u-boot -> Linux prompt assistance

alfman at austin.rr.com alfman at austin.rr.com
Tue Aug 21 11:34:15 EST 2007


I have u-boot running and am currently working on the Linux prompt. I 
store uImage at ffd00000. u-boot is at fff00000. 

* The Freescale BSP for the MPC8360 was used for the base. u-boot 
1.1.6 and Linux 2.6.19 are the versions selected and reported.
* u-boot is running at the prompt and seems stable for u-boot 
functionality.
* Enabling the 64M of DDR2 memory was quite an endeavor but it appears 
stable today. Nothing shows up with the basic u-boot memory test, and 
there are no crashes using the normal u-boot functions. i-cache is 
enabled. Memory problems have not been fully ruled out. Any 
suggestions on how to validate the memory better?
* Linux (~2.6.19) is compiled with no changes at this point. 

Executing "bootm ffd00000" results in a transfer to Linux that (due to 
the CodeWarrior USB TAP?) I can follow the code up to the RFI in 
head_32.S. It looks like it is executing farther than that. I have 
attached a dump of the memory found at log_buf. Also attached is 
System.map. 

My interpretation of the stack dump suggests that start_here is 
successfully executing in high memory and the crash is within 
MMU_init. The crashes that follow would seem to be as much a result of 
this first problem.

Any suggestions or observations about this crash would be appreciated.

Alfman

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: System_map_trimmed.txt
URL: <http://lists.ozlabs.org/pipermail/linuxppc-dev/attachments/20070820/907d5ec2/attachment.txt>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: oops.log
Type: application/octet-stream
Size: 6222 bytes
Desc: not available
URL: <http://lists.ozlabs.org/pipermail/linuxppc-dev/attachments/20070820/907d5ec2/attachment.obj>


More information about the Linuxppc-dev mailing list