AM625: ECAP

Hugh Yang

Part Number: AM625

Hi Dears,

We have customers who need to use the capture function of ecap based on the 08.03.00.07 SDK version, but this version only supports the pwm function of ecap, which cannot meet customer needs. Is there a way to adapt the ti-ecap-capture.c driver to this version of the SDK?

Br,

Hugh

over 2 years ago

0 Bin Liu over 2 years ago

TI__Guru**** 175131 points

Hi Hugh,

Please move to the latest SDK (v8.6) which has the ECAP capture function support.

0 Tony Tang over 2 years ago in reply to Bin Liu

TI__Mastermind 29852 points

Bin,

the user guide is not updated to tell how to use the ecap capture mode, would like to have it.

0 Mukul Bhatnagar over 2 years ago in reply to Tony Tang

TI__Guru* 85035 points

Thanks Tony. Bin is out of office, I have filed a bug against documentation to capture this.

Regards

Mukul

0 Bin Liu over 2 years ago in reply to Tony Tang

TI__Guru**** 175131 points

Hi Tony,

I am routing your query to our ECAP module expert for comments.

0 Nick Saulnier over 2 years ago in reply to Bin Liu

TI__Guru** 107090 points

Hello Tony & Hugh,

For reference of how to set up the ECAP capture in the Linux devicetree, please take a look at the sample devicetree file added in Linux SDK 9.0:
k3-am625-sk-ecap-capture.dtso

In terms of how to interact with the ECAP capture functionality from Linux userspace, I am asking the developers to see if they have any examples we can reference. It will probably take a couple of days before we hear back. Please feel free to ping the thread if I do not have another response by the middle of next week.

Regards,

Nick

0 Nick Saulnier over 2 years ago in reply to Nick Saulnier

TI__Guru** 107090 points

Hello Tony & Hugh,

Here is documentation around how to use the ECAP capture driver:

The Enhanced Capture (ECAP) module can be used to timestamp events
detected on signal input pin. It can be used for time measurements
of pulse train signals.

ECAP module includes 4 timestamp capture registers. For all 4 sequenced
timestamp capture events (0->1->2->3->0->...), edge polarity (falling/rising
edge) can be selected.

This driver leverages counter subsystem to :
- select edge polarity for all 4 capture events (event mode)
- log timestamps for each capture event
Event polarity, and CAP0/1/2/3 timestamps give all the information
about the input pulse train. Further information can easily be computed :
period and/or duty cycle if frequency is constant, elapsed time between
pulses, etc...

...

This series has already been posted (but not merged) to mainline linux :
https://lore.kernel.org/all/20220923142437.271328-1-jpanis@baylibre.com/
Some modifications were made to match TI v5.10 kernel.

This series must be applied on top of another series. For more info, reference
https://lore.kernel.org/all/20220923142437.271328-1-jpanis@baylibre.com

Userspace commands :
	### CLOCK SIGNAL ###
	cd /sys/bus/counter/devices/counter0/signal0

	# Get frequency
	cat frequency

	### INPUT SIGNAL ###
	cd /sys/bus/counter/devices/counter0/signal1

	# Get available polarities for each capture event
	cat polarity0_available
	cat polarity1_available
	cat polarity2_available
	cat polarity3_available

	# Get polarity for each capture event
	cat polarity0
	cat polarity1
	cat polarity2
	cat polarity3

	# Set polarity for each capture event
	echo positive > polarity0
	echo negative > polarity1
	echo positive > polarity2
	echo negative > polarity3

	### COUNT ###
	cd /sys/bus/counter/devices/counter0/count0

	# Get ceiling (counter max value)
	cat ceiling

	# Reset number of overflows & current timebase counter value
	echo 0 > num_overflows
	echo 0 > count

	# Run ECAP
	echo 1 > enable

	# Get number of overflows & current timebase counter value
	cat num_overflows
	cat count

	# Get captured timestamps
	cat capture0
	cat capture1
	cat capture2
	cat capture3

	# Note that counter watches can also be used to get
	# data from userspace application
	# -> see tools/counter/counter_example.c

	# Pause ECAP
	echo 0 > enable

How to test this driver ?
	1) Device tree modifications

	1.a) k3-am62-main.dtsi

	ecap2: capture@23120000 {
		compatible = "ti,am62-ecap-capture";
		reg = <0x00 0x23120000 0x00 0x100>;
		interrupts = <GIC_SPI 115 IRQ_TYPE_EDGE_RISING>;
		power-domains = <&k3_pds 53 TI_SCI_PD_EXCLUSIVE>;
		clocks = <&k3_clks 53 0>;
		clock-names = "fck";
	};

	Warning : ecap2 node already used as pwm by default
	-> must be replaced (as above) or overlayed for capture testing.

	1.b) k3-am625-sk.dts

	&main_pmx0 {
		main_ecap2_pins_default: main-ecap2-pins-default {
			pinctrl-single,pins = <
				AM62X_IOPAD(0x01a4, PIN_INPUT, 2) /* (B20) MCASP0_ACLKX.ECAP2_IN_APWM_OUT */
			>;
		};
	}

	&ecap2 {
		pinctrl-names = "default";
		pinctrl-0 = <&main_ecap2_pins_default>;
		status = "okay";
	};

	2) Kernel configuration

	Device Drivers  --->
		<*> Counter support  --->
			<*>   TI eCAP capture driver

	3) Once ECAP2 is probed, userspace commands can be used as explained above.

	ECAP2 input pin is available on connector J3 pin 11 for square waveform signal.

	For finest testing, this following userspace application code can be run :
	(stdout MUST be redirected to log file)

	// SPDX-License-Identifier: GPL-2.0-only
	/* Counter - example userspace application
	 *
	 * The userspace application opens /dev/counter0, configures the
	 * COUNTER_EVENT_CAPTURE event channels to gather count 0 capture
	 * events, and prints out the data as it becomes available on the
	 * character device node.
	 *
	 * Copyright (C) 2022 Julien Panis
	 */
	#include <errno.h>
	#include <fcntl.h>
	#include <linux/counter.h>
	#include <stdio.h>
	#include <string.h>
	#include <sys/ioctl.h>
	#include <unistd.h>

	/* assuming capture attributes are under the count0 directory */
	#define CAPTURE_WATCH(_id) \
	{ \
		.component.type = COUNTER_COMPONENT_EXTENSION, \
		.component.scope = COUNTER_SCOPE_COUNT, \
		.component.parent = 0, \
		.component.id = _id, \
		.event = COUNTER_EVENT_CAPTURE, \
		.channel = _id, \
	}

