[LinuxBIOS] flashrom: Doesn't seem to be able to write on ICH-2 (maybe other ICH-x) mainboards
corey.osgood at gmail.com
Sun Jun 24 05:33:22 CEST 2007
mmilburn at nmt.edu wrote:
> Hello Uwe,
> I have recently tried to use flashrom on two different ICH-6 based
> motherboards. As described in your original posting, detection and
> reading of the chips works. Each of the boards I've worked with (ASUS
> P5GD2-X, ASUS P5GD1-VM) use a winbond W83627EHF chip to communicate with
> the flash chip.
> I have not attempted to use uniflash on these boards for two reasons:
> A) the winbond chips supported (by uniflash) are not register compatible
> and do not use exactly the same communication scheme as the
> W83627EHF/W83627EHG chips.
This does not necessarily mean uniflash won't work. I'd try it before
discounting it entirely (uniflash is magic ;-) ). And if worst comes to
worst, asus's flash utility should work as well.
> B) the chipset enabling code seems to perform the same operation with the
> exception of not poking at the firmware hub decode enable register
> (uniflash seems to do this, though my pascal is shaky).
> I've been poking at the winbond chip and though I've sucessfully gotten it
> to hide the flash chip, I have yet to figure out what bit I'm forgetting
> to flip to allow writes to ROM.
> In reference to:
I'll have a peak at the ICH issues in the next few days, although I only
have ICH and ICH7 to test on (and I see you're using ICH6/6R). However,
it wouldn't surprise me to find out that the actual cause of your issues
is a lock put in place by the factory BIOS, Asus does this on a lot of
their boards, but uniflash knows how to work around it. Luc Verhaegen's
done quite a bit of work working around these with flashrom (see the
board specific fixups and this email:
http://article.gmane.org/gmane.linux.bios/24184 ), but he's pretty tied
up at the moment.
More information about the coreboot