Jump to content

Help installing Mojave on Xeon W-2175 and Asus WS C422 mobo

obus

By obus,
in OSx86 10.14 (Mojave)

 Share
852 posts in this topic

Recommended Posts

etorix
etorix
Posted
Posted (edited)
On 5/19/2021 at 11:42 PM, yapan4 said:

The first attempt did not give the desired result. I did not find a special 10.13.2 build for iMacPro and used the regular version 10.13.2 with iMac17.1 SMBIOS. Maybe that's why update 10.13.2 to 10.13.3(via 10.13.3 Combo Updater) was installed, but the Aquantia remained untouched.

Maybe the trick is to use iMacPro1,1 SMBIOS? @rolandino has not indicated which SMBIOS he uses.

 

On 5/20/2021 at 12:05 AM, yapan4 said:

@etorix

Currently, only some screenshots of the BIOS

Oh, it's the fancy "SuperO" graphical display, not the text-only interface used on server SuperMicro motherboards.

I see nothing I would have done differently but unfortunately, I do not understand most of these advanced options.

Edited by etorix
typing error
Link to comment
Share on other sites
rolandino
rolandino
Posted
Posted
21 hours ago, yapan4 said:

Currently used original BIOS, will do that later, thank you.

The first attempt did not give the desired result. I did not find a special 10.13.2 build for iMacPro and used the regular version 10.13.2 with iMac17.1 SMBIOS. Maybe that's why update 10.13.2 to 10.13.3(via 10.13.3 Combo Updater) was installed, but the Aquantia remained untouched.

You need iMacPro1,1 SMBIOS.

  • Like 1
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)
On 5/19/2021 at 6:50 PM, etorix said:

I think that npci=0x2000 is necessary on most/all C422/C62x platforms, in addition to Above 4G Decoding.

Apparently Asus WS C422 Pro/SE  is an exception, npci=0xX000 never used on this motherboard.

On 5/19/2021 at 6:50 PM, etorix said:

Alcid layout depends how outputs are wired to the codec, so different values may be needed for different boards, even with the same ALC chip.

I will remember this as a law. Thanks.

On 5/19/2021 at 6:50 PM, etorix said:

Your Clover EFI and bootlog were very informative (and, in some places, surprisingly familiar for an OpenCore user… ;)). Many quirks are off already, and if you can install Big Sur there must be native NVRAM.

I don't know what could be done from the Clover side, but based on your folder, here is what I would initially try with OpenCore:

EFI-OC-X11SRA.zip 2.48 MB · 0 downloads

This is based on your SSDTs and settings. Debug version, so it may take some time but even if it fails to boot at the first attempt, it will drop the ACPI tables for inspection and for making more SSDTs.

NVRAM is emulated at this point, because ACPI tables are necessary to try and make a SSDT.

I notice that you use EnableWriteUnprotector, but the bootlog says that there is MAT, so it may be worth trying with EnableWriteUnprotector=false + RebuildAppleMemoryMap, SyncRuntimePermissions=true. (This is the recommended setting for MAT, but on my Gigabyte boards EnableWriteUnprotector is more stable even when MAT is implemented.)

Both CPM quirks enabled, if you haven't applied @metacollin's patches to the BIOS.

Beware! Configuration is not finished. If you want to try, you need to open Sample.plist (TextEdit, or ProperTree are best) and add Serial, Board Serial (MlB) and UUID under PlatformInfo>Generic (possibly also set keyboard under NVRAM>7C436110-AB2A-4BBB-A880-FE41995C9F82>Add>prev-lang:kbd), then rename to config.plist.

Experiments with the High Sierra take a lot of time, I will consider this in detail after returning to "normal life".

On 5/20/2021 at 10:41 AM, etorix said:

I see nothing I would have done differently but unfortunately, I do not understand most of these advanced options.

Yes, BIOS setup has a few options that Asus or Gigabyte doesn't have and need extra attention. I will study ...

Edited by yapan4
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted
21 hours ago, rolandino said:

You need iMacPro1,1 SMBIOS.

