2021-03-08 09:18:20

by Daniel Lezcano

[permalink] [raw]
Subject: [PATCH v5 1/4] PM / devfreq: Register devfreq as a cooling device on demand

Currently the default behavior is to manually having the devfreq
backend to register themselves as a devfreq cooling device.

Instead of adding the code in the drivers for the thermal cooling
device registering, let's provide a flag in the devfreq's profile to
tell the common devfreq code to register the newly created devfreq as
a cooling device.

Suggested-by: Chanwoo Choi <[email protected]>
Signed-off-by: Daniel Lezcano <[email protected]>
---
V5:
- Changed subject prefix by: PM / devfreq
V4:
- Replaced thermal_cooling_device_unregister() by
devfreq_cooling_unregister()
V3:
- Rebased on linux-pm branch without units.h
- Set the cdev to NULL in case of error
- Added description for the cdev field in the devfreq structure
V2:
- Added is_cooling_device boolean in profile structure
- Register cooling device when the is_cooling_device boolean is set
- Remove devfreq cooling device registration in the backend drivers
V1:
- Register devfreq as a cooling device unconditionnally
---
drivers/devfreq/devfreq.c | 13 +++++++++++++
include/linux/devfreq.h | 8 ++++++++
2 files changed, 21 insertions(+)

diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
index bf3047896e41..8a535d4d6083 100644
--- a/drivers/devfreq/devfreq.c
+++ b/drivers/devfreq/devfreq.c
@@ -11,6 +11,7 @@
#include <linux/kmod.h>
#include <linux/sched.h>
#include <linux/debugfs.h>
+#include <linux/devfreq_cooling.h>
#include <linux/errno.h>
#include <linux/err.h>
#include <linux/init.h>
@@ -26,6 +27,7 @@
#include <linux/hrtimer.h>
#include <linux/of.h>
#include <linux/pm_qos.h>
+#include <linux/thermal.h>
#include "governor.h"

#define CREATE_TRACE_POINTS
@@ -935,6 +937,15 @@ struct devfreq *devfreq_add_device(struct device *dev,

mutex_unlock(&devfreq_list_lock);

+ if (devfreq->profile->is_cooling_device) {
+ devfreq->cdev = devfreq_cooling_em_register(devfreq, NULL);
+ if (IS_ERR(devfreq->cdev)) {
+ dev_info(dev, "Failed to register devfreq "
+ "cooling device\n");
+ devfreq->cdev = NULL;
+ }
+ }
+
return devfreq;

err_init:
@@ -960,6 +971,8 @@ int devfreq_remove_device(struct devfreq *devfreq)
if (!devfreq)
return -EINVAL;

+ devfreq_cooling_unregister(devfreq->cdev);
+
if (devfreq->governor) {
devfreq->governor->event_handler(devfreq,
DEVFREQ_GOV_STOP, NULL);
diff --git a/include/linux/devfreq.h b/include/linux/devfreq.h
index 26ea0850be9b..aba7ace11b72 100644
--- a/include/linux/devfreq.h
+++ b/include/linux/devfreq.h
@@ -98,11 +98,15 @@ struct devfreq_dev_status {
* @freq_table: Optional list of frequencies to support statistics
* and freq_table must be generated in ascending order.
* @max_state: The size of freq_table.
+ *
+ * @is_cooling_device: A self-explanatory boolean giving the device a
+ * cooling effect property.
*/
struct devfreq_dev_profile {
unsigned long initial_freq;
unsigned int polling_ms;
enum devfreq_timer timer;
+ bool is_cooling_device;

int (*target)(struct device *dev, unsigned long *freq, u32 flags);
int (*get_dev_status)(struct device *dev,
@@ -156,6 +160,7 @@ struct devfreq_stats {
* @suspend_count: suspend requests counter for a device.
* @stats: Statistics of devfreq device behavior
* @transition_notifier_list: list head of DEVFREQ_TRANSITION_NOTIFIER notifier
+ * @cdev: Cooling device pointer if the devfreq has cooling property
* @nb_min: Notifier block for DEV_PM_QOS_MIN_FREQUENCY
* @nb_max: Notifier block for DEV_PM_QOS_MAX_FREQUENCY
*
@@ -198,6 +203,9 @@ struct devfreq {

struct srcu_notifier_head transition_notifier_list;

