`kernel_task` is taking 100%+ CPU constantly

I found power metrics: https://developer.apple.com/library/mac/documentation/Darwin/Reference/ManPages/ man1/powermetrics.1.html

Here is some interesting output from my power metrics:

**** Interrupt distribution ****

CPU 0:

Vector 0x57(EHC2): 9.19 interrupts/sec

Vector 0x73(XHC1): 0.60 interrupts/sec

Vector 0x77(GFX0): 7.39 interrupts/sec

Vector 0x78(SATA): 8.79 interrupts/sec

Vector 0x79(ARPT): 7.39 interrupts/sec

Vector 0xdd(TMR): 97.71 interrupts/sec

Vector 0xde(IPI): 28.37 interrupts/sec

CPU 1:

Vector 0xdd(TMR): 100.11 interrupts/sec

Vector 0xde(IPI): 56.15 interrupts/sec

CPU 2:

Vector 0xdd(TMR): 12.59 interrupts/sec

Vector 0xde(IPI): 37.57 interrupts/sec

CPU 3:

Vector 0xdd(TMR): 11.79 interrupts/sec

Vector 0xde(IPI): 35.37 interrupts/sec

Interrupts are uninteresting (not any significantly high amount)

**** Processor usage ****

Core 0 C-state residency: 97.64% (C3: 0.00% C6: 97.64% C7: 0.00% )

Core 1 C-state residency: 0.00% (C3: 0.00% C6: 0.00% C7: 0.00% )

Core 2 C-state residency: 98.46% (C3: 0.00% C6: 98.46% C7: 0.00% )

Core 3 C-state residency: 97.18% (C3: 0.00% C6: 97.18% C7: 0.00% )

Note that one core is 0.0% across the board. I also notice that this changes to other cores.

I don't usually use 1/8" headphones (usually I use USB headphones). Plugging in headphones did not affect anything that I could see.

Without performing any reports, you can evaluate the basic problem by going to Applications/Utilities/Activity Monitor. When you look at the CPU and Network columns, you will find that an activity predominates. That's that's causing the excessive CPU usage.

Armed that that information, you can ask additional questions. And you can find a way to stop that CPU problem.

I know what is causing excessive CPU usage - the `kernel_task`. There is not significant disk or network activity. There is no increase in RAM usage. This problem is only with the CPU and seems to be limited to a single core.

This problem seems very much like a script or app that is stuck in a very tight infinite loop, or a very quick start/error/shutdown/restart loop. This problem is very difficult to debug because there are no messages being logged to the system log to indicate anything happening in a tight loop.

see if this helps you sort out the problem.

How to troubleshoot a kernel panic | Macworld

Thanks for the input Donald, but my computer is not having kernel panic attacks.

Hi TinusRuben,

Just wanted to update you and say that I might have solved my problem. I sat down and looked at the Console.app logs today and noticed I was seeing a lot of this error message:

12/11/15 3:34:44.161 PM WindowServer[168]: send_datagram_available_ping: pid 229 failed to act on a ping it dequeued before timing out.

12/11/15 3:34:53.880 PM xpcproxy[64942]: LaunchServices: received XPC_ERROR_CONNECTION_INTERRUPTED trying to map database

12/11/15 3:34:53.880 PM xpcproxy[64942]: LaunchServices: Database mapping failed with result -10822, retrying

12/11/15 3:34:53.897 PM com.apple.xpc.launchd[1]: (com.hp.devicemonitor[64942]) LaunchServices returned a bundle URL that does not match with the LoginItem's known association.

12/11/15 3:34:53.898 PM com.apple.xpc.launchd[1]: (com.hp.devicemonitor) Service only ran for 0 seconds. Pushing respawn out by 10 seconds.

The offending service was `com.hp.devicemonitor`. This service did not show up when looking at the output of the following command:

sudo launchctl list

I did some googling around and found this thread:

http://h30434.www3.hp.com/t5/Scanning-Faxing-and-Copying/OfficeJet-7612-Scanning -problem-in-Mac-Yosemite/td-p/4758673

It's worth noting that I only had the following file:

com.apple.print.custompresets.forprinter.HP_Officejet_Pro_8100.plist

I removed it and restarted my computer. It seems to have worked! After a couple hours I don't have a kernel_task taking 100%+ of my CPU. I will update again if it continues to happen.

Good Luck!

My 100% kernel_task came back a few days ago. I don't know what caused it - I have not used my computer much in the last few weeks.

I ran a spin dump while this was happening, here's the interesting result:

Process: kernel_task [0]

Path: /System/Library/Kernels/kernel

Architecture: x86_64

UID: 0

Version: Darwin Kernel Version 15.2.0: Fri Nov 13 19:56:56 PST 2015; root:xnu-3248.20.55~2/RELEASE_X86_64

Task size: 408374 pages

CPU Time: 10.062s

... threads removed ...

Thread 0x3ac 1000 samples (1-1000) priority 80 (base 80) cpu time 9.990s

*1000 call_continuation + 23 (kernel + 1884023) [0xffffff80003cbf77]

*1000 _IOConfigThread::main(void*, int) + 316 (kernel + 6895708) [0xffffff800089385c]

*1000 IOService::doServiceMatch(unsigned int) + 554 (kernel + 6871162) [0xffffff800088d87a]

*1000 IOService::probeCandidates(OSOrderedSet*) + 2358 (kernel + 6892486) [0xffffff8000892bc6]

*1000 IOService::startCandidate(IOService*) + 109 (kernel + 6893757) [0xffffff80008930bd]

*1000 AppleIntelCapriController::start(IOService*) + 3099 (AppleIntelFramebufferCapri + 81425) [0xffffff7f82f8ae11]

*1000 AppleIntelCapriController::initPMRegisters() + 161 (AppleIntelFramebufferCapri + 27927) [0xffffff7f82f7dd17]

*957 SafeForceWake(bool) + 176 (AppleIntelFramebufferCapri + 71022) [0xffffff7f82f8856e] (running)

*30 SafeForceWake(bool) + 172 (AppleIntelFramebufferCapri + 71018) [0xffffff7f82f8856a] (running)

*13 SafeForceWake(bool) + 179 (AppleIntelFramebufferCapri + 71025) [0xffffff7f82f88571] (running)

Now at least I know what is causing it, but I have no idea how to resolve it. Searching for "SafeForceWake" or "AppleIntelCapriController" does not yield interesting results.