int state_no; +unsigned int state_count; + +void (*notify) (struct cpuidle_state *state); int (*enter) (struct cpuidle_device *dev, struct cpuidle_driver *drv,
Re: [PATCH 1/1 v3] Honor state disabling in the cpuidle ladder governor - with sanitizer
Quoted 50 lines · Carsten Emde
Quoted 49 lines · Deepthi Dharwar
Quoted 35 lines · Carsten Emde
Quoted 26 lines · Deepthi Dharwar
Quoted 17 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism
is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the
C state was disabled via sysfs or not. The only way to select a specific
C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.Quoted 3 lines
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Yes, I agree that currently that disabling a particular C-state is not reflected in working of ladder governor. This patch is needed to fix it on ladder too.
Quoted 8 lines
Also wanted to clarify on the intended implementation here, if there are say 5 C-states on a system, disabling 2nd
state would also end by disabling all the remaining 3 deeper states too as ladder governor enters the lightest state first, and will only move on to the next deeper state if a idle period was long enough as per the implementation.
If one is disabling only the deepest state, then it would work as intended.Quoted 5 lines
Yes, the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect.
Agree, as per the ladder design.
Quoted 4 lines
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?
Quoted 7 lines
Quoted 4 lines
No, I dont think thats necessary, current code suffices it. The disable flag is knob we are giving to the user . So may be just document the intended use of disable flag working alongside design of ladder governor.
It's not necessary - but maybe better. Here comes v3 with a sanitizer. Is this too ugly?
The v2, with the documentation in place seems sufficient. Yup, this adds unnecessary fields which are not much use coz the same can be achieved with just disable flag check.
Also, any reason why the patch is being sent as an attachment ? Sending patches as an attachment is not a recommended practice.
Cheers,
Deepthi
Quoted 2 lines
-Carsten.
Re: [PATCH 1/1 v3] Honor state disabling in the cpuidle ladder governor - with sanitizer
Deepthi,
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
[..]
Quoted 4 lines
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?
No, I dont think thats necessary, current code suffices it. The disable flag is knob we are giving to the user . So may be just document the intended use of disable flag working alongside design of ladder governor.
Quoted 5 lines
Quoted 2 lines
It's not necessary - but maybe better. Here comes v3 with a sanitizer. Is this too ugly?
The v2, with the documentation in place seems sufficient. Yup, this adds unnecessary fields which are not much use coz the same can be achieved with just disable flag check.
ok, let's take v2.
Quoted 2 lines
Also, any reason why the patch is being sent as an attachment ? Sending patches as an attachment is not a recommended practice.
Sorry.
-Carsten.
Re: [PATCH 1/1 v3] Honor state disabling in the cpuidle ladder governor - with sanitizer
Hi,
Quoted 13 lines · Carsten Emde
Deepthi,
Quoted 11 lines
Quoted 11 lines
Quoted 9 lines
[..]
Quoted 4 lines
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?
No, I dont think thats necessary, current code suffices it. The disable flag is knob we are giving to the user . So may be just document the intended use of disable flag working alongside design of ladder governor.
It's not necessary - but maybe better. Here comes v3 with a sanitizer. Is this too ugly?
Quoted 8 lines
Quoted 3 lines
The v2, with the documentation in place seems sufficient. Yup, this adds unnecessary fields which are not much use coz the same can be achieved with just disable flag check.
ok, let's take v2.
Quoted 2 lines
Also, any reason why the patch is being sent as an attachment ? Sending patches as an attachment is not a recommended practice.
Sorry.
Can you please resend the version regarded as the current one?
And please send it inline this time.
Thanks,
Rafael
v2 cover cpuidle: allow to disable C states of the ladder governor
In reply to Carsten EmdeRafael,
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 5 lines
[..]
I could implement a sanitize mechanism of the ladder governor that takes care the "disable" variables of all deeper states are set to 1, if a state is disabled, and those of all lighter states are set to 0, if a state is enabled. Do you wish me to do that?No, I dont think thats necessary, current code suffices it. The disable flag is knob we are giving to the user . So may be just document the intended use of disable flag working alongside design of ladder governor.
Quoted 7 lines
Quoted 6 lines
Quoted 5 lines
Quoted 2 lines
It's not necessary - but maybe better. Here comes v3 with a sanitizer. Is this too ugly?
The v2, with the documentation in place seems sufficient. Yup, this adds unnecessary fields which are not much use coz the same can be achieved with just disable flag check.
ok, let's take v2.
Can you please resend the version regarded as the current one?
This is the version that was regarded as the current one (v2).
Changes in v2: A note in the documentation explains why the sysfs variable "disable" may not always reflect the current situation and why modifying it may not always work as expected.
-Carsten.
