Hallo there, I used my Pentium4 to build a LinuxBIOS image based on the jun30th snapshot, using FILO to boot directly into GRUB. Now, my problem is, that the Tyan is currently bricked, and yes, even a replacement bios from BIOSMAN did not do the trick I do not want to send the Mainboard into warranty just yet. I have my linuxbios.rom, and a brand new, shiny SST49LF080A chip to put the image into. However, since the Tyan is bricked, I cannot program with it's own flashing mechanism using the flashrom utility included in LinuxBios. I currently have a nice little Abit KV85 mainboard, which, in combination with uniflash and DOS is supported, however, uniflash cannot write the LinuxBIOS ELF image, and flashrom does not like this mainboard. Please advise on how to continue, ps. I have access to a TopMAX universal programmer. Thanks,
On Mon, Jul 03, 2006 at 06:54:37PM -0500, Arturo Mann wrote:
ps. I have access to a TopMAX universal programmer.
Big PS there. :)
TopMAX lists support for "Flash EPROM 256 Kbit to 32 Mbit 28Fxxx, 29Fxxx, 29Cxxx, 29BVxxx, 29LVxxx, 29Wxxx, 49Fxxx series(2.7, 3.3, 5.0, or 12 Volt)" so your LF may also be supported.
If you have a PLCC adapter for the programmer and the TopMAX software running on another system you're all set. If not, I guess your options are to build a DIP->PLCC adapter yourself or make flashrom work on your Abit board. The former is easy enough if TopMAX are communicative, the latter may be easy if Abit haven't done any tricks when connecting the flash chip to the chipset. (Some board makers connect the flash write signals to random GPIO pins and don't tell anyone which..)
//Peter
Yeah, I agree on the PS. :) I have access to the topMAX with another box, yes. <evil grin> :) However, do I need to take as a "bad sign" that UniFlash cannot verify the programmed LinuxBIOS ELF? Or is it just a sign that UniFlash cannot handle that sort of program? Oh, btw, the ABIT is PLCC it just seems to be unable to program with uniflash (or past 512KiB With the standard 1024KiB Tyan Flash)
On 7/3/06, Peter Stuge stuge-linuxbios@cdy.org wrote:
On Mon, Jul 03, 2006 at 06:54:37PM -0500, Arturo Mann wrote:
ps. I have access to a TopMAX universal programmer.
Big PS there. :)
TopMAX lists support for "Flash EPROM 256 Kbit to 32 Mbit 28Fxxx, 29Fxxx, 29Cxxx, 29BVxxx, 29LVxxx, 29Wxxx, 49Fxxx series(2.7, 3.3, 5.0, or 12 Volt)" so your LF may also be supported.
If you have a PLCC adapter for the programmer and the TopMAX software running on another system you're all set. If not, I guess your options are to build a DIP->PLCC adapter yourself or make flashrom work on your Abit board. The former is easy enough if TopMAX are communicative, the latter may be easy if Abit haven't done any tricks when connecting the flash chip to the chipset. (Some board makers connect the flash write signals to random GPIO pins and don't tell anyone which..)
//Peter
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios
On Mon, Jul 03, 2006 at 08:08:31PM -0500, Arturo Mann wrote:
Yeah, I agree on the PS. :) I have access to the topMAX with another box, yes. <evil grin> :)
And also a DIP->PLCC adapter so you can use it for the SST part? If you have one or can make one this is your best bet. I'd recommend using the standalone programmer if at all possible.
However, do I need to take as a "bad sign" that UniFlash cannot verify the programmed LinuxBIOS ELF? Or is it just a sign that UniFlash cannot handle that sort of program? Oh, btw, the ABIT is PLCC it just seems to be unable to program with uniflash (or past 512KiB With the standard 1024KiB Tyan Flash)
I'm not an expert on uniflash but it could have limits on flash size or require some particular checksumming algorithm to be used on the binary file and have a checksum in a special byte offset to consider the flash image valid, or maybe it checks this checksum on write verification so verify fails unless the checksum is correct, etc. Note that this is only speculation and wild guesses.
If uniflash or the Abit board is limited to 512KiB ROMs verification will (should) always fail - so maybe this is a matter of the chipset not decoding more than 512KiB to the LPC bus. This feels more likely than crazy, damaged code in uniflash.
//Peter
The problem is not quite the actual ABIT motherboard, remember, both the ABIT and Tyan use PLCC chips. The problem is that I think Uniflash is not quite meant to flash LinuxBios or ELF files, and it writes it, but every sector comes out as "BAD" in validation, and well, yeah, push comes to shove, the motherboard does not boot either with that program. I will flash with the topMAX and see if the program is good for the Mainboard, unless you lot can tell me about some sort of checksumming routine to make sure the linuxBios flash is actually a valid chunk of flash? BTW, is there anyone who has a VGA Bios for an ATi Rage 8mb XL (like on the Tyan), since my original bios is f****ed i can't get it out of there. :(
On 7/3/06, Peter Stuge stuge-linuxbios@cdy.org wrote:
On Mon, Jul 03, 2006 at 08:08:31PM -0500, Arturo Mann wrote:
Yeah, I agree on the PS. :) I have access to the topMAX with another box, yes. <evil grin> :)
And also a DIP->PLCC adapter so you can use it for the SST part? If you have one or can make one this is your best bet. I'd recommend using the standalone programmer if at all possible.
However, do I need to take as a "bad sign" that UniFlash cannot verify the programmed LinuxBIOS ELF? Or is it just a sign that UniFlash cannot handle that sort of program? Oh, btw, the ABIT is PLCC it just seems to be unable to program with uniflash (or past 512KiB With the standard 1024KiB Tyan Flash)
I'm not an expert on uniflash but it could have limits on flash size or require some particular checksumming algorithm to be used on the binary file and have a checksum in a special byte offset to consider the flash image valid, or maybe it checks this checksum on write verification so verify fails unless the checksum is correct, etc. Note that this is only speculation and wild guesses.
If uniflash or the Abit board is limited to 512KiB ROMs verification will (should) always fail - so maybe this is a matter of the chipset not decoding more than 512KiB to the LPC bus. This feels more likely than crazy, damaged code in uniflash.
//Peter
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios
-
Arturo Mann -
Peter Stuge