Skip to contentSkip to conversation

platform/chrome: Expose resume result via debugfs

PATCH In review Evan Green 4 participants 6 messages Jun 27, 2019 – Jul 12, 2019

Subsystems (3)
lore
Share & tools
lore

Evan Green

author
#0 lore raw

PATCH v3 1/1 platform/chrome: Expose resume result via debugfs

For ECs that support it, the EC returns the number of slp_s0 transitions and whether or not there was a timeout in the resume response. Expose the last resume result to usermode via debugfs so that usermode can detect and report S0ix timeouts.

---

Changes in v3:
- Expose the debugfs file on all EC types (Enric)

Changes in v2:
- Moved from sysfs to debugfs (Enric)
- Added documentation (Enric)

---
 Documentation/ABI/testing/debugfs-cros-ec | 22 ++++++++++++++++++++++
 drivers/mfd/cros_ec.c                     |  6 +++++-
 drivers/platform/chrome/cros_ec_debugfs.c |  5 +++++
 include/linux/mfd/cros_ec.h               |  1 +
 4 files changed, 33 insertions(+), 1 deletion(-)
Documentation/ABI/testing/debugfs-cros-ec +22 -0
index 573a82d23c89..1fe0add99a2a 100644
@@ -32,3 +32,25 @@ Description:
 		is used for synchronizing the AP host time with the EC
 		log. An error is returned if the command is not supported
 		by the EC or there is a communication problem.
+
+What:		/sys/kernel/debug/<cros-ec-device>/last_resume_result
+Date:		June 2019
+KernelVersion:	5.3
+Description:
+		Some ECs have a feature where they will track transitions to
+		the (Intel) processor's SLP_S0 line, in order to detect cases
+		where a system failed to go into S0ix. When the system resumes,
+		an EC with this feature will return a summary of SLP_S0
+		transitions that occurred. The last_resume_result file returns
+		the most recent response from the AP's resume message to the EC.
+
+		The bottom 31 bits contain a count of the number of SLP_S0
+		transitions that occurred since the suspend message was
+		received. Bit 31 is set if the EC attempted to wake the
+		system due to a timeout when watching for SLP_S0 transitions.
+		Callers can use this to detect a wake from the EC due to
+		S0ix timeouts. The result will be zero if no suspend
+		transitions have been attempted, or the EC does not support
+		this feature.
+
+		Output will be in the format: "0x%08x\n".
drivers/mfd/cros_ec.c +5 -1
index 5d5c41ac3845..2a9ac5213893 100644
@@ -102,12 +102,16 @@ static int cros_ec_sleep_event(struct cros_ec_device *ec_dev, u8 sleep_event)
 
 	/* For now, report failure to transition to S0ix with a warning. */
 	if (ret >= 0 && ec_dev->host_sleep_v1 &&
-	    (sleep_event == HOST_SLEEP_EVENT_S0IX_RESUME))
+	    (sleep_event == HOST_SLEEP_EVENT_S0IX_RESUME)) {
+		ec_dev->last_resume_result =
+			buf.u.resp1.resume_response.sleep_transitions;
+
 		WARN_ONCE(buf.u.resp1.resume_response.sleep_transitions &
 			  EC_HOST_RESUME_SLEEP_TIMEOUT,
 			  "EC detected sleep transition timeout. Total slp_s0 transitions: %d",
 			  buf.u.resp1.resume_response.sleep_transitions &
 			  EC_HOST_RESUME_SLEEP_TRANSITIONS_MASK);
+	}
 
 	return ret;
 }
drivers/platform/chrome/cros_ec_debugfs.c +5 -0
index 7ee060743844..967c78abfdd3 100644
@@ -447,6 +447,11 @@ static int cros_ec_debugfs_probe(struct platform_device *pd)
 	debugfs_create_file("uptime", 0444, debug_info->dir, debug_info,
 			    &cros_ec_uptime_fops);
 
+	debugfs_create_x32("last_resume_result",
+			   0444,
+			   debug_info->dir,
+			   &ec->ec_dev->last_resume_result);
+
 	ec->debug_info = debug_info;
 
 	dev_set_drvdata(&pd->dev, ec);
include/linux/mfd/cros_ec.h +1 -0
index 5ddca44be06d..45aba26db964 100644
@@ -155,6 +155,7 @@ struct cros_ec_device {
 	struct ec_response_get_next_event_v1 event_data;
 	int event_size;
 	u32 host_event_wake_mask;
+	u32 last_resume_result;
 };
 
 /**
Signature · 1 line
2.20.1
Acked-by Enric Balletbo i Serra on v3 1/1
#1 lore raw

Hi Evan, Lee,

Missatge de Evan Green <[email protected]> del dia dj., 27 de juny 2019 a les 22:46:

Quoted 7 lines

For ECs that support it, the EC returns the number of slp_s0 transitions and whether or not there was a timeout in the resume response. Expose the last resume result to usermode via debugfs so that usermode can detect and report S0ix timeouts.

