AM625: Device not able to wakeup from Deep sleep.

Nilesh Patil

Part Number: AM625

Hello Team,

We are trying to implement Sleep and Wakeup on our custom board, we are trying to use MCU GPIO for waking up the device from Sleep (Suspend-to-RAM (Deep Sleep)).

We have connected MCU_GPIO0_3 to an external switch on our board and we are using this to generate interrupt on device while in Deep Sleep. When we press the switch we get high to low signal on the MCU_GPIO but we observe that the device is not waking up from sleep. We have attached the changes we did in device tree.

We tried keeping the status of mcu_gpio0 node as both 'okay' and 'reserved' in device tree. but we did not get any success in waking the device from sleep.

Can you guide us how to wake up the device and let us know if any other changes need to be done for same.

Fullscreen

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
+++ b/arch/arm64/boot/dts/msc/am62xx/msc-sm2s-am62x-qc-14402C1I-module.dts
@@ -16,6 +16,8 @@
 #include <dt-bindings/leds/common.h>
 #include <dt-bindings/gpio/gpio.h>
 #include <dt-bindings/net/ti-dp83867.h>
+#include <dt-bindings/interrupt-controller/irq.h>
+#include "../../ti/k3-pinctrl.h"
 #include "../../ti/k3-am625.dtsi"
 
 / {
@@ -320,6 +322,22 @@
            };
        };
    }; */
+
+   mcu_gpio_key {
+           compatible = "gpio-keys";
+                   autorepeat;
+           pinctrl-names = "default";
+           pinctrl-0 = <&wake_mcugpio1_pins_default>;
+           interrupt-parent = <&mcu_gpio0>;
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

+++ b/arch/arm64/boot/dts/msc/am62xx/msc-sm2s-am62x-qc-14402C1I-module.dts
@@ -16,6 +16,8 @@
 #include <dt-bindings/leds/common.h>
 #include <dt-bindings/gpio/gpio.h>
 #include <dt-bindings/net/ti-dp83867.h>
+#include <dt-bindings/interrupt-controller/irq.h>
+#include "../../ti/k3-pinctrl.h"
 #include "../../ti/k3-am625.dtsi"
 
 / {
@@ -320,6 +322,22 @@
 			};
 		};
 	}; */
+
+	mcu_gpio_key {
+	        compatible = "gpio-keys";
+	                autorepeat;
+	        pinctrl-names = "default";
+	        pinctrl-0 = <&wake_mcugpio1_pins_default>;
+	        interrupt-parent = <&mcu_gpio0>;
+	        interrupts = <03 IRQ_TYPE_EDGE_RISING>;
+	        switch {
+	                label = "MCUGPIO";
+	                linux,code = <143>;
+	                gpios = <&mcu_gpio0 03 GPIO_ACTIVE_LOW>;
+	                wakeup-source;
+	        };
+	};
+
 };
 
 &mcu_pmx0 {
@@ -359,6 +377,13 @@
                         AM62X_IOPAD(0x03C, PIN_OUTPUT, 0) /* (E5) MCU_MCAN1_TX */
                 >;
         };
+
+	wake_mcugpio1_pins_default: wake-mcugpio1-pins-default {
+	        pinctrl-single,pins = <
+	                AM62X_MCU_IOPAD(0x00C, PIN_INPUT, 7) /* (D9) MCU_GPIO0_03 */
+	        >;
+	};
+
 };
 
 &main_pmx0 {
@@ -926,6 +951,13 @@
 		compatible= "ricoh,r2221tl";
 		reg=<0x32>;
 	};
+
+	tpm: st33tphf20@2e {
+                compatible = "st,st33htpm-i2c";
+                reg = <0x2e>;
+                status = "okay";
+        };
+
 };
 
 &main_i2c3 {

Thanks,
Nilesh

over 1 year ago

0 Swargam Anil over 1 year ago

TI__Mastermind 39521 points

Hello Nilesh,

I am suspecting that your MCU GPIO pin interrupt is not working.

Can you please confirm whether you are able to get a GPIO interrupt or not?

You are routing the MCU GPIO interrupt to the A53 core on Linux. So, I am forwarding this thread to Linux experts as well.

For the first level, we can verify whether the GPIO interrupt is working or not.

Regards,

S.Anil.

0 Nilesh Patil over 1 year ago in reply to Swargam Anil

Prodigy 60 points

Hello Swargam Anil,

Yes we are able to get the GPIO interrupt we tested it by getting output of /proc/interrupts entry. But the device is not able to wakeup when kept in deep sleep mode.