+ /* Pointer to the cooling device if used for thermal mitigation */
+ struct thermal_cooling_device *cdev;
+
struct notifier_block nb_min;
struct notifier_block nb_max;
};
--
2.17.1


2021-03-08 09:18:51

by Daniel Lezcano

[permalink] [raw]
Subject: [PATCH v5 4/4] PM / devfreq: lima: Use devfreq cooling device registration

The devfreq core code is able to register the devfreq device as a
cooling device if the 'is_cooling_device' flag is set in the profile.

Use this flag and remove the cooling device registering code.

Signed-off-by: Daniel Lezcano <[email protected]>
---
drivers/gpu/drm/lima/lima_devfreq.c | 14 +-------------
drivers/gpu/drm/lima/lima_devfreq.h | 2 --
2 files changed, 1 insertion(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/lima/lima_devfreq.c b/drivers/gpu/drm/lima/lima_devfreq.c
index 5686ad4aaf7c..86aea1bdc4f4 100644
--- a/drivers/gpu/drm/lima/lima_devfreq.c
+++ b/drivers/gpu/drm/lima/lima_devfreq.c
@@ -7,7 +7,6 @@
*/
#include <linux/clk.h>
#include <linux/devfreq.h>
-#include <linux/devfreq_cooling.h>
#include <linux/device.h>
#include <linux/platform_device.h>
#include <linux/pm_opp.h>
@@ -84,17 +83,13 @@ static struct devfreq_dev_profile lima_devfreq_profile = {
.polling_ms = 50, /* ~3 frames */
.target = lima_devfreq_target,
.get_dev_status = lima_devfreq_get_dev_status,
+ .is_cooling_device = true,
};

void lima_devfreq_fini(struct lima_device *ldev)
{
struct lima_devfreq *devfreq = &ldev->devfreq;

- if (devfreq->cooling) {
- devfreq_cooling_unregister(devfreq->cooling);
- devfreq->cooling = NULL;
- }
-
if (devfreq->devfreq) {
devm_devfreq_remove_device(ldev->dev, devfreq->devfreq);
devfreq->devfreq = NULL;
@@ -110,7 +105,6 @@ void lima_devfreq_fini(struct lima_device *ldev)

int lima_devfreq_init(struct lima_device *ldev)
{
- struct thermal_cooling_device *cooling;
struct device *dev = ldev->dev;
struct opp_table *opp_table;
struct devfreq *devfreq;
@@ -173,12 +167,6 @@ int lima_devfreq_init(struct lima_device *ldev)

ldevfreq->devfreq = devfreq;

- cooling = of_devfreq_cooling_register(dev->of_node, devfreq);
- if (IS_ERR(cooling))
- dev_info(dev, "Failed to register cooling device\n");
- else
- ldevfreq->cooling = cooling;
-
return 0;

err_fini:
diff --git a/drivers/gpu/drm/lima/lima_devfreq.h b/drivers/gpu/drm/lima/lima_devfreq.h
index 2d9b3008ce77..c43a2069e5d3 100644
--- a/drivers/gpu/drm/lima/lima_devfreq.h
+++ b/drivers/gpu/drm/lima/lima_devfreq.h
@@ -9,7 +9,6 @@

struct devfreq;
struct opp_table;
-struct thermal_cooling_device;

struct lima_device;

@@ -17,7 +16,6 @@ struct lima_devfreq {
struct devfreq *devfreq;
struct opp_table *clkname_opp_table;
struct opp_table *regulators_opp_table;
- struct thermal_cooling_device *cooling;

ktime_t busy_time;
ktime_t idle_time;
--
2.17.1

2021-03-08 09:20:18

by Daniel Lezcano

[permalink] [raw]
Subject: [PATCH v5 2/4] PM / devfreq: msm: Use devfreq cooling device registration

The devfreq core code is able to register the devfreq device as a
cooling device if the 'is_cooling_device' flag is set in the profile.