	/* get id from respective "captureX_component_id" attributes */
	#define NB_CAP 4
	static struct counter_watch watches[NB_CAP] = {
		CAPTURE_WATCH(0),
		CAPTURE_WATCH(1),
		CAPTURE_WATCH(2),
		CAPTURE_WATCH(3),
	};

	int main(void)
	{
		int fd;
		int ret;
		int i;
		struct counter_event event_data[NB_CAP];

		fd = open("/dev/counter0", O_RDWR);
		if (fd == -1) {
			perror("Unable to open /dev/counter0");
			return 1;
		}

		for (i = 0; i < NB_CAP; i++) {
			ret = ioctl(fd, COUNTER_ADD_WATCH_IOCTL, watches + i);
			if (ret == -1) {
				fprintf(stderr, "Error adding watches[%d]: %s\n", i,
					strerror(errno));
				return 1;
			}
		}
		ret = ioctl(fd, COUNTER_ENABLE_EVENTS_IOCTL);
		if (ret == -1) {
			perror("Error enabling events");
			return 1;
		}

		for (;;) {
			ret = read(fd, event_data, sizeof(event_data));
			if (ret == -1) {
				perror("Failed to read event data");
				return 1;
			}

			if (ret != sizeof(event_data)) {
				fprintf(stderr, "Failed to read event data\n");
				return -EIO;
			}

			/* [WARNING] For data consistency analysis, stdout must be redirected to log file !!! */
			puts("----------");
			printf("cap0: %llu\n", event_data[0].value);
			printf("cap1: %llu\n", event_data[1].value);
			printf("cap2: %llu\n", event_data[2].value);
			printf("cap3: %llu\n", event_data[3].value);
		}

		return 0;
	}

Julien Panis (3):
  dt-bindings: counter: add ti,am62-ecap-capture.yaml
  Documentation: ABI: sysfs-bus-counter: add capture items
  counter: ti-ecap-capture: capture driver support for ECAP

 Documentation/ABI/testing/sysfs-bus-counter   |  45 ++
 .../counter/ti,am62-ecap-capture.yaml         |  61 ++
 drivers/counter/Kconfig                       |  15 +
 drivers/counter/Makefile                      |   1 +
 drivers/counter/ti-ecap-capture.c             | 677 ++++++++++++++++++
 include/uapi/linux/counter.h                  |   2 +
 6 files changed, 801 insertions(+)
 create mode 100644 Documentation/devicetree/bindings/counter/ti,am62-ecap-capture.yaml
 create mode 100644 drivers/counter/ti-ecap-capture.c

From the developer: "For testing, I used the shell commands (also given in the cover letter) along with manually applying the input signal to pin 11 of J3 on the AM62x SK board"

Regards,

Nick

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

"Hello, TI expert! Following this guide, I compiled the corresponding patch and applied it, but I encountered the following error. Could you please advise on how to resolve this error?"

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

Are you still using Processor SDK 8.3, and trying to backport the ECAP drivers to that version of Linux?

Please note that we cannot support backporting drivers to earlier versions of Linux on the forums.

Whatever you are doing, it looks like there are some missing definitions, so the place where those variables / functions are defined either has not been added, or has not been updated.

If you want to backport the ECAP support on your own (without help from TI), I would probably:

1) check out the latest version of the ti-linux-kernel repo
kernel 6.1: https://git.ti.com/cgit/ti-linux-kernel/ti-linux-kernel/tree/?h=ti-linux-6.1.y-cicd
kernel 5.10: https://git.ti.com/cgit/ti-linux-kernel/ti-linux-kernel/tree/?h=ti-linux-5.10.y-cicd

2) search for the files where those variables / functions are defined, e.g. grep -r 'ecap_cnt_pm_ops'

3) check the git log for those files to make sure you are grabbing every single commit that impacts the ecap support

Even if you do all that, I cannot guarantee that it will work the way you expect in SDK 8.3, since nobody at TI has validated ECAP on that release.

Regards,

Nick

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

Thank you for your response. The software version I am currently using is SDK 08.06.00.42. The corresponding device tree for ecap is as follows. Without using the patch you mentioned above, when I output a PWM signal from a signal generator to ecap's input and follow the operation method you described, I always get a value of 0 when using 'cat capture0'.

What could be causing this issue and is there any solution?

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

1) Are you still running into build errors from your reply at https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1242640/am625-ecap/4773577#4773577 ? If so, please give me the terminal commands you are using so I can try to replicate

2) Are you getting any terminal output during boot time around the ECAP? Please attach it to the thread

3) Can you please confirm the signal generator settings?

4) Please attach the terminal output from when you are reading the capture inputs.

Regards,

Nick

0 Praneeth Bajjuri over 2 years ago in reply to Nick Saulnier

TI__Expert 6506 points

On 5.10/ SDK 8.6

Relevant patches in kernel to review and explore.

*	ti_config_fragments: connectivity: Enable TI_ECAP_CAPTURE
*	ti_config_fragments: connectivity: Enable TI_ECAP_CAPTURE
*	arm64: dts: ti: Add AM62x SK ECAP Capture overlay
*	arm64: dts: ti: k3-am62-main: Add ECAP capture nodes
*	counter: ti-ecap-capture: capture driver support for ECAP
*	Documentation: ABI: sysfs-bus-counter: add capture items
*	dt-bindings: counter: add ti,am62-ecap-capture.yaml
*	ti_config_fragments: v8_baseport: Enable counter support

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

To simplify the problem, I set ecap0 as the PWM output and ecap1 as the capture input. Additionally, my code does not include the patch file you provided, instead, it uses the original ecap code from SDK 08.06.00.42. Here is the device tree file and config file provided, along with the debugging log. However, even after using the "cat capture0" command, the value obtained is still 0.

1. dts file

2. config file

am62xx_config.zip

3. debug log

root@am62xx:/# uname -a
Linux am62xx 5.10.168-rt83-gc1a1291911 #5 SMP PREEMPT_RT Tue Aug 22 14:33:19 CST 2023 aarch64 GNU/Linux


root@am62xx:/# cat /etc/soc_info

Build Time         : Wed Aug 23 09:58:20 CST 2023

sdk_model          : am62xx-sdk-v1.0.0.0
sdk_uboot          : u-boot-2021.01+gitAUTOINC+2ee8efd654-g2ee8efd654
sdk_kernel         : linux-rt-5.10.168+gitAUTOINC+c1a1291911-gc1a1291911
sdk_atf            : trusted-firmware-a-2.8+gitAUTOINC+2fcd408bb3
sdk_tee            : optee_os-3.20.0
sdk_device_type    : hs

cpu_model          : AM6234ATCGHAALW,FCCSP-425,TI
cpu_security       : y
cpu_3d_graphics    : n
cpu_linux_dtb      : k3-am6234-sk-sec.dtb


root@am62xx:/# journalctl | grep cap
Jun 16 09:57:23 am62xx kernel: videodev: Linux video capture interface: v2.00
Jun 16 09:57:30 am62xx /usr/sbin/irqbalance[840]: No directory /sys/devices/platform/23110000.capture/: No such file or directory
Jun 16 09:57:30 am62xx /usr/sbin/irqbalance[840]: IRQ 23110000.capture(51) guessed as class 0
Jun 16 09:57:31 am62xx avahi-daemon[866]: Successfully dropped remaining capabilities.


