[U-Boot-Users] Re: Problem building U-boot for RpxLite with Mpc823e
Ccool
ccool at ninjatech.biz
Sat Nov 27 20:59:32 CET 2004
Thanks for the kind reply,
In this email, I tried to be a lot more specific. I tried to explain more
correctly the modification I have made. I'm sending this mail with 3
questions in it. The first one is about "saveenv" in U-boot. The second one
is about the silicon bug warning. And my last one is about mkimage.
I thank you for taking the time to answer my many questions.
First, here is the details about my board:
Rpxlite cw from embedded planet
MPC823 processor
16 Mb of RAM (0000 0000 - 00FF FFFF)
16 Mb of Flash (FF00 0000 - FFFF FFFF)
--> it is actually 4 AMD chips of 4 Mb each
128 Kb of NVRAM (FA00 0000 - FA02 0000)
The boot process, on the CW boards, start looking for a boot loader at the
adresses FFF0 0000 on the flash memory, so that's where I have put U-boot.
Now, in my last mail, I could not write into the flash with "saveenv". I
changed some of the configuration files, but as I'm a newbee, I kind of just
went into the files and did some random changes from the standard files in
the cvs until it worked. All my changes were based on what Sam Song has done
for the RPXlite_dw. My board seems to be somewhere in the middle of the
RPXlite and RPXlite_dw.
What I did, to make u-boot work for my board, I copied flash.c
from /board/RPXlite_dw to /board/RPXlite
I also modified slightly include/configs/RPXlite.h
in include/configs/RPXlite.h
43c43
< #define CONFIG_MPC850 1 /* This is a MPC850 CPU */
---
> #define CONFIG_MPC823 1 /* This is a MPC823 CPU */
123c123
< #define CFG_FLASH_BASE 0xFFC00000
---
> #define CFG_FLASH_BASE 0xFF000000
145c145
< #define CFG_MAX_FLASH_SECT 19 /* max number of sectors on one chip
*/
---
> #define CFG_MAX_FLASH_SECT 71 /* max number of sectors on one chip
*/
Q1: Also I really think the best thing would be to have the same kind of
configuration option like the ones you can find for the RPXlite_DW. I like
the fact that you just add the specific options for your board in the "make
RPXlite_XXX_config." I also like a lot the idea of using the NVRAM to save
the environment. I'm not certain what I would have to modify to use this on
my board?
Q2: Has I said in my last mail, I still have a "silicon bug warning"
CPU: PPC823ZTnnA at 48 MHz: 2 kB I-Cache 1 kB D-Cache
*** Warning: CPU Core has Silicon Bugs -- Check the Errata ***
I would like to know if that is a normal warning or it could really mess up
the stability of my system. When I built my linux kernel, I add the silicon
bugs patch. Is there a way in U-boot to apply a patch similar ?
Q3: This question is about the tool mkimage. It is the problem which is
affecting me the most. I've read the FAQ and my bug is similar to one of the
question there:
13.2.6. Linux hangs after uncompressing the kernel
I know it is not the "clocks_in_mhz" problem, because I've tested plenty. (I
removed it and try to change the 1 for the correct value of the speed of the
processor, just in case it would work)
I know my linux kernel work, because I have tested it before and it was
booting correctly. ( I tested it with the original boot loader of the
board )
So then, if I refer to the FAQ, the problem should be a bad definition of the
bd_info structure. All I found was:
in u-boot/include/asm-ppc/u-boot.h
unsigned long bi_immr_base; /* base of IMMR register */
in linux-2.4.22/include/asm-ppc/ppcboot.h
unsigned long bi_immr_base; /* base of IMMR register */
And I don't know what is the IMAP_ADDR, I don't know what to do. If you can
help me with only one question, please make it this one. I've been working
on this kernel for a long time. I tried the linux-2.4.25 from your site,
(with the mkimage tool included and working with mkimage), it's a great
kernel, but the mtd is not working at all, and it is working with my kernel.
I don't want to re-do all the work just for a little problem like this one
with "mkimage"
Thanks
Raphaël
More information about the U-Boot
mailing list