Why GPIO3_PI.01(NFC_INT) is busy?

Hi all,

My nfc_int_pi1 node is config as below,

nfc_en_pi0 {
                                nvidia,pins = "nfc_en_pi0";
                                nvidia,function = "rsvd0";
                                nvidia,pull = <0x0>;
                                nvidia,tristate = <0x0>;
                                nvidia,enable-input = <0x0>;
                        };

But when I use " root@linux:/home/sercomm# echo 64 > /sys/class/gpio/export" command on our customized board base on the Nano production moudle, it turns out failure.

root@linux:/home/sercomm# echo 64 > /sys/class/gpio/export
    bash: echo: write error: Device or resource busy

But with the same Image and dtb in our Nanno devKit, it is Okay.

What makes it happen?

hello garretzou,

since you’re develop a customized board, had you refer to Jetson Nano Product Design Guide for recommendations and guidelines.
you may also access Jetson Nano Pinmux to have customize and to update the board configuration files.
thanks

The Jetson Nano Pinmux 1.01 is not for R32.3.1.

Thanks

right, we’re now prepare the updated version of Nano pinmux spreadsheet.

Any solution to this yet? I am also facing the same issue (jetson nano + custom board, PI.0 is busy)

 gpio-64  (                    |i2c-mux-gpio        ) out hi

Hi shubhamp,

Please help to open a new topic for your issue. Thanks

Hi kayccc,

This issue is solved. My solution was to convert the dtb into dts, find out the node entry which used this property (“i2c-mux-gpio”), remove the entry and recompile the dts back into dtb. Some camera i2c driver was reading this node and since we do not have any camera hardware on the custom board, removing it made sense. Perhaps not the most elegant solution but worked for me.

Now I have a different problem, on the “production” module, the software is not able to drive PI.00 sufficiently low. Opening a new issue for the same.

1 Like