Use this flag and remove the cooling device registering code.

Tested on dragonboard 845c

Signed-off-by: Daniel Lezcano <[email protected]>
---
drivers/gpu/drm/msm/msm_gpu.c | 12 +-----------
drivers/gpu/drm/msm/msm_gpu.h | 2 --
2 files changed, 1 insertion(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/msm/msm_gpu.c b/drivers/gpu/drm/msm/msm_gpu.c
index ab7c167b0623..eade94271a60 100644
--- a/drivers/gpu/drm/msm/msm_gpu.c
+++ b/drivers/gpu/drm/msm/msm_gpu.c
@@ -14,7 +14,6 @@
#include <generated/utsrelease.h>
#include <linux/string_helpers.h>
#include <linux/devfreq.h>
-#include <linux/devfreq_cooling.h>
#include <linux/devcoredump.h>
#include <linux/sched/task.h>

@@ -82,6 +81,7 @@ static struct devfreq_dev_profile msm_devfreq_profile = {
.target = msm_devfreq_target,
.get_dev_status = msm_devfreq_get_dev_status,
.get_cur_freq = msm_devfreq_get_cur_freq,
+ .is_cooling_device = true,
};

static void msm_devfreq_init(struct msm_gpu *gpu)
@@ -112,14 +112,6 @@ static void msm_devfreq_init(struct msm_gpu *gpu)
}

devfreq_suspend_device(gpu->devfreq.devfreq);
-
- gpu->cooling = of_devfreq_cooling_register(gpu->pdev->dev.of_node,
- gpu->devfreq.devfreq);
- if (IS_ERR(gpu->cooling)) {
- DRM_DEV_ERROR(&gpu->pdev->dev,
- "Couldn't register GPU cooling device\n");
- gpu->cooling = NULL;
- }
}

static int enable_pwrrail(struct msm_gpu *gpu)
@@ -1056,6 +1048,4 @@ void msm_gpu_cleanup(struct msm_gpu *gpu)
if (gpu->worker) {
kthread_destroy_worker(gpu->worker);
}
-
- devfreq_cooling_unregister(gpu->cooling);
}
diff --git a/drivers/gpu/drm/msm/msm_gpu.h b/drivers/gpu/drm/msm/msm_gpu.h
index d7cd02cd2109..93419368bac8 100644
--- a/drivers/gpu/drm/msm/msm_gpu.h
+++ b/drivers/gpu/drm/msm/msm_gpu.h
@@ -155,8 +155,6 @@ struct msm_gpu {
struct msm_gpu_state *crashstate;
/* True if the hardware supports expanded apriv (a650 and newer) */
bool hw_apriv;
-
- struct thermal_cooling_device *cooling;
};

static inline struct msm_gpu *dev_to_gpu(struct device *dev)
--
2.17.1

2021-03-08 09:20:31

by Daniel Lezcano

[permalink] [raw]
Subject: [PATCH v5 3/4] PM / devfreq: panfrost: Use devfreq cooling device registration

The devfreq core code is able to register the devfreq device as a
cooling device if the 'is_cooling_device' flag is set in the profile.

Use this flag and remove the cooling device registering code.

Tested on rock960.

Signed-off-by: Daniel Lezcano <[email protected]>
---
drivers/gpu/drm/panfrost/panfrost_devfreq.c | 14 +-------------
drivers/gpu/drm/panfrost/panfrost_devfreq.h | 3 ---
2 files changed, 1 insertion(+), 16 deletions(-)

diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.c b/drivers/gpu/drm/panfrost/panfrost_devfreq.c
index 56b3f5935703..4d96edf1bc54 100644
--- a/drivers/gpu/drm/panfrost/panfrost_devfreq.c
+++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.c
@@ -3,7 +3,6 @@

#include <linux/clk.h>
#include <linux/devfreq.h>
-#include <linux/devfreq_cooling.h>
#include <linux/platform_device.h>
#include <linux/pm_opp.h>

