After long adventures blaming everything from HDD to next door's cat I now know I have an out of date BIOS and possibly even the wrong BIOS.

Trouble is... what have I got?

This is a 9 yr old Tiny and works on DSL altho not fast, have no trouble using ultimatebootcd or slackware installation as a live cd.

Stamped on the mainboard is MS6178 ver 1.1. I can't find that ver 1.1 anywhere.

ID string is 63-10b1-001169-00101111-071595-WHITNEY-1WHITNEY-H

I've ben looking in lots of places. Main one is http://global.msi.com.tw/index.php?f...3&maincat_no=1
and also

and
http://support.asus.com/technicaldoc...Language=en-us
(that's good for advice on using the flash utility) and http://www.wimsbios.com/faq.jsp is great for general BIOS stuff

That MS6178 wants an AWARD bios according to the website, latest 2.1, but I actually have AMI, v1.5. For the MS6183- I'm thinking the spec for the ver 1.1 may require the MS6183 bios??- latest is 1.7. There is also a MS6178E, but that has the same bios list AFAICT.

Checking against the ID string, it gets confusing. It's not listed where I checked http://www.wimsbios.com/faq/determinebiosupdate.jsp , closest reference I have so far is in http://www.wimsbios.com/phpBB2/topic8056.html.

2 questions really:
If anyone has experience with the BIOS flash tools... I tried the awdflash, in the "file name to program" I entered the name of the new BIOS, reply was "the program file's part number does not match with your system". I now know that I have to disable cache in BIOS & short across the lock terminals with jumper, adjacent to BIOS on mainboard. However it might be due to my identification problems. I think there's a way to override this but mainboards don't make very good ornaments...

And the main question- which BIOS should I try here?

(I have had loads of help on http://www.linuxquestions.org/questi...392/page2.html
and that shows where I have started to look at BIOS issues, even considered corebios. Any comments on this will be most helpful.)

If I end up exploding my mainboard, it's not my main computer anyway. But these boxes (I have several) can be quite good so I'd prefer to salvage this one if at all possible.

Thanks guys.

Can you explain what makes you certain that the BIOS is causing a problem? I ask this because once Linux is booted, the BIOS is no longer used. If a component on the board is being incorrectly configured pre-boot, it should be up to the kernel and/or respective driver module to make the accordant corrections. Perhaps a helping kernel argument is the proper remedy.
--- rod.

My last thread, http://www.linuxquestions.org/questi...events-649392/ contains the history of these investigations.

To summarize, trouble on both Debian and Slackware:
1 No trouble with installer, and either OS works initially. Then, after first few boots, there is a hang, "waiting for /dev to become fully populated..." (Debian) and "triggering udev events: /sbin/udevtrigger" (Slackware).
2 When the boot is finished (having reached udev timeout- and the previous thread contains link to my logs after "verbose" parameter in udev module, too long to post on LQ")- performance is very poor, with monitor showing CPU at almost full commitment. However, occasionally, either after a hardware change eg removing modem card, or for no reason, it just "settles"- the system works perfectly.

I really don't wish to ask anyone to go over old ground, hence starting a new thread on the subject of bios. So, the reason for suspecting the BIOS began when I was checking for web info on my mainboard. I found:

http://www.coreboot.org/News#2007.2F..._now_supported

The only corroboration I have for the idea that the MS6178 "never worked under Linux" (I don't have permission to get the article but this wording shows on Google: "MS6188 Linux") is that the comp was OK with W98 & XP.

In the previous thread, shadowsnipes (who provided me with valuable guidance throughout) was sceptical about the claims made by coreboot and encouraged me to look at MSI themselves, also believed that a BIOS update from the mainboard manufacturer was worth a try.

With nothing to lose, I could try the coreboot option, although would really appreciate further guidance before taking that risk.

Regards- Phil

Just to clarify, I wasn't skeptical about claims from coreboot. Rather, I had simply pointed out that their listed BIOS incompatibilities had nothing to do with your mainboard detection problems. The coreboot project provides replacement BIOS firmware. Your native BIOS firmware should not have those same incompatibilities.

As I said in the last thread, there shouldn't be a mainboard identity crisis. If you can't find the part number on there then comparing the specs on the MSI site to the actual hardware on the board should clue you in.

All the flash BIOS utilities I have used don't make me type in a name. They usually just make sure you actually want to flash the BIOS and then do it. Could you walk us through what it asks you and what you enter in as input? Please be detailed. You may just be able to accept any defaults it asks you.

Also the BIOS update README has this
So, you may not have to use your mainboard as an ornament even if you screw it up.

Edit: Apparently the link from the README is a rabbit hole.

A post on wimsbios has just provided 2 replies:
and
So that's the identifiaction problem. I think I see why awdflash refused to load the bios I had on the disk, it was detected as incompatible? I have uniflash now and expect I'll be able to provide the requested backup.

Going back to suggestion in http://www.linuxquestions.org/questi...events-649392/
That idea is looking much more attractive now, with any experiment using a non-specified bios including coreboot as a last resort

A post on wimsbios has just provided 2 replies:
and
So that's the identifiaction problem. I think I see why awdflash refused to load the bios I had on the disk, it was detected as incompatible? I have uniflash now and expect I'll be able to provide the requested backup.

Going back to suggestion in http://www.linuxquestions.org/questi...events-649392/
That idea is looking much more attractive now, with any experiment using a non-specified bios including coreboot as a last resort.