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

Romanesco

macrumors regular
Original poster
Jul 8, 2015
126
65
New York City
I received my Mac Pro (2019) on December 19 and it has been working fine since. Last week I added a PCIe card with a Samsung 960 EVO 500GB Solid State Drive for my Windows/ Bootcamp. Last night (Dec 28) I had my first issue -- the Mac Pro wouldn't turn back on after going to sleep (in Windows). Today, I had the same exact thing happen but in macOS. The only way to boot it back up was to pull out and reconnect the power cable.

Before blowing this up any further, has anyone experienced similar issues so far with or without add-ons?

Ordered configuration:
  • CPU: 16-Core 3.2GHz Intel Xeon W
  • GPU: 1xAMD Radeon Pro Vega II Duo
  • RAM: 32GB (to be replaced by 768.0GB OWC Memory Upgrade Kit)

Added parts:
The Mac Pro is hooked to the Pro Display XDR.
 
As an Amazon Associate we earn from qualifying purchases.
I have a bare bones base model, built in CN. I've had 1 unexpected shutdown (reboot) while it was asleep. Couldn't see anything interesting in the logs. I did have some issues with kernel extensions, at one point the machine hung at boot, but I had migrated from an ancient, cruft filled system. I did a reinstall and disabled lots of old garbage and it's been good since.
 
Funny I’ve restarted mine a few times and it sits for ages goes down after a few minutes then doesn’t restart automatically. I’ve added a PCI RME Madi card and a Samsung 860 pro in the Pegasus j2i cage. 16 core.
 
It looks like the two incidents were isolated. Hasn’t happened since and it’s been running flawlessly.
 
Last edited:
Just had it happen again. System was sleeping both times, pounded the keyboard to wake it up in the morning and no response, the system was powered down. No power failures, system set to restart after power interruption.
 
Doesn't that sound like 'Bridge OS' issues - related to that T2 chip?
 
  • Like
Reactions: Marekul and g75d3
RE: Samsung SSDs & M,2 drives.

Might be worth running Samsung Magician in Windows to check for firmware updates.
But of course all Samsung drives will need a Windows format partition to be seen by Magician.

For example when the Samsung 970 EVO Plus drives were released they would cause shutdowns.

Upgrading the firmware with Samsung Magician fixed that problem.
 
Last edited:
  • Like
Reactions: Romanesco
I was having random shutdown issues with my new Mac Pro.
  • CPU: 16-Core
  • GPU: 1xRadeon Pro Vega II
  • RAM: 32GB (Apple) + 2x32GB(Samsung)
Applecare seemed to think it was due to a faulty power supply. They sent me to my local 'Genius' bar to have the power supply replaced, but when I brought it in they said they had no ability to work on Mac Pros yet and did not have any parts. Then they told me it could be my RAM. Although I'm pretty sure the shutdowns were happening before I added the Samsung RAM. Plus, I was super careful to buy the correct RAM and install it in the proper slots.

I didn't wanna go down the road of troubleshooting hardware on a brand new $11k computer so I returned the damn thing.
 
  • Like
Reactions: Theophilos
I had a similar issue with my 16c Mac Pro that caused it to boot up in 5-6 minutes. After some troubleshooting, I discovered it was related to a couple of TB2 hard drives that I had connected via Apple's TB2 to TB3 adapter. Once I removed those drives, everything is much speedier. My humble suggestion is to remove the SSD and see if that (or the adapter) is the cause. Also, as someone mentioned, the Evo firmware may need to be upgraded to work reliable on MacOS, especially if it is an Eco Plus.
 
Yes, I found that dialogue box (hidden behind relaunched windows.) The issue is that hitting the report button just closed the dialogue box with no actual report being generated and I did a thorough search but there was no logging of the actual cause of the crash, just logging stopped at the time of the crash. It knew it was improperly shutdown, but was unable to record the reason (likely due to it being in sleep mode at the time.) I've disabled "power nap" as it seemed to cause mail to go into offline mode anyway and if the system is asleep I don't need or want it doing anything. It has been stable for the last week.
 
You could try this command in the terminal:

log show --predicate 'eventMessage contains "Previous shutdown cause"' --last 72h

and adjust the bolded, underlined part above for the number of hours within the last crash has occurred . This narrows down just how far the analysis of the logfiles will go back.

It should come back with a numeric code for the shutdown reason(s) you can use this document to refer to:
 
  • Like
Reactions: Marekul
Yep, I did that. No hits, like I said the logs were empty. The system only seemed to crash when asleep which is why I think it isn't getting logged.

if it crashes while asleep it could be the pcicard not understanding Mac sleep. set sleep time to 1 minute, see if it hangs. Then turn the system prefs > power settings to prevent drive spin down and prevent sleep. That should solve it.
 
