MP 1,1-5,1 Manually Configured OpenCore on the Mac Pro

[Macschrauber]

Hello,

I have run the Rom dump tool from Macschrauber and I have a few questions if I may:

- My dump has a CRC mismatch. What do I risk? Can I fix that without being a genius?
- The tool reports I have a high amount of VSS1. Shall I worry about it?
- I have an XFX RX570 and I am getting the annoying pink OC menu (common with XFX cards) which resolves when the login window shows up. Anything that can be done to fix that other than replacing the GPU?

Currently running OCLP but planning to switch to a manually configured OC file, though I need to get that Bios dump sorted out if necessary.

As you asked about my tool I sent you a PN.

Without telling where, a checksum error can be everything. Someone edited a serial number without checksum correction, an update failure, a chip problem, everything...



Try other outputs of your GPU for the false format until the driver kicks in.

 

[Dayo]

It doesn't work without OC, I'm currently looking at this, I think it should be possible in principle.

Was there any progress on getting BootKicker to work without OC?

 

[Bmju]

Was there any progress on getting BootKicker to work without OC?

The answer is no, but that's mainly because I was under the impression that, with EnableGop working, and BootKicker working under OpenCore, there was no remaining need for any other related functionality in this area. I know that you're a fan of getting things to work if you feel they should in principle, but since you're asking I'm guessing you have a use-case in mind, and it would help me to understand what. (EDIT: Perhaps to put it in simpler, or fairer, terms, I don't understand what you'd want it to do, or rather what all that I can imagine it doing, if it did work without OpenCore, would be any use for!)

 

[Dayo]

Well, it used to be incorporated into RefindPlus in the early days of trying to sort pre-boot screen issues out (2020ish) but it later got dropped because:

1. It never quite worked as it should back then
   • Seem to vaguely remember that it did work if blessed directly but might be mixed up ... need to test.
2. There was ultimately no need for it as the pre-boot screen issues were later solved by other means.

This is explained better here:

Manually Configured OpenCore on the Mac Pro

I don't think so as I can't see a role for it there. If you remember though, that back in mid 2020 when we were looking at MP51 Boot Screen stuff, BootKicker was an obvious angle to explore but it simply just didn't work: https://github.com/acidanthera/bugtracker/issues/1084 I still went...

forums.macrumors.com

Now looking at it leveraging it again as a potential way to work around issues with T2 chipped Macs:

MacPro 7.1 - bootloader

Hello ! Does someone use a custom bootloader (ideally OpenCore) on it Mac Pro 7.1 ? I have a macOS installed on the apple ssd and a Windows on a nvme drive. I've tried to install several bootloaders on my Mac Pro 7.1 (OCLP, Refind Plus, Clover) and tried to tweak several options but I always...

forums.macrumors.com

These cause a hard hang when trying to boot Mac OS via 3rd-party loaders.

Given the state of play as set out though, I can understand the conclusion about there not being a remaining need for expanding functionality from the OpenCore project's POV.

 

[Bmju]

@Dayo - There is a danger of going off topic for this thread, perhaps we need an acidanthera issue. However (and as ever) I've got nothing against working on an upgrade to OC code, if it's a natural extension to what the OC code does already, with a reasonable use case for it amongst OC users (Mac, hack, whoever). Ofc I (and I think I can say we, since others have supported me on this in the past) need a clear use case: it's hard to work on something, if I don't understand what it's for!! Unfortunately, I'm still not completely clear on how (whether) BootKicker can help you here, hence perhaps the need for an issue, or failing that further discussion by PM.

One thing you should know is that the 'bug' in the Apple boot picker mentioned in the previous discussions you've linked to was (IMO) a mismatch between what OC (the BootKicker code specifically) thought the Apple picker should do (for instance: boot something) and what Apple thought the Apple picker should do (namely: set the boot variables to specify the next (with Enter) or next and subsequent (with CTRL+Enter) boot item).

After realising that the latter is all the Apple picker thought it was meant to do, there were two obvious options - 1) add an additional code path to OC to handle these variables being set, in a way we needed to respond to, much later than normal (where 'normal' is before OC ever starts), or 2) just reboot, and let the changes take effect then. We, I, choose the latter. This has seemed to work very well up to now (for the very small number of people, AFAICT, who use BootKicker at all - cc @cdf ;-) ). Whether this change still matches your expectations, or needs, of what BootKicker should do to be useful to you now, I am less sure.

HTH!

 

[Dayo]

@Bmju ... The issue with the T2 Macs does not manifest when StartManager is in the boot chain and the very initial thought was to load StartManager on such boots on affected units based on a config setting in RefindPlus.

Was just re-orientating myself with where things stand on BootKicker as a candidate option and your initial response already provided the information I needed. Thanks.

An issue will be raised in the event this is determined to be worthwhile.

 

[tsialex]

OpenCore v1.0.0 released today:


v1.0.0

• Updated builtin firmware versions for SMBIOS and the rest
• Switched to Apple silicon GitHub runner for CI, thx @Goooler
• Added Apple Silicon support in all provided utilities
• Utilities now require macOS 10.9+ (OpenCore itself still supports macOS 10.4+)
• Added AllowRelocationBlock support for 32-bit version
• Enabled additional serial logging in non-RELEASE builds of OpenDuet
• Added missing DxeCore ImageContext HOB in OpenDuet
• Fixed assert caused by dependency ordering in OpenDuet
• Prevented assert in normal situation when freeing memory above 4GB in OpenDuet
• Prevented debug assert reporting that optional Hii protocols are not present in OpenDuet
• Fixed problem loading non-firmware runtime drivers (e.g. OpenRuntime.efi) in OpenDuet
• Resolved issue using NOOPT debugging in OpenDuet
• Fixed alphabetical ordering in Configuration.pdf, thx @leon9078

 

[cdf]

The guide has been updated to OpenCore version 1.0.0. No changes to your config are necessary if you are updating from the last version.

