Why Pinmux is exported to u-boot and the L4T Linux Kernel simultaneously?

Autonomous Machines Jetson & Embedded Systems Jetson Nano
1

Don’t know why some pins are initialized in uboot, and then in kernel.

In some case, uboot and kernel can use the same dts file, when the kernel boot, does those hardware initialized in uboot still work or not?

Thanks

2

I haven’t looked at the code in depth but I do know that u-boot does some part of the hardware initialization process. For the SPI enableing stuff that I did, it was absolutely necessary to disable the pins as GPIOs in u-boot or they would not be initialized correctly in the kernel. As far as I know, u-boot itself does nothing with the DTB other than take what it gets from cboot and pass it to the kernel. It’s not an ideal process to be sure and I believe the Nvidia folks are working on a better process for the next JetPack release.