MP 1,1-5,1 Mac OS 11.3 has broken support for older Mac Pros

[startergo]

Big Sur 11.4 released, same build as RC.

071-00696       11.4    20F71  2021-05-24  macOS Big Sur

Unfortunately that version has been tested already. No fix so far.

 

[eVasilis]

If at first you don't succeed, try, try again. I've had installs take the first time with no issues, I've had most fail 1/2/3 times requiring a power cycle, and I've had the system drive become corrupted during install, 5x now actually. Twice I was able to recover by booting into recovery and performing a reinstall, but I've also had to wipe the system drive and start over.

I dreaded the only answer would be that... Still, has anyone had any insight as to why it hangs?

 

[cdf]

Looks like we have another machine that hangs on the ethernet initialization with this patch applied.

Note that even without the patch, ethernet initialization is a common point at which the hanging occurs.

How can we deal with installation hanging continually?

If at first you don't succeed, try, try again.

Perseverance is great advice in general. But in this situation even if the installation eventually goes through, the hanging will just occur on the next reboot or so. If the purpose of the installation is just for testing, then it's easier to prepare a ≥11.3 bootable USB installer and try booting that.

 

[eVasilis]

then it's easier to prepare a ≥11.3 bootable USB installer and try booting that.

I tried both ways, System > Prefs update and usb 11.3 installer. In both cases the installer hung and kept on doing so after 8 forced reboots. I feel that if there were a way to determine the reason why installations hang we would be closer to finding a solution with regard to the behaviour of 11.3.x and up.

 

[cdf]

I tried both ways, System > Prefs update and usb 11.3 installer. In both cases the installer hung and kept on doing so after 8 forced reboots.

Did the USB installer even boot into the recovery environment? Because just booting into that will hang. The point is that simply trying to boot a USB installer is enough to test proposed patches.

 

[eVasilis]

Did the USB installer even boot into the recovery environment? Because just booting into that will hang. The point is that simply trying to boot a USB installer is enough to test proposed patches.

Yes it did. I was also able to start the terminal to to ensure SIP was off. When I clicked install macos, it started the installation, progressed to the first reboot where in OC you can choose the macos installer and then after a while in hung. Forced reboots allowed for a little progress of the bar to the point it came to a standstill.

 

[JohnD]

Note that even without the patch, ethernet initialization is a common point at which the hanging occurs.

Yes, however with the patch, it hangs at that point MUCH more frequently, at least in my case, and a couple others. It used to be much more random, but with the patch, it's near 100%. Syncretic was looking for change, and that's a change.

PS - I had the same experience booting with cpus=1 - hung on ethernet pretty much every time, rather than random places.

 

[eVasilis]

Too many mistypings... apologies

 

[MacForumsUser]

Big Sur 11.4 released, same build as RC.

071-00696       11.4    20F71  2021-05-24  macOS Big Sur

Is it safe to install?

 

[JoeOIVOV]

Is it safe to install?

Its not any better at all. Took me about 40 minutes to get two consecutive reboots as to enable Titan Ridge. If you don't need the two consecutive reboots then its not too bad, you could have to turn it off 10 times to get it to boot correcly once. Mostly its around 4-5 times. One thing is if you watch the status lights on NVMe drive, it will blink right away rapidly when its booting correctly, if it doesn't start rapidly blinking then just unplug an plug it back in, don't wait on a slow loading bar, its not doing anything if the status lights aren't blinking...
I leave a type-C usb plugged into the Titan-Ridge too and it doesn't seem to help/hurt. Holding down ESCAPE or EJECT while mashing on the enter key seems to work best for me but you'll have to be persistent to use it, just stay on 11.23 if your on it imo.

 

[Stex]

Its not any better at all.

FYI you left your serial number on the screenshot... might be better to remove it/cover it.

AND I am about to write a report on my tests of @Syncretic config patch on both 11.3.1 and 11.4... with several pics of verbose etc. hoping to post it in the next 1-2 hours.

 

[feinberj]

My 5,1 seems to be mostly OK with installing Big Sur 11.3.1 using OC 0.6.9 (and syncretic's unaltered config.plist) and then booting. One thing I learned the hard way was that I had to use a completely unaltered version of syncretic's config.plist file. My 5,1 crashed when I had altered the posted file for just legacy wifi and retina boot. I started with a SATA drive and moved on to an NVMe drive.

...

More info on my experience using JohnD's config.plist file built using syncretic's information. My 5,1 boots 11.3.1 reliably if and only if I use the unaltered config.plist file. One change that makes it crash every time is activating SIP ( search for csr-active-config, change dwgAAA== to AAAAAA== ). Then when I go back to SIP deactivated (the "default" setting of the unaltered config.plist file) 11.3.1 boots reliably again. I had initially thought that some spaces were creeping into the config.plist file, but that was not the case. Activating SIP always causes 11.3.1 to crash on boot, at least for my machine. I'd like to try a few other changes to the config.plist file, such as retina support for the bootpicker and enabling the legacy wifi card. But I suppose if this prebuilt config.plist works as long as it's not altered at all it would be good enough for me!