 

4 hours ago, etorix said:

I don't think there are special installers or updaters for iMacPro, just the regular 13.2 Installer and 13.3 Combo Updater. Rather, it is iMacPro1,1 SMBIOS which triggers the hardware-specific code. Run with iMacPro1,1 SMBIOS, and change back to MacPro7,1 afterwards.

Unfortunately regular macOS 10.13.2 does not work on iMacPro1,1 and on hackintosh with iMacPro1,1 SMBIOS. Can run regular 10.13.2 on iMacPro1,1 only with -no_comat_check boot argument.

Link to comment
Share on other sites
etorix
etorix
Posted
Posted

This is getting difficult…  :wallbash:I hope that @rolandino remembers exactly how he did.

 

All I can think of right now is: Boot 13.2 (-no_comat_check then), apply the 13.3 combo update, remove the -no_comat_check flag and re-apply the 13.3 combo update on 13.3. With some luck, it may update AQC firmware on the second run.

  • Like 1
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted
3 hours ago, etorix said:

This is getting difficult…  :wallbash:I hope that @rolandino remembers exactly how he did.

 

All I can think of right now is: Boot 13.2 (-no_comat_check then), apply the 13.3 combo update, remove the -no_comat_check flag and re-apply the 13.3 combo update on 13.3. With some luck, it may update AQC firmware on the second run.

The funny thing is that the combo update 10.13.3 as well as the full installer 10.13.3 are also obsolete for iMacPro1.1.
Then there were special builds for iMacPro type 17C2111, 17C2120, support for iMacPro1.1 in regular builds appeared later.

 

In the Aquantia topic there are reports of successful completion on iMac1X,X SMBIOS, but not in my case.

 

I doubt that anyone knows all the "system requirements" for successful completion, especially after 4 years full of events...

 

For now, I'm pausing experiments.

Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)

 

On 5/19/2021 at 6:50 PM, etorix said:

Both CPM quirks enabled, if you haven't applied @metacollin's patches to the BIOS.

Good(currently used original BIOS)

On 5/19/2021 at 6:50 PM, etorix said:

I notice that you use EnableWriteUnprotector, but the bootlog says that there is MAT, so it may be worth trying with EnableWriteUnprotector=false + RebuildAppleMemoryMap, SyncRuntimePermissions=true. (This is the recommended setting for MAT, but on my Gigabyte boards EnableWriteUnprotector is more stable even when MAT is implemented.)

Done.

 

On 5/19/2021 at 6:50 PM, etorix said:

Beware! Configuration is not finished. If you want to try, you need to open Sample.plist (TextEdit, or ProperTree are best) and add Serial, Board Serial (MlB) and UUID under PlatformInfo>Generic (possibly also set keyboard under NVRAM>7C436110-AB2A-4BBB-A880-FE41995C9F82>Add>prev-lang:kbd), then rename to config.plist.

Done.

 

 

Stop on this place

 

Spoiler

20210527_204508.thumb.jpg.b1dbc2f1998e20b85f73aa48c1d9e494.jpg

 

yapan4_OC_ Reports.zip

Edited by yapan4
  • Like 1
Link to comment
Share on other sites
etorix
etorix
Posted
Posted
22 hours ago, yapan4 said:

Stop on this place

Thanks for trying and giving me the opportunity to experiment with a C422 system.

 

From the logs all slides are good: Two quirks less!

DSDT: Processor objects (old style = good); no EC; a PCHRESV object which should provide native NVRAM with the trick from 300 series chipsets :); both an AWAC and a RTC which has the usual gap in its declared memory regions.

The last item can explain why boots stops at PCI initialisation: There is no valid clock for OS X— and neither the standard SSDT-AWAC nor the standard SSDT-RTC0-RANGE can work. I have adapted and combined the two tricks in order to disable the native AWAC, disable the native RTC and then provide a corrected RTC0.

 

Here is an EFI folder with custom SSDTs and less quirks. OpenCore release version because at this point issues are already within OS X, not with OpenCore.

