[U-Boot-Users] relocate_code problem
Hi all,
I try to get up U-Boot 1.1.1 on a "home brewed" board with MPC8280 and have problems with the relocation of code. It crashes after the call of in_ram. It seems that it calculates the address of in_ram correctly, but the copied code is somehow shifted. Is there anyone who can point me to the function that copies the code into RAM? Sorry, but the assembler code is not my favorite thing.
Thanks in advance, Oliver
---- some extra info:
symbol in_ram is linked to 0xfff03404 relocate_code is called: relocate_code( 03fbdf48, 03fbdf6c, 03fdf000) relocate_ram jumps to 0x03fe2404 // there is code, but not in_ram
In message 40EA7675.1087.8A1955@localhost you wrote:
I try to get up U-Boot 1.1.1 on a "home brewed" board with MPC8280 and have problems with the relocation of code. It crashes after the call of in_ram. It seems that it calculates the address of in_ram correctly, but the copied code is somehow shifted.
Maybe it just looks shifted or repeated?
RTFM - http://www.denx.de/twiki/bin/view/DULG/UBootCrashAfterRelocation
Is there anyone who can point me to the function that copies the code into RAM? Sorry, but the assembler code is not my favorite thing.
You don't need much knowledge about assembler code. "grep" will just do fine (hint: search for "relocate_code", and probably you want to searchin the relevant directory with CPU specific code, i. e. "cpu/mpc8260/").
Best regards,
Wolfgang Denk
From: Wolfgang Denk wd@denx.de Date sent: Tue, 06 Jul 2004 10:06:21 +0200
of in_ram correctly, but the copied code is somehow shifted.
Maybe it just looks shifted or repeated?
You are right. It was hard to accept the idea that the seen code was not within the RAM.
You don't need much knowledge about assembler code. "grep" will just do fine (hint: search for "relocate_code", and probably you want to searchin the relevant directory with CPU specific code, i. e. "cpu/mpc8260/").
It was not the problem to find relocate_code. I was not smart enough to recognize the nice little loop. Because the destination content did not change during relocate_code, I thought it was done elsewhere.
Best regards,
Wolfgang Denk
Thanks, Oliver Fuchs
In message 40EADF20.20424.AE0466@localhost you wrote:
of in_ram correctly, but the copied code is somehow shifted.
Maybe it just looks shifted or repeated?
You are right. It was hard to accept the idea that the seen code was not within the RAM.
Probably your SDRAM init sequence is broken.
It was not the problem to find relocate_code. I was not smart enough to recognize the nice little loop. Because the destination content did not change during relocate_code, I thought it was done elsewhere.
Well, check your RAM init code.
Best regards,
Wolfgang Denk
participants (2)
-
Oliver Fuchs -
Wolfgang Denk