Fullscreen

1
2
3
4
5
6
root@sm2s-am6254:~# 
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330:          0          0          0          0      GPIO  11 Edge    -davinci_gpio  MCUGPIO
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330:        401          0          0          0      GPIO  11 Edge    -davinci_gpio  MCUGPIO
root@sm2s-am6254:~# 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

root@sm2s-am6254:~# 
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330:          0          0          0          0      GPIO  11 Edge    -davinci_gpio  MCUGPIO
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330:        401          0          0          0      GPIO  11 Edge    -davinci_gpio  MCUGPIO
root@sm2s-am6254:~#

Can you let me know if any extra changes we need to do other than what is mentioned in https://software-dl.ti.com/processor-sdk-linux/esd/AM62X/09_00_00_03/exports/docs/linux/Foundational_Components/Kernel/Kernel_Drivers/Power_Management/pm_wakeup_sources.html#pm-wakeup-sources

Thanks

Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

Nilesh Patil said:
We have connected MCU_GPIO0_3

From the log you provided, the interrupt is set for MCU_GPIO0_11, not 3.

Also, the device tree pinmux (wake_mcugpio1_pins_default) is pointing to the wrong address. I believe you meant 0x000c not 0x00c.

Best Regards,

Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hello Anshu,

In last comment i forgot to mention that as per our hardware team on our device MCU_GPIO_3 is used for other function, so we tried with MCU_GPIO_11 and we observe same result with that GPIO also. Yes i also had tried with the address as you mentioned 0x002C for MCU_GPIO_11.

Below are the changes we did for using MCU_GPIO_11.

Fullscreen

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
@@ -320,6 +322,22 @@
                        };
                };
        }; */
+
+       mcu_gpio_key {
+               compatible = "gpio-keys";
+                       autorepeat;
+               pinctrl-names = "default";
+               pinctrl-0 = <&wake_mcugpio1_pins_default>;
+               interrupt-parent = <&mcu_gpio0>;
+               interrupts = <11 IRQ_TYPE_EDGE_FALLING>;
+               switch {
+                       label = "MCUGPIO";
+                       linux,code = <143>;
+                       gpios = <&mcu_gpio0 11 GPIO_ACTIVE_LOW>;
+                       wakeup-source;
+               };
+       };
+
 };
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

@@ -320,6 +322,22 @@
                        };
                };
        }; */
+
+       mcu_gpio_key {
+               compatible = "gpio-keys";
+                       autorepeat;
+               pinctrl-names = "default";
+               pinctrl-0 = <&wake_mcugpio1_pins_default>;
+               interrupt-parent = <&mcu_gpio0>;
+               interrupts = <11 IRQ_TYPE_EDGE_FALLING>;
+               switch {
+                       label = "MCUGPIO";
+                       linux,code = <143>;
+                       gpios = <&mcu_gpio0 11 GPIO_ACTIVE_LOW>;
+                       wakeup-source;
+               };
+       };
+
 };

 &mcu_pmx0 {
@@ -359,6 +377,13 @@
                         AM62X_IOPAD(0x03C, PIN_OUTPUT, 0) /* (E5) MCU_MCAN1_TX */
                 >;
         };
+
+       wake_mcugpio1_pins_default: wake-mcugpio1-pins-default {
+               pinctrl-single,pins = <
+                       AM62X_MCU_IOPAD(0x02C, PIN_INPUT, 7) /* (C6) MCU_GPIO0_11 */
+               >;
+       };
+
 };

 &main_pmx0 {

Thanks,
Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

Do you have a reference node to "&mcu_gpio0"? Under mcu_gpio_key, the interrupt parent is looking for <&mcu_gpio0> and in the switch node, but I don't see one in the code you provided.

A simple reference after &mcu_pmx0 should work since the other pinctrl parameters are done in mcu_gpio_key.

Fullscreen

1
2
3
&mcu_gpio0 {
    status = "okay";
};
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

&mcu_gpio0 {
	status = "okay";
};

Best Regards,

Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hello Anshu,

Yes i have this node in my dts file, i also tried changing the status to 'reserved' as well.

Fullscreen

1
2
3
4
5
6
7
&mcu_gpio_intr {
status = "okay";
};
&mcu_gpio0 {
status = "okay";
};
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

&mcu_gpio_intr {
status = "okay";
};

&mcu_gpio0 {
status = "okay";
};

Other than this nodes do i need to add any other node like below, i have added this too.

