POC for a new harness designed for power measurement purposes. #85130

gbarkadiusz · 2025-02-04T10:25:54Z

POC of Power twister harness.

This pr adds support for power measurements for twister and three example tests.
pm.states: Verifies the transition to all available power states on stm32l562e_dk.

During the test, pytest detects each step and calculates the RMS (Root Mean Square) current for every detected state. The test then compares the calculated RMS values against the expected values, which have been determined manually through experimental observation

pm.residency_time: Tests the residency time of each power state.
This test evaluates the residency time of different power states on the stm32l562e_dk board.

pm.wakeup_timer: Tests the RTC alarm’s ability to wake the CPU from the lowest power state.

Each test calculates RMS current values for each detected state and compares them to expected values, which were manually set based on experimental data.

Setup

To reproduce.:

Configure the stm32l562e_dk board according to the instructions provided above.
Connect both USB ports of the board to your host system. The board will appear as two separate targets:
Power Monitor: e.g., /dev/ttyACM0 or /dev/serial/by-id/usb-STMicroelectronics_PowerShield__Virtual_ComPort_in_FS_Mode__FFFFFFFEFFFF-if00
Target: e.g., /dev/ttyACM1 or /dev/serial/by-id/usb-STMicroelectronics_STLINK-V3_004##############39-if02
Run twister commend:

twister --device-testing --device-serial /dev/ttyACM1 -p stm32l562e_dk -T tests/subsys/pm/power_residency_time/ --west-flash="--dev-id=004800483137510D33333639" --west-runner=pyocd -vv  -ll DEBUG -xCONFIG_BOOT_DELAY=500 --pm-probe=stm_powershield --pm-probe-port=/dev/serial/by-id/usb-S
TMicroelectronics_PowerShield__Virtual_ComPort_in_FS_Mode__FFFFFFFEFFFF-if00

bjarki-andreasen · 2025-02-04T15:01:22Z

tests/subsys/pm/power_wakeup_timer/testcase.yaml

+      pytest_args: [--probe=stm_powershield,
+                    --probe-port=<path_to>/STMicroelectronics_PowerShield__Virtual_ComPort]


These two args (--probe and --probe-port ) should be provided via the command line, as they vary, as opposed to the pytest_dut_scope: session and the properties in test_expected_values.yaml file which are "portable"/"fixed" regardless of the probe and port used :)

Would it be possible to move

- name: "wakeup_timer" measurement_time: 2 num_of_transitions: 1 rms: [0.26, 140]

to

harness_config: measurements: - name: "wakeup_timer" measurement_time: 2 num_of_transitions: 1 rms: [0.26, 140]

instead of the separate file? not sure what having the commands in a separate file acheives :)

gbarkadiusz · 2025-02-17T15:26:42Z

@bjarki-andreasen Please take another look. I'm not sure that moving --probe and --probe-port to the Twister parameter is a good approach. They're more closely related to the harness's parameters rather than Twister's, so I suggest leaving them in the harness_config section.

bjarki-andreasen · 2025-02-17T15:46:10Z

@bjarki-andreasen Please take another look. I'm not sure that moving --probe and --probe-port to the Twister parameter is a good approach. They're more closely related to the harness's parameters rather than Twister's, so I suggest leaving them in the harness_config section.

Would you argue the same for the --serial option and its parameters? The probe and probe-port are instances of actual hardware right? If i ran the same test locally, would I not need to change the port to match my setup?

gbarkadiusz · 2025-02-18T12:27:48Z

@bjarki-andreasen Please take another look. I'm not sure that moving --probe and --probe-port to the Twister parameter is a good approach. They're more closely related to the harness's parameters rather than Twister's, so I suggest leaving them in the harness_config section.

Would you argue the same for the --serial option and its parameters? The probe and probe-port are instances of actual hardware right? If i ran the same test locally, would I not need to change the port to match my setup?

Yes, you need to change the port to match your setup.
Regarding the twister parameters. You right, so I've applied your suggestions, There's the example twister's command with these two parameters --pm-probe and --pm-probe-port.

twister --device-testing --device-serial /dev/ttyACM1 -p stm32l562e_dk -T tests/subsys/pm/power_residency_time/ --west-flash="--dev-id=004800483137510D33333639" --west-runner=pyocd -vv  -ll DEBUG -xCONFIG_BOOT_DELAY=500 --pm-probe=stm_powershield --pm-probe-port=/dev/serial/by-id/usb-S
TMicroelectronics_PowerShield__Virtual_ComPort_in_FS_Mode__FFFFFFFEFFFF-if00

@bjarki-andreasen , @nashif Please take a look.

PerMac · 2025-02-18T14:09:38Z

cc @nordic-piks

nordic-piks · 2025-02-19T06:50:38Z

scripts/pm/power_monitor_stm32l562e_dk/UnityFunctions.py

+            raise ValueError("Invalid unit. Use 'us', 'ms', or 's'.")
+
+    @staticmethod
+    def current_RMS(data, exclude_first_600=True, num_peaks=1):


IMHO, number of samples which needs to be excluded from analysis should be taken also from sample.yml,
so that this can be easily configures for given test.
Of course this is just a hint if you want to make this approach more general.

nordic-piks · 2025-02-19T06:52:05Z

scripts/pm/power_monitor_stm32l562e_dk/UnityFunctions.py

+        data = [float(x) for x in data]
+
+        # Find the peaks in the data using the `find_peaks` function
+        peaks = signal.find_peaks(data, distance=40, height=0.008)[0]


IMHO, minimal number of samples between peaks and its "hight" should be also taken from sample.yaml
Of course this is just a hint if you want to make this approach more general.

nordic-piks · 2025-02-19T06:56:43Z

tests/subsys/pm/power_states/testcase.yaml

+        - name: "power states"
+          measurement_time: 6
+          num_of_transitions: 4
+          rms: [56.0, 4.0, 1.3, 0.26, 140]


IMHO, it would be good to test/verify also the length of each period between peaks.
This will allow to verify if actual sleep time meets expectation (looking from current consumption measurements).

Also, you assume peaks of high current and measures current in between.
There is also beneficial to measure current during active time i.e. being able to specify requirements for that in test.

Of course this is just a hint if you want to make this approach more general :)

nordic-piks · 2025-02-19T07:01:35Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+        self.target_voltage = None
+        self.target_temperature = None
+        self.acqTimeoutThread = None
+        self.power_shield_conf = PowerShieldConf()


IMHO, parameters for PowerShieldConf e.g. voltage, sampling frequency should be taken from some board configuration file (here they just base on defaults). I do not see mechanisms to provide it externally.

Possibly, board parameters should be overwritable by some specific configuration in the test (sample.yaml)
Of course this is just a hint if you want to make this approach more general :)

nordic-piks · 2025-02-19T07:06:20Z

tests/subsys/pm/power_states/testcase.yaml

+        - name: "power states"
+          measurement_time: 6
+          num_of_transitions: 4
+          rms: [56.0, 4.0, 1.3, 0.26, 140]


Beside RMS, it would be also interesting to verify min and/or max value of current during some period.
This will allow to detect unexpected events (peaks) when you expect to have stable sleep current.

Of course this is just a hint if you want to make this approach more general.

gbarkadiusz · 2025-02-19T07:20:36Z

@nordic-piks Thank you for your feedback.
This can be improved in the next iteration.
This PR addresses the requirements outlined in PR #80692.
I would like to merge this initial approach now and improve it in subsequent iterations.
@bjarki-andreasen FYI.

nordic-bami · 2025-02-19T07:20:53Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+        self.target_temperature = self.get_temperature(self.power_shield_conf.temperature_unit)
+        self.target_voltage = self.get_voltage_level()
+
+    def connect(self):


Either declare all public methods in ABC base or make methods not referenced in ABC api private.

I wanted to limit the function in ABC to init measure and get_data.

In such case methods that do not override abstract methods should be private.

nordic-bami · 2025-02-19T07:22:32Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+        except Exception as e:
+            logging.error(f"Failed to reopen connection after reset: {e}")
+
+    def get_voltage_level(self) -> float:


Specify the unit of the return value (V, mV etc.)

nordic-bami · 2025-02-19T07:23:40Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+                return None
+        else:
+            logging.warning("No response for voltage command.")
+        return None


Please return float('nan') in case of 'no response', None will crash any further calculations.

nordic-bami · 2025-02-19T07:24:15Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+        Sends the temperature command and returns the temperature as a float.
+
+        :param unit: The unit to request the temperature in, either 'degc' or 'degf'.
+        :return: The temperature value as a float.


Same as above, please specify unit (C, K, F)

nordic-bami · 2025-02-19T07:25:57Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+        else:
+            logging.warning(f"Failed to set data format to {data_format}.")
+
+    def set_frequency(self, frequency: str):


Change type of this 'frequency' argument to either float or 'enum' type.
If this is not direct frequency in Hz then change the argument name to something like 'frequency_option'

nordic-bami · 2025-02-19T07:27:22Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+        else:
+            logging.warning(f"Failed to set sampling frequency to {frequency}.")
+
+    def set_acquire_time(self, acquire_time: str = '0'):


Nitpick: it should be: set_acquisition_time, acquire is a verb

Thank you, fixed.

nordic-bami · 2025-02-19T07:27:52Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShield.py

+        else:
+            logging.warning(f"Failed to set acquisition time to {acquire_time}.")
+
+    def set_voltage(self, voltage: str):


Same nitpick as for 'set_frequency'

nordic-bami · 2025-02-19T07:33:46Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShieldConfig.py

+        self.function_mode = function_mode
+        self.target_voltage = target_voltage
+        self.temperature_unit = temperature_unit
+        self.acquisition_time = None


Please annotate as 'Optional[float]'

nordic-bami · 2025-02-19T07:34:02Z

scripts/pm/power_monitor_stm32l562e_dk/PowerShieldData.py

+
+class PowerShieldData:
+    def __init__(self):
+        self.data = []


Please annotate 'data'

nordic-bami · 2025-02-19T07:34:27Z

scripts/pm/power_monitor_stm32l562e_dk/SerialHandler.py

+        self.port = port
+        self.baudrate = baudrate
+        self.timeout = timeout
+        self.serial_connection = None


Please annotate 'serial_connection'

What you mean?

nordic-bami · 2025-02-19T07:35:36Z

scripts/pm/power_monitor_stm32l562e_dk/SerialHandler.py

+                    "Connection to %s at %d baud opened successfully.", self.port, self.baudrate
+                )
+            except serial.SerialException as e:
+                logging.error("Error opening serial port %s: %s", self.port, str(e))


Why catch this error and not raise anything else. Test will continue with failed serial connection and will fail later.

Add simply raise should be enough?
Fixed.

nordic-bami · 2025-02-19T07:36:31Z

scripts/pm/power_monitor_stm32l562e_dk/SerialHandler.py

+
+        :return: The processed received data as a string.
+        """
+        s = ""


Please make meaningful variable name instead of 's'

nordic-bami · 2025-02-19T07:37:36Z

scripts/pm/power_monitor_stm32l562e_dk/SerialHandler.py

+        """Closes the connection if the object is deleted."""
+        self.close()
+        # Shut down logging
+        logging.shutdown()


Please don't call shutdown it will bring down entire logging system

nordic-bami · 2025-02-19T07:38:37Z

scripts/pm/power_monitor_stm32l562e_dk/UnityFunctions.py

+
+
+class UnityFunctions:
+    @staticmethod


If all methods are static class in not needed, they could be declared as functions in UnityFunctions.py module.