 

[Stex]

Here's my latest report:


BEFORE TESTS/UPGRADE

1. Replaced BR2032 battery (was in my to-do list for ages)
2. SIP enabled for all tests (always enabled for me)
3. Only WIFI Internet Connection
4. Stripped down system as follows:

• PCIe Slot 1 RX580 8GB with two monitors via DP
• PCIe Slot 2 PEXM2-130 Dual NVMe M.2 (with Samsung EVO 970 500GB Boot 11.2.3 + 1TB Samsung EVO 970 TEST BLADE with first Boot 11.3.1 then Boot 11.4)
• 128GB RAM (x8 16GB -- added two more DIMMS today -- yes against cMP wisdom!!!)
• HDD in Bay 4 with OpenCore and High Sierra
• Apple Wired USB Keyboard and Mouse plugged in USB2 in the back

ABOUT THE PROCESS

• I did a series of boot tests with @h9826790 OC 0.6.8 Config+ @Syncretic Patch on both 11.3.1 and 11.4
• I also did a series of boot tests with the "vanilla" @h9826790 OC 0.6.8 Config on 11.4
• The 11.4 upgrade process froze about x10 but eventually succeeded - I include 2 verbose pics of when it froze as examples (filename: 11_4 Install_Process Martin_Config PIC_01.jpeg, 11_4 Install_Process Martin_Config PIC_02.jpeg)
• What follows are the boot attempts/results for each test along with file names for verbose pics of where the boot froze.

RESULTS

11.3.1 BOOT (Syncretic Patch/SIP Enabled)	RESULT	VERBOSE PIC
Warm	✅
Warm	✅
Warm	❌
Cold (Forced Shutdown)	✅
Warm	✅
Warm	✅
Warm	❌	11_3_1 Boot_Test Syncretic_Patch PIC_01.jpeg
Cold (Forced Shutdown)	❌	11_3_1 Boot_Test Syncretic_Patch PIC_02.jpeg
Cold (Forced Shutdown)	✅
Warm	❌	11_3_1 Boot_Test Syncretic_Patch PIC_03.jpeg

11.4 BOOT (Martin OC Config 0.6.8/SIP Enabled)	RESULT	VERBOSE PIC
Warm	❌
Cold (Forced Shutdown)	✅
Cold (*Regular* Shutdown)	✅
Warm	❌	11_4 Boot_Test Martin_Config PIC_01.jpeg
Cold (Forced Shutdown)	❌	11_4 Boot_Test Martin_Config PIC_02.jpeg
Cold (Forced Shutdown)	✅
Warm	❌	11_4 Boot_Test Martin_Config PIC_03.jpeg
Cold (Forced Shutdown)	❌	11_4 Boot_Test Martin_Config PIC_04.jpeg
Cold (Forced Shutdown)	❌

11.4 BOOT (Syncretic Patch/SIP Enabled)	RESULT	VERBOSE PIC
Warm	❌	11_4 Boot_Test Syncretic_Patch PIC_01.jpeg
Cold (Forced Shutdown)	❌
Cold (Forced Shutdown)	❌
Cold (Forced Shutdown)	❌	11_4 Boot_Test Syncretic_Patch PIC_02.jpeg
Cold (Forced Shutdown)	❌	11_4 Boot_Test Syncretic_Patch PIC_03.jpeg
Cold (Forced Shutdown)	✅
Warm	✅
Warm	❌
Cold (Forced Shutdown)	✅

EDIT: Added "SIP Enabled" to table headers for clarity.

 

[Stex]

One change that makes it crash every time is activating SIP ( search for csr-active-config, change dwgAAA== to AAAAAA== ). Then when I go back to SIP deactivated (the "default" setting of the unaltered config.plist file) 11.3.1 boots reliably again.

Thanks for data point. I will try the boot tests on 11.4 with SIP disabled -- but tomorrow, tonight I am done with Big Sur and very much ready for Big Beer.

 

[eVasilis]

deleted

 

[lxmeta]

Just installed 11.4 as an update to my 11.3.1 that previously appeared to be fine.
System required about 3 cold starts to make the install run through.
Looks like my system has caught the flu as well. Cold start seems to work better than warm starts. (no hard testing)
System is up and running again, and is performing very well. I guess I may have to live with it and only restart when I absolutely must.
Still quite sad to see the end of the line for these systems!
Aren't there any Apple engineers that have these systems at home as well, and help with the debugging?
This is how this usually works out (hidden support)

 

[Tiyom]

hi, I´ve got a big problem with my Mac Pro 5.1, GeForce GTX 770, 32G ram, I managed to install Mojave with my new GPU on my principal drive, then I have used opencore legacy to install Big sur on a second drive, unfortunately it was the 11.3 with many boots problems and dual monitor not well handled, I managed to make everything work but after the last reboot it’s stay stuck on the progress bar on any partition I’ve got, even my principal drive with Mojave gets stuck, I can’t do anything, recovery mode gets stucked too, BigSur installer too... any clue ?

