On Thu, Mar 02, 2006 at 04:18:14PM +0000, M P wrote:

My V4 E3 boots using Mark's tools. There were a few issues when building :

• The buildroot "snapshots" fails to build, you need to explicitely

tell it not to use snapshots of uLibc and busybox. After than you still will get a hanging boot until you fix the inittab

• u-boot "examples" fails to build here, and break the build; I

commented them from the Makefile and had to make clean && make to get a working u-boot.bin

• The kernel config & patch as provided in the "v1" file works,

however the delta-20060227.diff generates a kernel that hangs just after decompression here, no way to make it work...

You need to patch u-boot to pass the right machine id if you've applied my delta-20060227.diff

Find the bit in board/E3/E3.c where it says "gd->bd->bi_arch_number =" and change the number after the =s to 862.

Looking forward to get that LCD + Console working :D

I'm trying to clean up my patch set and release an updated ready to go tarball that has this all sorted. Probably during the weekend.

Would it be easy to get the rootfs, u-boot and kernel in the flash ? we are fairly close to have a system that could be used to test stuff from a nfs partition or a usb flash drive; it would save a lot of time...

I have rootfs in flash at present. kernel in flash requires adding u-boot support for the flash, which probably isn't too hard. Getting u-boot into flash is trickier, but would be lovely. I'd hoped we could wrap it up in a LDR block so PBL would boot it, but I've found that writing my rootfs to the second 16MB of flash has stopped PBL completing a boot; it does its scan through the flash and stops onces it gets to the jffs2 bit.

I'm not sure if I can trick PBL into ignoring this, or if it's going to require replacing/patching PBL - I'd really rather not do this as if it goes wrong there's no easy way to fix it.

J.