root@am62xx:/# journalctl | grep pwm
Jun 16 09:57:23 am62xx kernel: pwm-beeper buzzer: supply amp not found, using dummy regulator
Jun 16 09:57:23 am62xx kernel: input: pwm-beeper as /devices/platform/buzzer/input/input0
Jun 16 09:57:23 am62xx kernel: pwm-backlight backlight: supply power not found, using dummy regulator


root@am62xx:/# lsmod | grep ecap
ti_ecap_capture        20480  0
counter                28672  1 ti_ecap_capture


root@am62xx:/# cd /sys/firmware/devicetree/base/bus@f0000
root@am62xx:/sys/firmware/devicetree/base/bus@f0000# ll *pwm*
pwm@23000000:
-r--r--r--    1 root     root             4 Jun 16 10:02 #pwm-cells
-r--r--r--    1 root     root            10 Jun 16 10:02 clock-names
-r--r--r--    1 root     root            20 Jun 16 10:02 clocks
-r--r--r--    1 root     root            30 Jun 16 10:02 compatible
-r--r--r--    1 root     root             4 Jun 16 10:02 name
-r--r--r--    1 root     root             4 Jun 16 10:02 phandle
-r--r--r--    1 root     root            12 Jun 16 10:02 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:02 reg
-r--r--r--    1 root     root             5 Jun 16 10:02 status

pwm@23010000:
-r--r--r--    1 root     root             4 Jun 16 10:02 #pwm-cells
-r--r--r--    1 root     root            10 Jun 16 10:02 clock-names
-r--r--r--    1 root     root            20 Jun 16 10:02 clocks
-r--r--r--    1 root     root            30 Jun 16 10:02 compatible
-r--r--r--    1 root     root             4 Jun 16 10:02 name
-r--r--r--    1 root     root             4 Jun 16 10:02 phandle
-r--r--r--    1 root     root            12 Jun 16 10:02 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:02 reg
-r--r--r--    1 root     root             5 Jun 16 10:02 status

pwm@23020000:
-r--r--r--    1 root     root             4 Jun 16 10:02 #pwm-cells
-r--r--r--    1 root     root            10 Jun 16 10:02 clock-names
-r--r--r--    1 root     root            20 Jun 16 10:02 clocks
-r--r--r--    1 root     root            30 Jun 16 10:02 compatible
-r--r--r--    1 root     root             4 Jun 16 10:02 name
-r--r--r--    1 root     root             4 Jun 16 10:02 phandle
-r--r--r--    1 root     root            12 Jun 16 10:02 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:02 reg
-r--r--r--    1 root     root             9 Jun 16 10:02 status

pwm@23100000:
-r--r--r--    1 root     root             4 Jun 16 10:02 #pwm-cells
-r--r--r--    1 root     root             4 Jun 16 10:02 clock-names
-r--r--r--    1 root     root            12 Jun 16 10:02 clocks
-r--r--r--    1 root     root            15 Jun 16 10:02 compatible
-r--r--r--    1 root     root             4 Jun 16 10:02 name
-r--r--r--    1 root     root             4 Jun 16 10:02 phandle
-r--r--r--    1 root     root             4 Jun 16 10:02 pinctrl-0
-r--r--r--    1 root     root             8 Jun 16 10:02 pinctrl-names
-r--r--r--    1 root     root            12 Jun 16 10:02 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:02 reg
-r--r--r--    1 root     root             5 Jun 16 10:02 status

pwm@23110000:
-r--r--r--    1 root     root             4 Jun 16 10:02 #pwm-cells
-r--r--r--    1 root     root             4 Jun 16 10:02 clock-names
-r--r--r--    1 root     root            12 Jun 16 10:02 clocks
-r--r--r--    1 root     root            15 Jun 16 10:02 compatible
-r--r--r--    1 root     root             4 Jun 16 10:02 name
-r--r--r--    1 root     root             4 Jun 16 10:02 phandle
-r--r--r--    1 root     root             4 Jun 16 10:02 pinctrl-0
-r--r--r--    1 root     root             8 Jun 16 10:02 pinctrl-names
-r--r--r--    1 root     root            12 Jun 16 10:02 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:02 reg
-r--r--r--    1 root     root             9 Jun 16 10:02 status

pwm@23120000:
-r--r--r--    1 root     root             4 Jun 16 10:02 #pwm-cells
-r--r--r--    1 root     root             4 Jun 16 10:02 clock-names
-r--r--r--    1 root     root            12 Jun 16 10:02 clocks
-r--r--r--    1 root     root            15 Jun 16 10:02 compatible
-r--r--r--    1 root     root             4 Jun 16 10:02 name
-r--r--r--    1 root     root             4 Jun 16 10:02 phandle
-r--r--r--    1 root     root             4 Jun 16 10:02 pinctrl-0
-r--r--r--    1 root     root             8 Jun 16 10:02 pinctrl-names
-r--r--r--    1 root     root            12 Jun 16 10:02 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:02 reg
-r--r--r--    1 root     root             9 Jun 16 10:02 status


root@am62xx:/sys/firmware/devicetree/base/bus@f0000# hexdump -C pwm@23100000/status
00000000  6f 6b 61 79 00                                    |okay.|
00000005

root@am62xx:/sys/firmware/devicetree/base/bus@f0000# hexdump -C pwm@23110000/status
00000000  64 69 73 61 62 6c 65 64  00                       |disabled.|
00000009

root@am62xx:/sys/firmware/devicetree/base/bus@f0000# hexdump -C pwm@23120000/status
00000000  64 69 73 61 62 6c 65 64  00                       |disabled.|
00000009


root@am62xx:/sys/firmware/devicetree/base/bus@f0000# cat /sys/kernel/debug/pwm

platform/23010000.pwm, 2 PWM devices
 pwm-0   ((null)              ): period: 0 ns duty: 0 ns polarity: normal
 pwm-1   (backlight           ): requested enabled period: 25000 ns duty: 19607 ns polarity: normal

platform/23000000.pwm, 2 PWM devices
 pwm-0   (buzzer              ): requested period: 366166 ns duty: 183083 ns polarity: normal
 pwm-1   ((null)              ): period: 0 ns duty: 0 ns polarity: normal

platform/23100000.pwm, 1 PWM device
 pwm-0   ((null)              ): period: 0 ns duty: 0 ns polarity: normal




