[context: The story from the IRC channel is like this: Jordan Farrell did a BIOS upgrade where flashrom showed success, entered Setup after reboot, and the system behaved strange after the following reset: It turned itself off after some seconds. After that, he was back to the old BIOS version.]
Am Samstag, den 06.02.2010, 00:51 +0100 schrieb Michael Karcher:
Am Freitag, den 05.02.2010, 22:42 +0100 schrieb Jordan Farrell: I did not commit this patch as I did not yet get around to find out what mechanism undoes the successfull flashing you observed.
Now I took apart the whole BIOS image for your board, and I was unable to find any reflash code in it. What I did find was code that rewrites blocks 0 and 8 of your BIOS. Block zero contains ESCD stuff and I didn't work out what's exactly in block 8, but I the BIOS code itself is stored in blocks 1 to 7, CPU microcode updates are stored in blocks 12-14, and BIOS initialization code in blocks 14 and 15. Blocks 9 to 11 are unused.
If your BIOS chip (the 32 pin PLCC one) really is the chip detected by flashrom and not a bigger chip that masquerades as the smaller one and contains the reflash code in the hidden part of it, I have no idea which code could reflash it, where the old ROM contents might be stored and how to prevent that. The standard AWDFLASH utility provided with your board does not seem to contain board-specific code, as this type of code is put to the BIOS instead.
Could you please try again and double-check that the image you write is really the image you wanted to write, and that this image you want to write really is a newer version of the BIOS than the one you currently have in your flash chip?
Regards, Michael Karcher
-
Michael Karcher