Again, to use this EFI folder you need to add serial/MLB/UUID under PlatformInfo and rename to config.plist.

EFI-OC-X11SRA_v1.zip

 

Source code for the customs SSDTs: SSDT-X11SRA.zip

  • Thanks 1
Link to comment
Share on other sites
etorix
etorix
Posted
Posted

@yapan4 Is this with my last OpenCore EFI folder? Right away?

I hoped it would work, but I'm still pleasantly surprised by the result!

 

You definitely need to make a USB map. To remove the USB quirk, and improve stability. And because it may solve the sleep issue.

 

  • Thanks 1
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted (edited)
44 minutes ago, etorix said:

Is this with my last OpenCore EFI folder? Right away?

Yes, your EFI v1 without any changes(only IntelMausi.kext was removed because onboard i210 is natively supported)

Thank you very much

Edited by yapan4
Link to comment
Share on other sites
  • 2 months later...
  • 2 weeks later...
amjsez
amjsez
Posted
Posted
On 8/2/2021 at 9:39 PM, khe91 said:

New BIOS 3406

https://www.asus.com/de/Motherboards-Components/Motherboards/All-series/WS-C422-PRO-SE/HelpDesk_BIOS/

Works for me

Hi khe91, Is your USB tree for the GC-Titan Ridge showing in IORegistryExplorer?? With Bios1202 all is showing. Any bios after that, it does not show...

Screen Shot 2021-08-14 at 4.12.36 PM copy.jpg

Link to comment
Share on other sites
khe91
khe91
Posted
Posted (edited)
2 hours ago, amjsez said:

Hi khe91, Is your USB tree for the GC-Titan Ridge showing in IORegistryExplorer?? With Bios1202 all is showing. Any bios after that, it does not show...

 

No it's not showing since that older bios versions. But I didn't bother with special AML for it. Had an eGPU connected to it and it always worked.

Edited by khe91
Link to comment
Share on other sites
obus
obus
Posted
  • Author
Posted (edited)
On 8/14/2021 at 2:18 PM, amjsez said:

Hi  khe91, Is your USB tree for the GC-Titan Ridge showing in IORegistryExplorer?? With Bios1202 all is showing. Any bios after that, it does not show...

Screen Shot 2021-08-14 at 4.12.36 PM copy.jpg

Hi @amjsez

You can use attached bios 3303 (without ASUS splash screen) and with this setting in OC:

Spoiler

968994494_Screenshot2021-08-15at19_30_27.thumb.png.43b440f6cc98f66f6c874d2c84244c09.png

This bios will fix your problems with CPU cores not showing up correct in IORegistryExplorer. Could you please upload a copy of your IORegistryExplorer with the latest bios.👍

WSCSE.CAP.zip

Edited by obus
Link to comment
Share on other sites
obus
obus
Posted
  • Author
Posted (edited)
1 hour ago, khe91 said:

 

The USB might not match yours and there is nothing for the Thunderbolt in it.

OpenCore 7.2 release version

EFI-20210807-7_2.zip 717.07 kB · 0 downloads

could you please upload your IORegistryExplorer file for 3406 bios😀

My bad. Booting without problems with the new 3406 bios. I made a silly mistake in bios settings. Thunderbolt port is not enumerated in IORegistryExplorer with this bios so I have to revert back to 3304.

Edited by obus
Link to comment
Share on other sites
  • 5 weeks later...
yapan4
yapan4
Posted
Posted (edited)

Supermicro X11SRA-F Part Two(part 1 - on page 31)😀:

1) In search why Sleep don't work I found this:

 

Spoiler

866829511_Screenshot2021-09-13at23_23_35.thumb.png.0baab33d52d99e77b0a511926b4f9034.png

This is official Supermicro FAQ.

 

2) Problem with black screen during POST solved by flash another BIOS into my WX4100 GPU. This card has some incompatibility with X11SRA-F in UEFI mode and this is strange because no problem on ASUS WS C422 Pro/SE. Now after flashing OEM DELL BIOS work fine on SuperO but not sure on ASUS😀...