root@am62xx:/sys/firmware/devicetree/base/bus@f0000# ll /sys/class/pwm
lrwxrwxrwx    1 root     root             0 Jun 16 09:57 pwmchip0 -> ../../devices/platform/bus@f0000/23100000.pwm/pwm/pwmchip0
lrwxrwxrwx    1 root     root             0 Jun 16 09:57 pwmchip1 -> ../../devices/platform/bus@f0000/23000000.pwm/pwm/pwmchip1
lrwxrwxrwx    1 root     root             0 Jun 16 09:57 pwmchip3 -> ../../devices/platform/bus@f0000/23010000.pwm/pwm/pwmchip3



root@am62xx:/sys/class/pwm/pwmchip0# base/bus@f0000# cd /sys/class/pwm/pwmchip0/
root@am62xx:/sys/class/pwm/pwmchip0# cat npwm
1
root@am62xx:/sys/class/pwm/pwmchip0# echo 0 > export
root@am62xx:/sys/class/pwm/pwmchip0# cd pwm0
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 100000000 > period
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 30000000 > duty_cycle
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo "inversed" > polarity
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 1 > enable

root@am62xx:/sys/class/pwm/pwmchip0/pwm0# cat /sys/kernel/debug/pwm
platform/23010000.pwm, 2 PWM devices
 pwm-0   ((null)              ): period: 0 ns duty: 0 ns polarity: normal
 pwm-1   (backlight           ): requested enabled period: 25000 ns duty: 19607 ns polarity: normal

platform/23000000.pwm, 2 PWM devices
 pwm-0   (buzzer              ): requested period: 366166 ns duty: 183083 ns polarity: normal
 pwm-1   ((null)              ): period: 0 ns duty: 0 ns polarity: normal

platform/23100000.pwm, 1 PWM device
 pwm-0   (sysfs               ): requested enabled period: 100000000 ns duty: 30000000 ns polarity: inverse
 


root@am62xx:/sys/class/pwm/pwmchip0/pwm0# cd /sys/firmware/devicetree/base/bus@f0000

root@am62xx:/sys/firmware/devicetree/base/bus@f0000# ll *capture*
capture@23100000:
-r--r--r--    1 root     root             4 Jun 16 10:04 clock-names
-r--r--r--    1 root     root            12 Jun 16 10:04 clocks
-r--r--r--    1 root     root            21 Jun 16 10:04 compatible
-r--r--r--    1 root     root            12 Jun 16 10:04 interrupts
-r--r--r--    1 root     root             8 Jun 16 10:04 name
-r--r--r--    1 root     root             4 Jun 16 10:04 phandle
-r--r--r--    1 root     root             4 Jun 16 10:04 pinctrl-0
-r--r--r--    1 root     root             8 Jun 16 10:04 pinctrl-names
-r--r--r--    1 root     root            12 Jun 16 10:04 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:04 reg
-r--r--r--    1 root     root             9 Jun 16 10:04 status

capture@23110000:
-r--r--r--    1 root     root             4 Jun 16 10:04 clock-names
-r--r--r--    1 root     root            12 Jun 16 10:04 clocks
-r--r--r--    1 root     root            21 Jun 16 10:04 compatible
-r--r--r--    1 root     root            12 Jun 16 10:04 interrupts
-r--r--r--    1 root     root             8 Jun 16 10:04 name
-r--r--r--    1 root     root             4 Jun 16 10:04 phandle
-r--r--r--    1 root     root             4 Jun 16 10:04 pinctrl-0
-r--r--r--    1 root     root             8 Jun 16 10:04 pinctrl-names
-r--r--r--    1 root     root            12 Jun 16 10:04 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:04 reg
-r--r--r--    1 root     root             5 Jun 16 10:04 status

capture@23120000:
-r--r--r--    1 root     root             4 Jun 16 10:04 clock-names
-r--r--r--    1 root     root            12 Jun 16 10:04 clocks
-r--r--r--    1 root     root            21 Jun 16 10:04 compatible
-r--r--r--    1 root     root            12 Jun 16 10:04 interrupts
-r--r--r--    1 root     root             8 Jun 16 10:04 name
-r--r--r--    1 root     root             4 Jun 16 10:04 phandle
-r--r--r--    1 root     root             4 Jun 16 10:04 pinctrl-0
-r--r--r--    1 root     root             8 Jun 16 10:04 pinctrl-names
-r--r--r--    1 root     root            12 Jun 16 10:04 power-domains
-r--r--r--    1 root     root            16 Jun 16 10:04 reg
-r--r--r--    1 root     root             9 Jun 16 10:04 status

/status62xx:/sys/firmware/devicetree/base/bus@f0000# hexdump -C capture@23100000/
00000000  64 69 73 61 62 6c 65 64  00                       |disabled.|
00000009
root@am62xx:/sys/firmware/devicetree/base/bus@f0000#
/status62xx:/sys/firmware/devicetree/base/bus@f0000# hexdump -C capture@23110000/
00000000  6f 6b 61 79 00                                    |okay.|
00000005
root@am62xx:/sys/firmware/devicetree/base/bus@f0000#
/status62xx:/sys/firmware/devicetree/base/bus@f0000# hexdump -C capture@23120000/
00000000  64 69 73 61 62 6c 65 64  00                       |disabled.|
00000009



root@am62xx:/sys/firmware/devicetree/base/bus@f0000# cat /proc/interrupts | grep capture
 51:          0          0          0          0     GICv3 146 Edge      23110000.capture


root@am62xx:/sys/firmware/devicetree/base/bus@f0000# ll /dev/counter0
crw-------    1 root     root      236,   0 Jun 16 09:57 /dev/counter0



/oot@am62xx:/sys/firmware/devicetree/base/bus@f0000# ll /sys/bus/counter/devices/
lrwxrwxrwx    1 root     root             0 Jun 16 09:57 counter0 -> ../../../devices/platform/bus@f0000/23110000.capture/counter0


root@am62xx:/sys/firmware/devicetree/base/bus@f0000# cd /sys/bus/counter/devices/counter0
root@am62xx:/sys/bus/counter/devices/counter0# ll
drwxr-xr-x    2 root     root             0 Jun 16 10:05 count0
-r--r--r--    1 root     root          4096 Jun 16 10:05 dev
-rw-r--r--    1 root     root          4096 Jun 16 10:05 events_queue_size
-r--r--r--    1 root     root          4096 Jun 16 10:05 name
-r--r--r--    1 root     root          4096 Jun 16 10:05 num_counts
-r--r--r--    1 root     root          4096 Jun 16 10:05 num_signals
lrwxrwxrwx    1 root     root             0 Jun 16 10:05 of_node -> ../../../../../firmware/devicetree/base/bus@f0000/capture@23110000
drwxr-xr-x    2 root     root             0 Jun 16 10:05 power
drwxr-xr-x    2 root     root             0 Jun 16 10:05 signal0
drwxr-xr-x    2 root     root             0 Jun 16 09:58 signal1
lrwxrwxrwx    1 root     root             0 Jun 16 09:57 subsystem -> ../../../../../bus/counter
-rw-r--r--    1 root     root          4096 Jun 16 09:57 uevent