Lee, actually this patch depends on some patches from chrome-platform to apply cleanly. Once is fine with you and if you're happy to have this merged for 5.3, I can just carry the patch with me, shouldn't be any conflicts with your current changes or if you prefer I can create an immutable branch for you.

Thanks,
~ Enric

Quoted 100 lines

---

Changes in v3:
- Expose the debugfs file on all EC types (Enric)

Changes in v2:
- Moved from sysfs to debugfs (Enric)
- Added documentation (Enric)

---
 Documentation/ABI/testing/debugfs-cros-ec | 22 ++++++++++++++++++++++
 drivers/mfd/cros_ec.c                     |  6 +++++-
 drivers/platform/chrome/cros_ec_debugfs.c |  5 +++++
 include/linux/mfd/cros_ec.h               |  1 +
 4 files changed, 33 insertions(+), 1 deletion(-)
Documentation/ABI/testing/debugfs-cros-ec +22 -0
index 573a82d23c89..1fe0add99a2a 100644
@@ -32,3 +32,25 @@ Description:
                is used for synchronizing the AP host time with the EC
                log. An error is returned if the command is not supported
                by the EC or there is a communication problem.
+
+What:          /sys/kernel/debug/<cros-ec-device>/last_resume_result
+Date:          June 2019
+KernelVersion: 5.3
+Description:
+               Some ECs have a feature where they will track transitions to
+               the (Intel) processor's SLP_S0 line, in order to detect cases
+               where a system failed to go into S0ix. When the system resumes,
+               an EC with this feature will return a summary of SLP_S0
+               transitions that occurred. The last_resume_result file returns
+               the most recent response from the AP's resume message to the EC.
+
+               The bottom 31 bits contain a count of the number of SLP_S0
+               transitions that occurred since the suspend message was
+               received. Bit 31 is set if the EC attempted to wake the
+               system due to a timeout when watching for SLP_S0 transitions.
+               Callers can use this to detect a wake from the EC due to
+               S0ix timeouts. The result will be zero if no suspend
+               transitions have been attempted, or the EC does not support
+               this feature.
+
+               Output will be in the format: "0x%08x\n".
drivers/mfd/cros_ec.c +5 -1
index 5d5c41ac3845..2a9ac5213893 100644
@@ -102,12 +102,16 @@ static int cros_ec_sleep_event(struct cros_ec_device *ec_dev, u8 sleep_event)
 
        /* For now, report failure to transition to S0ix with a warning. */
        if (ret >= 0 && ec_dev->host_sleep_v1 &&
-           (sleep_event == HOST_SLEEP_EVENT_S0IX_RESUME))
+           (sleep_event == HOST_SLEEP_EVENT_S0IX_RESUME)) {
+               ec_dev->last_resume_result =
+                       buf.u.resp1.resume_response.sleep_transitions;
+
                WARN_ONCE(buf.u.resp1.resume_response.sleep_transitions &
                          EC_HOST_RESUME_SLEEP_TIMEOUT,
                          "EC detected sleep transition timeout. Total slp_s0 transitions: %d",
                          buf.u.resp1.resume_response.sleep_transitions &
                          EC_HOST_RESUME_SLEEP_TRANSITIONS_MASK);
+       }
 
        return ret;
 }
drivers/platform/chrome/cros_ec_debugfs.c +5 -0
index 7ee060743844..967c78abfdd3 100644
@@ -447,6 +447,11 @@ static int cros_ec_debugfs_probe(struct platform_device *pd)
        debugfs_create_file("uptime", 0444, debug_info->dir, debug_info,
                            &cros_ec_uptime_fops);
 
+       debugfs_create_x32("last_resume_result",
+                          0444,
+                          debug_info->dir,
+                          &ec->ec_dev->last_resume_result);
+
        ec->debug_info = debug_info;
 
        dev_set_drvdata(&pd->dev, ec);
include/linux/mfd/cros_ec.h +1 -1
index 5ddca44be06d..45aba26db964 100644
@@ -155,6 +155,7 @@ struct cros_ec_device {
        struct ec_response_get_next_event_v1 event_data;
        int event_size;
        u32 host_event_wake_mask;
+       u32 last_resume_result;
 };
 
 /**
--

2.20.1

Acked-by
#2 lore raw
Quoted 12 lines · Enric Balletbo Serra

Hi Evan, Lee,

Missatge de Evan Green <[email protected]> del dia dj., 27 de juny 2019 a les 22:46:

Quoted 7 lines

For ECs that support it, the EC returns the number of slp_s0 transitions and whether or not there was a timeout in the resume response. Expose the last resume result to usermode via debugfs so that usermode can detect and report S0ix timeouts.