@@ -80,6 +79,7 @@ static struct devfreq_dev_profile panfrost_devfreq_profile = {
.polling_ms = 50, /* ~3 frames */
.target = panfrost_devfreq_target,
.get_dev_status = panfrost_devfreq_get_dev_status,
+ .is_cooling_device = true,
};

int panfrost_devfreq_init(struct panfrost_device *pfdev)
@@ -90,7 +90,6 @@ int panfrost_devfreq_init(struct panfrost_device *pfdev)
struct device *dev = &pfdev->pdev->dev;
struct devfreq *devfreq;
struct opp_table *opp_table;
- struct thermal_cooling_device *cooling;
struct panfrost_devfreq *pfdevfreq = &pfdev->pfdevfreq;

opp_table = dev_pm_opp_set_regulators(dev, pfdev->comp->supply_names,
@@ -139,12 +138,6 @@ int panfrost_devfreq_init(struct panfrost_device *pfdev)
}
pfdevfreq->devfreq = devfreq;

- cooling = devfreq_cooling_em_register(devfreq, NULL);
- if (IS_ERR(cooling))
- DRM_DEV_INFO(dev, "Failed to register cooling device\n");
- else
- pfdevfreq->cooling = cooling;
-
return 0;

err_fini:
@@ -156,11 +149,6 @@ void panfrost_devfreq_fini(struct panfrost_device *pfdev)
{
struct panfrost_devfreq *pfdevfreq = &pfdev->pfdevfreq;

- if (pfdevfreq->cooling) {
- devfreq_cooling_unregister(pfdevfreq->cooling);
- pfdevfreq->cooling = NULL;
- }
-
if (pfdevfreq->opp_of_table_added) {
dev_pm_opp_of_remove_table(&pfdev->pdev->dev);
pfdevfreq->opp_of_table_added = false;
diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.h b/drivers/gpu/drm/panfrost/panfrost_devfreq.h
index db6ea48e21f9..470f5c974703 100644
--- a/drivers/gpu/drm/panfrost/panfrost_devfreq.h
+++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.h
@@ -9,14 +9,11 @@

struct devfreq;
struct opp_table;
-struct thermal_cooling_device;
-
struct panfrost_device;

struct panfrost_devfreq {
struct devfreq *devfreq;
struct opp_table *regulators_opp_table;
- struct thermal_cooling_device *cooling;
bool opp_of_table_added;

ktime_t busy_time;
--
2.17.1

2021-03-08 10:18:21

by Steven Price

[permalink] [raw]
Subject: Re: [PATCH v5 1/4] PM / devfreq: Register devfreq as a cooling device on demand

On 08/03/2021 09:16, Daniel Lezcano wrote:
> Currently the default behavior is to manually having the devfreq
> backend to register themselves as a devfreq cooling device.
>
> Instead of adding the code in the drivers for the thermal cooling
> device registering, let's provide a flag in the devfreq's profile to
> tell the common devfreq code to register the newly created devfreq as
> a cooling device.
>
> Suggested-by: Chanwoo Choi <[email protected]>
> Signed-off-by: Daniel Lezcano <[email protected]>
> ---
> V5:
> - Changed subject prefix by: PM / devfreq
> V4:
> - Replaced thermal_cooling_device_unregister() by
> devfreq_cooling_unregister()
> V3:
> - Rebased on linux-pm branch without units.h
> - Set the cdev to NULL in case of error
> - Added description for the cdev field in the devfreq structure
> V2:
> - Added is_cooling_device boolean in profile structure
> - Register cooling device when the is_cooling_device boolean is set
> - Remove devfreq cooling device registration in the backend drivers
> V1:
> - Register devfreq as a cooling device unconditionnally
> ---
> drivers/devfreq/devfreq.c | 13 +++++++++++++
> include/linux/devfreq.h | 8 ++++++++
> 2 files changed, 21 insertions(+)
>
> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> index bf3047896e41..8a535d4d6083 100644
> --- a/drivers/devfreq/devfreq.c
> +++ b/drivers/devfreq/devfreq.c
> @@ -11,6 +11,7 @@
> #include <linux/kmod.h>
> #include <linux/sched.h>
> #include <linux/debugfs.h>
> +#include <linux/devfreq_cooling.h>
> #include <linux/errno.h>
> #include <linux/err.h>
> #include <linux/init.h>
> @@ -26,6 +27,7 @@
> #include <linux/hrtimer.h>
> #include <linux/of.h>
> #include <linux/pm_qos.h>
> +#include <linux/thermal.h>
> #include "governor.h"
>
> #define CREATE_TRACE_POINTS
> @@ -935,6 +937,15 @@ struct devfreq *devfreq_add_device(struct device *dev,
>
> mutex_unlock(&devfreq_list_lock);
>
> + if (devfreq->profile->is_cooling_device) {
> + devfreq->cdev = devfreq_cooling_em_register(devfreq, NULL);
> + if (IS_ERR(devfreq->cdev)) {
> + dev_info(dev, "Failed to register devfreq "
> + "cooling device\n");