Fullscreen

1
2
3
&cbass_mcu {
status = "okay";
};
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

&cbass_mcu {
status = "okay";
};

Thanks,
Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

I was able to confirm that this setup works for MCU_GPIO0_15 and 16 for the TI EVM I am using.

I noticed that the interrupt is set for IRQ_TYPE_EDGE_FALLING, instead of IRQ_TYPE_EDGE_RISING. I don't think it matters since it works for both. Just to verify, change it to rising edge and see if there are any differences.

Are there any errors in the logs when the device moves into deep sleep?

Best Regards,

Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hi Anshu,

We tried changing the interrupt from IRQ_TYPE_EDGE_FALLING to IRQ_TYPE_EDGE_RISING. There is no effect, device does not wake up from sleep.

Here is the logs we get when we put the device in sleep with the command

# echo mem > /sys/power/state

Fullscreen

1
2
3
4
5
6
7
8
9
10
11
root@sm2s-am6254:~# echo mem > /sys/power/state
[  497.983406] PM: suspend entry (deep)
[  497.992091] Filesystems sync: 0.004 seconds
[  501.133982] k3-m4-rproc 5000000.m4fss: k3_m4_rproc_stop: timedout waiting for rproc completion event
[  501.143306] remoteproc remoteproc0: can't stop rproc: -16
[  501.149006] Freezing user space processes
[  501.156178] Freezing user space processes completed (elapsed 0.003 seconds)
[  501.163253] OOM killer disabled.
[  501.166557] Freezing remaining freezable tasks
[  501.173149] Freezing remaining freezable tasks completed (elapsed 0.002 seconds)
[  501.180699] printk: Suspending console(s) (use no_console_suspend to debug)
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

root@sm2s-am6254:~# echo mem > /sys/power/state
[  497.983406] PM: suspend entry (deep)
[  497.992091] Filesystems sync: 0.004 seconds
[  501.133982] k3-m4-rproc 5000000.m4fss: k3_m4_rproc_stop: timedout waiting for rproc completion event
[  501.143306] remoteproc remoteproc0: can't stop rproc: -16
[  501.149006] Freezing user space processes
[  501.156178] Freezing user space processes completed (elapsed 0.003 seconds)
[  501.163253] OOM killer disabled.
[  501.166557] Freezing remaining freezable tasks
[  501.173149] Freezing remaining freezable tasks completed (elapsed 0.002 seconds)
[  501.180699] printk: Suspending console(s) (use no_console_suspend to debug)

Below we get interrupt in normal state.

Fullscreen

1
2
3
4
5
6
root@sm2s-am6254:~#
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330: 0 0 0 0 GPIO 11 Edge -davinci_gpio MCUGPIO
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330: 418 0 0 0 GPIO 11 Edge -davinci_gpio MCUGPIO
root@sm2s-am6254:~#
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

root@sm2s-am6254:~#
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330: 0 0 0 0 GPIO 11 Edge -davinci_gpio MCUGPIO
root@sm2s-am6254:~# cat /proc/interrupts | grep "MCUGPIO"
330: 418 0 0 0 GPIO 11 Edge -davinci_gpio MCUGPIO
root@sm2s-am6254:~#

Thanks,

Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

Thank you for the logs.

One noticeable difference in the logs you provided is the CPUs are not turning off as can be seen in the example output: software-dl.ti.com/.../pm_low_power_modes.html

Just to confirm, does the CPUs turn off when going into Deep Sleep?

Best Regards,

Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hi Anshu,

We tried to test this on EVM with default image of 9.0 SDK in that we were are getting below logs

