Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.
Don't all those "Invalid" entries count as free space? I think UEFITool should try to identify the contents of those entries and mark them as Deleted? That way you can tell what NVRAM variables are churning more frequently.
It's not really deleted until the garbage collection erase the full VSS store (remember the 4KB sector) and re-writes just the valid entries from the circular log. Garbage collection sometimes take up to 11 reboots to fully complete.

The wording obviously could be better, instead of invalid could be something like superseded.
 
  • Like
Reactions: joevt
Got it. I do have a MATT card with my reconstructed never booted BootROM, but it's there on the ready for that tragic day when my cMP might brick. I don't see a problem in doing the flashing twice, adds a bit of extra work of course, but it's not a big deal to be on the safe side of things.
You might want to install your MATT card before your Mac bricks to preserve your original flash chip.
 
@tsialex Hi, I was informed about BootROM in the OpenCore thread.

I checked the full size value of Free Space in BootROM's VSS Store and it was "50285". Is this possibly dangerous?
Can I fix this?
No, your value right now is very good, but your issue is usually related to NVIDIA web drivers saving a binary blob that makes modern AMD GPUs crazy under certain conditions.
 
  • Like
Reactions: ZNDK
Hi Alex,

As always, thank you for your contributions and we appreciate your work!

So I thought, and seeing that I had 12.3 crash on me twice before booting normally, I might check my cMP's rom and noticed 2 kernel panic dumps. The nvram free space was about 28K which I think is normal, no?

Anyway, I decided to carry out a triple nvram reset and thankfully both the kernel panic dumps and memory config resolved on their own.

Before NVRAM reset

Screen Shot 2022-03-19 at 6.33.29 PM.png



2022-03-19_19-26-09.png


and after NVRAM reset

Screen Shot 2022-03-19 at 6.32.59 PM.png


2022-03-19_19-28-43.png



What I found strange was that OpenCore was automatically detected and loaded after nvram reset which I usually had to manually bless (the EFI partition) in order for it to be recognized and loaded. I have OC installed on the internal EFI partition of my macOS drive (not using USB anymore).

Edit: Added UEFITool screen grabs just in case I may have missed out on something.

So everything checks out?

Cheers
 
Last edited:
Hi Alex,

As always, thank you for your contributions and we appreciate your work!

So I thought, and seeing that I had 12.3 crash on me twice before booting normally, I might check my cMP's rom and noticed 2 kernel panic dumps. The nvram free space was about 28K which I think is normal, no?
For a dual CPU MacPro after SoftwareUpdates, yes.
Anyway, I decided to carry out a triple nvram reset and thankfully both the kernel panic dumps and memory config resolved on their own.

Before NVRAM reset

View attachment 1976524


and after NVRAM reset

View attachment 1976525


What I found strange was that OpenCore was automatically detected and loaded after nvram reset which I usually had to manually bless (the EFI partition) in order for it to be recognized and loaded. I have OC installed on the internal EFI partition of my macOS drive (not using USB anymore).
I've should written a clarification on this before, I've talked about this before years ago with @joevt and never discussed it again.

There are two very different types of PanicInfoLogs with MacPro5,1, more with newer Macs. I've not yet found the Apple documentation about the different PanicInfo logs, so for differentiation here let's call the one that just points to the crash dump inside ~/Library/Logs/DiagnosticReports as PanicInfoLogs.A, and the second type where the crash dump itself is saved inside the VSS store, let's call it PanicInfoLogs.B, it's what I use on my diagnostic signatures to differentiate one from the other:

Type of PanicInfoLogtype Atype B
Content:Just a pointer to a file saved on /Library/Logs/DiagnosticReportsFull crash log, compressed, saved inside the NVRAM entry
Erasable:Yes, with deep NVRAM reset.No, permanent or at least not removable by deep NVRAM reset. Not yet know if normal garbage collection removes it over time.
Can fill the VSS store if happening in a loop:Usually no, it's a very tiny entry.Yes, some entries are over 3KB and one dump had 23 saved over the two VSS stores.
Often found or not:This is the everyday crash dump, frequently found on BootROM image dumps.It was rare to be found present in a dump before BigSur, now is becoming common with Monterey.

Since you got it removed from the VSS store with a NVRAM reset, your crashes probably were the first type. Usually you can confirm looking at SystemPreferences/Software/Logs of the disk that was blessed.
 
  • Like
Reactions: w1z
@tsialex:
can you point me to an example of Type B so I can include counting them in my dumper?
Sure, I'll send it on Monday. This is the signature to detect it:

Code:
#
# AAPL,PanicInfoLog signature B
# 
0    string    \x41\x00\x41\x00\x50\x00\x4C\x00\x2C\x00\x50\x00\x61\x00\x6E\x00\x69\x00\x63\x00\x49\x00\x6E\x00\x66\x00\x6F\x00\x30\x00\x30\x00\x30    NVRAM PanicInfo Log B
 
You might want to install your MATT card before your Mac bricks to preserve your original flash chip.