Please don't split strings across multiple lines (it makes grepping for
them harder). With that fixed:

Reviewed-by: Steven Price <[email protected]>

> + devfreq->cdev = NULL;
> + }
> + }
> +
> return devfreq;
>
> err_init:
> @@ -960,6 +971,8 @@ int devfreq_remove_device(struct devfreq *devfreq)
> if (!devfreq)
> return -EINVAL;
>
> + devfreq_cooling_unregister(devfreq->cdev);
> +
> if (devfreq->governor) {
> devfreq->governor->event_handler(devfreq,
> DEVFREQ_GOV_STOP, NULL);
> diff --git a/include/linux/devfreq.h b/include/linux/devfreq.h
> index 26ea0850be9b..aba7ace11b72 100644
> --- a/include/linux/devfreq.h
> +++ b/include/linux/devfreq.h
> @@ -98,11 +98,15 @@ struct devfreq_dev_status {
> * @freq_table: Optional list of frequencies to support statistics
> * and freq_table must be generated in ascending order.
> * @max_state: The size of freq_table.
> + *
> + * @is_cooling_device: A self-explanatory boolean giving the device a
> + * cooling effect property.
> */
> struct devfreq_dev_profile {
> unsigned long initial_freq;
> unsigned int polling_ms;
> enum devfreq_timer timer;
> + bool is_cooling_device;
>
> int (*target)(struct device *dev, unsigned long *freq, u32 flags);
> int (*get_dev_status)(struct device *dev,
> @@ -156,6 +160,7 @@ struct devfreq_stats {
> * @suspend_count: suspend requests counter for a device.
> * @stats: Statistics of devfreq device behavior
> * @transition_notifier_list: list head of DEVFREQ_TRANSITION_NOTIFIER notifier
> + * @cdev: Cooling device pointer if the devfreq has cooling property
> * @nb_min: Notifier block for DEV_PM_QOS_MIN_FREQUENCY
> * @nb_max: Notifier block for DEV_PM_QOS_MAX_FREQUENCY
> *
> @@ -198,6 +203,9 @@ struct devfreq {
>
> struct srcu_notifier_head transition_notifier_list;
>
> + /* Pointer to the cooling device if used for thermal mitigation */
> + struct thermal_cooling_device *cdev;
> +
> struct notifier_block nb_min;
> struct notifier_block nb_max;
> };
>

2021-03-08 10:19:29

by Lukasz Luba

[permalink] [raw]
Subject: Re: [PATCH v5 1/4] PM / devfreq: Register devfreq as a cooling device on demand

Hi Daniel,

In general the approach is good. If there is a special GPU driver, which
would like to provide 'struct devfreq_cooling_power *' it would leave
'is_cooling_device=false' and register manually:
devfreq_cooling_em_register(df, dfc_power);

Please find only a few minor comments below.


