I installed Flashrom on an older single board computer and read the BIOS to a file which seemed to work fine.
I noticed that both chips were listed on the wiki with ?'s
Atmel AT29C020 256 Parallel ? ? ? ?
VIA VT82C686A/B 1106:0686 ?
I then tried to write that file to an identical SBC and got the following feedback:
I thought it was strange that the -wv failed to verify, but -v verified.
However after rebooting and checking the BIOS setup, nothing had changed.
Is there anything else I might try to get Flashrom to write the new BIOS?
# ./sbin/flashrom -wv NEWBIOS.bin
Calibrating delay loop... OK.
No coreboot table found.
Found chipset "VT82C686", enabling flash write... OK.
At29C020 found at physical address 0xfffc0000.
Flash part is At29C020 (256 KB).
Flash image seems to be a legacy BIOS. Disabling checks.
Programming page: 1023 at address: 0x0003ff00
Verifying flash... FAILED!
# ./sbin/flashrom -v NEWBIOS.bin
Calibrating delay loop... OK.
No coreboot table found.
Found chipset "VT82C686", enabling flash write... OK.
At29C020 found at physical address 0xfffc0000.
Flash part is At29C020 (256 KB).
Flash image seems to be a legacy BIOS. Disabling checks.
Verifying flash... VERIFIED.
Some background of system I'm using:
Phoenix-AwardBIOS v6.00PG 1984-2003
Platform is running Fedora version 9
The EPX-C3-G is a VIA C3-based, EPIC-compatible single board computer (SBC). C3 processors have extremely low power dissipation which allows fanless operation, making them ideal for industrial applications. The board is configured with a 733MHz or 1GHz MMXcompatible CPU with up to 512MB of PC133 SDRAM plus a CompactFlash socket. Also, a 10/100 Ethernet controller, USB 2.0, 4X AGP video with CRT/flat panel/LVDS interfaces, four serial COM channels, 24 digital I/O lines, AC97 audio, and the standard AT peripheral feature set are included.
The board measures 4.5 x 6.5-inches (115 mm x 165 mm) and is compliant with the EPIC (Embedded Platform for Industrial Computing) standard. It supports expansion options with PC/104 and PC/104-Plus modules or with two high-speed USB 2.0 channels. The board will operate from - 40° to +85°C for rugged applications requiring an embedded PC. Its x86-PC software compatibility assures a wide range of tools to aid in your application's program development and checkout.
Hi Jay,
thanks for your report. It seems the flashrom version you're using is ancient, and you might be hit by a few bugs which were present in versions before 0.9.0. Could you try flashrom 0.9.2 or even latest svn and send the output (as root) of flashrom -V lspci -nnvvvxxx superiotool -deV so we can take a look at your configuration and figure out what's going on? You can skip the superiotool step if superiotool is not installed and/or won't compile.
On 04.09.2010 00:16, Wells, Jay wrote:
I installed Flashrom on an older single board computer and read the BIOS to a file which seemed to work fine.
I noticed that both chips were listed on the wiki with ?'s
Atmel AT29C020 256 Parallel ? ? ? ? VIA VT82C686A/B 1106:0686 ?
I then tried to write that file to an identical SBC and got the following feedback:
I thought it was strange that the -wv failed to verify, but -v verified.
Absolutely, but this might be an artifact of the old flashrom version which didn't have the paranoid checks modern flashrom has.
However after rebooting and checking the BIOS setup, nothing had changed.
Is there anything else I might try to get Flashrom to write the new BIOS?
Our goal is to have flashrom run fine everywhere, and reports like yours are crucial for this. Once we have the requested info from you, we can guide you further and/or directly give you a fix.
# ./sbin/flashrom -wv NEWBIOS.bin
Calibrating delay loop... OK.
No coreboot table found.
Found chipset "VT82C686", enabling flash write... OK. At29C020 found at physical address 0xfffc0000. Flash part is At29C020 (256 KB). Flash image seems to be a legacy BIOS. Disabling checks. Programming page: 1023 at address: 0x0003ff00 Verifying flash... FAILED!
# ./sbin/flashrom -v NEWBIOS.bin Calibrating delay loop... OK. No coreboot table found. Found chipset "VT82C686", enabling flash write... OK. At29C020 found at physical address 0xfffc0000. Flash part is At29C020 (256 KB). Flash image seems to be a legacy BIOS. Disabling checks. Verifying flash... VERIFIED.
Yes, this is very odd. I definitely want to know what happened there, and if this also happens with latest flashrom.
Some background of system I'm using:
Phoenix-AwardBIOS v6.00PG 1984-2003
Platform is running Fedora version 9
The EPX-C3-G is a VIA C3-based [...] compliant with the EPIC (Embedded Platform for Industrial Computing) standard.
Regards, Carl-Daniel
Hi Carl-Daniel,
Here is the output you asked for. Actually, I did install the 0.9.2 version on the compact flash for the system we are building. I used the Live CD though to get these files since all I loaded on the CF was the flashrom application anyway. I already had the Live CD so it was probably even less trouble. I'm also attaching the datasheet for the SBC.
Thanks, Jay
-----Original Message----- From: Carl-Daniel Hailfinger [mailto:c-d.hailfinger.devel.2006@gmx.net] Sent: Friday, September 03, 2010 5:42 PM To: Wells, Jay Cc: flashrom@flashrom.org Subject: Re: [flashrom] BIOS doesn't change
Hi Jay,
thanks for your report. It seems the flashrom version you're using is ancient, and you might be hit by a few bugs which were present in versions before 0.9.0. Could you try flashrom 0.9.2 or even latest svn and send the output (as root) of flashrom -V lspci -nnvvvxxx superiotool -deV so we can take a look at your configuration and figure out what's going on? You can skip the superiotool step if superiotool is not installed and/or won't compile.
On 04.09.2010 00:16, Wells, Jay wrote:
I installed Flashrom on an older single board computer and read the
BIOS to a file which seemed to work fine.
I noticed that both chips were listed on the wiki with ?'s
Atmel AT29C020 256 Parallel ? ? ? ? VIA VT82C686A/B 1106:0686 ?
I then tried to write that file to an identical SBC and got the
following feedback:
I thought it was strange that the -wv failed to verify, but -v
verified.
Absolutely, but this might be an artifact of the old flashrom version which didn't have the paranoid checks modern flashrom has.
However after rebooting and checking the BIOS setup, nothing had
changed.
Is there anything else I might try to get Flashrom to write the new
BIOS?
Our goal is to have flashrom run fine everywhere, and reports like yours are crucial for this. Once we have the requested info from you, we can guide you further and/or directly give you a fix.
# ./sbin/flashrom -wv NEWBIOS.bin
Calibrating delay loop... OK.
No coreboot table found.
Found chipset "VT82C686", enabling flash write... OK. At29C020 found at physical address 0xfffc0000. Flash part is At29C020 (256 KB). Flash image seems to be a legacy BIOS. Disabling checks. Programming page: 1023 at address: 0x0003ff00 Verifying flash... FAILED!
# ./sbin/flashrom -v NEWBIOS.bin Calibrating delay loop... OK. No coreboot table found. Found chipset "VT82C686", enabling flash write... OK. At29C020 found at physical address 0xfffc0000. Flash part is At29C020 (256 KB). Flash image seems to be a legacy BIOS. Disabling checks. Verifying flash... VERIFIED.
Yes, this is very odd. I definitely want to know what happened there, and if this also happens with latest flashrom.
Some background of system I'm using:
Phoenix-AwardBIOS v6.00PG 1984-2003
Platform is running Fedora version 9
The EPX-C3-G is a VIA C3-based [...] compliant with the EPIC (Embedded
Platform for Industrial Computing) standard.
Regards, Carl-Daniel
Hi Carl-Daniel,
I sent these files as .rtf some time ago, but I haven't seen any reply that they were received or other feedback in reference to our issue. I had included a PDF datasheet and wonder if the size of the attachments created a problem for the email. Judging from all the other activity on the message board, it sounds like you are quite busy. So, I am sending only the data files as .txt again. I had installed what I thought to be 0.9.2 on the compact flash for this system, but for expediency I used the latest version of Flashrom Live CD to get these files. Please let me know if you get the data ok.
Jay T. Wells Automation Test Engineer IV TransCore - AmTech Technology Center
-----Original Message----- From: Carl-Daniel Hailfinger [mailto:c-d.hailfinger.devel.2006@gmx.net] Sent: Friday, September 03, 2010 5:42 PM To: Wells, Jay Cc: flashrom@flashrom.org Subject: Re: [flashrom] BIOS doesn't change
Hi Jay,
thanks for your report. It seems the flashrom version you're using is ancient, and you might be hit by a few bugs which were present in versions before 0.9.0. Could you try flashrom 0.9.2 or even latest svn and send the output (as root) of flashrom -V lspci -nnvvvxxx superiotool -deV so we can take a look at your configuration and figure out what's going on? You can skip the superiotool step if superiotool is not installed and/or won't compile.
On 04.09.2010 00:16, Wells, Jay wrote:
I installed Flashrom on an older single board computer and read the
BIOS to a file which seemed to work fine.
I noticed that both chips were listed on the wiki with ?'s
Atmel AT29C020 256 Parallel ? ? ? ? VIA VT82C686A/B 1106:0686 ?
I then tried to write that file to an identical SBC and got the
following feedback:
I thought it was strange that the -wv failed to verify, but -v
verified.
Absolutely, but this might be an artifact of the old flashrom version which didn't have the paranoid checks modern flashrom has.
However after rebooting and checking the BIOS setup, nothing had
changed.
Is there anything else I might try to get Flashrom to write the new
BIOS?
Our goal is to have flashrom run fine everywhere, and reports like yours are crucial for this. Once we have the requested info from you, we can guide you further and/or directly give you a fix.
# ./sbin/flashrom -wv NEWBIOS.bin
Calibrating delay loop... OK.
No coreboot table found.
Found chipset "VT82C686", enabling flash write... OK. At29C020 found at physical address 0xfffc0000. Flash part is At29C020 (256 KB). Flash image seems to be a legacy BIOS. Disabling checks. Programming page: 1023 at address: 0x0003ff00 Verifying flash... FAILED!
# ./sbin/flashrom -v NEWBIOS.bin Calibrating delay loop... OK. No coreboot table found. Found chipset "VT82C686", enabling flash write... OK. At29C020 found at physical address 0xfffc0000. Flash part is At29C020 (256 KB). Flash image seems to be a legacy BIOS. Disabling checks. Verifying flash... VERIFIED.
Yes, this is very odd. I definitely want to know what happened there, and if this also happens with latest flashrom.
Some background of system I'm using:
Phoenix-AwardBIOS v6.00PG 1984-2003
Platform is running Fedora version 9
The EPX-C3-G is a VIA C3-based [...] compliant with the EPIC (Embedded
Platform for Industrial Computing) standard.
Regards, Carl-Daniel
-
Carl-Daniel Hailfinger -
Wells, Jay