Pre Release Tests

[dakanji]

@MarioG-X and @startergo

Can you kindly help give the X222 file attached in the link below a once over?
#104 (comment)

X222 is more or less the final state for the next release with only some minor tidy up planned.

This is subject to resolution of the issue raised, assuming the poster does get back and is willing/able to work through at attempt to find a fix.

@startergo ... please also check UEFI Windows as some additional stuff is now blocked. Just want to be sure there is no adverse impact.

Might be better to post your outcomes here and to leave that item alone for now ... unless you get the same issue that is!

Thanks!

[startergo]

Both boot Windows and Mac OS
Archive.zip

[dakanji]

Thanks. Any issues with REL file not showing RP screen?

[startergo]

No issues. It is good

[dakanji]

Hello both or any other BTW,

Please give X229 the once over when you have a min
X229-BOOTx64.zip

UEFI Windows test appreciated

Thanks

[startergo]

EFI.zip
Both DBG and REL boot OC and WIN UEFI

[MarioG-X]

X229 Results: All partitions on 2 Macs booted fine, attached one log, all internal HD/SSD unless specified:

Mac mini 2014 - RP DBG
HD BTRFS OpenSUSE 1
HD BTRFS OpenSUSE 2
HD EXT4 Elementary OS
HD APFS Monterey
SSD APFS Catalina
SSD APFS Big Sur
SSD APFS Monterey
External HD Monterey
22h10t4203.log

iMac 2012 RP REL
HD BTRFS OpenSUSE 1
HD EXT4 Elementary OS
HD APFS Catalina
SSD BTRFS OpenSUSE 2
SSD APFS Catalina

[dakanji]

Can you guys please give X230 a try and share the log?
REMOVED ... See Updated

Only looking for MacOS (Monterey mainly) and UEFI Windows as trying to get info on post-boot NVRAM writes.
There is a good chance it will hang, so be prepared to recover
Let me know if it does not hang btw, as it includes some mitigation but not sure if will be effective.

Thanks

[MarioG-X]

Hung on Apple logo same as before AP fix:
22h22u1001.log

[MarioG-X]

I want to mention something that has never worked: My understanding is that after a reboot the highlighted system should be the last one booted if not otherwise overridden. If you look at the log 'Boot MacOS from FS Monterey Mini' is highlighted, but I haven't booted that in weeks. While testing, I have booted most others. Thought you should know, I don't think it should keep highlighting a system that has not been use in a long time. You may want to look into it, it tough, save a fix for a later date.

The last booted system was the first one, but it highlights the wrong one: First is on SSD second on HD if that helps.
Configured Default Loader:- 'Boot MacOS from FS Monterey'
Highlighted Screen Option:- 'Boot MacOS from FS Monterey Mini'

[dakanji]

Thanks. It was the same file as earlier with additional logging, so was expected to hang. Just did a mitigation which obviously was not enough was anticipated.

On the highlighted options, the first string is a search string and it selects the first entry it finds that contains that substring.
That's just how it works and you need to give your volumes unique enough names to avoid such mismatches.

[dakanji]

On the highlighted options, I will adjust the search pattern to first look for an exact match and to fallback to looking for a substring only if an exact match is not found

[dakanji]

Can you share a log from X231? Curious about Catalina or older as well if possible.
REMOVED ... See Updated

@startergo ... can you also try with UEFI Windows when you have a min?

Same considerations on hanging btw.
Thanks

[startergo]

X231 hang on Monterey boot.
22h23m3607.log

[dakanji]

Thanks ... please try X232 if possible
X232-BOOTx64.zip

[startergo]

This one boots fine both Monterey and Windows.

[dakanji]

Excellent ... can you share logs?

[dakanji]

Essentially decided to bring BootKicker back but will not set it up to run automatically but based on a config token, show a window asking the user whether they want to boot directly or go through StartManager when they try to load Mac OS. Users without the token set will be booted directly as usual while those with the token can choose which option they want.

A bit fidly and not ideal but seems better than the current situation where things hang. Users will just need to remember to set the token if on a T2 Mac and the next ConfigFactory can now identify these. Any Thoughts?

@startergo ... I believe the issues with BootKicker were ultimately solved. Not sure if you know for certain. Suppose I need to confirm it.

EDIT:
Seems there might be an issue with running BootKicker outside OpenCore:
https://forums.macrumors.com/posts/31785591

[MarioG-X]

Just to ensure clarity: Only the 2018 Mini is booted into RP using the Option key to boot from an external SSD. I no longer keep Linux on the internal. Makes life easier for me. No OC on the machine (yet).

