kernel_task is using 100% CPU when waking from sleep

Question

Level 1

7 points

kernel_task is using 100% CPU when waking from sleep

Since updating my 13" mid-2013 MacBook Air to Mavericks, there's this issue that's been happening when I wake the computer up from sleep. A proccess called kernel_task uses 100% CPU (a whole core) and hangs there until the computer is restarted. This doesn't make the computer unresponsive, but it kills battery life.

So far, I've tried the following:

1. Resetting SMC, NVRAM and PRAM: no effect.

2. Safe Boot: the issue doesn't happen.

3. Deleting old apps/plugins: issue persists.

4. Closing all apps doesn't solve issue after it takes place.

5. Repairing permissions: no effect.

6. Apple Diagnostics: no issue found in hardware.

7. No kernel extensions from third-parties are installed (there are 120 kext, according to Terminal, but they all come from Apple).

8. Everything is up to date.

9. Only external plugin installed is Flash.

Given these, I reckon it's something software-related. Any tips on solving this would be welcome — things to look for in Console logs would be greatly appreciated.

MacBook Air (13-inch Mid 2013), OS X Mavericks (10.9)

Posted on Oct 28, 2013 12:03 PM

Reply

Answer 1

twinn888

Level 1

0 points

Oct 28, 2013 2:00 PM in response to HalexPereira

I also have this issue on the new Macbook Pro Retina 13. Only a restart would help. Its killing battery life and annoying to restart after sleep.

Reply

Answer 2

Oct 28, 2013 4:07 PM in response to HalexPereira

I'm in the same boat. No issues before installing Mavericks.

I did the same tests, even disabled the 3rd party kext I had (pace driver, related to some vst plugins I have) with no luck.

I ran sysdiagnostics and can see the following interrupts, taken from powermetrics.txt:

**** Interrupt distribution ****

CPU 0:

Vector 0x49(MacBookAir6,2): 31.67 interrupts/sec

Vector 0x92(IGPU): 408.07 interrupts/sec

Vector 0x96(HDEF): 216491.54 interrupts/sec

Vector 0x98(ARPT): 11.06 interrupts/sec

Vector 0x9e(SSD0): 25.10 interrupts/sec

Vector 0xdd(TMR): 367.54 interrupts/sec

Vector 0xde(IPI): 161.86 interrupts/sec

CPU 1:

Vector 0xdd(TMR): 32.17 interrupts/sec

Vector 0xde(IPI): 56.18 interrupts/sec

CPU 2:

Vector 0xdd(TMR): 934.68 interrupts/sec

Vector 0xde(IPI): 1103.50 interrupts/sec

CPU 3:

Vector 0xdd(TMR): 165.04 interrupts/sec

Vector 0xde(IPI): 242.83 interrupts/sec

I feel like that is not normal(see bold) 🙂 Anyone know what HDEF is referring to, might help tracking this down.

Reply

Answer 3

HalexPereira Author

Level 1

7 points

Oct 28, 2013 4:45 PM in response to espressodre

This smells like a good lead. 😎 Here's a snapshot of when kernel_task is acting up:

For comparison, here's what's going on when everything is normal:

Now, what the heck is (HDEF)? 😕

Reply

Answer 4

Oct 28, 2013 5:40 PM in response to HalexPereira

Hard to find anything concrete, but found some references to it on insanelymac.com and tonymacosx86 which seem to indicate that it's related to handling the inbuilt audio device - AppleHDA.kext is the kernel extension responsible (/System/Library/Extensions/AppleHDA.kext/)

Google search results http://goo.gl/hXKnNQ

Might be entirely wrong though!

Reply

Answer 5

HalexPereira Author

Level 1

7 points

Oct 28, 2013 5:52 PM in response to espressodre

Regarding what espressodre said, do you have a headphone connected when this issue comes up? I have an old Apple In-Ear always connected. Might try to see what happens if I pull it out.

It's interesting to note that Safe Boot disables sound, so it might be related, after all (I would never have guessed, though).

Reply

Answer 6

Stan_

Level 1

5 points

Oct 28, 2013 6:36 PM in response to HalexPereira

I hope Apple does something about this. The bug is definitely there. I get the exact same super high number of interrupts for HDEF after waking up on OS X Mavericks. Mid-2013 Macbook Air 13". It's not what I paid for, Apple!

