BSODs (mostly with Skype) after switching to 10 Pro from Insider Preview

Status
Not open for further replies.

Skelebob9

Reputable
Oct 18, 2014
10
0
4,510
I've recently started getting very randomly-timed BSODs with the error WHEA_UNCORRECTABLE_ERROR (STOP: 0x124). Sometimes minutes after logging in, sometimes it can be hours after logging in. This never happened before I upgraded to Windows 10 Pro 64-bit from the Insider Preview. Before that I was using Windows 8.1.

First of all, my specs:
CPU: AMD FX-4300 Quad-Core @ 3.8GHz (8MB Level 3 Cache, 4MB Level 2 Cache)
GPU: AMD R9 270X dual-fan (2GB GDDR3)
Memory: 2x 2.00 GB @ 1333 MHz and 1x 2.00 GB @ 1600 MHz (speed lowered to 1333 MHz)
PSU: Corsair CX750 Builder Series 750W Bronze 80+
HDD: 500 GB unbranded

I usually get the crashes while browsing or gaming and they are a lot more frequent when Skype is open, but not exclusive to when Skype is open. When browsing, the crashes occur when switching to a new page, when switching to a different video on YouTube after watching one already, or when an animation is played on the page (last crash was when the small window popped up in the bottom right on this website), which makes me wonder if my graphics card is linked as every crash has happened during something that would utilise my GPU.

My DxDiag.txt

I've done some digging & testing:
• CPU stress tests don't crash the system
• Memtests don't show any errors
• chkdsk says no errors
• sfc /scannow says no error
• All hardware drivers have been updated/checked for updates
• CPU temps remain below 114°F (46°C)

I've checked the latest dump file (link here) for errors and it says the following:
BUGCHECK_STR: 0x124_AuthenticAMD
CUSTOMER_CRASH_COUNT: 1
DEFAULT_BUCKET_ID: WIN8_DRIVER_FAULT
PROCESS_NAME: csrss.exe

..then further down, it says:
MODULE_NAME: AuthenticAMD
IMAGE_NAME: AuthenticAMD
DEBUG_FLR_IMAGE_TIMESTAMP: 0
IMAGE_VERSION:
FAILURE_BUCKET_ID: 0x124_AuthenticAMD_PROCESSOR_CACHE
BUCKET_ID: 0x124_AuthenticAMD_PROCESSOR_CACHE
ANALYSIS_SOURCE: KM
FAILURE_ID_HASH_STRING: km:0x124_authenticamd_processor_cache
FAILURE_ID_HASH: {b533751b-676e-7546-bcdc-24e6c40d0064}

I've also noticed that dllhost (COM Surrogate) often exceeds 50% of my total memory (I have 6 GB memory) and it continues to use more until I kill the task. This has led me to believe that some of my crashes were caused by all the memory being used up until it crashed. After killing the process, my system 'lives' for a longer period of time, but eventually crashes again.

BlueScreenView says the faulty modules are hal.dll and ntoskrnl.exe

I've done some research and a lot of the times the solutions point to removing AODDriver2.sys, however I don't want to just remove it in case something else stops loading (I do not have AMD OverDrive installed but it has a section in Catalyst Control Centre). I would try to disable any AOD-related options in CCC but after the last driver update, CCC stopped loading (I have tried uninstalling and reinstalling). The only thing I have left to try is to reset the BIOS settings (which I doubt will have any effect as the system doesn't crash when in safe mode). One thing I have noticed is that the thermal paste is 'used up' (most of it is just gone, as though it evaporated away) however I've read up on thermal paste and a lot of people say it should only ever have to be applied once. I haven't tried putting on any more thermal paste yet, but like was said before, temperatures remain decent.

I'm really hoping it's not a faulty processor, I've had the processor for under a year now and I've had no problems with heat or stress at all. It's curious that it crashes more often with Skype open (I recently helped out a friend whose computer kept crashing with Skype open, but I don't have any USB 3.0 ports to conflict with Skype & controllers so there goes that idea).

In case it was missed, my latest dump is here.

EDIT: Where Windows 10 drivers were unavailable, I used 8.1 (like in case of graphics drivers)
 
Solution
A BIOS update may also fix that stop error. Since its CPU related

But that stop error can also mean it's overheating / overclocked / faulty, or a core / the cache on it is stuffed

Sometimes disabling cores on a CPU will fix it. If this is what the prob is. Thats if you dont replace it

A BIOS update may also fix that stop error. Since its CPU related

But that stop error can also mean it's overheating / overclocked / faulty, or a core / the cache on it is stuffed

Sometimes disabling cores on a CPU will fix it. If this is what the prob is. Thats if you dont replace it

 
Solution

Skelebob9

Reputable
Oct 18, 2014
10
0
4,510


Thanks, I'll try a BIOS update and I'll try the thermal paste. If they don't work I'll see if I can disable any cores, and if that still doesn't work then I guess I'll have to get a new one.

 

Skelebob9

Reputable
Oct 18, 2014
10
0
4,510
For some reason, my CPU clock speeds were frequently going above the maximum (max is 3800 MHz, they were hitting ~4100 MHz). I turned down the maximum in the Windows settings to 3400 MHz and it seems to have stopped the crashing. I guess I was unintentionally overclocking after all? I only found this out when I tried disabling some of my cores, so thanks Paul NZ!
 
Status
Not open for further replies.