root@am62xx:/sys/bus/counter/devices/counter0# ll /sys/bus/counter/devices/counter0/count0
-rw-r--r--    1 root     root          4096 Jun 16 10:05 capture0
-r--r--r--    1 root     root          4096 Jun 16 10:05 capture0_component_id
-rw-r--r--    1 root     root          4096 Jun 16 10:05 capture1
-r--r--r--    1 root     root          4096 Jun 16 10:05 capture1_component_id
-rw-r--r--    1 root     root          4096 Jun 16 10:05 capture2
-r--r--r--    1 root     root          4096 Jun 16 10:05 capture2_component_id
-rw-r--r--    1 root     root          4096 Jun 16 10:05 capture3
-r--r--r--    1 root     root          4096 Jun 16 10:05 capture3_component_id
-r--r--r--    1 root     root          4096 Jun 16 10:05 ceiling
-r--r--r--    1 root     root          4096 Jun 16 10:05 ceiling_component_id
-rw-r--r--    1 root     root          4096 Jun 16 10:05 count
-rw-r--r--    1 root     root          4096 Jun 16 10:05 enable
-r--r--r--    1 root     root          4096 Jun 16 10:05 enable_component_id
-r--r--r--    1 root     root          4096 Jun 16 10:05 function
-r--r--r--    1 root     root          4096 Jun 16 10:05 function_available
-r--r--r--    1 root     root          4096 Jun 16 10:05 name
-rw-r--r--    1 root     root          4096 Jun 16 10:05 num_overflows
-r--r--r--    1 root     root          4096 Jun 16 10:05 num_overflows_component_id
-r--r--r--    1 root     root          4096 Jun 16 10:05 signal0_action
-r--r--r--    1 root     root          4096 Jun 16 10:05 signal0_action_available
-r--r--r--    1 root     root          4096 Jun 16 10:05 signal0_action_component_id
-r--r--r--    1 root     root          4096 Jun 16 10:05 signal1_action
-r--r--r--    1 root     root          4096 Jun 16 10:05 signal1_action_available
-r--r--r--    1 root     root          4096 Jun 16 10:05 signal1_action_component_id


root@am62xx:/sys/bus/counter/devices/counter0# ll /sys/bus/counter/devices/counter0/signal0/
-r--r--r--    1 root     root          4096 Jun 16 10:05 frequency
-r--r--r--    1 root     root          4096 Jun 16 10:05 frequency_component_id
-r--r--r--    1 root     root          4096 Jun 16 10:05 name
----------    1 root     root          4096 Jun 16 10:05 signal

root@am62xx:/sys/bus/counter/devices/counter0# ll /sys/bus/counter/devices/counter0/signal1/
-r--r--r--    1 root     root          4096 Jun 16 09:58 name
-rw-r--r--    1 root     root          4096 Jun 16 09:58 polarity0
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity0_available
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity0_component_id
-rw-r--r--    1 root     root          4096 Jun 16 09:58 polarity1
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity1_available
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity1_component_id
-rw-r--r--    1 root     root          4096 Jun 16 09:58 polarity2
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity2_available
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity2_component_id
-rw-r--r--    1 root     root          4096 Jun 16 09:58 polarity3
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity3_available
-r--r--r--    1 root     root          4096 Jun 16 09:58 polarity3_component_id
----------    1 root     root          4096 Jun 16 09:58 signal


root@am62xx:/sys/bus/counter/devices/counter0# cd /sys/bus/counter/devices/counter0/signal0
root@am62xx:/sys/bus/counter/devices/counter0/signal0# cat frequency
125000000

root@am62xx:/signal1s/counter/devices/counter0/signal0# cd /sys/bus/counter/devices/counter0/signal1
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity0_available
positive
negative
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity1_available
positive
negative
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity2_available
positive
negative
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity3_available
positive
negative
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity0
positive
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity1
positive
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity2
positive
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity3
positive
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo positive > polarity0
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo negative > polarity1
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo positive > polarity2
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo negative > polarity3
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity0
positive
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity1
negative
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity2
positive
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cat polarity3
negative


root@am62xx:/sys/bus/counter/devices/counter0/signal1# cd /sys/bus/counter/devices/counter0/count0
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat ceiling
4294967295
root@am62xx:/sys/bus/counter/devices/counter0/count0# echo 0 > num_overflows
root@am62xx:/sys/bus/counter/devices/counter0/count0# echo 0 > count
root@am62xx:/sys/bus/counter/devices/counter0/count0# echo 1 > enable
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat num_overflows
0
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat count
1398114814
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture0
0
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture1
0
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture2
0
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture3
0

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

Have you verified that you are actually seeing an output on the ECAP PWM output?

When I reference other devicetree files like k3-am642-evm.dts, it looks like the pinmux settings for ECAP PWM output are still set to PIN_INPUT, even though it is an output signal. I have not had time to check on my side whether PIN_INPUT vs PIN_OUTPUT makes a difference in this usecase.

Regards,

Nick

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

Hello Nick，

I measured the ECAP PWM output using an oscilloscope, and the measurement results are shown in the following figure.

root@am62xx:/sys/class/pwm/pwmchip0# base/bus@f0000# cd /sys/class/pwm/pwmchip0/
root@am62xx:/sys/class/pwm/pwmchip0# cat npwm
1
root@am62xx:/sys/class/pwm/pwmchip0# echo 0 > export
root@am62xx:/sys/class/pwm/pwmchip0# cd pwm0
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 100000000 > period
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 30000000 > duty_cycle
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo "inversed" > polarity
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 1 > enable

root@am62xx:/sys/class/pwm/pwmchip0/pwm0# cat /sys/kernel/debug/pwm
platform/23010000.pwm, 2 PWM devices
 pwm-0   ((null)              ): period: 0 ns duty: 0 ns polarity: normal
 pwm-1   (backlight           ): requested enabled period: 25000 ns duty: 19607 ns polarity: normal

platform/23000000.pwm, 2 PWM devices
 pwm-0   (buzzer              ): requested period: 366166 ns duty: 183083 ns polarity: normal
 pwm-1   ((null)              ): period: 0 ns duty: 0 ns polarity: normal

platform/23100000.pwm, 1 PWM device
 pwm-0   (sysfs               ): requested enabled period: 100000000 ns duty: 30000000 ns polarity: inverse

Taking into account the above commands, the analysis reveals that the PWM has output a waveform that meets the expected requirements.

So why is the ecap still not working properly even though the pwm is outputting correctly?

