Hi Guys,

I'm somewhat confused here..

We're booting QNX using UBoot from a paged device (NAND flash).
This means that I have to integrate parts of the IPL (image scan, setup, start) into UBoot.

Now, what exactly is the idea of the inner workings of this IPL-and-Image-Header-concept?
From this page
http://qnx.com/developers/docs/6.3.2/neutrino/building/load_process.html
it seems that all I need to do is call the according functions - but this will not work out due to the paged nature of 
flash.
Another IPL I have as an example (of ppc8360) in contrast seems to do much more than that,
putting IMMR and Clock settings into the header and copying the image manually, not using the setup function...

For now, I just copy the whole image to address 0x200000 in RAM and jump there, just as would happen when i download it 
via UBoot/tFTP.
But like that compression won't work, and it does not seem the "official" way to do it. Also I don't set IMMR, clock etc
, and this has to be good for something, right?

So:
- do I actually need to modify the header, or is the header designed to be generated at build time and then read at boot
 time?
- who copies the image from the paged device to the RAM? It seems the image needs to be placed at a fixed location...
- In case compression is used, it seems the image needs to be copied twice: one time from flash to RAM, then 
decompressed from RAM to another location in RAM

What would be the proper way to do all of this?

Greetings,
 Marc