Fullscreen

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
root@am62xx-evm:~# cat /proc/interrupts
           CPU0       CPU1       CPU2       CPU3
11:       4648       3975       4599       4287     GICv3  30 Level     arch_timer
14:          8          0          0          0     GICv3 108 Level     mbox-m4-0, mbox-r5-0
15:          0          0          0          0     GICv3  23 Level     arm-pmu
16:          0          0          0          0     GICv3 130 Level     pinctrl
17:       3742          0          0          0     GICv3  66 Level     4d000000.mailbox thr_012
26:        586          0          0          0     GICv3 193 Level     20000000.i2c
27:        280          0          0          0     GICv3 194 Level     20010000.i2c
28:          0          0          0          0     GICv3 195 Level     20020000.i2c
29:          2          0          0          0     GICv3 171 Level     fc40000.spi
30:       1464          0          0          0  MSI-INTA 1713152 Level     485c0100.dma-controller chan0
48:          0          0          0          0  MSI-INTA 1714176 Edge      485c0100.dma-controller chan1
60:          1          0          0          0  MSI-INTA 1714688 Level     485c0100.dma-controller chan1
78:          0          0          0          0  MSI-INTA 1715718 Edge      485c0100.dma-controller chan2
96:          1          0          0          0  MSI-INTA 1716230 Level     485c0100.dma-controller chan2
118:          0          0          0          0  MSI-INTA 1970707 Level     8000000.ethernet-tx0
190:          0          0          0          0  MSI-INTA 1970779 Level     485c0000.dma-controller chan2
208:          0          0          0          0  MSI-INTA 1971731 Level     8000000.ethernet
232:          0          0          0          0  MSI-INTA 1971755 Level     485c0000.dma-controller chan0
233:          0          0          0          0  MSI-INTA 1971756 Level     485c0000.dma-controller chan1
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

root@am62xx-evm:~# cat /proc/interrupts
           CPU0       CPU1       CPU2       CPU3
11:       4648       3975       4599       4287     GICv3  30 Level     arch_timer
14:          8          0          0          0     GICv3 108 Level     mbox-m4-0, mbox-r5-0
15:          0          0          0          0     GICv3  23 Level     arm-pmu
16:          0          0          0          0     GICv3 130 Level     pinctrl
17:       3742          0          0          0     GICv3  66 Level     4d000000.mailbox thr_012
26:        586          0          0          0     GICv3 193 Level     20000000.i2c
27:        280          0          0          0     GICv3 194 Level     20010000.i2c
28:          0          0          0          0     GICv3 195 Level     20020000.i2c
29:          2          0          0          0     GICv3 171 Level     fc40000.spi
30:       1464          0          0          0  MSI-INTA 1713152 Level     485c0100.dma-controller chan0
48:          0          0          0          0  MSI-INTA 1714176 Edge      485c0100.dma-controller chan1
60:          1          0          0          0  MSI-INTA 1714688 Level     485c0100.dma-controller chan1
78:          0          0          0          0  MSI-INTA 1715718 Edge      485c0100.dma-controller chan2
96:          1          0          0          0  MSI-INTA 1716230 Level     485c0100.dma-controller chan2
118:          0          0          0          0  MSI-INTA 1970707 Level     8000000.ethernet-tx0
190:          0          0          0          0  MSI-INTA 1970779 Level     485c0000.dma-controller chan2
208:          0          0          0          0  MSI-INTA 1971731 Level     8000000.ethernet
232:          0          0          0          0  MSI-INTA 1971755 Level     485c0000.dma-controller chan0
233:          0          0          0          0  MSI-INTA 1971756 Level     485c0000.dma-controller chan1
240:        567          0          0          0     GICv3 210 Level     2800000.serial
241:          0          0          0          0   pinctrl 456 Edge      2800000.serial:wakeup
242:          0          0          0          0     GICv3 134 Level     8000000.ethernet
245:        281          0          0          0     GICv3 165 Level     mmc0
324:          7          0          0          0      GPIO  72 Edge    -davinci_gpio  wl18xx
371:          0          0          0          0      GPIO  23 Edge    -davinci_gpio  1-0022
401:      14622          0          0          0     GICv3 115 Level     mmc1
402:       2498          0          0          0     GICv3 114 Level     mmc2
403:          0          0          0          0     GICv3 269 Level     2b10000.audio-controller_rx
404:          0          0          0          0     GICv3 270 Level     2b10000.audio-controller_tx
409:          0          0          0          0 1-0022  16 Edge      1-003b
413:          0          0          0          0     GICv3 132 Level     2b1f0000.rtc
414:          0          0          0          0     GICv3 116 Level     tidss
415:         50          0          0          0     GICv3 118 Level     pvrsrvkm
429:          0          0          0          0     GICv3 258 Level     xhci-hcd:usb1
IPI0:      1156       2939       1379       2715       Rescheduling interrupts
IPI1:      3013       6902       5082       5097       Function call interrupts
IPI2:         0          0          0          0       CPU stop interrupts
IPI3:         0          0          0          0       CPU stop (for crash dump) interrupts
IPI4:         0          0          0          0       Timer broadcast interrupts
IPI5:       231        152        284        185       IRQ work interrupts
IPI6:         0          0          0          0       CPU wake-up interrupts
Err:          0

root@am62xx-evm:~# echo mem > /sys/power/state
[   67.841531] PM: suspend entry (deep)
[   67.862049] Filesystems sync: 0.016 seconds
[   67.912336] remoteproc remoteproc0: stopped remote processor 5000000.m4fss
[   67.919707] Freezing user space processes
[   67.926054] Freezing user space processes completed (elapsed 0.002 seconds)
[   67.933099] OOM killer disabled.
[   67.936331] Freezing remaining freezable tasks
[   67.942360] Freezing remaining freezable tasks completed (elapsed 0.001 seconds)
[   67.949788] printk: Suspending console(s) (use no_console_suspend to debug)

