Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

bxs

macrumors 65816
Original poster
Oct 20, 2007
1,180
553
Seattle, WA
Subject: My iMP's T2 chip ARM coprocessor has crashed 4 times this past week (I think - speculation)

In Console I've found 4 instances of what I think maybe the T2 chip crashing as the crash/panic report file name such as panic-full-2018-01-25-072847.151.ips has a suffix of ips and some of the data presented in the crash report shows "Bridge OS".

This came to my attention as twice now in past week when I've booted my iMP a display window pops up saying the iMP encountered a problem and do I want to report it.
 
I hate to be that guy but anything T2 related isn't going to be solvable by anyone here and should be absolutely passed on to engineering for future fixes in macOS.


Oh absolutely. But so far it's just speculation. We can indeed help parsing the console results to figure out more about if it really is the T2.

Speaking of which, could we get more details about those references to "BridgeOS"?
 
I wouldn't mind taking a peek at those panic reports ... I look at a lot of crash reports for a living. I can't promise anything, but it would be helpful to have two of these panics to compare against to see if it's a common cause.
 
I wouldn't mind taking a peek at those panic reports ... I look at a lot of crash reports for a living. I can't promise anything, but it would be helpful to have two of these panics to compare against to see if it's a common cause.


Agreed. And if so, what more specifically could be causing it. A specific system call or whatever.
 
Agreed. And if so, what more specifically could be causing it. A specific system call or whatever.
Oh man if that's the case I'm writing up a vulnerability submission to Apple on the spot. I kinda doubt we'll find that's the case though. I think it's more likely it'll be a lost connection or something.
 
Oh man if that's the case I'm writing up a vulnerability submission to Apple on the spot. I kinda doubt we'll find that's the case though. I think it's more likely it'll be a lost connection or something.

Yeah - but you never know! The T2 is brand new.... Hmmm. Apple pays you for finding vulnerabilities...Think we can make a million on this one? hehe
 
Yeah - but you never know! The T2 is brand new.... Hmmm. Apple pays you for finding vulnerabilities...Think we can make a million on this one? hehe
According to their bug bounty system, finding a flaw in the secure boot system (T2 chip) would be a 200k payout. They also potentially double the payout if you submit a code fix, but obviously that's not really an option here. So at best it's 200k. Unless we weaponized the vulnerability by accessing the API that causes the kernel panic from java script and targeting the safari web browser ... and also proved that the kernel panic could be controlled in some fashion. That'd be worth a million to someone like Zerodium, or ZDI.

Edit: Again, pipe dream, 99.9% sure it wont happen here.
 
If the entire T2 chip crashed you'd have a kernel panic, as access to the SSDs would be lost entirely, as well as the decryption/encryption engine

Well, when this happened twice it was during a boot Startup and the iMP hung at the white Apple logo and went no further until I powered it off and back on.

Also, I believe based on the date & time stamp of one of the crash reports it appeared the T2 crashed (again speculation on my part) while my iMP was shutdown over night. When I booted it up in the morning I was presented with a window message saying the computer had an error and did I want to report it.

If the T2 crashes it could very simply stall the boot process presumably and not present macOS with a kernel panic.

BTW.... I do have Full Security enabled.
 
Well, when this happened twice it was during a boot Startup and the iMP hung at the white Apple logo and went no further until I powered it off and back on.

Also, I believe based on the date & time stamp of one of the crash reports it appeared the T2 crashed (again speculation on my part) while my iMP was shutdown over night. When I booted it up in the morning I was presented with a window message saying the computer had an error and did I want to report it.

If the T2 crashes it could very simply stall the boot process presumably and not present macOS with a kernel panic.

BTW.... I do have Full Security enabled.
Interesting ... so while the computer was shut down, the T2 was still powered on and doing ... what exactly? Not sure what security features need to take place while a machine is powered down. Now I'm REALLY interested to see that crash log, if only to see the stack trace.
 
  • Like
Reactions: casperes1996
I'd call apple and exchange the computer, given the high price, I would be packing it up and sending it back
 
  • Like
Reactions: old-wiz
I'd call apple and exchange the computer, given the high price, I would be packing it up and sending it back

It has been reported to Apple and I allowed them to gather up all the diagnostic data they wanted on my iMP and it then was uploaded to them for engineering to inspect. It will be of interest to hear back from them as to what this is all about and if there's any need for concern.

My iMP is running pretty darn smoothly apart from this peculiarity and the Space Gray KB taking on a life of its own at times by suddenly echoing weird and unwanted characters when I type. It will suddenly start echoing characters that I've not typed and I have to quickly try and stop it. If I try stopping it with me pressing "delete" it will stop but then start deleting backwards at its own pace and if left alone would delete all my entered text. I have to quickly grab the KB and hold it up and turn it off and then back on to have the KB return to normal behavior. It is VERY weird to watch it happening.
 