Thanks for the suggestion. My approach, which I sense is not ideal based on your response, was to let my original flash chip on the backplane run its course until it dies and then start using the MATT card, the latter being new with no use (other the first time I installed it to flash my never booted BootROM). If I understand correctly, your suggestion implies that it would be better to start using the MATT card now and have/keep a still functional flash chip on the backplane? Any further insights on this would be greatly appreciated. Thanks!
 
Sure, I'll send it on Monday. This is the signature to detect it:

Code:
#
# AAPL,PanicInfoLog signature B
# 
0    string    \x41\x00\x41\x00\x50\x00\x4C\x00\x2C\x00\x50\x00\x61\x00\x6E\x00\x69\x00\x63\x00\x49\x00\x6E\x00\x66\x00\x6F\x00\x30\x00\x30\x00\x30    NVRAM PanicInfo Log B

thanks a bunch :)

I added differing the kernel panics. So the dumper now counts both of them separately.

I found some of them in my dumps collection, I remembered I have seen those AAPL,PanicInfo000x.
 
Thanks for the suggestion. My approach, which I sense is not ideal based on your response, was to let my original flash chip on the backplane run its course until it dies and then start using the MATT card, the latter being new with no use (other the first time I installed it to flash my never booted BootROM). If I understand correctly, your suggestion implies that it would be better to start using the MATT card now and have/keep a still functional flash chip on the backplane? Any further insights on this would be greatly appreciated. Thanks!
Yes, that’s what I do simply because I like knowing that my logic board is still fully functional even without the MATT card. It makes no difference in practice, though.
 
  • Like
Reactions: Stex
For a dual CPU MacPro after SoftwareUpdates, yes.

I've should written a clarification on this before, I've talked about this before years ago with @joevt and never discussed it again.

There are two very different types of PanicInfoLogs with MacPro5,1, more with newer Macs. I've not yet found the Apple documentation about the different PanicInfo logs, so for differentiation here let's call the one that just points to the crash dump inside ~/Library/Logs/DiagnosticReports as PanicInfoLogs.A, and the second type where the crash dump itself is saved inside the VSS store, let's call it PanicInfoLogs.B, it's what I use on my diagnostic signatures to differentiate one from the other:

Type of PanicInfoLogtype Atype B
Content:Just a pointer to a file saved on /Library/Logs/DiagnosticReportsFull crash log, compressed, saved inside the NVRAM entry
Erasable:Yes, with deep NVRAM reset.No, permanent or at least not removable by deep NVRAM reset. Not yet know if normal garbage collection removes it over time.
Can fill the VSS store if happening in a loop:Usually no, it's a very tiny entry.Yes, some entries are over 3KB and one dump had 23 saved over the two VSS stores.
Often found or not:This is the everyday crash dump, frequently found on BootROM image dumps.It was rare to be found present in a dump before BigSur, now is becoming common with Monterey.

Since you got it removed from the VSS store with a NVRAM reset, your crashes probably were the first type. Usually you can confirm looking at SystemPreferences/Software/Logs of the disk that was blessed.

Thanks Alex. I believe it may have been type B as I could not find any panic/crash dump in ~/Library/Logs/DiagnosticReports or /Library/Logs/DiagnosticReports

2022-03-20_10-14-23.png
 
Until yesterday @Macschrauber tool didn't detected the second type at all, so you couldn't had it shown on his tool.

it had detected both types of Panic Logs but did not differ them.

It now counts Type A and B and shows both.

 
Last edited:
  • Like
Reactions: w1z and Stex
Until yesterday @Macschrauber tool didn't detected the second type at all, so you couldn't had it shown on his tool.

Confirmed - type A

Screen Shot 2022-03-21 at 11.00.41 PM.png


it had detected both types of Panic Logs but did not differ them.

It now counts Type A and B and shows both.


Thanks for updating the tool ?
 
I don't know if I'm contributing anything useful here, but given this thread prompted me to take a look at my ROM for the first time in a while after updating to 12.3 (without VMM) and I have the results, here we go.

In an attempt to take what I thought would be a clean ROM dump, I triple-reset and booted into Mojave. Got this:

Code:
Firmware 144.0.0.0 (latest)
old Bootblock of MP51.007F.B03
12 Memory Configs (ok)
1 xml (ok)
0 iCloud Tokkens (ok)
0 Microsoft Certificates (ok)
1 BluetoothActiveControllerInfos (ok)
1 BluetoothInternalControllerInfos (ok)
0 current-network (ok)
4 AAPL Path Properties (ok)
Invalid VSS Store 1 header (very bad!)
VSS2 is empty (ok after triple nvram reset or recent firmware rebuilt)
34688 Bytes free space of 65472

Oh dear. Flashed the backup I took a couple of years ago, came back to this:

Code:
Firmware 144.0.0.0 (latest)
old Bootblock of MP51.007F.B03
25 + 3 Memory Configs (caution)
0 + 0 xml (ok)
0 + 0 iCloud Tokkens (ok)
0 + 0 Microsoft Certificates (ok)
1 + 1 BluetoothActiveControllerInfos (ok)
1 + 1 BluetoothInternalControllerInfos (ok)
1 + 1 current-network (ok)
5 + 1 AAPL Path Properties (ok)
3520 Bytes free space of 65472

Okay... I see a caution on my memory configs. Gave it the triple-reset treatment again:

Code:
Firmware 144.0.0.0 (latest)
old Bootblock of MP51.007F.B03
12 Memory Configs (ok)
1 xml (ok)
0 iCloud Tokkens (ok)
0 Microsoft Certificates (ok)
1 BluetoothActiveControllerInfos (ok)
1 BluetoothInternalControllerInfos (ok)
0 current-network (ok)
4 AAPL Path Properties (ok)
VSS2 is empty (ok after triple nvram reset or recent firmware rebuilt)
34688 Bytes free space of 65472

And that's where I am now. Blessed my OC EFI and I'm back in Monterey.

So does that last entry look fine now? I guess so from what I have read, so I'll keep it as my "clean ROM".

Only other question is with regard to the "old Bootblock of MP51.007F.B03". Should I be doing anything about that? I see others have "Bootblock of 144.0.0.0 - rebuilt Firmware", but I assume this is following flashing a hand-rebuilt firmware image?
 
I don't know if I'm contributing anything useful here, but given this thread prompted me to take a look at my ROM for the first time in a while after updating to 12.3 (without VMM) and I have the results, here we go.

In an attempt to take what I thought would be a clean ROM dump, I triple-reset and booted into Mojave. Got this:

Code:
Firmware 144.0.0.0 (latest)
old Bootblock of MP51.007F.B03
12 Memory Configs (ok)
1 xml (ok)
0 iCloud Tokkens (ok)
0 Microsoft Certificates (ok)
1 BluetoothActiveControllerInfos (ok)
1 BluetoothInternalControllerInfos (ok)
0 current-network (ok)
4 AAPL Path Properties (ok)
Invalid VSS Store 1 header (very bad!)
VSS2 is empty (ok after triple nvram reset or recent firmware rebuilt)
34688 Bytes free space of 65472

Oh dear.

This is bad, when the secondary VSS store header is overrun, your Mac Pro NVRAM volume is already corrupt.

Flashed the backup I took a couple of years ago, came back to this:

Code:
Firmware 144.0.0.0 (latest)
old Bootblock of MP51.007F.B03
25 + 3 Memory Configs (caution)
0 + 0 xml (ok)
0 + 0 iCloud Tokkens (ok)
0 + 0 Microsoft Certificates (ok)
1 + 1 BluetoothActiveControllerInfos (ok)
1 + 1 BluetoothInternalControllerInfos (ok)
1 + 1 current-network (ok)
5 + 1 AAPL Path Properties (ok)
3520 Bytes free space of 65472

Okay... I see a caution on my memory configs. Gave it the triple-reset treatment again:

Code:
Firmware 144.0.0.0 (latest)
old Bootblock of MP51.007F.B03
12 Memory Configs (ok)
1 xml (ok)
0 iCloud Tokkens (ok)
0 Microsoft Certificates (ok)
1 BluetoothActiveControllerInfos (ok)
1 BluetoothInternalControllerInfos (ok)
0 current-network (ok)
4 AAPL Path Properties (ok)
VSS2 is empty (ok after triple nvram reset or recent firmware rebuilt)
34688 Bytes free space of 65472

And that's where I am now. Blessed my OC EFI and I'm back in Monterey.

So does that last entry look fine now? I guess so from what I have read, so I'll keep it as my "clean ROM".

It will work fine if it's not a cross-flashed early-2009.

Only other question is with regard to the "old Bootblock of MP51.007F.B03".

BootBlock is not upgraded by Apple efiflasher, so it kept the same version the factory flashed (or if is a cross-flashed early-2009, the version that was installed by the cross-flashing process) forever.

Should I be doing anything about that? I see others have "Bootblock of 144.0.0.0 - rebuilt Firmware", but I assume this is following flashing a hand-rebuilt firmware image?

