Kernel Panic on 16" i9 MBPwRD

Had the same problem, solved it 4 days ago.

Status here

My GPU Panic sleep problem is still solved. I have since installed my video-rendering tools, my developer tools, all my other stuff.

My energy saving stuff is back to normal.

Having it gooing to sleep on power, on batteries no problems so far.

Running 10.15.4

Solution for me was

My brand new macbook pro 16 (Shipped 2. april 2020) also restarted (crashed) over nigth - actually it crashed (Kernel Panic GPU) everytime it went to sleep mode. I had nothing installed except default OSX with updates.

Solution details

1. Reset: Pram & NVRAM (still crashed)
2. Reset: SMC (still crashed)
3. Internet Recovery Command+Option+R (when booting) (Crashed under installation, but continued afterwards)
4. Disabled "Prevent computer from sleeping automatically when the display is off." & "PowerNap.." in Energy Saver settings (Problems disappeared)
5. Two days later I enabled "Prevent computer from sleeping...." and "Powernap..." again (Still no problems)

I don't know if step 1-3 is necessary, but I think so.

So a macbook pro 16 that I almost returned, is now working perfectly - and I'm very happy:-)

Looks like 10.15.5 fixes that issue:

https://forums.macrumors.com/threads/some-users-experiencing-system-crashes-on-macos-10-15-4-especially-during-large-file-transfers.2229749/post-28341955

https://www.macrumors.com/2020/04/06/macos-10-5-4-kernel-panic-crashing-issues/

Finally in the news...

I finally heard back from Apple support today. The rep acknowledged that Engineering is aware of this issue, and asked a series of questions (some quite general, so it seems they haven't made it very far debugging the issue).

I briefly perused the XNU source mentioned in the kernel Panics we're all getting; the logic is related to a fairly general IOKit abstraction that manages the power state transitions of the mentioned driver (AppleIntelFramebuffer) or a parent of that drive. This appears to be the same code path used by any driver that handles power state transitions (cf. IOService::ackTimerTick() in IOServicePM.cpp).

Given the the actual problem is that the Intel driver isn't changing it's state within 45 seconds (which seems beyond ample for these machines), I fear this may take some time for the worst of reasons. Unless apple actually has a license to the driver's source and/or detailed and accurate hardware specs from Intel, Apple engineers might be facing a binary blob that requires Intel engineers to actually address (and Apple to integrate into some future update). But I'm not blaming Intel either; the bad power state could easily have come from Apple missing their API. Regardless, we're likely going to have to wait for two companies to debug and oddly intermittent issue. FWIW, I opened FB7658531 and offered to give full access to my machine to help them diagnose, but it seems that Apple has an *insane* policy that prevents their engineers from working directly with customers.

I know this information is only useful to the most technical users, but hopefully it saves somebody some time. If I get any additional information or status updates from Apple, I'll be sure to share, but again because of an insane policy, they don't give any updates until *after* the fix is released.

Here's the relevant code from IOServicePM.cpp in the XNU kernel that led to the above conclusions:

bool IOService::ackTimerTick( void )
{
    IOPMinformee *      nextObject;
    bool                done = false;

    PM_ASSERT_IN_GATE();
    switch (fMachineState) {
        case kIOPM_OurChangeWaitForPowerSettle:
        case kIOPM_ParentChangeWaitForPowerSettle:
            // are we waiting for controlling driver to acknowledge?
            if ( fDriverTimer > 0 )
            {
                // yes, decrement timer tick
                fDriverTimer--;
                if ( fDriverTimer == 0 )
                {
                    // controlling driver is tardy
                    uint64_t nsec = computeTimeDeltaNS(&fDriverCallStartTime);
                    OUR_PMLog(kPMLogCtrlDriverTardy, 0, 0);
                    setProperty(kIOPMTardyAckSPSKey, kOSBooleanTrue);
                    PM_ERROR("%s::setPowerState(%p, %lu -> %lu) timed out after %d ms\n",
                        fName, OBFUSCATE(this), fCurrentPowerState, fHeadNotePowerState, NS_TO_MS(nsec));

// <snip>

if (gIOKitDebug & kIOLogDebugPower)
                    {
                        panic("%s::setPowerState(%p, %lu -> %lu) timed out after %d ms",
                            fName, this, fCurrentPowerState, fHeadNotePowerState, NS_TO_MS(nsec));

HTH...

Same Kernel Panic and Sleep issues here with my 16" Macbook Pro that I purchased in December of 2019. Apple final asked me to bring in for repair. The replaced the entire Logic Board. I have now gone 3 weeks without an issue. Looks like this model has a Logic Board issue and it's not software related.

