Well, I think with the couple of tips and extra documentation (I didn't know about the JEDEC docs) I'll keep chugging away at this until the boss says to give it up. I would much rather use LinuxBIOS on our product than a proprietary bios.
I did try using the register values from a known BIOS. However, that didn't work as it isn't a static configuration. The docs do outline many registers in the memory controller, but I have notice that when I manipulate some undocumented ones to the values seen in a working bios I get further (although not much) in the boot, leaving me to believe not all the registers are documented :(.
Anyway I haven't given up yet, but at the same time, I won't be too surprised if this doesn't work out.
Jon
On 11/29/06, Drew Lundsten drew.lundsten@ccpu.com wrote:
This is depressing, definitely. You did try copying the register values seen in a booted board?
And I assume you do have the datasheet that's on Intel's public site? At least the registers are defined there, although it doesn't give you the sequence needed to use them optimally.
Drew Lundsten drew.lundsten@ccpu.com +1 858-882-8828/619-895-1764 cell Continuous Computing Corporation, 9380 Carroll Park Dr, San Diego 92121
-----Original Message----- From: linuxbios-bounces+drew.lundsten=ccpu.com@linuxbios.org [mailto:linuxbios-bounces+drew.lundsten=ccpu.com@linuxbios.org] On Behalf Of Jon Dufresne Sent: Wednesday, November 29, 2006 5:12 PM To: ron minnich Cc: linuxbios@linuxbios.org Subject: Re: [LinuxBIOS] Very confused about ram initialization
Well that is a bummer, I was having fun (accompanied by frustration) trying to get this to work. Thanks for all the help I received up to this point.
Jon
On 11/29/06, ron minnich rminnich@gmail.com wrote:
I hate to say it, but you are going to probably have to give up on that chipset absent docs. I have the "DOCS" as intel ships them from developer.intel.com, and, unlike what you could get in 1999, the current set of docs are totally worthless for writing a BIOS. I am told this is intentional. I have had it put to me that this policy simplifies EFI lockin -- no competition.
Sorry
ron
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios