BIOS Update Disaster Intel S2600CP2J

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

Lennong

Member
Jun 8, 2017
119
24
18
47
I assume it's *nix derivative code in the binary, as most embedded type of applications are these days.. Yes, it might be so. It seems a bit exhausted now to use the regular recovery process.. In this stage I would resort to actually desolder the eeprom chips and reprogram the in a eeprom writer, but I understand your situation there.
 

Son of Homer

Member
May 9, 2016
171
15
18
48
Thanks for all your help and patience. I will try one last flash stick tomorrow, and consider options after that.
 

Terry Kennedy

Well-Known Member
Jun 25, 2015
1,140
594
113
New York City
www.glaver.org
Thanks for your reply. I have been doing it on another computer. Out of an abundance of caution, I was doing a full format, not quick. You are saying the the recovery files are *nix? The recovery files are: iFlash32.efi, ipmi.efi, R02.06.0005Rec.cap, rml.rom, Startup.nsh.
Note that I have no experience with this board, but...

Those are UEFI files. I wouldn't expect the recovery mode to use UEFI, as it is very complex and requires most of a working BIOS already. I'd expect the recovery mode to use a single file, with a "magic" filename that the recovery looks for, on a FAT-formatted USB stick. If that's the case, the few blips you're getting on the USB stick are from the recovery mode looking at the stick and going "Nope, nothing for me here".
 
  • Like
Reactions: Son of Homer

Lennong

Member
Jun 8, 2017
119
24
18
47
Yes, my contention is that the BIOS is ready to do it's rescue but either the media (USB stick) or files are not found/rejected.
 
  • Like
Reactions: Son of Homer

Son of Homer

Member
May 9, 2016
171
15
18
48
What is strange is that when I attempt to boot, without the USB stick, it complains that recovery media is not present, but with the stick it does not complain, but goes nowhere. It seems like the recovery files are not good. The bios version that shows when attempting to boot is the new version I was trying to update to, and that new version is what also shows in the recovery files (R02.06.0005Rec.cap).
 

s2600gl

Member
Apr 1, 2016
33
11
8
50
Finland
A little late to the party but... Did you try the recovery with only the bare essentials on the motherboard ? E.g. the absolutely minimum setup to boot the motherboard (1 CPU / 1 stick of RAM, both tested to be good) and no PCIe cards or hard drives connected. BIOS updating / recovery can
be very finicky stuff so you'll want as little on / connected to your MB as possible.

Also did the diagnostic / status leds on the rear of the MB give you any indication at which point the boot process failed ? The MB flashes the leds in a sequence during boot-up and if it fails at some point you can use the leds to debug just where things went wrong.
 
  • Like
Reactions: Son of Homer

Son of Homer

Member
May 9, 2016
171
15
18
48
Thanks. I have not tried the minimalist approach, but I will give it a try when I can find some time next week, and also try to figure out the led light sequence.

Those are UEFI files. I wouldn't expect the recovery mode to use UEFI, as it is very complex and requires most of a working BIOS already. I'd expect the recovery mode to use a single file, with a "magic" filename that the recovery looks for, on a FAT-formatted USB stick. If that's the case, the few blips you're getting on the USB stick are from the recovery mode looking at the stick and going "Nope, nothing for me here".
Terry Kennedy and Lennong both flagged a likely problem with the recovery files which is a further concern.

After fumbling around with this last weekend and earlier in the week, I may have caused another problem when reinserting a RAM stick, with the little clip at the end of the RAM socket breaking loose. While it feels like it is properly inserted, and the clip does click in, I no longer get anything on the screen. I think the beep sequence and leds pattern has changed as well.

I have to focus for a few days on my growing work backlog, before diving back into this.
 

JoeSixpack

New Member
Jul 15, 2017
2
2
3
74
I attempted to do the update package (Sept 21, 2016) which includes System BIOS 02.06.0005, ME firmware 02.01.07.328, BMC firmware 01.27.9958, and FRUSDR 1.11. As each component was installed, it confirmed that it was successful.
I have the exact same problem going on for two months now. Next time be sure you read all of the release notes from your current BIOS version and upwards and don't do large jumps to the latest as sometimes you need minimum versions of components to avoid these kind of bricking issues. That was Intel's brilliant "solution" after a lot of back and forth. They're useless.