 

[startergo]

Aren't there any Apple engineers that have these systems at home as well, and help with the debugging?

First of all why would Apple engineers keep/test very old vintage equipment?
Second, are you 100 percent sure this "bug" is not by design?

 

[VitaminK]

can’t do anything, recovery mode gets stucked too, BigSur installer too... any clue ?

Do an NVRAM reset, has happen to mine before.

 

[eVasilis]

Second, are you 100 percent sure this "bug" is not by design?

It seems that with every update, the possibility to get 11.3.x (now it is 11.4) to work on 5,1 CMPs diminishes. I guess, sooner rather than later it will be null. I am not knowledgeable to the point I can ascertain this is a bug but I do not think so; I feel the problems cMPs have with said updates are by design. Hardware has changed a lot in the past ten years and, probably, whatever is inside our cMPs cannot fit in with the "modern crowd". I'd be curious to see Apple's new Mac Pro, the one with apple silicon in, how it will compete with modern threadrippers (64 cores - 128 threads, 128 PCI lines) paired with Nvidia's two or more RTX A6000. We'll see.

 

[Stex]

Here's my last report to test @feinberj apparent success case described at post #912:


BEFORE TESTS

1. SIP disabled for all tests
2. Only WIFI Internet Connection
3. Stripped down system as follows:

• PCIe Slot 1 RX580 8GB with two monitors via DP
• PCIe Slot 2 PEXM2-130 Dual NVMe M.2 (with Samsung EVO 970 500GB Boot 11.2.3 + 1TB Samsung EVO 970 TEST BLADE with Boot 11.4)
• 128GB RAM (x8 16GB)
• HDD in Bay 4 with OpenCore and High Sierra
• Apple Wired USB Keyboard and Mouse plugged in USB2 in the back

ABOUT THE PROCESS

• Tested @h9826790 OC 0.6.8 Config + @Syncretic Patch on 11.4 with SIP disabled
• Tested "vanilla" @h9826790 OC 0.6.8 Config on 11.4 with SIP disabled
• Boot attempts/results follow with one verbose pic for each batch

RESULTS

11.4 BOOT (Martin OC Config 0.6.8/SIP Disabled)	RESULT	VERBOSE PIC
Cold (*Regular* Shutdown)	✅
Warm	❌	11_4 Boot_Test Martin_Config SIP_Disabled.jpeg
Cold (Forced Shutdown)	✅
Warm	✅
Warm	❌
Cold (Forced Shutdown)	✅
Warm	✅
Warm	❌
Cold (Forced Shutdown)	✅
Warm	❌

11.4 BOOT (Syncretic Patch/SIP Disabled)	RESULT	VERBOSE PIC
Cold (*Regular* Shutdown)	❌
Cold (Forced Shutdown)	✅
Warm	❌
Cold (Forced Shutdown)	❌	11_4 Boot_Test Syncretic_Patch SIP_Disabled.jpeg
Cold (Forced Shutdown)	❌
Cold (Forced Shutdown)	❌
Cold (Forced Shutdown)	❌
Cold (Forced Shutdown)	❌
Cold (Forced Shutdown)	✅
Warm	❌

 

[cdf]

Hardware has changed a lot in the past ten years and, probably, whatever is inside our cMPs cannot fit in with the "modern crowd".

But the Mac Pro 3,1 has even older hardware and can boot fine.

I feel the problems cMPs have with said updates are by design.

Maybe not by design (because we can verify the code), but convenient for Apple and therefore maintained (especially if it's been reported).

 

[eVasilis]

But the Mac Pro 3,1 has even older hardware and can boot fine.

Maybe this is a rather happy coincidence!

Maybe not by design (because we can verify the code), but convenient for Apple and therefore maintained (especially if it's been reported).

When I say by design I do not mean on purpose but a design to accommodate newer and newest technologies. It is definitely convenient, though.

 

[Tiyom]

Do an NVRAM reset, has happen to mine before.

I tried but nothing better, I just tested with my old original nvidia radeon, I can start without problem even with Mojave, it’s seems to be a GPU problem, but I´m as stuck as my MP, it use to work well before, something go wrong but I can't figure what, is there a debug mode or boot logs I could check...?
could it be a corrupted MP EFI ?

 

[cdf]

@Stex's comprehensive results illustrate a case where the current patch and SIP have no obvious effect on the issue. These results line up with what I've observed.