Thanks for all that Plutomaniac, hopefully we can now find out if it’s in the ME or not, hopefully not! And thanks about the microcode info too
Do any of you guys with these CPU’s know of a similar working Asus model for these CPU’s that I can pull modules from for inspection or swap testing?
@malucul - thanks for testing and report back! From this test, and what plutomaniac says, I bet this is either ME or within the BIOS itself, which now once you can test his files, we’ll know if it’s in the ME or not.
Thanks @plutomaniac - I’ve given these a shot. Here are the results:
1. WS-C621E-SAGE-ASUS-0303_HAP - No change, same issue as before.
2. WS-C621E-SAGE-ASUS-0303_MOD - Will not power-up.
Interestingly, the contents of the FLASH changes after power up with the HAP image. It may have been doing this before, but I only just noticed it.
Here’s a link to the modified version: https://www.dropbox.com/s/959t19wdba0oz7…hanged.zip?dl=0
Great, thanks for testing @malucul - so now we know it’s not in the ME. I will dig around more in BIOS and see if I can find restriction or CPU allow list etc.
Alright, it seems that disabling the CSSPS did not yield any change. So, as LOST said, you need to focus your efforts in the BIOS region where the restriction is probably placed by ASUS. The changes you noticed are related to some PCH Soft Strap values which were made on the fly when the system attempted to boot and I can see that they are the same at your previous dumps and thus not related to the recent CSSPS HAP modification. When/if you guys can get a working dump, as explained at the first paragraph of post #120, do tag me.
Plutomaniac - hopefully soon Maxdata will be able to provide you a dump, he has working CPU’s and a programmer, but not correct cable yet.
@Mov_AX_0xDEAD @CodeRush @DeathBringer - hope you guys can spare a second, thanks! Can any of you tell if this is what is blocking 50654 ES/QS CPUID? Or do you know from past experience what other modules I need to be looking at? Thanks!
I also find 3D 50 06 05 in a few other modules too (CPUMpDXE, a Raw and padding section, MemCacheInit, and VTF) , so probably need to edit those if this is it.
This is from PlatformCPUpolicy module, here is IDA ready copy if you need to look deeper - http://s000.tinyupload.com/index.php?fil…482887293448936
I could be talking absolute nonsense here, but it would seem to me that these bits of code would only catch cases where the CPUID did not start with 5065x, which (based on my understanding of the CPUID on ES parts) would never be true with the devices we have, or on the retail parts. The second screenshot does not show where r11d is populated from, but I presume it gets the same bitmask as the other screenshot, 0xFFFF FFF0, which would mask the final byte in the CPUID value.
Again, please correct me if I am wrong! My x86 assembly knowledge is very rusty.
My assembly knowledge is extremely limited @malucul - so you may be 100% correct. It caught my eye due to being 5065x and splitting one way or the other, and 5065/0/1/2/4 are all in the compatible list/microcodes, so I thought maybe this is go left/red if 50653 or 50654 ES/QS etc (Whatever is blocked)
I’m not sure how to check what you mentioned about r11d, text view looks the same basically, and I couldn’t figure out anything much else to do there See, I told you I know squat on assembly . Do you have any guesses with your knowledge on what I should try to look for?
I found these by searching 50 06 05, no results in this file for 54 06 05, and only two results overall in the BIOS for that and it’s a USB module so not what were looking for I’m sure.
Would a more extended look at either of those images give you a better view of what’s going on? I can try to get a scrolled screenshot for you if you think it will help.
why do you think stoping on postcode means "blocking" ?
usualy this mean what bios doesnt know some capabilies/features of inserted cpu/pci subsystem/memory controller/system agent/other_device_in_ cpu. cpupolicy does many check of cpuid families, but as i understand this engineer version belong to same family like usual cpus, so family check is positive. checks maybe different ways, for examling 4-bit shifting of cpuid 0x00050654=>0x00005065, then check like "cmp r9d, 5065h" (see adress 0x144A in PlatformCpuPolicy)
Thanks for your thoughts @Mov_AX_0xDEAD - I just assumed blocked, since many B0 SKU’s all have similar problems on this particular board but not others. So do you think maybe it’s not blocked, but there is some other issue/compatibility?
