PM / Domains: enable domain idle state accounting

PATCH Superseded Lina Iyer 2 participants 7 messages Oct 15, 2020 – Oct 16, 2020

Subsystems (6)

Arm Qualcomm mailing list Driver core Suspend to RAM Hibernation +2 more subsystems

lore

Share & tools

lore

Lina Iyer

author

2020-10-15 20:47 UTC #0 lore raw

PATCH v3 1/1 PM / Domains: enable domain idle state accounting

To enable better debug of PM domains, let's keep a track of the success and rejections in entering each domain idle state.

This statistics is exported in debugfs when reading the idle_states node, associated with each PM domain.

Signed-off-by Lina Iyer <[email protected]>

---
Changes in v3:
- Rebased on top of mainline

---
 drivers/base/power/domain.c | 10 +++++++---
 include/linux/pm_domain.h   |  2 ++
 2 files changed, 9 insertions(+), 3 deletions(-)

drivers/base/power/domain.c +7 -3

index 2cb5e04cf86c..20ecc1a7113c 100644
@@ -559,12 +559,15 @@ static int genpd_power_off(struct generic_pm_domain *genpd, bool one_dev_on,
 		 * happen very often).
 		 */
 		ret = _genpd_power_off(genpd, true);
-		if (ret)
+		if (ret) {
+			genpd->states[genpd->state_idx].rejected++;
 			return ret;
+		}
 	}
 
 	genpd->status = GPD_STATE_POWER_OFF;
 	genpd_update_accounting(genpd);
+	genpd->states[genpd->state_idx].usage++;
 
 	list_for_each_entry(link, &genpd->child_links, child_node) {
 		genpd_sd_counter_dec(link->parent);

@@ -2932,7 +2935,7 @@ static int idle_states_show(struct seq_file *s, void *data)
 	if (ret)
 		return -ERESTARTSYS;
 
-	seq_puts(s, "State          Time Spent(ms)\n");
+	seq_puts(s, "State          Time Spent(ms) Usage          Rejected\n");
 
 	for (i = 0; i < genpd->state_count; i++) {
 		ktime_t delta = 0;

@@ -2944,7 +2947,8 @@ static int idle_states_show(struct seq_file *s, void *data)
 
 		msecs = ktime_to_ms(
 			ktime_add(genpd->states[i].idle_time, delta));
-		seq_printf(s, "S%-13i %lld\n", i, msecs);
+		seq_printf(s, "S%-13i %-14lld %-14llu %llu\n", i, msecs,
+			      genpd->states[i].usage, genpd->states[i].rejected);
 	}
 
 	genpd_unlock(genpd);

include/linux/pm_domain.h +2 -0

index ee11502a575b..1d9a8de6b7ac 100644
@@ -82,6 +82,8 @@ struct genpd_power_state {
 	s64 power_off_latency_ns;
 	s64 power_on_latency_ns;
 	s64 residency_ns;
+	u64 usage;
+	u64 rejected;
 	struct fwnode_handle *fwnode;
 	ktime_t idle_time;
 	void *data;

Signature · 2 lines

The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project

Rafael J. Wysocki

maintainer

2020-10-16 15:55 UTC #1 lore raw

Quoted 11 lines · Lina Iyer

To enable better debug of PM domains, let's keep a track of the success and rejections in entering each domain idle state.
This statistics is exported in debugfs when reading the idle_states node, associated with each PM domain.
Signed-off-by Lina Iyer <[email protected]>
---
Changes in v3:
- Rebased on top of mainline

I still needed to rebase it to apply it to my tree, so please double check the result in the bleeding-edge branch.

Thanks!

Quoted 62 lines

---
 drivers/base/power/domain.c | 10 +++++++---
 include/linux/pm_domain.h   |  2 ++
 2 files changed, 9 insertions(+), 3 deletions(-)

drivers/base/power/domain.c +7 -3

index 2cb5e04cf86c..20ecc1a7113c 100644
@@ -559,12 +559,15 @@ static int genpd_power_off(struct generic_pm_domain *genpd, bool one_dev_on,
                 * happen very often).
                 */
                ret = _genpd_power_off(genpd, true);
-               if (ret)
+               if (ret) {
+                       genpd->states[genpd->state_idx].rejected++;
                        return ret;
+               }
        }
 
        genpd->status = GPD_STATE_POWER_OFF;
        genpd_update_accounting(genpd);
+       genpd->states[genpd->state_idx].usage++;
 
        list_for_each_entry(link, &genpd->child_links, child_node) {
                genpd_sd_counter_dec(link->parent);