As mentioned in section 3.2.2.11.4.2. Suspend-to-RAM (Deep Sleep) of https://software-dl.ti.com/processor-sdk-linux/esd/AM62X/09_00_00_03/exports/docs/linux/Foundational_Components/Kernel/Kernel_Drivers/Power_Management/pm_low_power_modes.html#suspend-to-ram-deep-sleep we checked TP25/PMIC_LPM_EN pin on the EVM if the pin is 3.3V when active and 0V when in deep sleep. but it's voltage is High at 3.3V.

Is it that we need to set it as mentioned (after programming the PMCTRL_SYS register (0x43018080) to 0x15) or it is done by default in 9.0 SDK

Thanks,
Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

Thanks for checking the test point. That indicates the device isn't moving into deep sleep and thus the wakeup sources are not working.

Nilesh Patil said:
Is it that we need to set it as mentioned (after programming the PMCTRL_SYS register (0x43018080) to 0x15) or it is done by default in 9.0 SDK

The PMCTRL_SYS register can be found in Table 14-4667 in the TRM which is linked here: https://www.ti.com/lit/ug/spruiv7b/spruiv7b.pdf

It should be done by default, but if not, a Linux Command you can run is "devmem2 0x43018080" which will read the register and "devmem2 0x43018080 w 0x15" to write to the register.

Based on the logs you provided, I think the device is hanging at some point before the CPUs are turned off in the Sleep Sequence.

Can you try using/provide logs for RTC deep sleep? This command should put the device into deep sleep then wakes up in 30 seconds:

rtcwake -s 30 -m mem

On a different point, there is no mention of the "MCUGPIO" or MCU_GPIO0_11 in the cat /proc/interrupts log you provided.

Best Regards,
Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hi Anshu,

The logs in above comment was from EVM and i have used default image from 9.0 SDK provided by TI. By following the steps mentioned in below ticket i was able to do the sleep and wakeup on EVM kit. And also using the RTC deep sleep on EVM it is working fine. https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1261500/am623-gpio-wake-up-time-from-linux-suspend-to-ram

While in our device when i run the commands for enabling deep sleep mode i get the following error logs

[ 316.049822] k3-m4-rproc 5000000.m4fss: k3_m4_rproc_stop: timedout waiting for rproc completion event
[ 316.059242] remoteproc remoteproc0: can't stop rproc: -16

Fullscreen

1
2
3
4
5
6
7
8
9
10
11
root@sm2s-am6254:~# echo mem > /sys/power/state
[  312.934988] PM: suspend entry (deep)
[  312.943663] Filesystems sync: 0.004 seconds
[  316.049822] k3-m4-rproc 5000000.m4fss: k3_m4_rproc_stop: timedout waiting for rproc completion event
[  316.059242] remoteproc remoteproc0: can't stop rproc: -16
[  316.064939] Freezing user space processes
[  316.072190] Freezing user space processes completed (elapsed 0.003 seconds)
[  316.079283] OOM killer disabled.
[  316.082565] Freezing remaining freezable tasks
[  316.089331] Freezing remaining freezable tasks completed (elapsed 0.002 seconds)
[  316.096900] printk: Suspending console(s) (use no_console_suspend to debug)
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

root@sm2s-am6254:~# echo mem > /sys/power/state
[  312.934988] PM: suspend entry (deep)
[  312.943663] Filesystems sync: 0.004 seconds
[  316.049822] k3-m4-rproc 5000000.m4fss: k3_m4_rproc_stop: timedout waiting for rproc completion event
[  316.059242] remoteproc remoteproc0: can't stop rproc: -16
[  316.064939] Freezing user space processes
[  316.072190] Freezing user space processes completed (elapsed 0.003 seconds)
[  316.079283] OOM killer disabled.
[  316.082565] Freezing remaining freezable tasks
[  316.089331] Freezing remaining freezable tasks completed (elapsed 0.002 seconds)
[  316.096900] printk: Suspending console(s) (use no_console_suspend to debug)