The only way out now seems to be by externally flashing the BIOS and/or BMC chips. The catch is you won't find the full raw firmware from Intel as they don't provide that for security reasons. Using the update firmware that Intel does provide, I've managed to successfully flash the ME part to the latest (so obviously that was never upgraded) but no change. Replacing the BIOS part did not work at all (no bootup at all) but I have to double check to make sure it was inserted properly as theres a lot of sections. I'm currently examining the BMC, but wasn't able to find any similarities between the original and the update for proper placement.

So can someone please provide the full raw S2600CP Rev F BIOS and BMC?
 
Last edited:

Lennong

Member
Jun 8, 2017
119
24
18
47
As for 'raw image' my experience is that the upgrade files comes in that shape, just some times packaged into a bigger package. I have with a hex-editor sucessfully compared resident bios with ones in update package and then 'cut' it in proper size and headers. After that it's a just a matter of burning the eeprom. Bios is simply a linear script.
 

JoeSixpack

New Member
Jul 15, 2017
2
2
3
74
As for 'raw image' my experience is that the upgrade files comes in that shape, just some times packaged into a bigger package. I have with a hex-editor sucessfully compared resident bios with ones in update package and then 'cut' it in proper size and headers. After that it's a just a matter of burning the eeprom. Bios is simply a linear script.
Could you give us the correct offset information? Because I tried pasting the Intel provided ME and BIOS into a bad copy read from the BIOS and while the ME seemed to update its version successfully, replacing the BIOS part resulted in no power on. It's not a trivial task to figure out where to put the BIOS because of the header junk that Intel adds on beyond the cap stuff. And the bad BIOS copy may be corrupted anyway. What would have helped is if the darn partition layout file was available.
 

Lennong

Member
Jun 8, 2017
119
24
18
47
Could you give us the correct offset information? Because I tried pasting the Intel provided ME and BIOS into a bad copy read from the BIOS and while the ME seemed to update its version successfully, replacing the BIOS part resulted in no power on. It's not a trivial task to figure out where to put the BIOS because of the header junk that Intel adds on beyond the cap stuff.
I have not done it on this specific board (or any Intel server board). I'm sorry. I should have been more clear.
 

DarkKnight

New Member
Jul 21, 2016
19
8
3
74
I really, *really* wish I had seen this flipping thread before I ran my BIOS update just now. Board is stuck at first screen showing BIOS version info. WTF Intel? No sanity checks for incompatible BIOS versions?!
 
Last edited:

DarkKnight

New Member
Jul 21, 2016
19
8
3
74
FYI, Good news. I unplugged all of my add-in cards, reset BIOS recovery AND ME force update jumpers to 1-2 positions (4 jumpers total) and the damn thing booted up! I'm sitting at a BIOS screen now.

Keep in mind, I cycled through several different 1-2/2-3 jumper settings and pulled the CMOS battery during this period.

System information confirms that BIOS is the 02.06.0006 version.

Also want to report that previously in IPMI SDR version info was garbled, like the data was corrupted. It wasn't regular Alphanumeric characters. Now everything reads correctly. I was even able to customize the SDR to finally get the fans to spin down.

Info in IPMI as below:

BIOS ID : SE5C600.86B.02.06.0006.032420170950
BMC FW Rev : 01.28.10603
Boot FW Rev : 01.17
SDR Package Version : SDR Package 1.11
Mgmt Engine (ME) FW Rev : 02.01.07.328

This is on a board with PN PBA G56797-510

Feeling like:

 
Last edited:
  • Like
Reactions: Son of Homer

Lennong

Member
Jun 8, 2017
119
24
18
47
@Son of Homer I guess you in US. I probably could save the board for you with my access to lab in university. Consider this an invitation to last resort. (do some calculations on shipping fees)
 
Last edited:
  • Like
Reactions: Son of Homer

Son of Homer

Member
May 9, 2016
171
15
18
48
Thank you, that is so amazingly kind. I have a couple of things to try before going to the last resort.
 

DarkKnight

New Member
Jul 21, 2016
19
8
3
74
Keep in mind that this board takes a long time to initialize. What looked like a hang on the screen was actually it cycling through silent POST. Make sure to watch the LEDs In different places on the board and give it at least 15 seconds per LED state to see if it's stuck. It took a long time each time it cycled through the memory to move on. Also, it seems to cycle through everything more than once.

Lastly, do like I did and I unplug all your addin cards. If it gets stuck on memory, unplug all but 2 modules, try again. The recovery jumper settings did make a difference. It seemed like the combination of these things was what I needed to get it to boot.
 
  • Like
Reactions: Son of Homer