SK-AM62-SIP: Enabling UART5 from device tree compilation issue

Follow up. For the missing header file  #include <dt-bindings/pinctrl/k3.h> is this supposed to be generated by using the sysconfig tool? 

Hello Bin Lui, thank for looking into this. 

I removed the #includes I had and replaced them with the one you mentioned. Also, make linux-dtbs is helpful and faster than the steps I had. 

The following was built with make linux-dtbs

/dts-v1/;
/plugin/;

#include "k3-pinctrl.h"

&main_pmx0 {
    user_exp_uart5_pins_default: user-exp-uart5-pins-default {
        pinctrl-single,pins = <
            AM62X_IOPAD(0x01dc, PIN_OUTPUT, 1) /* E15 -> UART5_TXD */
            AM62X_IOPAD(0x01d8, PIN_INPUT,  1) /* C15 -> UART5_RXD */
        >;
    };
};

&main_uart5 {
    pinctrl-names = "default";
    pinctrl-0 = <&user_exp_uart5_pins_default>;
    status = "okay";
};

/*
&{/} {
    aliases { serial5 = &main_uart5; };
}
*/

/* power the 40-pin 3V3 rail 
&main_i2c1 {
    gpio@22 { /* TCA6424 at 0x22 *\/
        p05-hog {
            gpio-hog;
            gpios = <5 GPIO_ACTIVE_HIGH>; /* EXP_PS_3V3_EN *\/
            output-high;
            line-name = "EXP_PS_3V3_EN";
        };
    };
};
*/

I commented out the alias since that was producing syntax errors. 

After moving the .dtbo to dtb in boot/dtb/ti/ and adding ti to uEnv.txt Linux booted but I do not see the uart5 as active as shown here: 

root@sk-am62-sip:~# cat /proc/tty/driver/serial 
serinfo:1.0 driver revision:
0: uart:unknown port:00000000 irq:0
1: uart:unknown port:00000000 irq:0
2: uart:8250 mmio:0x02800000 irq:238 tx:23461 rx:153 RTS|DTR|DSR
3: uart:unknown port:00000000 irq:0
4: uart:unknown port:00000000 irq:0
5: uart:unknown port:00000000 irq:0
6: uart:unknown port:00000000 irq:0
7: uart:unknown port:00000000 irq:0
8: uart:unknown port:00000000 irq:0
9: uart:unknown port:00000000 irq:0
10: uart:unknown port:00000000 irq:0
11: uart:unknown port:00000000 irq:0

I believe it is not active because I did not assign an alias to uart5 as mentioned in this post 

https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1372383/sk-am62-how-to-enable-userexpension-uart?utm_source=chatgpt.com 

Yes so at this point it should just be a syntax issue. I just need to enable the alias section without a syntax error. Will update soon..

Hi Bin Liu

Update: I just need to find the correct path for serial5 alias

I compiled the following dtso

/dts-v1/;
/plugin/;


#include "k3-pinctrl.h"

&main_pmx0 {
    user_exp_uart5_pins_default: user-exp-uart5-pins-default {
        pinctrl-single,pins = <
            AM62X_IOPAD(0x01dc, PIN_OUTPUT, 1) /* E15 -> UART5_TXD */
            AM62X_IOPAD(0x01d8, PIN_INPUT,  1) /* C15 -> UART5_RXD */
        >;
    };
};

&main_uart5 {
    pinctrl-names = "default";
    pinctrl-0 = <&user_exp_uart5_pins_default>;
    status = "okay";
};


&{/} {
    aliases {
        /* UART-5 lives at /bus@f0000/serial@2850000 in k3-am62-main.dtsi */
        serial5 = "/bus@f0000/serial@2850000";
    };
};


/*
&{/} {
    aliases { serial5 = &main_uart5; 
    };
};
*/

The results show the path I used for the alias is wrong

  DTC     arch/arm64/boot/dts/ti/k3-am62x-sk-uart5.dtbo
arch/arm64/boot/dts/ti/k3-am62x-sk-uart5.dtso:26.9-47: Warning (alias_paths): /fragment@2/__overlay__/aliases:serial5: aliases property is not a valid node (/bus@f0000/serial@2850000)

Where is one able to find the correct path for this alias?

I see under the k3-am62* examples  

k3-am62a7-sk-ethernet-dco1.dtso

has the following snippet

&{/} {

    aliases {

        ethernet1 = "/bus@f0000/ethernet@8000000/ethernet-ports/port@2";

    };

};

From the example, I see we need to find the correct path to complete a correct alias. 

Is the path generated from sysconfig tool?

Thank you again for taking the time. 