Reply

Answer 7

miwelc

Level 1

0 points

Oct 28, 2013 9:36 PM in response to HalexPereira

Hi guys. I was experiencing exactly the same issue, after sleep kernel_task always used 100% cpu. While reading HalexPereira post I realized I was using earphones! I recreated the problem and then fixed it just unplugging the earphones, closing the lid and opening it again... The problem must be related with AppleHDA for sure...

Reply

Answer 8

Stan_

Level 1

5 points

Oct 28, 2013 11:50 PM in response to HalexPereira

I'm not using any headphones btw and I'm sure there is no reliable way of fixing it even temporarily. It's a complete mess this one, at least for me. Fix it, Apple, now! Why push out an OS X update which breaks things for God's sake huh? I'm not a darned beta tester.

Reply

Answer 9

HalexPereira Author

Level 1

7 points

Oct 29, 2013 1:51 AM in response to Stan_

I feel your pain, Stan_. 😢

Until further investigation, I advise sending bug reports to Apple with this info and a link to this thread. A close friend started experiencing this same issue, so I guess it's not a random problem.

Let's document the issue, send reports and hope for a fix in a 10.9.x update.

Reply

Answer 10

Stan_

Level 1

5 points

Oct 29, 2013 3:12 AM in response to HalexPereira

One more thing that would confirm that something is wrong with the HDA driver - today the sound stopped working without any apparent reason. The audio enable/disable and volume keys on the keyboard show a greyed out HUD on the screen. But the kernel_task issue on wake up is gone!!! Now I think I'm not gonna restart it, hehe I prefer it without sound and without kernel_task lol.

Reply

Answer 11

Oct 29, 2013 1:47 PM in response to HalexPereira

I have the same issue happening, kernel_task ramps to 100% and only a restart will fix it. Seems to only happen after my MBA (mid-13) has had teh lid down and comes back from sleep.

I do use headphones but I have not nailed it done to that fully. I am also getting an issue where sound will competly stop if the system goes into deep sleep.

praying for a fix soon...very very annoying.

Reply

Answer 12

HalexPereira Author

Level 1

7 points

Oct 29, 2013 2:56 PM in response to Stan_

This doesn't sound good (pun not intended). If things here escalate to this, I won't have any option other than taking it to technical assistance (we don't have Apple Stores in my country).

The addiotional evidence that things related to sound might be awry in Mavericks is interesting, though. Keep us updated on how things evolve, Stan_.

Reply

Answer 13

Oct 29, 2013 10:44 PM in response to HalexPereira

I have the same issue. But for me, the weird thing is. It happens only when I keep my headphones plugged in. WHen i remove my headphones, it stabilizes i.e Kernel_task goes back to using just 4-5% CPU. I resetted PRAM SMC etc but to no avail. Now I dont feel like using my headphones a lot because of this when I am travelling because battery life reduces to 6 hours

Reply

Answer 14

Cyber Wu

Level 1

0 points

Oct 30, 2013 7:33 AM in response to HalexPereira

Exact same problem for me. And I unplug the headphone, sleep, then wake up, it's gone!

I really hope apple can solve this quickly.

**** Interrupt distribution ****

CPU 0:

Vector 0x92(IGPU): 1241.09 interrupts/sec

Vector 0x94(XHC1): 71.54 interrupts/sec

Vector 0x96(HDEF): 216026.79 interrupts/sec

Vector 0x98(ARPT): 20.87 interrupts/sec

Vector 0xdd(TMR): 460.07 interrupts/sec

Vector 0xde(IPI): 469.01 interrupts/sec

CPU 1:

Vector 0xdd(TMR): 60.61 interrupts/sec

Vector 0xde(IPI): 227.55 interrupts/sec

CPU 2:

Vector 0xdd(TMR): 372.63 interrupts/sec

Vector 0xde(IPI): 1658.43 interrupts/sec

CPU 3:

Vector 0xdd(TMR): 185.82 interrupts/sec

Vector 0xde(IPI): 763.14 interrupts/sec

Reply

Answer 15

Oct 30, 2013 8:57 PM in response to HalexPereira

Im in the same boat ladies and gentlemen. If my 2013 macbook air wakes from sleep, kernel task will run 1 core 100% for at least 45 minutes. Thats all I could stand before I just reset. I never had this problem in mountain lion. I have tried a re-install, and a clean install. No effect.

Reply