I am trying to understand a complication with using a custom device tree blob while using the prebuilt Pulsar image for microZed (http://microzed.org/sites/default/files/design/Pulsar-SoM-Image-Creation_0.zip).
I do not know too much about U-boot, but I have added my own .dtb file in the OVERCBOOT partition and I stop u-boot to add my own commands (build from petalinux tools 2015.2). Taking a look at the getting started guide (http://zedboard.org/sites/default/files/documentations/Wind-River-Getting-Started-with-Pulsar-Linux-OS_2.pdf), i can boot using the commands shown there using the provided blob
> fatload mmc 0 0x3000000 uImage
> fatload mmc 0 0x2A00000 dtb
> setenv bootargs 'console=ttyPS0,115200 root=/dev/mmcblk0p3
rw rootfstype=btrfs softdog.nowayout=0 rootwait'
> bootm 0x3000000 - 0x2a00000
If I change dtb to my own system.dtb, the boot seems to always hang after uncompressing the image.
Ultimately, I would like to add my own seperate .bin file for the PL during u-boot, as well (can do this, and will boot if I use the default dtb file).
I would like to solve this as it seems pretty much the easiest way to get rolling with the Pulsar persistent image. I would like to use my own device tree as I have created my own DMA drivers.
