jmp INSTRUCTION IN boot.c fails and resets again , what couldbethe reason.

[Mark Wilkinson]

Hi All,
	Further to my previous message, I've modified the raminit.inc file to setup 
the RAM banks as the Award BIOS configured them, and let the sizeram routing 
do it's stuff as it should.

The romimage, again reports the 127M, but this time jumps to the elf image 
correctly.

Does anybody know how to correctly setup the RAM bank registers so that these 
values are not hard coded ?

Ron,
	I've included a patch to modify the raminit.inc file (northbridge/via/vt8601) 
for 128M of ram.


Regards
Mark.


On Thursday 08 May 2003 20:41, you wrote:
> Hi All,
> 	I think a bug has been introduced somewhere in the Via vt8601 ram
> initialisation which is causing this symptom.
>
> I have a VIA Epia 5000 mainboard with 128M ram, and using a fresh checkout
> from the CVS (when sourceforge finally let me :-) I built a fresh romimage.
>
> While this linuxbios image correctly reported 127M of ram (I assume that 1M
> grabbed by the vga chip) I kept seeing the fail, reset problem, where
> linuxbios would reset before getting a chance to jump to the loaded elf
> image.
>
> After commenting out the bit in the northbridge sizeram routine which sets
> the correct memory size, and just let it return the good old 64M, not only
> does the Image boot and load correctly, but memtest86 (v2.9) loaded and
> started running.
>
> Does this give anyone any ideas as to the problem ?
>
> Regards,
> Mark.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 128Mram.patch
Type: text/x-diff
Size: 1150 bytes
Desc: not available
URL: <http://www.coreboot.org/pipermail/coreboot/attachments/20030508/b2316003/attachment.bin>