Regards,

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

I am going to try to replicate your results on my side tomorrow.

Are you testing on a TI EVM, or on a custom board?

Regards,

Nick

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

on a custom board

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

any updates today？

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

I apologize for the continued delayed response. Will try to run tests tomorrow.

In the meantime, can you please clarify for me exactly how the ecap-pwm, ecap-capture, and production.dtsi device tree files are related to each other?

Please include each file, and the "includes" in each file. Also, I assume there is a 4th file that ends in .dts that includes even production.dtsi?

For each device tree file, please type out:

filename
includes
pwm settings

Please remember that devicetrees are read from top to bottom. e.g., if your top-level .dts file includes the 3 .dtsi files like this:
ecap-pwm.dtsi
ecap-capture.dtsi
production.dtsi

then the settings in production.dtsi would overwrite the settings in the earlier included files. However, if you did it like this in your top-level .dts file:

production.dtsi
ecap-capture.dtsi
ecap-pwm.dtsi

Then the settings in ecap-pwm would overwrite the settings in the earlier files, and the PWM output would be enabled for every ecap instance instead of enabling PWM for one instance, and Capture for another instance.

Regards,

Nick

0 Nick Saulnier over 2 years ago in reply to Nick Saulnier

TI__Guru** 107090 points

Just wanted to let you know that I have been spending time on getting a hardware setup going that I can run tests on. I do not have working code yet. Will provide another update on Thursday or Friday.

Regards,

Nick

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

root@am62xx:/sys/bus/counter/devices/counter0/count0# cat count
1398114814
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture0
42461133
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture1
51211134
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture2
54961134
root@am62xx:/sys/bus/counter/devices/counter0/count0# cat capture3
88711137

Hello Nick，

If I have the above output, how do I calculate the duty cycle and period of PWM?

Regards,

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

Glad to see that you are able to capture inputs now! Did anything change?

For hardware setup, I connected ECAP0 PWM output to ECAP2 capture on the AM62-SKB on the J3 header. See pinmuxing below, and search for the pin numbers on the EVM schematic to find the J3 pins.

Devicetree changes:

&main_pmx0 {
...
        main_ecap0_pwm_pins_default: main-ecap0-pwm-pins-default {
                pinctrl-single,pins = <
                        AM62X_IOPAD(0x01b8, PIN_OUTPUT, 3) /* (C13) SPI0_CS1.ECAP0_IN_APWM_OUT */
                >;
        };
        main_ecap2_capture_pins_default: main-ecap2-capture-pins-default {
                pinctrl-single,pins = <
                        AM62X_IOPAD(0x01a4, PIN_INPUT, 2) /* (B20) MCASP0_ACLKX.ECAP2_IN_APWM_OUT */
                >;
        };

};

&ecap0_pwm {
        pinctrl-names = "default";
        pinctrl-0 = <&main_ecap0_pwm_pins_default>;
        status = "okay";
};

&ecap2_pwm {
        status = "disabled";
};

&ecap2_capture {
        pinctrl-names = "default";
        pinctrl-0 = <&main_ecap2_capture_pins_default>;
        status = "okay";
};

When I finished the test on my side, I was able to see the inputs as expected:

root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# uname -a
Linux am62xx-evm 5.10.168-g2c23e6c538 #1 SMP PREEMPT Fri Feb 24 05:59:04 UTC 2023 aarch64 aarch64 aarch64 GNU/Linux

// check ECAP PWM output and ECAP capture input
root@am62xx-evm:~# cat /proc/device-tree/__symbols__/ecap0_pwm
/bus@f0000/pwm@23100000
root@am62xx-evm:~# cat /proc/device-tree/bus@f0000/pwm@23100000/status
okay
root@am62xx-evm:~#
cat /proc/device-tree/__symbols__/ecap2_capture
/bus@f0000/capture@23120000root@am62xx-evm:~# cat /proc/device-tree/bus@f0000/capture@23120000/status
okay

// set PWM output exactly as you did

// set capture input exactly as described in the nodes at the top of this e2e thread

root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# echo 0 > num_overflows
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# echo 0 > count
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# echo 1 > enable
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat num_overflows
0
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat count
1057336545
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat count
1361843282
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat count
1508350869
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat count
1629343053
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat count
1756358067
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat capture0
2102288297
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat capture1
2464788326
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat capture2
2777288351
root@am62xx-evm:/sys/bus/counter/devices/counter0/count0# cat capture2
3227288387

You will not be able to get useful duty cycle/PWM period information from hand-typing out stuff. I would use a program to read all 4 entries at once, as shown above: https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1242640/am625-ecap/4766153#4766153

If the captures are set to trigger on alternating polarity, that tells you the time for the high part of the duty cycle, and the low part of the duty cycle. Then you can calculate the duty cycle and the frequency from those 4 datapoints.

Regards,

Nick

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

Hello Nick，

I obtained the following information from this link: https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1242640/am625-ecap/4766153#4766153

The Enhanced Capture (ECAP) module can be used to timestamp events
detected on signal input pin. It can be used for time measurements
of pulse train signals.

ECAP module includes 4 timestamp capture registers. For all 4 sequenced
timestamp capture events (0->1->2->3->0->...), edge polarity (falling/rising
edge) can be selected.

This driver leverages counter subsystem to :
- select edge polarity for all 4 capture events (event mode)
- log timestamps for each capture event
Event polarity, and CAP0/1/2/3 timestamps give all the information
about the input pulse train. Further information can easily be computed :
period and/or duty cycle if frequency is constant, elapsed time between
pulses, etc...

...

There is no explanation here about the method to calculate the duty cycle and period of PWM from the values of capture0~capture3.

Could you provide an example program that calculates the duty cycle and period of PWM from the values of capture0~capture3?

Regards,

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

We are starting to get outside the scope of what you can ask TI to support. But here is a quick reminder about how duty cycle and period are defined. You can write your own code.

Let's say capture0 rising edge, capture1 falling edge, capture2 rising edge, capture3 falling edge.

Then period = capture2 - capture0 (or average that with capture3 - capture1)

The signal is high for capture1 - capture0

the high part of the duty cycle is signal_high_time / period

Regards,

Nick

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

Hi Nick,

I have test ecap with the application you mentioned above, and the code of the application is as follows:

	// SPDX-License-Identifier: GPL-2.0-only
	/* Counter - example userspace application
	 *
	 * The userspace application opens /dev/counter0, configures the
	 * COUNTER_EVENT_CAPTURE event channels to gather count 0 capture
	 * events, and prints out the data as it becomes available on the
	 * character device node.
	 *
	 * Copyright (C) 2022 Julien Panis
	 */
	#include <errno.h>
	#include <fcntl.h>
	#include <linux/counter.h>
	#include <stdio.h>
	#include <string.h>
	#include <sys/ioctl.h>
	#include <unistd.h>

	/* assuming capture attributes are under the count0 directory */
	#define CAPTURE_WATCH(_id) \
	{ \
		.component.type = COUNTER_COMPONENT_EXTENSION, \
		.component.scope = COUNTER_SCOPE_COUNT, \
		.component.parent = 0, \
		.component.id = _id, \
		.event = COUNTER_EVENT_CAPTURE, \
		.channel = _id, \
	}

	/* get id from respective "captureX_component_id" attributes */
	#define NB_CAP 4
	static struct counter_watch watches[NB_CAP] = {
		CAPTURE_WATCH(0),
		CAPTURE_WATCH(1),
		CAPTURE_WATCH(2),
		CAPTURE_WATCH(3),
	};

	int main(void)
	{
		int fd;
		int ret;
		int i;
		struct counter_event event_data[NB_CAP];

		fd = open("/dev/counter0", O_RDWR);
		if (fd == -1) {
			perror("Unable to open /dev/counter0");
			return 1;
		}

		for (i = 0; i < NB_CAP; i++) {
			ret = ioctl(fd, COUNTER_ADD_WATCH_IOCTL, watches + i);
			if (ret == -1) {
				fprintf(stderr, "Error adding watches[%d]: %s\n", i,
					strerror(errno));
				return 1;
			}
		}
		ret = ioctl(fd, COUNTER_ENABLE_EVENTS_IOCTL);
		if (ret == -1) {
			perror("Error enabling events");
			return 1;
		}

		for (;;) {
			ret = read(fd, event_data, sizeof(event_data));
			if (ret == -1) {
				perror("Failed to read event data");
				return 1;
			}

			if (ret != sizeof(event_data)) {
				fprintf(stderr, "Failed to read event data\n");
				return -EIO;
			}

			/* [WARNING] For data consistency analysis, stdout must be redirected to log file !!! */
			puts("----------");
			printf("cap0: %llu\n", event_data[0].value);
			printf("cap1: %llu\n", event_data[1].value);
			printf("cap2: %llu\n", event_data[2].value);
			printf("cap3: %llu\n", event_data[3].value);
		}

		return 0;
	}

The commands used during testing are as follows:

root@am62xx:/opt# cd /sys/class/pwm/pwmchip0
root@am62xx:/sys/class/pwm/pwmchip0#
root@am62xx:/sys/class/pwm/pwmchip0# echo 0 > export
root@am62xx:/sys/class/pwm/pwmchip0#
root@am62xx:/sys/class/pwm/pwmchip0# cd pwm0/
root@am62xx:/sys/class/pwm/pwmchip0/pwm0#
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 100000000 > period
root@am62xx:/sys/class/pwm/pwmchip0/pwm0#
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 30000000 > duty_cycle
root@am62xx:/sys/class/pwm/pwmchip0/pwm0#
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo "inversed" > polarity
root@am62xx:/sys/class/pwm/pwmchip0/pwm0#
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# echo 1 > enable
root@am62xx:/sys/class/pwm/pwmchip0/pwm0#
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# cat /sys/kernel/debug/pwm
platform/23010000.pwm, 2 PWM devices
 pwm-0   ((null)              ): period: 0 ns duty: 0 ns polarity: normal
 pwm-1   (backlight           ): requested enabled period: 25000 ns duty: 19607 ns polarity: normal

platform/23000000.pwm, 2 PWM devices
 pwm-0   (buzzer              ): requested period: 366166 ns duty: 0 ns polarity: normal
 pwm-1   ((null)              ): period: 0 ns duty: 0 ns polarity: normal

platform/23100000.pwm, 1 PWM device
 pwm-0   (sysfs               ): requested enabled period: 100000000 ns duty: 30000000 ns polarity: inverse
root@am62xx:/sys/class/pwm/pwmchip0/pwm0#
root@am62xx:/sys/class/pwm/pwmchip0/pwm0# cd /sys/bus/counter/devices/counter0/signal1
root@am62xx:/sys/bus/counter/devices/counter0/signal1#
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo positive > polarity0
root@am62xx:/sys/bus/counter/devices/counter0/signal1#
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo positive > polarity1
root@am62xx:/sys/bus/counter/devices/counter0/signal1#
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo positive > polarity2
root@am62xx:/sys/bus/counter/devices/counter0/signal1#
root@am62xx:/sys/bus/counter/devices/counter0/signal1# echo positive > polarity3
root@am62xx:/sys/bus/counter/devices/counter0/signal1#
root@am62xx:/sys/bus/counter/devices/counter0/signal1# cd /sys/bus/counter/devices/counter0/count0/
root@am62xx:/sys/bus/counter/devices/counter0/count0#
root@am62xx:/sys/bus/counter/devices/counter0/count0# echo 0 > num_overflows
root@am62xx:/sys/bus/counter/devices/counter0/count0#
root@am62xx:/sys/bus/counter/devices/counter0/count0# echo 0 > count
root@am62xx:/sys/bus/counter/devices/counter0/count0#
root@am62xx:/sys/bus/counter/devices/counter0/count0# echo 1 > enable
root@am62xx:/sys/bus/counter/devices/counter0/count0#
root@am62xx:/sys/bus/counter/devices/counter0/count0# cd /opt
root@am62xx:/opt#
root@am62xx:/opt# ./counter_example > counter_log.txt
Failed to read event data
root@am62xx:/opt#
root@am62xx:/opt#
root@am62xx:/opt# cat counter_log.txt
----------
cap0: 1588965858
cap1: 1601465859
cap2: 1613965860
cap3: 1626465861
----------
cap0: 1638965862
cap1: 1651465863
cap2: 1663965864
cap3: 1676465865
----------
cap0: 1688965866
cap1: 1701465867
cap2: 1713965868
cap3: 1726465869
----------
cap0: 1738965870
cap1: 1751465871
cap2: 1763965872
cap3: 1776465873
----------
cap0: 1788965874
cap1: 1801465875
cap2: 1813965876
cap3: 1826465877
----------
cap0: 1838965878
cap1: 1851465879
cap2: 1863965880
cap3: 1876465881
----------
cap0: 1888965882
cap1: 1901465883
cap2: 1913965884
cap3: 1926465885
root@am62xx:/opt#
root@am62xx:/opt#

From the above log, it can be seen that there are the following two issues:

Question 1: During the test, there was an error that prevented the test from continuing. May I know the reason behind this error?

root@am62xx:/opt# ./counter_example > counter_log.txt
Failed to read event data
root@am62xx:/opt#

Question 2:The most important thing is that, according to the method you mentioned above, the calculated PWM period and duty cycle based on the log results are inconsistent with the actual settings.