The older than 144.0.0.0.0 BootBlocks have a serious issue with PCIe switched cards, sometimes when you power on the Mac Pro stops at POST. It was one of the things that was finally solved with 144.0.0.0.0, thanks to @crjackson2134 insisting with Apple support, but can only be installed with a reconstructed BootROM, since Apple firmware tools don't touch it.
 
  • Like
Reactions: macsforme
BootBlock is not upgraded by Apple efiflasher, so it kept the same version the factory flashed (or if is a cross-flashed early-2009, the version that was installed by the cross-flashing process) forever.

The older than 144.0.0.0.0 BootBlocks have a serious issue with PCIe switched cards, sometimes when you power on the Mac Pro stops at POST. It was one of the things that was finally solved with 144.0.0.0.0, thanks to @crjackson2134 insisting with Apple support, but can only be installed with a reconstructed BootROM, since Apple firmware tools don't touch it.
Ah, interesting. I’ve not experienced that issue - are you still offering a BootROM reconstruction, Alex? Should probably just DM. :)
 
I recently acquired a MacPro 5,1 8 core incl. 23" Cinema Display for really small money that was primarily used for video editing by a professional company. I did upgrade the firmware to 144.0.0.0.0, did NVRAM deep clean and upgraded CPUs to dual X5680. It has now run without issues since October 2021.
I am currently running macOS 11.6.5 and was considering to upgrade to 12.3 when I spotted the warnings regarding the NVRAM being filled-up by firmware staging. I am currently using OCLP 0.43 (based on OpenCore 0.7.8) to boot.
So I ran @Macschrauber RomDump from BigSur booted from OCLP because I cannot boot into Mojave recovery partition for some reason, it always boots into regular Mojave install. Maybe this is because of me using a RX480 that does not offer boot screen ?
Anyway, I got the below report and noticed that I also have the old boot block and for some strange reason have multiple BluetoothActiveControllerInfos :oops:. Is that critical and how do I fix this ?

Everything else seems ok also when viewing it in UEFI Tool (59).
@tsialex: Should I still opt to have it reconstructed incl. the latest boot block before I upgrade to Monterey ?
How do I go about flashing it if I cannot boot into Mojave recovery to disable SIP ?

Firmware 144.0.0.0 (latest) old Bootblock of MP51.007F.B03 Boot0001 is EFI\OC\OpenCore.efi 10 Memory Configs (ok) 0 xml (ok) 2 Kernel Panic Dumps Type A: Pointer Type 0 iCloud Tokkens (ok) 0 Microsoft Certificates (ok) 3 BluetoothActiveControllerInfos (not ok) 2 BluetoothInternalControllerInfos (ok) 0 current-network (ok) 2 AAPL Path Properties (ok) 42368 Bytes free space of 65472
 
Sip must have been disabled by OpenCore, if not, flashrom had not dumped your firmware.

DirectHW.kext loading needs sip disabled for kext. Flashrom needs this kext for communication with the Southbridge reading the spi flash chip.

Also you can use the dumper (by holding the alt-key while starting the tool) to write to the chip. Firmware write lock needs to be deactivated by long holding the power button like with every firmware update.

I know Alex has concerns flashing and reading by an openCore powered OS, but as we talk directly to the chip via flashrom I have seen no problem yet with flashing a _rebuilt_ rom file.

But, if you can, better plug in an old GPU and do it natively via Mojave.

It is always a good idea to get the firmware reconstruction by @tsialex. You will get a perfect firmware setup.

If you flash the rebuilt firmware, bless OpenCore and boot just once you can use this dump as your personal, already set up firmware what you can reflash regulary.
 
Last edited:
Thank you for your contribution to this forum @tsialex I finally got a power supply to repair my 5,1 and preparing to move to Big Sur with OCLP. Did a romdump to my surprise there was little space left. #4916 Here is the before and after 4 x reset.

Screen Shot 2022-04-03 at 1.26.21 PM.png
Screen Shot 2022-04-03 at 2.07.30 PM.png
 
Thank you for your contribution to this forum @tsialex I finally got a power supply to repair my 5,1 and preparing to move to Big Sur with OCLP. Did a romdump to my surprise there was little space left. #4916 Here is the before and after 4 x reset.

View attachment 1985920View attachment 1985921
After the forced garbage collection you have the expected value for a long running NVRAM volume with a similar config of the one on your signature. The NVRAM is now the Achilles hell of MacPro5,1 and you have track it for some time, if your available space is frequently in the low tens or even worse (OCLP have more entries on the NVRAM than pure OC) time to think on getting a upgraded/never booted BootROM image.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.