Sadly, I also need to see why 50652 B0 CPU not working too, this is why I thought blocked, two different CPUID’s already in mainly compatible list but with B0 sample do not work. I don’t know assembly much at all, so was looking blindly
CPU’s stop on debug code 66 for most users here I think, CPU DXE from what I gather, not much help to narrow down
I’ve now compared what I asked about above to a known working Supermicro board’s same module, and the same instructions are there and right above is same cmp r9d, 5065h you pointed out too, so this isn’t a block at all anyway since same is found in known working board
@everyone else testing - you all have tried single CPU correct? Does anyone know any confirmed Asus boards these B0 work in? If yes, I can build some module swapped BIOS for you to test, if not I may end up doing that with some of the supermicro modules.
When I initially had the board, I did try with one CPU and it wouldn’t post either.
@Lost_N_BIOS my board actually stops at POST code BB, not 66. According to ASUS this refers to a memory error.
In my opinion the X11SPL-F Bios would be the way to go, cause it seems one of the only working boards with c621 chipset, or am i wrong?
Makes sense given that its the same chipset. Anyone else have insight into this?
And @Lost_N_BIOS , I don’t believe there are any confirmed Asus boards that work, even the ebay resellers for these chips note that none of the asus boards work as mentioned here:
Found here for reference:
https://www.ebay.com/itm/Intel-Xeon-Scal…a0q3p:rk:4:pf:1
However I did find a couple of other asus boards that exist:
ASUS Z11PA-U12
ASUS Z11PA-D8
But it seems like most folks are interested in the ASUS WS C621E Sage due to its capabilities (in my case multi-core enhancement, I want to run all of the cores at the turbo frequency of 3.2 GHz vs being limited to 2.2 GHz).
One last thing I found was that it seems like the QS chips do seem to be supported across the board as seen in the image below:
Found here:
https://www.ebay.com/itm/Intel-Xeon-Gold…bK3Bc:rk:1:pf:1
Given that it seems like the ES chips are only working in Supermicro boards, maybe its something that supermicro is implementing that other manufacturers are not?
I’ve seen that as well @mitdal369 assumed it was obviously due to still no compatible BIOS created by someone, or of course none given out by Asus yet.
Do B0 CPU’s work in the other two Asus boards you mentioned, ASUS Z11PA-U12 - ASUS Z11PA-D8? MCE is simply a BIOS setting to be enabled or disabled (it’s hidden in BIOS from user view when disabled)
I do not believe so, mind I’m only making that inference from the compatibility sheet that the ebay sellers are listing since I don’t know anyone that has tested it personally.
Slightly off-topic, could MCE be enabled on the supermicro board (and potentially enable voltage control)?
EDIT: Oh I just noticed that the chips apparently work in the Gen10 HP servers. Anyone happen to have one of them around?
Hi. Link in the post #120 is dead. I have c621e sage with stock bios, I also have compatible and working cpu for it as well as uncompatible b0 cpu. And i have spi programmer, so i probably can help you. Please clarify what exactly do you need me to do.
Thanks @avd - but I think we are at a standstill here for now - I do not know how else to check for a block, or to verify if should be compatible
@avd
The HAP/MOD tests I had provided at post #120 were tested at post #122 and did not yield any results. I can verify though what I said at post #117: The CSSPS was not initialized properly at these B0 SPI dumps because the platform never managed to boot properly.
This is not really related to the B0 problem but, due to the rarity of such platforms, I’d appreciate if you could do me a favor and dump your SPI chip after flashing the latest ASUS 3501 SPI/BIOS w/o Capsule using the programmer and running the system once using the non-B0 (working) CPU.
My Mobo didn’t power on with this BIOS image. Nothing happens on pressing the power button and BMC LED did not start to blink as it should after about 40 seconds of stand-by mode, so it looks to me like BMC FW is incompatible with 3501 or something. Also the Asus website says that I need to flash 0501 before the newer versions - https://www.asus.com/us/Commercial-Serve…/HelpDesk_BIOS/
My stock bios was 0401. I have created bios dump from the mobo right after this failed attempt to start 3501.
Also after I recovered my stock 0401 bios I have boot up the system to windows with xeon 6130 and after shutting down made another bios dump. Hope this will help.
As a bonus I have recorded a video of post-code sequences for b0 and retail cpus, maybe this would help as well. They include B0 fail start, the first start of compatible cpu after b0 and the second start which goes much faster.
The link - https://mega.nz/#F!qb4V2aAD!jRTdFI-AnIf_g-rk2UtqiQ
sorry for my english, not a native speaker