The 2013 iMac 27 is booted normally. Both RP and OC are on the internal SSD, not internal HD (was a Fusion Drive). Depending on when I updated RP or OC one or the other starts, then I can launch the other one if needed. Right now, RP starts at power up, then boot into Linux, or OC if I want Sonoma, etc.

Everything boots fine using the latest RP and OC.

[dakanji]

@MarioG-X ... Can you check whether you are able to load RefindPlus from StartManager (OPT key) on the 2018 Mini if you disable your internet connection before starting the unit?

[startergo]

Shouldn’t he first boot to macOS through OPT key and then try through RP?

[dakanji]

Not relevant to this test. Checking whether an internet connection is needed by StartManager on such units as Recovery does.

[startergo]

I think it is only needed if the security files needed for boot are somehow tampered with.

[startergo]

No such issue with OC.
You can test my OCLP MacMini8,1 folder
Archive.zip

[startergo]

After I clean shutdown or an SMC reset the Mac mini it takes me to the login screen and after entering the password it just freezes. After a forced shutdown it takes me to a password reset screen. After reboot I select macOS through OPT button and I realize ithe NVRAM was reset based on the loud chime I hear. All subsequent reboots or shutdowns through the dedicated shutdown button are ok. To summarize I only see the issue after a forced shutdown or an SMC reset. Somehow in the sequence the NVRAM gets reset too.

[MarioG-X]

Unplugged ethernet cable, powered up, and option key got to all normal systems including RP, RP to Linux, nothing different with internet disconnected.

[dakanji]

Noted. Thanks.

[dakanji]

@MarioG-X ... can you kindly try it after also cutting the power at the electrical socket?

That is shut down, cut power at socket, disconnect the internet connection, give it about a minute, reconnect power leaving the internet disconnected, wait a few seconds, then restart.

Thanks

[MarioG-X]

Good:
Unplugged ethernet cable -> unplugged power 1 min -> powered up with option key -> normal options -> boot to RP -> boot to Linux. Normal.

Not good:
Unplugged ethernet cable -> unplugged power 1 min -> powered up with option key -> normal options -> boot to RP -> boot to Sonoma. Displays white screen and stops.

Edit: Also tried Sonoma test with ethernet plugged in, still white screen.

[startergo]

In this case ethernet cable has no influence on the white screen and freeze. You can verify that by leaving ethernet cable plugged in and disconnecting the power cable for more than 30 sec. You will get the same results. If you tamper with the im4m files and security is enabled.
https://michaellynn.github.io/2018/07/27/booting-secure/

[MarioG-X]

I only reported what I was asked to test on the 2018 Mini. I don't care about the white screen since I do not use RP on this machine to boot macOS. I only OPT into RP on external SSD to boot external SSD Linux systems which works perfectly.

[dakanji]

@MarioG-X ... Should have been clearer. The issue is basically Mac OS related and is not typically manifested by Linux.

[MarioG-X]

Been very busy with personal stuff but had a chance to test some things you may find useful, and shocking:

On the 2012 iMac, booting from the same external SSD as the Mini with RP on it, these two tests were good:
• Start iMac with option key -> boot RP from external SSD -> boot Catalina on internal HD. Works.
• Start iMac with option key -> boot RP from external SSD -> boot Opencore -> boot Sonoma on internal SSD. Works.

Now the really interesting part on the 2018 Mac mini:
• Start Mini with option key -> boot RP from external SSD -> boot Ventura on internal SSD. Works.
I had not previously tested Ventura since Sonoma gave white screen. So I thought it was an RP/Sonoma issue. But when I tried to reproduce the Sonoma white screen:
• Start Mini with option key -> boot RP from external SSD -> boot Sonoma on internal SSD. Works! This did not previously work, with or without ethernet. So retested with ethernet unplugged and lo and behold it worked!

I did not change anything on the 2018 at all. No idea why Sonoma gave consistent white screen before and today it works. Didn't even power it up between 5 days ago and today.

[dakanji]

Noted. Gained enough insight and you can skip further tests. Thanks.

[MarioG-X]

14.1.AA All good for me:

2018 Mac mini (RP on external in /boot/BOOTx64.efi):
• Start option key -> boot RP from external SSD -> boot Ventura on internal SSD. Good.
• Start option key -> boot RP from external SSD -> boot Sonoma on internal SSD. Good.
• Start option key -> boot RP from external SSD -> boot Opensuse, etc. on external SSD. Good.

On the 2012 iMac, RP autostarts from internal SSD:
• Start iMac -> boot RP from internal SSD -> boot Catalina on internal HD. Good.
• Start iMac -> boot RP from internal SSD -> boot Opencore -> boot Sonoma on internal SSD. Good.
• Start iMac -> boot RP from internal SSD -> boot Opensuse on internal HD. Good.

[dakanji]

Thanks for the update