Question relating to EVGA X58 RAID option ROM mod

[dejanh]

I was thinking about removing some of the display code. I actually do intend to look into it, but I can almost promise you that my skills won't be good enough for anything to come of it.

Maybe we can collaborate on something...not sure though how useful it will be. Crucially, we lack any information as to what modules need to be addressed to break that barrier. We only know that some Phoenix-Award BIOS work fine (like Gigabyte) and others that are the same version of Phoenix-Award BIOS do not work at all (like EVGA and some Asus).

I wish I had AMI BIOS. So much easier to work with than Phoenix-Award.

Hmm, can't make much sense of what IDA is telling me, unfortunately.

Now, I wonder, what file in this BIOS is responsible for loading OROMs? Also, does anyone know an AMIBIOS X58 board that does work with one of the newer Matrix RAID OROMs like 9.x/10.x?

Yeah, IDA output is a mess, and like I said I have no clue and nobody seems to know which modules are actually responsible for invoking the OROMs. If we knew that we could at least look for the "problem" in the right place. I just sent a mod of an AMIBIOS to a guy with an ASRock X58 board that includes the newest Intel RAID OROM over on EVGA forums. I will PM him to see whether it was a success.

[Dansolo]

Asus is mostly using AMI and that's got the 64KB limit. I actually prefer working with Award by far.

[dejanh]

Asus is mostly using AMI and that's got the 64KB limit. I actually prefer working with Award by far.

I think it's hit and miss. The good thing is the AMIBIOS of the ASRock X58 I just mentioned in my edit of the previous post is that it shipped with the 74K OROM (I believe 8.9.x.x) instead of the common 64K OROM. I modded it to include the latest 10.1.x.x 84K OROM but considering that it was accepting 74K before I think it will accept 84K OROM as well without issues.

I will PM the guy to see what he says was the result of his attempt to flash the board.

[dejanh]

Asus is mostly using AMI and that's got the 64KB limit. I actually prefer working with Award by far.

I think it's hit and miss. The good thing is the AMIBIOS of the ASRock X58 I just mentioned in my edit of the previous post is that it shipped with the 74K OROM (I believe 8.9.x.x) instead of the common 64K OROM. I modded it to include the latest 10.1.x.x 84K OROM but considering that it was accepting 74K before I think it will accept 84K OROM as well without issues.

I will PM the guy to see what he says was the result of his attempt to flash the board.

Yeah, the ASRock X58 Extreme accepted the newest OROM without issues. He messaged me to say thanks and that everything is working perfectly. So that is now a confirmed AMIBIOS X58 board that accepts newest OROM.

Still no progress though on the EVGA X58 boards at all

[Dansolo]

It'd be good if we could find an Award BIOS that does work with > 64KB and has the same internal structure (as in files) as EVGA's. Gigabyte has a lot of different files so I don't feel comfortable swapping stuff over. Comparing Asrock and Asus may lead to using Asrock code to make Asus work.

[dejanh]

It'd be good if we could find an Award BIOS that does work with > 64KB and has the same internal structure (as in files) as EVGA's. Gigabyte has a lot of different files so I don't feel comfortable swapping stuff over. Comparing Asrock and Asus may lead to using Asrock code to make Asus work.

To be honest I do not think that directly swapping files will be possible in any circumstance. If you happen to swap over a component that controls the power on the board by accident you can completely brick the board (I have done this before by accident trying to cross flash R2G and R2E boards). That was the only time I did something like this and won't repeat the same mistake. Just trust me...is very risky.

I wonder if anyone would know which module is responsible for actually invoking the option ROMs to load. It would almost certainly have to be that specific BIN that limits the executable code to load at only 64K...

Oh and by the way, I did a bit more testing on this in the last few days...here is something else I observed (c&p from a post on EVGA forums, in response to the EVGA Product Manager Jacob commenting on the RAID issue; first his comment them my reply).

6. Intel RAID ROM Needs updating to latest ROM. If not possible because of BIOS size, replace bios chip for larger size and send to customers. BIOS size is the wall we have hit now, but we continue to look into alternate solutions.

I posted this in the other thread but will post it here as well...

I did a quick test by manipulating the 8.5.0.1030 RAID option ROM that is originally 64K in size to be 84K in size by padding it with FFh at the end to force it to be 84K in size (FFh has no functional impact on the ROM). In both cases I ensured that the checksum is 00h so that the option ROM remains valid. With the ROM padded, I produced two versions of it, one that states that the executable code size is 64K and the other 84K. I loaded first the 84K ROM with 64K executable code size (determined by 3rd byte in option ROM) and the option ROM loaded fine and worked fine. Then I again loaded the same ROM this time with the 3rd byte indicating that executable code size is 84K, flashed, and the option ROM would no longer work. So, to reiterate, this is the same exact code base padded to 84K in both tests, I just modified the size and the 3rd byte to indicate how much of the ROM is executable and in both cases the ROM substituted fine, flashed fine, but it only worked when I was addressing 64K, no more.

So, with that said, what kind of size limitation are we talking about here? It really looks to me like the BIOS simply cannot address a ROM larger than 64K. Can you please elaborate?

Edit: To add to this, if say theoretically we drop the PXE boot ROM that frees up 28K of compressed space and 50K of executable size (at run-time the PXE loads at 50K) we will gain 28K of compressed space and 50K of executable space. Difference between the 84K RAID ROM and 64K RAID ROM is 20K, 20K of executable code (only about 10K compressed). 20K executable size is < 50K executable size freed up by releasing the PXE ROM. It is even less than the 28K compressed size of PXE ROM. And you guys are still saying size is the issue...

I am really failing to see how size is an issue...

Oh and by the way, here is the link to the modded ASRock BIOS http://www.mediafire.com/?h57qvvdvluzo8l2

[Dansolo]

I'm glad I didn't go EVGA when I was looking at X58 boards.

[dejanh]

I'm glad I didn't go EVGA when I was looking at X58 boards.

Well, to be fair, E759 is still probably one of the best, if not the best board ever released for the X58 platform. It's just these couple of issues that are putting a damper on things. I know that there has to be a way to resolve it. I wonder if you know anyone who has a contact for a Gigabyte BIOS engineer that we could ask which module they had to update to handle larger OROM.

[Dansolo]

No board with 8.0 RAID is even in the running for best board.

[dejanh]

No board with 8.0 RAID is even in the running for best board.

There are other criteria you know ahhahahahaha

I totally agree though! Completely unacceptable, especially when 5-7 year old boards can run 10.0 RAID. What a joke