Hi David,

there are multiple ways to find the path. I guess the easy one for you is to search for *main_uart5* on your board under /sys/firmware/devicetree/fdt/base/ with your uart5 overlay loaded. 

sorry I am out of office and don’t have access to my EVM and kernel code to tell you the exact path. 

Hi Bin,

My mistake, the path in the above dtso is correct. I was doing unrelated testing that caused the warning.

So to recap the following dtso worked for enabling uart5 on expansion connector J3.

/dts-v1/;
/plugin/;


#include "k3-pinctrl.h"

&main_pmx0 {
    user_exp_uart5_pins_default: user-exp-uart5-pins-default {
        pinctrl-single,pins = <
            AM62X_IOPAD(0x01dc, PIN_OUTPUT, 1) /* E15 -> UART5_TXD */
            AM62X_IOPAD(0x01d8, PIN_INPUT,  1) /* C15 -> UART5_RXD */
        >;
    };
};

&main_uart5 {
    pinctrl-names = "default";
    pinctrl-0 = <&user_exp_uart5_pins_default>;
    status = "okay";
};


&{/} {
    aliases {
        serial5 = "/bus@f0000/serial@2850000";
    };
};

here is the confirmation of the port 

cat /proc/tty/driver/serial
serinfo:1.0 driver revision:
0: uart:unknown port:00000000 irq:0
1: uart:unknown port:00000000 irq:0
2: uart:8250 mmio:0x02800000 irq:238 tx:69575 rx:1320 RTS|DTR|DSR
3: uart:unknown port:00000000 irq:0
4: uart:unknown port:00000000 irq:0
5: uart:8250 mmio:0x02850000 irq:240 tx:230 rx:230 RTS|DTR|DSR
6: uart:unknown port:00000000 irq:0
7: uart:unknown port:00000000 irq:0
8: uart:unknown port:00000000 irq:0
9: uart:unknown port:00000000 irq:0
10: uart:unknown port:00000000 irq:0
11: uart:unknown port:00000000 irq:0

I did find the path to confirm uart5 in the location you mentioned. 

/sys/firmware/devicetree/base/bus@f0000

we see serial@2850000

root@sk-am62-sip:/sys/firmware/devicetree/base/bus@f0000# ls
'#address-cells'            ecc@25010000                   phandle          spi@20110000
 audio-controller@2b00000   esm@420000                     phy@30110000     spi@20120000
 audio-controller@2b10000   ethernet@8000000               pinctrl@f4000    spinlock@2a000000
 audio-controller@2b20000   gpio@600000                    pruss@30040000   sram@70000000
 bootph-all                 gpio@601000                    pwm@23000000     system-controller@44043000
 bus@100000                 gpu@fd00000                    pwm@23010000     ticsi2rx@30102000
 bus@4000000                i2c@20000000                   pwm@23020000     timer@2400000
 bus@48000000               i2c@20010000                   pwm@23100000     timer@2410000
 bus@b00000                 i2c@20020000                   pwm@23110000     timer@2420000
 bus@fc00000                i2c@20030000                   pwm@23120000     timer@2430000
 can@20701000               interrupt-controller@1800000   ranges           timer@2440000
 compatible                 interrupt-controller@a00000    serial@2800000   timer@2450000
 counter@23200000           mailbox@29000000               serial@2810000   timer@2460000
 counter@23210000           mailbox@43600000               serial@2820000   timer@2470000
 counter@23220000           memory-controller@3b000000     serial@2830000   watchdog@e000000
 crc@30300000               mmc@fa00000                    serial@2840000   watchdog@e010000
 crypto@40900000            mmc@fa10000                    serial@2850000   watchdog@e020000
 dss@30200000               mmc@fa20000                    serial@2860000   watchdog@e030000
 dwc3-usb@f900000           mux-controller@a40000         '#size-cells'     watchdog@e0f0000
 dwc3-usb@f910000           name                           spi@20100000

Thank you again for taking the time to help with this issue!

Glad the issue is resolved. Thanks for sharing the details. 

Hi Bin Liu,

I still get the following warning but everything seems to work just fine as expected. 

I am using PSDK 10x on Armbian which was made on PSDK 11X. not sure if that is the cause. But in your experience, these warnings should not be there, correct? 

  DTC     arch/arm64/boot/dts/ti/k3-am62x-sk-uart5.dtbo
arch/arm64/boot/dts/ti/k3-am62x-sk-uart5.dtso:25.9-47: Warning (alias_paths): /fragment@2/__overlay__/aliases:serial5: aliases property is not a valid node (/bus@f0000/serial@2850000)

Hi, I am out of office for a week, please expect delayed response.

No worries. Might delete this follow up since the issue is resolved but was just curious.