I just got some new information. When I did the -Ewv earlier, you can see that it verified. However, a -v right after it would fail. Also reading the bios image off of the chip, I get a different file than the image.
I retried with just -Ew and a manual -v afterward. That failed.
Then I found the good news. I did separate -E, -w, and -v runs. That consistently works. Reading the image off the chip consistently match the image file copied to the chip. Is it possible that the delay between the erase, write, and verify cycles are too short when run as one command?
Thanks, wt
On 2009-06-07, Warren Turkal wt@penguintechs.org wrote:
I have good news. I took a huge chance and flashed my Gigabyte EX58-UD4P[1] bios with flashrom. I used the Gigabyte bios, which seems to be an LHA archive, strangely enough. I used the tip of the SVN.
wt@mediakitchen:~/projects/coreboot/coreboot-v3/util/flashrom$ svn info Path: . URL: svn://coreboot.org/flashrom/trunk Repository Root: svn://coreboot.org/flashrom Repository UUID: 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1 Revision: 580 Node Kind: directory Schedule: normal Last Changed Author: hailfinger Last Changed Rev: 580 Last Changed Date: 2009-06-05 13:53:07 -0700 (Fri, 05 Jun 2009)
I had a successful flash, but flashrom didn't seem to be friends with my chips.
I attached the output of flashrom -V as flashrom_V.txt.
Here's the output of flashrom -wv ex58ud4p.f8:
root@mediakitchen:~/test# ../projects/coreboot/coreboot-v3/util/flashrom/flashrom -wv ex58ud4p.f8 flashrom v0.9.0-r580 No coreboot table found. Found chipset "Intel ICH10R", enabling flash write... OK. Calibrating delay loop... OK. Found chip "SST SST25VF080B" (1024 KB) at physical address 0xfff00000. Flash image seems to be a legacy BIOS. Disabling checks. Verifying flash... FAILED at 0x000ec000! Expected=0xff, Read=0x45
A second time yielded the same FAILED. However, this worked.
root@mediakitchen:~/test# ../projects/coreboot/coreboot-v3/util/flashrom/flashrom -Ewv ex58ud4p.f8 flashrom v0.9.0-r580 No coreboot table found. Found chipset "Intel ICH10R", enabling flash write... OK. Calibrating delay loop... OK. Found chip "SST SST25VF080B" (1024 KB) at physical address 0xfff00000. Erasing flash chip... SUCCESS. Verifying flash... VERIFIED.
The difference is that I manually told it to erase the flash chip. The system boots properly with the new Gigabyte bios.
BTW, this mobo has the Gigabyte Dual Bios technology. AFAICT, I only flashed the primary bios, but it definitely works.
I also found some docs that may be useful in getting the Intel X58 chipset support at [2]. Please let me know if these are detailed enough to do anything useful.
Thanks for working so hard to make a open source bios a reality.
[1]http://www.gigabyte.us/Products/Motherboard/Products_Spec.aspx?ClassValue=Mo... [2]http://www.intel.com/products/desktop/chipsets/x58/x58-technicaldocuments.ht...
Thanks, wt