Bios Mods -The Best BIOS Update and Modification Source

Full Version: HP DL145 Opteron G3/G4 MicroCode Updates
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Pages: 1 2 3 4 5 6
System in question:

Seeking a customized bios that will enable support for quad and possibly six core processors.

Possibly enabling AMD-V as well (though not as important)

Willing to pay fwiw. Or give a nice donation to the site.

Believe its using Phoenix. SST SST49LF080A is the flashchip installed.

Sorry but as the HP BIOS Title states , this is a ROMPAQ bios - a special type of bios made by HP which we currently are unable to modify Sad

Its actually a Phoenix Bios. I have it opened up in it now in the Phoenix editor. I think they used Rompaq for a generic term for all their bioses. I did note the new ones from their other servers are actually a custom bios (as I was downloading them to see if I could grab the microcode) which cannot be opened with any of the bios editors I have tried.

Since that wont work trying to look at like a supermicro or other that have a phoenix bios with the same serverworks chipset and support for g3/g4 to again attempt grabbing the updated microcode. Though I honestly have no idea what Im doing alone.

I can upload the rom so you can see for yourself?
Ok the best I can find is this Tyan board.

Which even has support for Istanbul cores with its update. Same chipset. However its using AMI bios. I could not find (at least so far any other serverworks chipset boards that have a bios recent enough to see at least Shanghai cores). I see there is a way to extract the microcode and hex it into the Phoenix from some info on the Rebel Haven forums. Though again not a clue where to start.
OK , il look at this once more Smile
please upload the unmodified rom for me as i cant extract it!


In fact I came across that same Tyan MB and in the zip you can find it along with the HP.

I admit I have been working at another forum also. Don't want to link per bad etiquette. (Will PM you the links so you know where Im getting my info).

What I have been able to do (at least I thought)

-Extract the microcode section (11 P6 MicroCode) from the AMI.
-Hexed it into the Phoenix Update00 (Microcode).
-Made a new bios and I >thought< flashed with phlash16.

However, I see no difference shown on the build date of the bios in the posting header (I thought would be based on file versus a string in bios). And the Shanghai (Opteron 8376 HE) still wont post.

So this leads me to believe I either.
1. Need something more. Like Agesa. Agesa is in the Phoenix in the update00 section at least according to what I read elsewhere. In the AMI tyan however It sounds like it may be in a different place (not the id11 but b1 section). So in short while I might have updated the microcodes I may have not updated to the necessary Agesa.

2. The phflash16 did not actually flash. It just emulated it did. Or there is a jumper on the MB that locks the flash. I need to look in manual to make sure. One would think though that phflash16 would have not been able to even attempt flashing then?

3. Even with Tyan the updates are only for 2000 series processors since a 2way system. While again using a 8000 which can be used up to 8 way series since got a smoking deal. In theory should work. In practice if they require different cpu patches between the 2000 and 8000 beyond the core revision Tyan might have not bothered since normally a 8000 would cost a greater amount more than a 2000 version. If I knew how to interpret the CPU REV. ID numbers in the tyans with mmtool and compare it to the Opteron list I could see what it does and does not cover.

4. Tyan was blowing smoke. As the last update I see in the cpu patch section on mmtool is from December of 08 while the Shanghai series I'm using *HE* were released Jan 2009.

5. Sold Bad Chip. However no way to test.

Question: Are microcode updates chipset dependent as well? Like I'm thinking if its #3 by chance. Using a 4 way bios from another chipset with socket f to get its microcodes.
With the AMD platform, literally all the microcodes are the same and do not change from platform to platform (maybe a bit of a different layout between award, ami, and phoenix but they are all the same values)

Yes, it is possible that AGESA and or ACPITBL need to be updated. I would first check though to make sure the microcode entries really were updated (check windows timestamp)

from my experience with AMD CPU upgrades, the microcode is merely bug fixes for the newer stuff and doesn't help with detection, so an AGESA update should be considered.

I'll also start working on this Big Grin
Hmm. I thought it was being too easy.

Anyhow in fact the timestamp is newer and I checked the hex for the microcode and it matches the modded one I made.

I read somewhere else that AGESA is in the update00 section. However it is someplace else in the AMI bios. So in other words. I only updated the microcode and not the AGESA.

Is the AGESA also chipset dependant? It might be easier for me to find another Phoenix bios with a newer flash if I don't limit to serverwerks ht1000/ht2100 chipsets. As 1234 already discovered there seems to be only that Tyan with the same chipset with any newer support (well not true the new HP servers do as well but they now DO use a custom bios so Phoenix no longer works).
Ok new files for you all. Readme included. I have thrown all files and mods I have done in it. With the original and compiled firmwares thus far.

I was told elsewhere the update00 section should be chipset independent. Or at least the microcode and AGESA. With that in mind I found a Gigabyte Server that uses the Phoenix bios with support for 6 core and a Nvidia chipset (HP is Broadcom/Serverworks) . I used its update00 to replace the HP and compiled. Before attempting to flash I hope someone in the know can notify me if the update00 section can be replaced entirely if for same processor line without worries about chipset or other? Hex comparison showed the Gigabyte update00 being completely different than stock HP. Though again final compile was fine.

The Tyan if possible would be best since it uses 3.5 AGESA while this Gigabyte has only v. 3.17. However I have no idea of how to strip out of the ami tyan bios to insert into the HP.

Pages: 1 2 3 4 5 6