On 3/8/21 9:16 AM, Daniel Lezcano wrote:
> Currently the default behavior is to manually having the devfreq
> backend to register themselves as a devfreq cooling device.
>
> Instead of adding the code in the drivers for the thermal cooling
> device registering, let's provide a flag in the devfreq's profile to
> tell the common devfreq code to register the newly created devfreq as
> a cooling device.
>
> Suggested-by: Chanwoo Choi <[email protected]>
> Signed-off-by: Daniel Lezcano <[email protected]>
> ---
> V5:
> - Changed subject prefix by: PM / devfreq
> V4:
> - Replaced thermal_cooling_device_unregister() by
> devfreq_cooling_unregister()
> V3:
> - Rebased on linux-pm branch without units.h
> - Set the cdev to NULL in case of error
> - Added description for the cdev field in the devfreq structure
> V2:
> - Added is_cooling_device boolean in profile structure
> - Register cooling device when the is_cooling_device boolean is set
> - Remove devfreq cooling device registration in the backend drivers
> V1:
> - Register devfreq as a cooling device unconditionnally
> ---
> drivers/devfreq/devfreq.c | 13 +++++++++++++
> include/linux/devfreq.h | 8 ++++++++
> 2 files changed, 21 insertions(+)
>
> diff --git a/drivers/devfreq/devfreq.c b/drivers/devfreq/devfreq.c
> index bf3047896e41..8a535d4d6083 100644
> --- a/drivers/devfreq/devfreq.c
> +++ b/drivers/devfreq/devfreq.c
> @@ -11,6 +11,7 @@
> #include <linux/kmod.h>
> #include <linux/sched.h>
> #include <linux/debugfs.h>
> +#include <linux/devfreq_cooling.h>
> #include <linux/errno.h>
> #include <linux/err.h>
> #include <linux/init.h>
> @@ -26,6 +27,7 @@
> #include <linux/hrtimer.h>
> #include <linux/of.h>
> #include <linux/pm_qos.h>
> +#include <linux/thermal.h>

Since the unregister is now called from devfreq_cooling.h, this header
here is not needed. It might rather be moved into the devfreq.h
where there is this new field: 'cdev' if we are going to use something
from it.

> #include "governor.h"
>
> #define CREATE_TRACE_POINTS
> @@ -935,6 +937,15 @@ struct devfreq *devfreq_add_device(struct device *dev,
>
> mutex_unlock(&devfreq_list_lock);
>
> + if (devfreq->profile->is_cooling_device) {
> + devfreq->cdev = devfreq_cooling_em_register(devfreq, NULL);
> + if (IS_ERR(devfreq->cdev)) {
> + dev_info(dev, "Failed to register devfreq "
> + "cooling device\n");

I would skip this print. Not every kernel config has DEVFREQ_THERMAL set
and the value ERR_PTR(-EINVAL) from header would be returned in such
case.
Inside the devfreq_cooling_em_register() (and others) there are prints
showing that something wrong happen and would be triggered if compiled.

> + devfreq->cdev = NULL;
> + }
> + }
> +
> return devfreq;
>
> err_init:
> @@ -960,6 +971,8 @@ int devfreq_remove_device(struct devfreq *devfreq)
> if (!devfreq)
> return -EINVAL;
>
> + devfreq_cooling_unregister(devfreq->cdev);
> +
> if (devfreq->governor) {
> devfreq->governor->event_handler(devfreq,
> DEVFREQ_GOV_STOP, NULL);
> diff --git a/include/linux/devfreq.h b/include/linux/devfreq.h
> index 26ea0850be9b..aba7ace11b72 100644
> --- a/include/linux/devfreq.h
> +++ b/include/linux/devfreq.h
> @@ -98,11 +98,15 @@ struct devfreq_dev_status {
> * @freq_table: Optional list of frequencies to support statistics
> * and freq_table must be generated in ascending order.
> * @max_state: The size of freq_table.
> + *
> + * @is_cooling_device: A self-explanatory boolean giving the device a
> + * cooling effect property.
> */
> struct devfreq_dev_profile {
> unsigned long initial_freq;
> unsigned int polling_ms;
> enum devfreq_timer timer;
> + bool is_cooling_device;
>
> int (*target)(struct device *dev, unsigned long *freq, u32 flags);
> int (*get_dev_status)(struct device *dev,
> @@ -156,6 +160,7 @@ struct devfreq_stats {
> * @suspend_count: suspend requests counter for a device.
> * @stats: Statistics of devfreq device behavior
> * @transition_notifier_list: list head of DEVFREQ_TRANSITION_NOTIFIER notifier
> + * @cdev: Cooling device pointer if the devfreq has cooling property
> * @nb_min: Notifier block for DEV_PM_QOS_MIN_FREQUENCY
> * @nb_max: Notifier block for DEV_PM_QOS_MAX_FREQUENCY
> *
> @@ -198,6 +203,9 @@ struct devfreq {
>
> struct srcu_notifier_head transition_notifier_list;
>
> + /* Pointer to the cooling device if used for thermal mitigation */
> + struct thermal_cooling_device *cdev;