@@ -2932,7 +2935,7 @@ static int idle_states_show(struct seq_file *s, void *data)
        if (ret)
                return -ERESTARTSYS;
 
-       seq_puts(s, "State          Time Spent(ms)\n");
+       seq_puts(s, "State          Time Spent(ms) Usage          Rejected\n");
 
        for (i = 0; i < genpd->state_count; i++) {
                ktime_t delta = 0;

@@ -2944,7 +2947,8 @@ static int idle_states_show(struct seq_file *s, void *data)
 
                msecs = ktime_to_ms(
                        ktime_add(genpd->states[i].idle_time, delta));
-               seq_printf(s, "S%-13i %lld\n", i, msecs);
+               seq_printf(s, "S%-13i %-14lld %-14llu %llu\n", i, msecs,
+                             genpd->states[i].usage, genpd->states[i].rejected);
        }
 
        genpd_unlock(genpd);

include/linux/pm_domain.h +2 -1

index ee11502a575b..1d9a8de6b7ac 100644
@@ -82,6 +82,8 @@ struct genpd_power_state {
        s64 power_off_latency_ns;
        s64 power_on_latency_ns;
        s64 residency_ns;
+       u64 usage;
+       u64 rejected;
        struct fwnode_handle *fwnode;
        ktime_t idle_time;
        void *data;
--

The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project

Lina Iyer

author

2020-10-16 16:07 UTC #2 lore raw

Quoted 8 lines · Rafael J. Wysocki

Quoted 7 lines · Lina Iyer
To enable better debug of PM domains, let's keep a track of the success and rejections in entering each domain idle state.
This statistics is exported in debugfs when reading the idle_states node, associated with each PM domain.

Quoted 8 lines

Quoted 4 lines
Signed-off-by Lina Iyer <[email protected]>
---
Changes in v3:
- Rebased on top of mainline
I still needed to rebase it to apply it to my tree, so please double check the result in the bleeding-edge branch.

Okay, will rebase on that branch and post shortly. Thanks Rafael.

Quoted 64 lines

Thanks!

Quoted 62 lines

---
 drivers/base/power/domain.c | 10 +++++++---
 include/linux/pm_domain.h   |  2 ++
 2 files changed, 9 insertions(+), 3 deletions(-)

drivers/base/power/domain.c +7 -3

index 2cb5e04cf86c..20ecc1a7113c 100644
@@ -559,12 +559,15 @@ static int genpd_power_off(struct generic_pm_domain *genpd, bool one_dev_on,
                 * happen very often).
                 */
                ret = _genpd_power_off(genpd, true);
-               if (ret)
+               if (ret) {
+                       genpd->states[genpd->state_idx].rejected++;
                        return ret;
+               }
        }
 
        genpd->status = GPD_STATE_POWER_OFF;
        genpd_update_accounting(genpd);
+       genpd->states[genpd->state_idx].usage++;
 
        list_for_each_entry(link, &genpd->child_links, child_node) {
                genpd_sd_counter_dec(link->parent);

@@ -2932,7 +2935,7 @@ static int idle_states_show(struct seq_file *s, void *data)
        if (ret)
                return -ERESTARTSYS;
 
-       seq_puts(s, "State          Time Spent(ms)\n");
+       seq_puts(s, "State          Time Spent(ms) Usage          Rejected\n");
 
        for (i = 0; i < genpd->state_count; i++) {
                ktime_t delta = 0;

@@ -2944,7 +2947,8 @@ static int idle_states_show(struct seq_file *s, void *data)
 
                msecs = ktime_to_ms(
                        ktime_add(genpd->states[i].idle_time, delta));
-               seq_printf(s, "S%-13i %lld\n", i, msecs);
+               seq_printf(s, "S%-13i %-14lld %-14llu %llu\n", i, msecs,
+                             genpd->states[i].usage, genpd->states[i].rejected);
        }
 
        genpd_unlock(genpd);

include/linux/pm_domain.h +2 -1

index ee11502a575b..1d9a8de6b7ac 100644
@@ -82,6 +82,8 @@ struct genpd_power_state {
        s64 power_off_latency_ns;
        s64 power_on_latency_ns;
        s64 residency_ns;
+       u64 usage;
+       u64 rejected;
        struct fwnode_handle *fwnode;
        ktime_t idle_time;
        void *data;
--