Enjoy your lean and mean manual OC setup!

 

[crjackson2134]

No changes to your config are necessary if you are updating from the last version.

I skipped the last version, but may I presume that there would still be no changes to the config needed?

 

[vrds]

may I presume

Who are we to deny you? 😄 First post has a history. Just check it for config changes.
AFAIK, there were no changes required starting from OC 0.9.6.

 

[cdf]

I skipped the last version, but may I presume that there would still be no changes to the config needed?

Yes, that’s right.

 

[Gustav Holdoff]

Just updated macos
The update was installed simply (with the exception of my individual adventures - as always, I then had to reset SMC and nvram several times to remove the kernel task, which occupied 70% of the CPU)
and wait another 2.5 hours until the MDS process indexes all disks
Now works ok

 

[mrkapqa]

Hello , this has probably already been answerered.

Have the following issue:
Upgraded the Win11 from 22H2 to 23H2 , now with with the newer OpenCore M.Lo 0.98 Package i am no longer able to bootinto Windows11; when i downgrade OpenCore to say package 0.88 it would again functin to boot into Windows 11 "correctly"; if i use OpenCore 0.98 there is just Windows Logo and no BootProgress icon, practically the Bootprocess is like frozen.

Can use 0.88 aswell, as it seems also the Monterey function, but would rather not loose functionality.
Thank you very much.

 

[Gustav Holdoff]

Hello , this has probably already been answerered.

Have the following issue:
Upgraded the Win11 from 22H2 to 23H2 , now with with the newer OpenCore M.Lo 0.98 Package i am no longer able to bootinto Windows11; when i downgrade OpenCore to say package 0.88 it would again functin to boot into Windows 11 "correctly"; if i use OpenCore 0.98 there is just Windows Logo and no BootProgress icon, practically the Bootprocess is like frozen.

Can use 0.88 aswell, as it seems also the Monterey function, but would rather not loose functionality.
Thank you very much.

This question should be asked to Martin, in his section
or maybe to the authors of the opencore sections who are present in this chat
every time a new version of opencore is released, you need to very carefully collect folders, transferring current kexts, drivers and other tools, although it is clear from the number in the name on the forum that you are not a newbie
To start Windows you need to check whether there is ExFatDxeLegacy.efi in the drivers folder and whether the corresponding lines are written in config
and check if you accidentally deleted the microsoft folder in efi?
Martin, of course, has EFI compiled, but nevertheless, check what is the difference in the configuration of folders in the efi between your two versions
In my personal experience, the Windows startup process froze several times - I fixed it by resetting SMS and nvram.

 

[Bmju]

Hello , this has probably already been answerered.

Have the following issue:
Upgraded the Win11 from 22H2 to 23H2 , now with with the newer OpenCore M.Lo 0.98 Package i am no longer able to bootinto Windows11; when i downgrade OpenCore to say package 0.88 it would again functin to boot into Windows 11 "correctly"; if i use OpenCore 0.98 there is just Windows Logo and no BootProgress icon, practically the Bootprocess is like frozen.

Can use 0.88 aswell, as it seems also the Monterey function, but would rather not loose functionality.
Thank you very much.

You could try to set up OpenCore again following the instructions on p.1 of this thread. You might want to back up your existing EFI folder first, e.g. using a natively bootable Mojave install as suggested on p.1.

OpenCore set up this way might hopefully just fix Windows booting (in which case, there would just be some config change needed in the package you are currently using to get it to boot Windows - but there are several possible candidates, see OpenCore documentation). But if not you could then get further support on this thread.

 

[mrkapqa]

This question should be asked to Martin, in his section

To start Windows you need to check whether there is ExFatDxeLegacy.efi in the drivers folder and whether the corresponding lines are written in config
and check if you accidentally deleted the microsoft folder in efi?
Martin, of course, has EFI compiled, but nevertheless, check what is the difference in the configuration of folders in the efi between your two versions
In my personal experience, the Windows startup process froze several times - I fixed it by resetting SMS and nvram.

Thanks for the help,
i had previously more folders in the RootSectionEFI of the Harddrive where i have OpenCore installed , because at one point had also installed ReFindPlus or howitiscalled;
since had some problems which have not been fullsy solved, decided to delete them, now i have just left the folders EFI (BOOT and OC respectively) plus SYSTEM plus one file which seems native to the HDD (Bay 3 Sata Toshiba DT 01ACA200)

the
ExFatDxeLegacy.efi
seems also present in the OpenCore 0.98 Package, but something must have changed in the settings or so.

dont know what you mean with the Microsoft Folder in EFI section?
if you mean SYSTEM, then it is present.



EDIT_: Thank you all, this has been solved now;
re-downloaded the OpenCore 0.98 https://forums.macrumors.com/thread...n.2180095/page-53?post=28255048#post-28255048
and applied from inside MontereyBooTdisk; now it would again boot into Windows 11 23H2 ;

maybe it made a difference if the Operation was made from Mojave BootDisk instead from MontereyBootDisk? Strange Things happen. But seems also other people had similar issue, glad now functions.