The linux/thermal.h for 'cdev' would be needed in this header.

> +
> struct notifier_block nb_min;
> struct notifier_block nb_max;
> };
>

With this small changes:

Reviewed-by: Lukasz Luba <[email protected]>

Regards,
Lukasz

2021-03-08 11:22:17

by Steven Price

[permalink] [raw]
Subject: Re: [PATCH v5 3/4] PM / devfreq: panfrost: Use devfreq cooling device registration

On 08/03/2021 09:16, Daniel Lezcano wrote:
> The devfreq core code is able to register the devfreq device as a
> cooling device if the 'is_cooling_device' flag is set in the profile.
>
> Use this flag and remove the cooling device registering code.
>
> Tested on rock960.
>
> Signed-off-by: Daniel Lezcano <[email protected]>

Reviewed-by: Steven Price <[email protected]>

> ---
> drivers/gpu/drm/panfrost/panfrost_devfreq.c | 14 +-------------
> drivers/gpu/drm/panfrost/panfrost_devfreq.h | 3 ---
> 2 files changed, 1 insertion(+), 16 deletions(-)
>
> diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.c b/drivers/gpu/drm/panfrost/panfrost_devfreq.c
> index 56b3f5935703..4d96edf1bc54 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_devfreq.c
> +++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.c
> @@ -3,7 +3,6 @@
>
> #include <linux/clk.h>
> #include <linux/devfreq.h>
> -#include <linux/devfreq_cooling.h>
> #include <linux/platform_device.h>
> #include <linux/pm_opp.h>
>
> @@ -80,6 +79,7 @@ static struct devfreq_dev_profile panfrost_devfreq_profile = {
> .polling_ms = 50, /* ~3 frames */
> .target = panfrost_devfreq_target,
> .get_dev_status = panfrost_devfreq_get_dev_status,
> + .is_cooling_device = true,
> };
>
> int panfrost_devfreq_init(struct panfrost_device *pfdev)
> @@ -90,7 +90,6 @@ int panfrost_devfreq_init(struct panfrost_device *pfdev)
> struct device *dev = &pfdev->pdev->dev;
> struct devfreq *devfreq;
> struct opp_table *opp_table;
> - struct thermal_cooling_device *cooling;
> struct panfrost_devfreq *pfdevfreq = &pfdev->pfdevfreq;
>
> opp_table = dev_pm_opp_set_regulators(dev, pfdev->comp->supply_names,
> @@ -139,12 +138,6 @@ int panfrost_devfreq_init(struct panfrost_device *pfdev)
> }
> pfdevfreq->devfreq = devfreq;
>
> - cooling = devfreq_cooling_em_register(devfreq, NULL);
> - if (IS_ERR(cooling))
> - DRM_DEV_INFO(dev, "Failed to register cooling device\n");
> - else
> - pfdevfreq->cooling = cooling;
> -
> return 0;
>
> err_fini:
> @@ -156,11 +149,6 @@ void panfrost_devfreq_fini(struct panfrost_device *pfdev)
> {
> struct panfrost_devfreq *pfdevfreq = &pfdev->pfdevfreq;
>
> - if (pfdevfreq->cooling) {
> - devfreq_cooling_unregister(pfdevfreq->cooling);
> - pfdevfreq->cooling = NULL;
> - }
> -
> if (pfdevfreq->opp_of_table_added) {
> dev_pm_opp_of_remove_table(&pfdev->pdev->dev);
> pfdevfreq->opp_of_table_added = false;
> diff --git a/drivers/gpu/drm/panfrost/panfrost_devfreq.h b/drivers/gpu/drm/panfrost/panfrost_devfreq.h
> index db6ea48e21f9..470f5c974703 100644
> --- a/drivers/gpu/drm/panfrost/panfrost_devfreq.h
> +++ b/drivers/gpu/drm/panfrost/panfrost_devfreq.h
> @@ -9,14 +9,11 @@
>
> struct devfreq;
> struct opp_table;
> -struct thermal_cooling_device;
> -
> struct panfrost_device;
>
> struct panfrost_devfreq {
> struct devfreq *devfreq;
> struct opp_table *regulators_opp_table;
> - struct thermal_cooling_device *cooling;
> bool opp_of_table_added;
>
> ktime_t busy_time;
>