Taking one set of data from the aforementioned log as an example:

----------
cap0: 1688965866
cap1: 1701465867
cap2: 1713965868
cap3: 1726465869

cap2 - cap0 = 1713965868 - 1688965866 = 25000002

cap1 - cap0 = 1701465867 - 1688965866 = 12500001

Inconsistent with the actual settings， as follows:

platform/23100000.pwm, 1 PWM device
 pwm-0   (sysfs               ): requested enabled period: 100000000 ns duty: 30000000 ns polarity: inverse

Regards,

0 li weiyu over 2 years ago in reply to Nick Saulnier

Expert 1100 points

Hello Nick，

It seems that there is a slight issue with the eCAP in PROCESSOR-SDK-LINUX-RT-AM62X-08.06.00.42.

I am using the following test script to test eCAP:

root@am62xx:~# cat ecap.sh
#!/bin/bash

./ecap_init.sh
./ecap_echo.sh


root@am62xx:~# cat ecap_init.sh
#!/bin/bash

#set -x

cd /sys/class/pwm/pwmchip0
echo 0 > export
cd pwm0/
echo 100000000 > period
echo 30000000 > duty_cycle
echo "normal" > polarity
echo 1 > enable
cat /sys/kernel/debug/pwm

cd /sys/bus/counter/devices/counter0/signal1
echo positive > polarity0
echo negative > polarity1
echo positive > polarity2
echo negative > polarity3

cd /sys/bus/counter/devices/counter0/count0/
echo 0 > num_overflows
echo 0 > count
echo 1 > enable

sleep 1




root@am62xx:~# cat ecap_echo.sh
#!/bin/bash

cd /sys/bus/counter/devices/counter0/count0/

freq=$(cat /sys/bus/counter/devices/counter0/signal0/frequency)
ceiling=$(cat /sys/bus/counter/devices/counter0/count0/ceiling)

cap0_val=$(cat capture0)
cap1_val=$(cat capture1)
cap2_val=$(cat capture2)
cap3_val=$(cat capture3)


if [ $cap0_val -lt $cap2_val ]; then
  period_cnt=$(( cap2_val - cap0_val - 1 ))
else
  period_cnt=$(( ceiling + cap2_val - cap0_val - 1 ))
fi
period=$(echo "scale=2; $period_cnt / $freq" | bc)


if [ $cap0_val -lt $cap1_val ]; then
  high_cnt=$(( cap1_val - cap0_val ))
else
  high_cnt=$(( ceiling + cap1_val - cap0_val ))
fi
duty=$(bc <<< "scale=2; $high_cnt / $period_cnt * 100")


echo
echo "cap0_val:   $cap0_val"
echo "cap1_val:   $cap1_val"
echo "cap2_val:   $cap2_val"
echo "cap3_val:   $cap3_val"

echo
echo "high_cnt:   $high_cnt"
echo "period_cnt: $period_cnt"
echo "freq:       ${freq}hz"

echo
echo "period:     $(printf "%.2f" $period)s"
echo "duty:       $duty%"
echo

The log output during testing is as follows:

root@am62xx:~# ./ecap.sh
platform/23010000.pwm, 2 PWM devices
 pwm-0   ((null)              ): period: 0 ns duty: 0 ns polarity: normal
 pwm-1   (backlight           ): period: 0 ns duty: 0 ns polarity: normal

platform/23000000.pwm, 2 PWM devices
 pwm-0   (buzzer              ): period: 0 ns duty: 0 ns polarity: normal
 pwm-1   ((null)              ): period: 0 ns duty: 0 ns polarity: normal

platform/23100000.pwm, 1 PWM device
 pwm-0   (sysfs               ): requested enabled period: 100000000 ns duty: 30000000 ns polarity: normal



root@am62xx:~# ./ecap_echo.sh

cap0_val:   3768832424
cap1_val:   3772582424
cap2_val:   3781332425
cap3_val:   3785082425

high_cnt:   3750000
period_cnt: 12500000
freq:       125000000hz

period:     0.10s
duty:       30.00%

root@am62xx:~# ./ecap_echo.sh

cap0_val:   3843832430
cap1_val:   3847582430
cap2_val:   3831332429
cap3_val:   3835082429

high_cnt:   3750000
period_cnt: 4282467293
freq:       125000000hz

period:     34.25s
duty:       0%

root@am62xx:~# ./ecap_echo.sh

cap0_val:   3893832434
cap1_val:   3897582434
cap2_val:   3906332435
cap3_val:   3885082433

high_cnt:   3750000
period_cnt: 12500000
freq:       125000000hz

period:     0.10s
duty:       30.00%

From the above testing log, it can be seen that this set of data is correct:

root@am62xx:~# ./ecap_echo.sh

cap0_val:   3768832424
cap1_val:   3772582424
cap2_val:   3781332425
cap3_val:   3785082425

high_cnt:   3750000
period_cnt: 12500000
freq:       125000000hz

period:     0.10s
duty:       30.00%

However, there is a set of data in the testing log that is clearly incorrect, as shown below:

root@am62xx:~# ./ecap_echo.sh

cap0_val:   3843832430
cap1_val:   3847582430
cap2_val:   3831332429
cap3_val:   3835082429

high_cnt:   3750000
period_cnt: 4282467293
freq:       125000000hz

period:     34.25s
duty:       0%

Further analysis of the above dataset leads to the following conclusions:

high_cnt：
cap3_val - cap2_val = 3835082429 - 3831332429     = 3750000

period_cnt：
cap0_val - cap2_val = 3843832430 - 3831332429 - 1 = 12500000

But in reality, the formulas for calculating high_cnt and period_cnt are as follows:

high_cnt=$(( cap1_val - cap0_val ))
period_cnt=$(( cap2_val - cap0_val - 1 ))

Based on comprehensive analysis, it can be inferred that the erroneous set of data exhibits the following correlation.

cap0_val:   3843832430  --> cap2_val
cap1_val:   3847582430  --> cap3_val
cap2_val:   3831332429  --> cap0_val
cap3_val:   3835082429  --> cap1_val

The captured values were not saved into the correct device node.

How can we resolve this issue?

Regards,

0 Nick Saulnier over 2 years ago in reply to li weiyu

TI__Guru** 107090 points

Hello Li,

It looks like the capture values are constantly updated. i.e., if they update like this:

cap0 cap1 cap2 cap3 cap0 cap1 cap2 cap3 cap0 cap1 cap2 cap3

Then you just sampled the 4 readings here:

cap0 cap1 cap2 cap3 cap0 cap1 cap2 cap3 cap0 cap1 cap2 cap3 
                             ^
                             sampled between cap1 and cap2

That is not a bug. Your code will just need to check to see which counter value is the latest, and then do the math based on that order. Without looking into it, I am not sure whether there is a 32 bit counter rollover or not, so you'll want to check for that.

Regards,

Nick

Processors

Processors forum

AM625: ECAP