Firstly, it drops the compiled FDT file into /boot/kernel on the UFS file system, and configures u-boot to pass the name of the DTB file to ubldr. This leverages some recent changes to ubldr and the fact that the Wandboard kernel no longer uses a statically linked FDT
Secondly, I was looking for a way to simplify the patch sets we apply to u-boot. I've added a function to uboot.sh that uses the u-boot tool "mkimage" to pass boot instructions to u-boot. This relieves us of the numerous patches that were applied to wandboard.h to configure the boot. In the new system, boot instructions in /files/boot.txt are compiled using mkimage to boot.scr. boot.scr is put on the FAT filesystem where u-boot finds it and executes the instructions. My hope is that this method of configuring u-boot can be used for other boards, and that it will make it easier to configure u-boot and to update to new versions.
This patchset does two things
Firstly, it drops the compiled FDT file into /boot/kernel on the UFS file system, and configures u-boot to pass the name of the DTB file to ubldr. This leverages some recent changes to ubldr and the fact that the Wandboard kernel no longer uses a statically linked FDT
Secondly, I was looking for a way to simplify the patch sets we apply to u-boot. I've added a function to uboot.sh that uses the u-boot tool "mkimage" to pass boot instructions to u-boot. This relieves us of the numerous patches that were applied to wandboard.h to configure the boot. In the new system, boot instructions in /files/boot.txt are compiled using mkimage to boot.scr. boot.scr is put on the FAT filesystem where u-boot finds it and executes the instructions. My hope is that this method of configuring u-boot can be used for other boards, and that it will make it easier to configure u-boot and to update to new versions.