System using a full core after resume from sleep, sometimes USB not functional

I'm having a problem with my new notebook -- whenever I resume from sleep, something in the System process gets stuck and I end with a single (hyperthreaded) Core being 100% used. Additionally, sometimes all the USB ports are non-functioning after the sleep, and any system task that I try to do (install a program, shut down, etc) will never complete. Other than this, it's totally usable -- apps/games/graphics/audio all work fine. Just any system-level operation hangs, and I've got a full core used.

Obviously something in the OS is getting stuck, presumably a driver somewhere, but I'm having a hard time tracking down what it is. I've system restored to a checkpoint from last month, and it's still happening. I've tried using Xperf to drill into what's using CPU in that process, but when it tries to write the trace file it hangs just like all other system processes. Process Explorer doesn't let me drill down into that System process sufficiently to figure out what's causing it.

I'm at the point where I'm pretty much going to have to do a full OS install to fix this (and hope it doesn't come back), which is pretty frustrating -- I just got this computer all set up and installed with all my software, and I'd really rather not do it all again.

Anyone else seen this? Any ideas of how I could debug it?

Some system specs:
Clevo P170SM
Windows 8 64 bit
i7-4900MQ
16GB RAM
Nvidia 780M
1TB of Raid-0 SSDs (two Samsung 840s)

 

Interesting... the W350ST that I had did the same thing. And after several tests, I concluded that it's either that the RAM slots that were faulty or they just hate the stock RAM that came. It's almost like the RAM had bad contact, yet it's completely random if the system likes it or not and thus either doesn't give a POST screen at all or boots with all of the problems that you listed (and it liked all of my older gen RAM except for the Patriot memory that the system came with). So presumably I had a bad stick of RAM because memtest detected an error with one of them.
During the rearrangements, I encountered Core 1 being stuck at 100%, fingerprint reader disappears, USB ports non-functional, and failed to shutdown without hard powering off.

I got that replaced and ...

the SAME thing had happened when I got a brand new stick from Sager. After inserting a brand new stick onto one of the slots and testing it out with memtest, I found zero errors for like 2 passes, but that hung at the second pass. So I decided to boot into Windows and guess what... I had 100% core usage on core 1, could not shut down without hard shutting down, fingerprint reader disappeared, and USB ports completely non-functional. The test proved repeatable without removing the RAM. I booted it a second time and the same result.
I reinserted the memory and all problems disappeared.

And to add onto that,
It took me forever to get both 8GB sticks to work together. The W350ST just did NOT like those sticks. Took me over 20 boots and many reinserts to get it two Patriot Memory sticks to work with one another and get past POST screen. Furthermore, each stick passed memtest of 8 full passes, so individually they seemed to work.

 

Huh, interesting. I'll definitely try out memtest and see what happens.

The one difference is that the system works 100% when I do a normal boot, literally never had a problem there. It's exclusively coming out of Sleep that it goes into this full-core-no-usb-won't-shutdown-cleanly mode.

 

But it seems like both machines have the same issues once it gets into that "full-core-no-usb-won't-shutdown-cleanly mode" state even though yours seems triggered by sleep. If you want to be sure, try booting it up a couple times in a row to see if it reproduces a similar state from plain boot up. Of course, it's not going to happen on the same spot on the RAM... the system uses different parts of the RAM at different times, and was more easily diagnosed when I used only one of the 8GB sticks in the slots.

It's possible that if you do what I did and just reinsert ram... that it gets seemingly fixed completely. (but only after all else fails)
And don't forget to test ONE stick at a time... because you never know which one it is.

 

Have the same issue on my W230ST....

 

I ended up doing a full Windows 8 reinstall and haven't seen this since. I suspect this was caused by installing uxstyle, which I'd recommend people avoid. Particularly because when I uninstalled it it causes a bug where your computer isn't bootable (even in safe mode).

 

Usually if you have an unknown application that is using 100% that application is either a trojan/virus/whatever or that application is infected. That is not normal behavior.

I would suggest trying to fix it rather than doing a full reinstall. Hitman Pro for example free for 30 days. Run that. Malwarebytes is always free, run that. Roguekiller is also free, run that.

If I had to guess: You installed an infected Adobe Flash, rss.exe. New install, new browser, saw message update flash and you did it. Even if it came from a popup on your browser. Just my suspicion. That rss.exe willl use 100% of your CPU, plenty of threads out there people trying to get rid of it. My solution above will fix this trojan after you uninstall Adobe Flash from Program Files.

Also have never had issues with uxstyle and I've been running Windows 8 since it was released.

 

Intel have listed a problem with the USB ports and resuming from sleeps in the current chipset batch.

 

Well I already did a full reinstall, so that ship has kind of sailed.

I really did try everything to fix it -- scanned it with everything out there, did many system restores, etc. It wasn't a strange app using a lot of CPU, it was literally the kernel process, which led me to believe that it was either a bad driver (possible, though surprising that system restore didn't fix it), or something core to the OS had broken. If it was something as simple as an Adobe trojan I definitely would have found it. I mean, I was going through xperf traces of my system, and it was turning up nothing suspicious.

The reason I suspect uxstyle is twofold: One, that it does a fairly non-standard install/modification of windows, and Two, when I tried uninstalling it it basically bricked my OS -- there's an known issue on uninstall that causes your machine to alternate flashing something my display couldn't sync and a black background. This happens in safe mode too, and as far as I know is completely unrecoverable (google around for it, it's pretty well known) short of a complete OS reinstall.

It might not be uxstyle, like I said it was just a suspicion on my part. But I'm definitely not installing it again, and I've had no issues so far (after reinstalling basically everything else I had before the issue started). Given the severity of the issue, I'd advise people to avoid it unless you really want what it's offering.

As far as the Intel problem, I don't think that's what this was -- the lack of USB issue I think was just a side-effect of whatever in the OS was failing to resume from sleep properly. Especially since lots of other things just didn't work either -- I couldn't install/uninstall programs when it was in this state, I couldn't shut down cleanly, etc. And since it's been working fine post-reinstall, it definitely seems like a software issue.

 

For what it's worth, this is happening again and it's beyond frustrating. I've done everything I can to debug it but can't track it down to anything. Half of the tools I'd use don't work when it's in this state because something's broken in the system process.

Guess I'm going to have to reinstall Windows 8. Again.

 

I had the same problem with my P151SM. For me it always occurred after resume from hibernate but the behavior was the same as you described - one core running at 100%, shutdown/hibernate would hang, some installations would never complete. It turned out the cause was the Intel graphics driver that came with my notebook. After I downloaded and installed the newest one from the Intel website the problem hasn't occurred since.