AppleSupportOntario wrote:

NOTE: THIS ADVICE IS NOT FROM ME, I FOUND THIS FROM ANOTHER APPLE SUPPORT FORM. USE AT YOUR OWN RISK

Sleep issue are often sorted by

resetting the SMC https://support.apple.com/en-us/HT201295

If no resolve—

Try a SafeBoot https://support.apple.com/en-us/HT201262

Takes noticeable longer to get to the login screen, does a 5-15 minute disk repair before it fully boots up, and certain system caches get cleared and rebuilt, including dynamic loader cache, etc. Login and test. Reboot and test

This test will tell you if third party interference; extensions etc are not loaded in safe boot mode.

If you still need further assistance—

To get a good look at your System config. for conflicts or issues, you can run this trusted utility http://etrecheck.com

If you need help interpreting the report you can post it here in its entirety in the "Additional Text" box in the editing toolbar below, in your reply.

You can set its preference to "Allow full Disk Access", with this you get a digest of issues from the last 7 days that are saved in your system.

I'd be curious to know this other " ANOTHER APPLE SUPPORT FORM." and if you could site the specific link.

I have the same kernel panics on pretty much the same configuration. Tried all kinds of resets, fresh install, .. nothing helped.

The latest thing I have tried so far prevented crashes: I turned off Power Nap in the energy settings. Can someone confirm whether they also see crashes without Power Nap? I would still expect Apple to fix that of course.. but at least if that reliably prevents the kernel panic, I could live with that workaround for a while.

Yes.. given that the panic only happens during/while entering sleep mode, deactivating it will definitely help as well. I didn't want to deactivate it completely, though. Now, with power nap off and no crash since, the next step obviously would be to figure out what exactly during power nap causes it (could be anything like checking mail, update photos or syncing iCloud). Then again, even if we did know, none of them can be turned off individually anyway.. so I might just as well keep power nap off in full.

As I mentioned before, the problems disappeared with SMC/VRAM-Reset plus to shot down and after a view minutes to turn on the Mac.

Now the device is working fine for two days at least. No crashes at all. I will let you know if the wind is changing again.

It is definitely the powernap option. I have an OWC Thunderbolt 3 dock with two external EIZO screens. Without powernap no Kernel Panic with powernap always Kernel Panic. As I do a timed restart all mornings 4:00 am I see the result right after login at 7:15am. Thanks for the tip with powernap.

I did not reset nvram nor smc.

Very helpful - brand new MacBook Pro fresh out of the box, no external devices and updated to latest showing the same problem - the default 'power nap' option is on under the power adapter settings - turning that off has so far stopped it.

I have a cal digit hook up and an external monitor attached to it. I had this same issue. I called the apple specialist first. We did a SMC and NVRAm reset. It didn't help.

What I did was...

Sys preferences - energy saver - CHECK prevent computer from sleeping automatically when the display is off - UNCHECK put hard disks to sleep when possible.

Haven't had an issue since. Hope this helps!

OK.... MacbookPro 16 (from march/april): Friday my macbook updated from 10.15.4 to 10.15.5 (19F96). That reintroduced the Sleep-Nap-Kernel-Panic (Kernel Panic when sleeping and on power). The disable "Power Nap" trick solved the problem in the weekend.

Then this morning (in EU), my computer did a huge update again. 15 minutes. Did not ask questions (none). Bum... New version of 10.15.5 was installed: Going from build 10.15.5 (19F96) to 10.15.5 (19F101).

2020-06-02 09:56:35+02 CaspersMBP16 softwareupdated[366]: softwareupdated: Starting with build 10.15.5 (19F101)

2020-06-02 09:56:50+02 CaspersMBP16 softwareupdated[366]: Previous System Version : 10.15.5 (19F96), Current System Version : 10.15.5 (19F101).

Talking about panic.. Apple releases a brand new 10.15.5 build three days after the previous one.

Lets hope that 10.15.5 (19F101) solves the problems, because 10.15.5 (19F96) did not!!!

So I have Gen 1 of Macbook Pro 16 (base model). I had this problem (see comments above). After I turned off Power Nap in the earlier versions, the problem went away. When 10.15.5 was released I turned PowerNap on again. I haven't had a problem since. I do have a my TimeMachine drive plugged into machine, also overnight. No problems. I haven't updated .6 yet.

I have now gone 7-8 weeks without a Kernel Panic issue since Apple replaced my Logic Board. I no longer have to turn off any of the system preferences listed in many of the posts in this thread (ie. Automatic Graphics Switching, Enable Power Nap, Put Hard Drive to Sleep when Possible, Prevent Computer from Sleeping Automatically when the Display is off). Appears to be a Logic Board Issue with the 2019 16" MacBook Pro.

Yes!! Powernap disable and today morning no restart! Problem solved for now... thanks!