2021-03-08 12:02:07

by Daniel Lezcano

[permalink] [raw]
Subject: Re: [PATCH v5 1/4] PM / devfreq: Register devfreq as a cooling device on demand

On 08/03/2021 10:53, Lukasz Luba wrote:
> Hi Daniel,
>
> In general the approach is good. If there is a special GPU driver, which
> would like to provide 'struct devfreq_cooling_power *' it would leave
> 'is_cooling_device=false' and register manually:
> devfreq_cooling_em_register(df, dfc_power);
>
> Please find only a few minor comments below.
>
>
> On 3/8/21 9:16 AM, Daniel Lezcano wrote:
>> Currently the default behavior is to manually having the devfreq
>> backend to register themselves as a devfreq cooling device.
>>
>> Instead of adding the code in the drivers for the thermal cooling
>> device registering, let's provide a flag in the devfreq's profile to
>> tell the common devfreq code to register the newly created devfreq as
>> a cooling device.
>>
>> Suggested-by: Chanwoo Choi <[email protected]>
>> Signed-off-by: Daniel Lezcano <[email protected]>
>> ---

[ ... ]

>> +    struct thermal_cooling_device *cdev;
>
> The linux/thermal.h for 'cdev' would be needed in this header.

May be just a forward declaration ?

struct thermal_cooling_device;

>> +
>>       struct notifier_block nb_min;
>>       struct notifier_block nb_max;
>>   };
>>
>
> With this small changes:
>
> Reviewed-by: Lukasz Luba <[email protected]>
>
> Regards,
> Lukasz


--
<http://www.linaro.org/> Linaro.org │ Open source software for ARM SoCs

Follow Linaro: <http://www.facebook.com/pages/Linaro> Facebook |
<http://twitter.com/#!/linaroorg> Twitter |
<http://www.linaro.org/linaro-blog/> Blog

2021-03-08 12:08:00

by Lukasz Luba

[permalink] [raw]
Subject: Re: [PATCH v5 1/4] PM / devfreq: Register devfreq as a cooling device on demand



On 3/8/21 12:00 PM, Daniel Lezcano wrote:
> On 08/03/2021 10:53, Lukasz Luba wrote:
>> Hi Daniel,
>>
>> In general the approach is good. If there is a special GPU driver, which
>> would like to provide 'struct devfreq_cooling_power *' it would leave
>> 'is_cooling_device=false' and register manually:
>> devfreq_cooling_em_register(df, dfc_power);
>>
>> Please find only a few minor comments below.
>>
>>
>> On 3/8/21 9:16 AM, Daniel Lezcano wrote:
>>> Currently the default behavior is to manually having the devfreq
>>> backend to register themselves as a devfreq cooling device.
>>>
>>> Instead of adding the code in the drivers for the thermal cooling
>>> device registering, let's provide a flag in the devfreq's profile to
>>> tell the common devfreq code to register the newly created devfreq as
>>> a cooling device.
>>>
>>> Suggested-by: Chanwoo Choi <[email protected]>
>>> Signed-off-by: Daniel Lezcano <[email protected]>
>>> ---
>
> [ ... ]
>
>>> +    struct thermal_cooling_device *cdev;
>>
>> The linux/thermal.h for 'cdev' would be needed in this header.
>
> May be just a forward declaration ?
>
> struct thermal_cooling_device;

Make sense

Regards,
Lukasz