That first one is crashing due to incompatibility with a logitech device.

The second one is crashing due to the CPU / T2 processor becoming unresponsive to each other. This is likely a firmware issue. This is likely similar to the issue OP is discussing.

The third is due to the configuration of SMB settings, as handled by 'Automatic' - there is a workaround and Apple has been notified of the problem, and is working on a fix for this particular issue.
 
I’ve been having nightly kernel panics for the past few weeks on my iMac Pro. Other than that it’s been running like a dream. The only weird behavior has been that I can’t wake the display sleep or from screensaver with the keyboard (space gray) or by moving the mouse. I have to actually click the mouse. Wondering if you wouldn’t mind looking at my crash reports to see if there’s an obvious cause?
I wouldn't mind taking a peek at those panic reports ... I look at a lot of crash reports for a living. I can't promise anything, but it would be helpful to have two of these panics to compare against to see if it's a common cause.
 
I'm wondering if they included T2 updates in this 10.13.6 supplemental.

I've noticed some weird behavior before the update as well.
 
I'm wondering if they included T2 updates in this 10.13.6 supplemental.

I've noticed some weird behavior before the update as well.

I have had no KPs since 10.13.6 was installed. I think the fixes were in the 10.13.6, not the supplemental which iMac Pro didn't get?
 
I had my first BridgeOS panic yesterday:

Code:
{"caused_by":"macos","macos_system_state":"running","bug_type":"210","os_version":"Bridge OS 2.3.2 (15P6059)","timestamp":"2018-08-01 19:56:24.94 +0000","incident_id":"FC91D77B-C77B-4FAA-A20F-2806EE39A00F"}
  "build" : "Bridge OS 2.3.2 (15P6059)",
  "crashReporterKey" : "c0dec0dec0dec0dec0dec0dec0dec0dec0de0001",
  "date" : "2018-08-01 19:56:24.81 +0000",
  "incident" : "FC91D77B-C77B-4FAA-A20F-2806EE39A00F",
  "kernel" : "Darwin Kernel Version 17.7.0: Fri Jun 15 23:12:25 PDT 2018; root:xnu-4570.71.1~3\/RELEASE_ARM64_T8010",
  "macOSOtherString" : "\n** In Memory Panic Stackshot Succeeded ** Bytes Traced 165344 **\n",
  "macOSPanicFlags" : "0x404",
  "macOSPanicString" : "panic(cpu 14 caller 0xffffff801d0b51c4): \"Invalid waitq: 0\"@\/BuildRoot\/Library\/Caches\/com.apple.xbs\/Sources\/xnu\/xnu-4570.71.2\/osfmk\/kern\/waitq.c:4979\nBacktrace (CPU 14), Frame : Return

First I suspected it was a power outage, but then I noticed the crash.

It happen suddenly with no substantial load (light editing of source in the Xcode). I did not lose any data, but it is supper annoying to have your reliable professional Mac suddenly rebooting during work.

I noticed there is a thread in the Macbook Pro forum. I was wondering if anyone had any useful suggestions from Apple Support before I go to them. I am not keen to spent my time on chat just to hear I should reinstall the system from scratch, reset PRAM or bring it in for service....
 
Well, when this happened twice it was during a boot Startup and the iMP hung at the white Apple logo and went no further until I powered it off and back on.

Also, I believe based on the date & time stamp of one of the crash reports it appeared the T2 crashed (again speculation on my part) while my iMP was shutdown over night. When I booted it up in the morning I was presented with a window message saying the computer had an error and did I want to report it.

If the T2 crashes it could very simply stall the boot process presumably and not present macOS with a kernel panic.

BTW.... I do have Full Security enabled.
A year later and this is still a problem. Same thing here and I do have security disabled and not using file vault nor siri. Latest Mojave not helping at all. Only happens on startup. frequently crashes to "off", like the post I quoted. Bridge OS crash in the report. Whatever it is they have not fixed anything as far as I can tell. Maybe TB3 peripheral related? But isn't that the point of TB3? Useless to me without that.
 
After the update to Mojave I haven't had any issues except from my Razer peripherals. I am looking to replace them sadly, due to the instability they bring to my system.
 
I think this is mostly occurring with TB3 connections of one sort or another coupled with using some of the T2 functions that are related to the Bridge OS that it controls. Not everyone would see or notice this issue otherwise. That said, I suppose it is possible that only certain vintage T2 chips are suspect, but I'm just guessing. It isn't exactly easy to swap out a 10 core custom imac pro and quickly test it out. Apple forums have pages and pages on this subject for the macbook and imac pros with the T2.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.