PATCH v2 1/1 Honor state disabling in the cpuidle ladder governor
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if CONFIG_CPU_IDLE is selected, the menu governor additionally requires CONFIG_NO_HZ.
A particular C state can be disabled by writing to the sysfs file /sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system where CONFIG_NO_HZ is not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to /dev/cpu_dma_latency and keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Note that the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect. A related section has been added to the documentation.
--- Documentation/cpuidle/sysfs.txt | 10 +++++++++- drivers/cpuidle/governors/ladder.c | 4 +++- 2 files changed, 12 insertions(+), 2 deletions(-)
Documentation/cpuidle/sysfs.txt +9 -1
@@ -76,9 +76,17 @@ total 0 * desc : Small description about the idle state (string) -* disable : Option to disable this idle state (bool) +* disable : Option to disable this idle state (bool) -> see note below * latency : Latency to exit out of this idle state (in microseconds) * name : Name of the idle state (string) * power : Power consumed while in this idle state (in milliwatts) * time : Total time spent in this idle state (in microseconds) * usage : Number of times this state was entered (count) + +Note: +The behavior and the effect of the disable variable depends on the +implementation of a particular governor. In the ladder governor, for +example, it is not coherent, i.e. if one is disabling a light state, +then all deeper states are disabled as well, but the disable variable +does not reflect it. Likewise, if one enables a deep state but a lighter +state still is disabled, then this has no effect.
drivers/cpuidle/governors/ladder.c +3 -1
@@ -88,6 +88,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !drv->states[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;
@@ -100,7 +101,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (drv->states[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
Re: [PATCH 1/1 v2] Honor state disabling in the cpuidle ladder governor
Quoted 18 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Quoted 8 lines
Note that the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect. A related section has been added to the documentation.
Signed-off-by Carsten Emde <[email protected]>
Your patch doesn't seem to take this linux-next commit:
into account, does it?
Rafael
Quoted 53 lines
--- Documentation/cpuidle/sysfs.txt | 10 +++++++++- drivers/cpuidle/governors/ladder.c | 4 +++- 2 files changed, 12 insertions(+), 2 deletions(-)Index:
linux-3.4.4-rt14-rc2-64/Documentation/cpuidle/sysfs.txt
===================================================================Documentation/cpuidle/sysfs.txt +9
-1@@ -76,9 +76,17 @@ total 0 * desc : Small description about the idle state (string) -* disable : Option to disable this idle state (bool) +* disable : Option to disable this idle state (bool) -> see note below * latency : Latency to exit out of this idle state (in microseconds) * name : Name of the idle state (string) * power : Power consumed while in this idle state (in milliwatts) * time : Total time spent in this idle state (in microseconds) * usage : Number of times this state was entered (count) + +Note: +The behavior and the effect of the disable variable depends on the +implementation of a particular governor. In the ladder governor, for +example, it is not coherent, i.e. if one is disabling a light state, +then all deeper states are disabled as well, but the disable variable +does not reflect it. Likewise, if one enables a deep state but a lighter +state still is disabled, then this has no effect.Index:
linux-3.4.4-rt14-rc2-64/drivers/cpuidle/governors/ladder.c
===================================================================drivers/cpuidle/governors/ladder.c +3
-1@@ -88,6 +88,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !drv->states[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;@@ -100,7 +101,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (drv->states[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
v3 cover cpuidle: allow to disable C states of the ladder governor
In reply to Carsten EmdeRafael,
Quoted 5 lines
Your patch doesn't seem to take this linux-next commit:
into account, does it?
Hmm, oops, you're right. This one came in after I checked it last time.
Changes in v2: A note in the documentation explains why the sysfs variable "disable" may not always reflect the current situation and why modifying it may not always work as expected.
Changes in v3: The patch now applies to the current linux-next and uses the per-cpu "disable" field.
-Carsten.
PATCH v3 1/1 Honor state disabling in the cpuidle ladder governor
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if CONFIG_CPU_IDLE is selected, the menu governor additionally requires CONFIG_NO_HZ.
A particular C state can be disabled by writing to the sysfs file /sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system where CONFIG_NO_HZ is not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to /dev/cpu_dma_latency and keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.
With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Note that the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect. A related section has been added to the documentation.