Quoted 7 lines

Lee, actually this patch depends on some patches from chrome-platform to apply cleanly. Once is fine with you and if you're happy to have this merged for 5.3, I can just carry the patch with me, shouldn't be any conflicts with your current changes or if you prefer I can create an immutable branch for you.

I won't be taking any more patches this cycle, so if you're sure that it does not conflict, there is no need for an immutable branch.

Signature · 4 lines
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
#3 lore raw
Quoted 22 lines · Lee Jones
Quoted 12 lines · Enric Balletbo Serra

Hi Evan, Lee,

Missatge de Evan Green <[email protected]> del dia dj., 27 de juny 2019 a les 22:46:

Quoted 7 lines

For ECs that support it, the EC returns the number of slp_s0 transitions and whether or not there was a timeout in the resume response. Expose the last resume result to usermode via debugfs so that usermode can detect and report S0ix timeouts.

Quoted 7 lines

Lee, actually this patch depends on some patches from chrome-platform to apply cleanly. Once is fine with you and if you're happy to have this merged for 5.3, I can just carry the patch with me, shouldn't be any conflicts with your current changes or if you prefer I can create an immutable branch for you.

Quoted 5 lines

I won't be taking any more patches this cycle, so if you're sure that it does not conflict, there is no need for an immutable branch.

Thanks Lee.

I think the patch is simple enough, I queued for the autobuilders to play with, and if all goes well I'll add for 5.3 via chrome-platform.

Thanks,
~ Enric

author
#4 lore raw
Quoted 29 lines · Enric Balletbo i Serra
Quoted 25 lines · Lee Jones
Quoted 12 lines · Enric Balletbo Serra

Hi Evan, Lee,

Missatge de Evan Green <[email protected]> del dia dj., 27 de juny 2019 a les 22:46:

Quoted 7 lines

For ECs that support it, the EC returns the number of slp_s0 transitions and whether or not there was a timeout in the resume response. Expose the last resume result to usermode via debugfs so that usermode can detect and report S0ix timeouts.

Quoted 7 lines

Lee, actually this patch depends on some patches from chrome-platform to apply cleanly. Once is fine with you and if you're happy to have this merged for 5.3, I can just carry the patch with me, shouldn't be any conflicts with your current changes or if you prefer I can create an immutable branch for you.

I won't be taking any more patches this cycle, so if you're sure that it does not conflict, there is no need for an immutable branch.

Quoted 8 lines
Quoted 2 lines

Thanks Lee.

I think the patch is simple enough, I queued for the autobuilders to play with, and if all goes well I'll add for 5.3 via chrome-platform.

I was hoping to pick this patch from a maintainer's tree, has it landed anywhere?
-Evan

author
#5 lore raw
Quoted 37 lines · Evan Green
Quoted 34 lines · Enric Balletbo i Serra
Quoted 22 lines · Lee Jones
Quoted 12 lines · Enric Balletbo Serra

Hi Evan, Lee,

Missatge de Evan Green <[email protected]> del dia dj., 27 de juny 2019 a les 22:46:

Quoted 7 lines

For ECs that support it, the EC returns the number of slp_s0 transitions and whether or not there was a timeout in the resume response. Expose the last resume result to usermode via debugfs so that usermode can detect and report S0ix timeouts.

Quoted 7 lines

Lee, actually this patch depends on some patches from chrome-platform to apply cleanly. Once is fine with you and if you're happy to have this merged for 5.3, I can just carry the patch with me, shouldn't be any conflicts with your current changes or if you prefer I can create an immutable branch for you.

Quoted 5 lines

I won't be taking any more patches this cycle, so if you're sure that it does not conflict, there is no need for an immutable branch.

Thanks Lee.

Quoted 7 lines
Quoted 3 lines

I think the patch is simple enough, I queued for the autobuilders to play with, and if all goes well I'll add for 5.3 via chrome-platform.

I was hoping to pick this patch from a maintainer's tree, has it landed anywhere?
-Evan

Nevermind, it's all the way in Linus' tree. Sorry for the spam, and thanks Benson for the heads up!
-Evan