Well disabling power nap was not helpful for avoiding the panic. She crashed again :-( The system has no third party cards or hardware, it is 100% stock. It is able to sleep and wake with no issues most of the time, just seems to kernel panic about once every week or two while sleeping. Never when in use. The Report Panic dialogue box after restart always just disappears when I hit the "Report..." button, which doesn't bode well for Apple getting relevant crash reports.

The difference is that this time there does seem to be a log entry for the crash:

❯ log show --predicate 'eventMessage contains "Previous shutdown cause"' --last 72h --info --debug
Filtering the log data using "composedMessage CONTAINS "Previous shutdown cause""
Timestamp Thread Type Activity PID TTL

2020-01-25 12:17:03.065309+0900 0x102 Default 0x0 0 0 kernel: (AppleSMC) Previous shutdown cause: 5

This time stamp is when I attempted to wake her (don't judge me, it is Saturday over here.) The prior crashes were at random times in the middle of the night (likely due to power nap trying to operate, which would also explain the lack of prior logs.) Based on similar reports I'm willing to write this off as a Catalina issue and wait for the next release to see if they include a fix for this.

Just to be clear, this isn't a serious issue for me, I'm simply reporting my experience to the community to help assist with identifying emergent issues with new hardware.
 
  • Like
Reactions: dclivejazz
according to those, cause number 5 seems to be an "no issues shutdown". Strange.
 
Well disabling power nap was not helpful for avoiding the panic. She crashed again :-( The system has no third party cards or hardware, it is 100% stock. It is able to sleep and wake with no issues most of the time, just seems to kernel panic about once every week or two while sleeping. Never when in use. The Report Panic dialogue box after restart always just disappears when I hit the "Report..." button, which doesn't bode well for Apple getting relevant crash reports.

The difference is that this time there does seem to be a log entry for the crash:

❯ log show --predicate 'eventMessage contains "Previous shutdown cause"' --last 72h --info --debug
Filtering the log data using "composedMessage CONTAINS "Previous shutdown cause""
Timestamp Thread Type Activity PID TTL

2020-01-25 12:17:03.065309+0900 0x102 Default 0x0 0 0 kernel: (AppleSMC) Previous shutdown cause: 5

This time stamp is when I attempted to wake her (don't judge me, it is Saturday over here.) The prior crashes were at random times in the middle of the night (likely due to power nap trying to operate, which would also explain the lack of prior logs.) Based on similar reports I'm willing to write this off as a Catalina issue and wait for the next release to see if they include a fix for this.

Just to be clear, this isn't a serious issue for me, I'm simply reporting my experience to the community to help assist with identifying emergent issues with new hardware.

not power nap. Turn the setting that says prevent computer from sleeping and prevent the drives from spinning down. Power nap will not help.
 
Dude... LOL, I know the difference. I want it to sleep, I will risk the occasional panic. I'm reporting my experience in case others are hitting the same issue so we can find a commonality and potentially a workaround.
 
Dude... LOL, I know the difference. I want it to sleep, I will risk the occasional panic. I'm reporting my experience in case others are hitting the same issue so we can find a commonality and potentially a workaround.

Sorry, my bad then. Sorry to hear it's not working right for you.
 
Continuing my crash "reports..."

Updated to 10.15.3 last night. This morning it was powered down, booting showed the "Unexpected Shutdown" dialogue. This time clicking report didn't silently close, it gave this:

BAD MAGIC! (flag set in iBoot panic header), no macOS panic log available

Progress? :p
 
Bakafish

Did you shutdown & remove the PCIe NVME adapter along with the ( now old ) 960 EVO ?

And then do a NVRAM reset 3 times consecutively ?
 
Bakafish

Did you shutdown & remove the PCIe NVME adapter along with the ( now old ) 960 EVO ?

And then do a NVRAM reset 3 times consecutively ?

As I mentioned above, this is a 100% stock Pro, no additional PCI cards or storage. The issues I am experiencing seem to be T2 related bugs coupled with Catalina and are affecting the iMac Pro and new 16" machines as well. The last update is now at least able to report the crash happened back to Apple, to they will hopefully start collecting analytics to the extent of the issue.

I'm also not asking for any advice on what to do, this is just reporting in case other owners are seeing the same thing so we can get a handle on how frequent it is or if it is tied to a particular configuration.
 
  • Like
Reactions: thexash
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.