Add support for powerShield of stm32l562e_dk board. Signed-off-by: Arkadiusz Cholewinski <arkadiuszx.cholewinski@intel.com>

Add three tests. - power_states - power_wakeup_timer - power_residency_time Signed-off-by: Arkadiusz Cholewinski <arkadiuszx.cholewinski@intel.com>

gbarkadiusz added the DNM This PR should not be merged (Do Not Merge) label Feb 4, 2025

gbarkadiusz mentioned this pull request Feb 4, 2025

tests: pm: Adding tests based on new approach to PM testing. #80692

Closed

gbarkadiusz force-pushed the topic/power_management_harness branch 2 times, most recently from 679de9e to c474ca5 Compare February 4, 2025 14:12

bjarki-andreasen reviewed Feb 4, 2025

View reviewed changes

gbarkadiusz force-pushed the topic/power_management_harness branch 4 times, most recently from 4109da0 to 1424b32 Compare February 17, 2025 15:21

gbarkadiusz force-pushed the topic/power_management_harness branch 2 times, most recently from 380d8c0 to 91fa4eb Compare February 18, 2025 12:24

gbarkadiusz force-pushed the topic/power_management_harness branch 7 times, most recently from f202877 to c0e83bd Compare February 18, 2025 13:47

gbarkadiusz marked this pull request as ready for review February 18, 2025 13:51

gbarkadiusz removed the DNM This PR should not be merged (Do Not Merge) label Feb 18, 2025

zephyrbot added area: Power Management area: Twister Twister labels Feb 18, 2025

zephyrbot requested review from ceolin, gchwier, golowanow, hakehuang and JordanYates February 18, 2025 13:52

zephyrbot requested review from KamilxPaszkiet, LukaszMrugala, nashif, PerMac, teburd and tmleman February 18, 2025 13:52

zephyrbot assigned ceolin and bjarki-andreasen Feb 18, 2025

gbarkadiusz force-pushed the topic/power_management_harness branch from c0e83bd to 8bab44b Compare February 18, 2025 13:56

gbarkadiusz requested a review from bjarki-andreasen February 18, 2025 14:29

nordic-piks reviewed Feb 19, 2025

View reviewed changes

nordic-bami reviewed Feb 19, 2025

View reviewed changes

gbarkadiusz force-pushed the topic/power_management_harness branch 4 times, most recently from 50c4d38 to a46de65 Compare February 19, 2025 11:54

gbarkadiusz added 2 commits February 19, 2025 13:12

Twister: Add power measurements support for Twister.

27d1df3

Add support for powerShield of stm32l562e_dk board. Signed-off-by: Arkadiusz Cholewinski <arkadiuszx.cholewinski@intel.com>

CI: Add power_states, power_wakeup_timer and power_residency_time tests

3ecc5a6

Add three tests. - power_states - power_wakeup_timer - power_residency_time Signed-off-by: Arkadiusz Cholewinski <arkadiuszx.cholewinski@intel.com>

gbarkadiusz force-pushed the topic/power_management_harness branch from a46de65 to 3ecc5a6 Compare February 19, 2025 12:12

		pytest_args: [--probe=stm_powershield,
		--probe-port=<path_to>/STMicroelectronics_PowerShield__Virtual_ComPort]

POC for a new harness designed for power measurement purposes. #85130

Are you sure you want to change the base?

POC for a new harness designed for power measurement purposes. #85130

Conversation

gbarkadiusz commented Feb 4, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gbarkadiusz commented Feb 17, 2025

bjarki-andreasen commented Feb 17, 2025 • edited Loading

gbarkadiusz commented Feb 18, 2025

PerMac commented Feb 18, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

nordic-piks Feb 19, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gbarkadiusz commented Feb 19, 2025

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gbarkadiusz Feb 19, 2025 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gbarkadiusz Feb 19, 2025 • edited Loading

Choose a reason for hiding this comment

gbarkadiusz commented Feb 4, 2025 •

edited

Loading

bjarki-andreasen commented Feb 17, 2025 •

edited

Loading

nordic-piks Feb 19, 2025 •

edited

Loading

gbarkadiusz Feb 19, 2025 •

edited

Loading

gbarkadiusz Feb 19, 2025 •

edited

Loading