PROCESSOR-SDK-AM62X: AM62X: Power Management: Main GPIO as an I/O daisy chain wakeup source

Hi Pier,

Are you using a TI EVM or custom hardware?

Have you update the main_pmx node in the device tree with the change?

Thanks,

Anshu

Hi Anshu Madwesh,

I'm not using TI EVK, but GPIO1_30  is correctly working as interrupt as gpio-keys, iIt's just not working as wakeup source.

The gpio-keys node has been changed as follow

gpio-keys {
    compatible = "gpio-keys";
    status = "okay";
    pinctrl-names = "default";
    pinctrl-0 = <&main_wkgpio_pins_default>;
    switch007 {
        label = "WKGPIO0_07";
        linux,code = <KEY_WAKEUP>;
        interrupts-extended = <&main_gpio0 7 IRQ_TYPE_EDGE_RISING>,
            <&main_pmx0 0x01c>;
        interrupt-names = "irq", "wakeup";
    };
    switch110 {
        label = "WKGPIO1_10";
        linux,code = <KEY_WAKEUP>;
        interrupts-extended = <&main_gpio1 10 IRQ_TYPE_EDGE_RISING>,
            <&main_pmx0 0x1a0>;
        interrupt-names = "irq", "wakeup";
    };
    switch130 {
        label = "WKGPIO1_30";
        linux,code = <KEY_WAKEUP>;
        interrupts-extended = <&main_gpio1 30 IRQ_TYPE_EDGE_RISING>,
            <&main_pmx0 0x1f0>; /* pin 69 => J18.2 */
        interrupt-names = "irq", "wakeup";
    };
};

and the main_pmx has been changed as follow

main_wkgpio_pins_default: main-wkgpio-pins-default {
    pinctrl-single,pins = <
        AM62X_IOPAD(0x1a0, PIN_INPUT_PULLUP, 7) /* (E18) MCASP0_AXR0.GPIO1_10 */
        AM62X_IOPAD(0x01c, PIN_INPUT_PULLUP, 7) /* (J23) OSPI0_D4.GPIO0_7 */
        AM62X_IOPAD(0x01f0, PIN_INPUT_PULLUP, 7) /* (A18) EXT_REFCLK1.GPIO1_30 */
    >;
};

Can you please try the same on your side ?

Thanks

Best Regards

Pier

Hi Pier,

I will try it on my end and get back to you. Please ping this thread if you haven't received a response by Friday.

Thanks,

Anshu

Thanks Anshu Madwesh

Best Regards

Pier

Hi Anshu Madwesh,

any update on this ?

Can you reproduce the same behavior on your side ?

Thanks

Best Regards

Pier

Hi Pier,

Thank you for your patience on this.

I also tested this on my end and also noticed that GPIO1_30 does not wakeup.

After digging through the design specifications, it looks like the PAD EXT_REFCLK1 (GPIO1_30 when MUXMODE = 7) is not capable of being a wakeup source because its not connected to the IO daisy chain.

Since its not connected to the daisy chain, its not capable of being a wakeup source (also won't support state retention in sleep). This is likely caused by the pad used by various other peripherals as a source clock.

Thanks,

Anshu

Hi Anshu Madwesh,

thanks for confirming the behavior.

Given a specific pad, how can I know if it's connected or not to the IO daisy chain ?

Thanks

Best Regards

Pier

Thanks Anshu Madwesh !

Hi Pier,

This thread will be closed. Please create a new E2E thread for any follow ups.

Thanks,

Anshu