The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project

Rafael J. Wysocki

maintainer

2020-10-16 16:13 UTC #3 lore raw

Quoted 15 lines · Lina Iyer

Quoted 13 lines · Rafael J. Wysocki
Quoted 11 lines · Lina Iyer
To enable better debug of PM domains, let's keep a track of the success and rejections in entering each domain idle state.
This statistics is exported in debugfs when reading the idle_states node, associated with each PM domain.
Signed-off-by Lina Iyer <[email protected]>
---
Changes in v3:
- Rebased on top of mainline

Quoted 4 lines

Quoted 3 lines
I still needed to rebase it to apply it to my tree, so please double check the result in the bleeding-edge branch.
Okay, will rebase on that branch and post shortly.

No need, it's been applied already, just please check the result in bleeding-edge.

Thanks!

Lina Iyer

author

2020-10-16 16:21 UTC #4 lore raw

Quoted 16 lines · Rafael J. Wysocki

Quoted 15 lines · Lina Iyer
Quoted 13 lines · Rafael J. Wysocki
Quoted 11 lines · Lina Iyer
To enable better debug of PM domains, let's keep a track of the success and rejections in entering each domain idle state.
This statistics is exported in debugfs when reading the idle_states node, associated with each PM domain.
Signed-off-by Lina Iyer <[email protected]>
---
Changes in v3:
- Rebased on top of mainline

Quoted 8 lines

Quoted 4 lines
Quoted 3 lines
I still needed to rebase it to apply it to my tree, so please double check the result in the bleeding-edge branch.
Okay, will rebase on that branch and post shortly.
No need, it's been applied already, just please check the result in bleeding-edge.

I think I may have the wrong bleeding-edge branch. Is it not https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/log/?h=bleeding-edge

Apologize for the confusion.

--Lina

Rafael J. Wysocki

maintainer

2020-10-16 16:32 UTC #5 lore raw

Quoted 23 lines · Lina Iyer

Quoted 21 lines · Rafael J. Wysocki
Quoted 19 lines · Lina Iyer
Quoted 8 lines · Rafael J. Wysocki
Quoted 7 lines · Lina Iyer
To enable better debug of PM domains, let's keep a track of the success and rejections in entering each domain idle state.
This statistics is exported in debugfs when reading the idle_states node, associated with each PM domain.
Quoted 8 lines
Quoted 4 lines
Signed-off-by Lina Iyer <[email protected]>
---
Changes in v3:
- Rebased on top of mainline
I still needed to rebase it to apply it to my tree, so please double check the result in the bleeding-edge branch.
Okay, will rebase on that branch and post shortly.

Quoted 7 lines

Quoted 3 lines
No need, it's been applied already, just please check the result in bleeding-edge.
I think I may have the wrong bleeding-edge branch. Is it not
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/log/?h=bleeding-edge
Apologize for the confusion.

Yes, it is now. :-)

It was not pushed out before, sorry.

Lina Iyer

author

2020-10-16 16:44 UTC #6 lore raw

Quoted 27 lines · Rafael J. Wysocki

Quoted 26 lines · Lina Iyer
Quoted 24 lines · Rafael J. Wysocki
Quoted 15 lines · Lina Iyer
Quoted 13 lines · Rafael J. Wysocki
Quoted 11 lines · Lina Iyer
To enable better debug of PM domains, let's keep a track of the success and rejections in entering each domain idle state.
This statistics is exported in debugfs when reading the idle_states node, associated with each PM domain.
Signed-off-by Lina Iyer <[email protected]>
---
Changes in v3:
- Rebased on top of mainline
Quoted 4 lines
Quoted 3 lines
I still needed to rebase it to apply it to my tree, so please double check the result in the bleeding-edge branch.
Okay, will rebase on that branch and post shortly.
No need, it's been applied already, just please check the result in bleeding-edge.

Quoted 8 lines

Quoted 4 lines
I think I may have the wrong bleeding-edge branch. Is it not
https://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git/log/?h=bleeding-edge
Apologize for the confusion.
Yes, it is now. :-)
It was not pushed out before, sorry.

Thanks, now I see it and looks good to me.

--Lina