3) New v2.4 BIOS is not good - macOS(any versions) don't boot, so I back to v2.3

4) Supermicro allow flash previous version of BIOS just in the same manner as BIOS Update, no need recovery mode.

5) I do not understand how Recovery Mode works on this mobo and how to flash patched BIOS? If somebody has experience, I will be grateful for a tip...

Edited by yapan4
  • Like 2
Link to comment
Share on other sites
etorix
etorix
Posted
Posted (edited)

1) So for Supermicro a workstation board works all day and does not sleep. Nice find!

 

It reminds me of a joke:

"A train station is where the train stops. A bus station is where the bus stops. On my desk, there is a workstation."

(I hope this translates well…)

 

3) Then carefully keep a copy of BIOS v2.3! Supermicro is notorious for preventing the download of old firmware versions. Some people have developed special scripts to try and extract old versions anyway:

https://forums.servethehome.com/index.php?threads/supermicro-file-index.27148/#post-250426

 

4, 5) Boards with a BMC generally allows flashing BIOS without any restriction. Because another BIOS can always be flashed through the BMC, it is impossible to brick the board with a defective BIOS. So it should be possible to flash a patched BIOS just like a normal BIOS.

The safest and most convenient procedure would be through the BMC: Log on to IPMI, go to Maintenance>BIOS Update, upload file. But Supermicro requires a paid licence for that, which is a shame. (I have not bough a licence for my NAS boards.)

The UEFI method is free. Put the BIOS file and the flash.nsh utility on a FAT32 USB drive, boot to UEFI Shell, 'map' to list the drives, 'fs#:' to go to drive fs#, 'dir' to list the files and confirm you've found the right drive, 'flash.nsh BIOS.FILE'. I suppose that's the method you used to update to v2.4 and revert to v2.3.

BIOS recovery mode would be to put the original BIOS file in the root directory of a FAT32 USB drive and rename it to 'SUPER.ROM' (no need for flash.nsh), plug in the USB drive and power on. The board should automatically detect a BIOS file with the special name SUPER.ROM and enter recovery mode. Supermicro still recommends updating the BIOS again through the normal procedure. (I have never used this procedure, so I'm just rephrasing the manual here.) Obviously, this requires that the previous BIOS is at least partially functional… If the patched BIOS is completely defective and cannot enter recovery mode, you would have to buy a $27 licence key to restore the BIOS through IPMI.

https://store.supermicro.com/out-of-band-sft-oob-lic.html?utm=newsm

Good luck with the patched BIOS!

 

Edit. I have checked Supermicro eStore (link above). X11SRA-F is one of the few boards in the drop down menu with the mention "(No OOB BIOS Flash)". So no recovery through the BMC. :wallbash:

Edited by etorix
should have FULLY checked before posting :(
  • Thanks 1
Link to comment
Share on other sites
yapan4
yapan4
Posted
Posted

I don't know reason for not supporting S3 Sleep mode on this mobo and also reason to immediately deleting old BIOS after new one released    ...well, Supermicro has personal Quirks😀

I know about the inability to update the BIOS through IPMI and this is another one "-" for this motherboard.

In addition, I was unable to enter Recovery Mode, although I did as you wrote. Motherboard just starts as usual. 

But I still decided to flash the patched bios, and everything went well. Thanks for the hint, the motherboard really flashes the modified bios just like the native! One little or big?  "+"😀

 

Still opened two questions:

1) What is wrong with BIOS v2.4?

2) Why don't work Recovery Mode? (I'll try again later)

 

bootlog.txt EFI.zip

  • Like 1
Link to comment
Share on other sites
etorix
etorix
Posted
Posted

A2SDi-H-TF does not have the BIOS recovery with "SUPER.ROM" so I cannot even test.

 

Have you dumped the DSDT or the full ACPI tables while on v2.4? I could try to have a look and compare with v2.3 to see what Supermicro have changed.

  • Like 1
Link to comment
Share on other sites
 Share
Go to topic listing
×
×
  • Create New...