I'm not able to change the state of the remoteproc extermally from sysfs by using below command

Fullscreen

1
echo stop > /sys/class/remoteproc/remoteproc0/state
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

echo stop > /sys/class/remoteproc/remoteproc0/state

it gives the same error as above

Fullscreen

1
2
3
4
5
root@sm2s-am6254:~# echo stop > /sys/class/remoteproc/remoteproc0/state 
[   62.605993] k3-m4-rproc 5000000.m4fss: k3_m4_rproc_stop: timedout waiting for rproc completion event
[   62.615312] remoteproc remoteproc0: can't stop rproc: -16
echo: write error: device or resource busy
root@sm2s-am6254:~# 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX

root@sm2s-am6254:~# echo stop > /sys/class/remoteproc/remoteproc0/state 
[   62.605993] k3-m4-rproc 5000000.m4fss: k3_m4_rproc_stop: timedout waiting for rproc completion event
[   62.615312] remoteproc remoteproc0: can't stop rproc: -16
echo: write error: device or resource busy
root@sm2s-am6254:~#

Thanks,

Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

Just to clarify my understanding,

With the TI EVM and SDK 9.0 Default Image, Deep Sleep is working.
With your custom hardware and SDK 9.0 Default Image, Deep Sleep hangs due to the remoteproc issue stated earlier.

Please correct me if my understanding is wrong.

Remoteproc is not a topic I am familiar with but the two errors point to the following files: remoteproc_core.c (can't stop rproc) & ti_k3_m4_remoteproc.c (timedout ...). Looks like the peripheral maybe busy with another process. Does your custom hardware have an M4 Core and is it being used by another device driver or process?

Thanks,

Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hi Anshu,

Yes for first point you are wright

With the TI EVM and SDK 9.0 Default Image, Deep Sleep is working.

For second point

With your custom hardware and SDK 9.0 Default Image, Deep Sleep hangs due to the remoteproc issue stated earlier.

For our custom board we have made changes in the Linux BSP it is not the Default image from TI.

Does your custom hardware have an M4 Core and is it being used by another device driver or process?

Yes our board has M4 Core and we have not made any changes reated to it in our BSP.

Thanks,

Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

Can you try using MCU Only mode? The documentation can be seen here: https://software-dl.ti.com/processor-sdk-linux/esd/AM62X/09_00_00_03/exports/docs/linux/Foundational_Components/Kernel/Kernel_Drivers/Power_Management/pm_low_power_modes.html#mcu-only-mode

I understand that its not within the scope of what you need, but it may give some insight on the M4's response in LPMs. Please share the logs of MCU Only Mode.

Thanks,

Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hi Anshu,

Thanks for your reply, i looked into MCU only mode, it looks like it needs MCU UART for waking device form sleep mode and in our device we have used MCU UART with other interface.

I tried MCU only mode commands mentioned, but device is not going into Sleep mode. I checked PMIC_LPM_EN pin it is high means device is not going into sleep mode.

Can you point the changes need to be done in Device Tree, or do we need to enable any drivers from defconfig for enabling Deep Sleep mode.

Thanks,

Nilesh

0 Anshu Madwesh over 1 year ago in reply to Nilesh Patil

TI__Genius 14725 points

Hi Nilesh,

The config file should support the PM Options. By default, the SDK should have this enabled. If there are doubts about if the all needed PM Options are not enabled, we can go deeper into that topic.

The relevant device tree changes are for wakeup sources, not necessarily enabling Deep Sleep.

Does the R5F Core perform any other applications other than run the DM Firmware? If so, there will be errors entering Deep Sleep. Here is the documentation for this in the Introduction section: https://software-dl.ti.com/mcu-plus-sdk/esd/AM62X/09_01_00_39/exports/docs/api_guide_am62x/DEVELOP_AND_DEBUG_DMR5.html

Best Regards,

Anshu

0 Nilesh Patil over 1 year ago in reply to Anshu Madwesh

Prodigy 60 points

Hi Anshu,

As you mentioned that the changes we did are for wakeup source, are there any changes need to be done for enabling Deep Sleep.

> Does the R5F Core perform any other applications other than run the DM Firmware?

We checked for this we see same firmware on our device as of the EVM.

We could not find the '-CONFIG_LPM_DM=y' in our source as mentioned in the link you provided. Can you help let us know if any specific source needed for R5F Core.

And also if any changes required to enable Deep sleep mode.

Thanks,

Nilesh