--- Documentation/cpuidle/sysfs.txt | 10 +++++++++- drivers/cpuidle/governors/ladder.c | 4 +++- 2 files changed, 12 insertions(+), 2 deletions(-)
Documentation/cpuidle/sysfs.txt +9 -1
@@ -76,9 +76,17 @@ total 0 * desc : Small description about the idle state (string) -* disable : Option to disable this idle state (bool) +* disable : Option to disable this idle state (bool) -> see note below * latency : Latency to exit out of this idle state (in microseconds) * name : Name of the idle state (string) * power : Power consumed while in this idle state (in milliwatts) * time : Total time spent in this idle state (in microseconds) * usage : Number of times this state was entered (count) + +Note: +The behavior and the effect of the disable variable depends on the +implementation of a particular governor. In the ladder governor, for +example, it is not coherent, i.e. if one is disabling a light state, +then all deeper states are disabled as well, but the disable variable +does not reflect it. Likewise, if one enables a deep state but a lighter +state still is disabled, then this has no effect.
drivers/cpuidle/governors/ladder.c +3 -1
@@ -88,6 +88,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !dev->states_usage[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;
@@ -100,7 +101,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (dev->states_usage[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
Re: [PATCH 1/1 v3] Honor state disabling in the cpuidle ladder governor
Quoted 18 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Quoted 8 lines
Note that the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect. A related section has been added to the documentation.
Signed-off-by Carsten Emde <[email protected]>
This looks fine to me, but it's too late for v3.6. I can queue it up for v3.7, though.
Thanks,
Rafael
Quoted 53 lines
--- Documentation/cpuidle/sysfs.txt | 10 +++++++++- drivers/cpuidle/governors/ladder.c | 4 +++- 2 files changed, 12 insertions(+), 2 deletions(-)Index:
linux-3.4.4-rt14-rc2-64/Documentation/cpuidle/sysfs.txt
===================================================================Documentation/cpuidle/sysfs.txt +9
-1@@ -76,9 +76,17 @@ total 0 * desc : Small description about the idle state (string) -* disable : Option to disable this idle state (bool) +* disable : Option to disable this idle state (bool) -> see note below * latency : Latency to exit out of this idle state (in microseconds) * name : Name of the idle state (string) * power : Power consumed while in this idle state (in milliwatts) * time : Total time spent in this idle state (in microseconds) * usage : Number of times this state was entered (count) + +Note: +The behavior and the effect of the disable variable depends on the +implementation of a particular governor. In the ladder governor, for +example, it is not coherent, i.e. if one is disabling a light state, +then all deeper states are disabled as well, but the disable variable +does not reflect it. Likewise, if one enables a deep state but a lighter +state still is disabled, then this has no effect.Index:
linux-3.4.4-rt14-rc2-64/drivers/cpuidle/governors/ladder.c
===================================================================drivers/cpuidle/governors/ladder.c +3
-1@@ -88,6 +88,7 @@ static int ladder_select_state(struct cp/* consider promotion */ if (last_idx < drv->state_count -1 && + !dev->states_usage[last_idx +1 ].disable && last_residency > last_state->threshold.promotion_time && drv->states[last_idx +1 ].exit_latency <= latency_req) { last_state->stats.promotion_count++;@@ -100,7 +101,8 @@ static int ladder_select_state(struct cp/* consider demotion */ if (last_idx >CPUIDLE_DRIVER_STATE_START && - drv->states[last_idx].exit_latency > latency_req) { + (dev->states_usage[last_idx].disable || + drv->states[last_idx].exit_latency > latency_req)) {int i;for (i = last_idx -1 ; i >CPUIDLE_DRIVER_STATE_START ; i--) {
Re: [PATCH 1/1 v3] Honor state disabling in the cpuidle ladder governor
Quoted 26 lines · Rafael J. Wysocki
Quoted 25 lines · Carsten Emde
There are two cpuidle governors ladder and menu. While the ladder governor is always available, if
CONFIG_CPU_IDLEis selected, the menu governor additionally requiresCONFIG_NO_HZ.A particular C state can be disabled by writing to the sysfs file
/sys/devices/system/cpu/cpuN/cpuidle/stateN/disable, but this mechanism is only implemented in the menu governor. Thus, in a system whereCONFIG_NO_HZis not selected, the ladder governor becomes default and always will walk through all sleep states - irrespective of whether the C state was disabled via sysfs or not. The only way to select a specific C state was to write the related latency to/dev/cpu_dma_latencyand keep the file open as long as this setting was required - not very practical and not suitable for setting a single core in an SMP system.With this patch, the ladder governor only will promote to the next C state, if it has not been disabled, and it will demote, if the current C state was disabled.
Note that the patch does not make the setting of the sysfs variable "disable" coherent, i.e. if one is disabling a light state, then all deeper states are disabled as well, but the "disable" variable does not reflect it. Likewise, if one enables a deep state but a lighter state still is disabled, then this has no effect. A related section has been added to the documentation.
Quoted 4 lines
Quoted 1 line
Signed-off-by Carsten Emde<[email protected]>This looks fine to me, but it's too late for v3.6. I can queue it up for v3.7, though.
Yes, please.
Thanks,
-Carsten.