WHEA uncorrectable error

[Yukiko ~Snow Child~]

So I got this error at startup about a month ago. From what I can find, this is usually a cpu or pcie error? Thing is, my SSD died about a week after getting this error. I replaced the SSD and the error hasn't come back (yet.) Now I'm wondering if the cpu is faulty or if it was the ssd. (Or both!)

[plat]

Well, it's a grey area--will it some back? Or not?

Do you have an Intel cpu? If so, you can run the Intel Processor Diagnostic Tool, which if showing a "PASS" may not always be reliable, but if "FAIL" I consider that outcome to be valid. In other words, just b/c it "passes" the test, there may be a faulty core with like one instruction off or something.

https://www.intel.com/content/www/us/en/download/15951/intel-processor-diagnostic-tool.html

Just for good housekeeping, if you're not going to use your machine for a few hours, you can consider to run the memory test. I used this one recently, I have two modules and tested each one indiv.

https://www.memtest86.com/index.html

It's free to use. I ran it recently, it took a long time (hours) but it was worth the piece of mind. My machine was blue-sceening (Critical_Process_Died) when just sitting there and it turned out I had my Power Plan set to "power saving" so it just kind of fainted away.

[plat]

Also: you can look for the faulting device in Event Viewer/System logs. Look under the 'friendly version" and find the device and vendor id/s. For example: 8086 is Intel's vendor id. the device id is prefaced with a 0x. If it's part of your chipset, see if there's a later version of your chipset driver on the manuf. website.

[Yukiko ~Snow Child~]

Oprindeligt skrevet af plat:

Well, it's a grey area--will it some back? Or not?

Do you have an Intel cpu? If so, you can run the Intel Processor Diagnostic Tool, which if showing a "PASS" may not always be reliable, but if "FAIL" I consider that outcome to be valid. In other words, just b/c it "passes" the test, there may be a faulty core with like one instruction off or something.

https://www.intel.com/content/www/us/en/download/15951/intel-processor-diagnostic-tool.html

Just for good housekeeping, if you're not going to use your machine for a few hours, you can consider to run the memory test. I used this one recently, I have two modules and tested each one indiv.

https://www.memtest86.com/index.html

It's free to use. I ran it recently, it took a long time (hours) but it was worth the piece of mind. My machine was blue-sceening (Critical_Process_Died) when just sitting there and it turned out I had my Power Plan set to "power saving" so it just kind of fainted away.

I have an AMD cpu. I also already ran memtest86. It took like 7 hours to finish. No errors, so that's good.

Oprindeligt skrevet af plat:

Also: you can look for the faulting device in Event Viewer/System logs. Look under the 'friendly version" and find the device and vendor id/s. For example: 8086 is Intel's vendor id. the device id is prefaced with a 0x. If it's part of your chipset, see if there's a later version of your chipset driver on the manuf. website.

I don't have access to that Windows installation. I'm not seeing any faults in Event Viewer with the new installation of Windows on the new SSD.

[Yukiko ~Snow Child~]

Oprindeligt skrevet af Midnight Aurais:

go on the microsoft store and download this "WinDbg Preview"
go to C:\Windows\Minidump there should be a file in there that was created after your last crash open it with windbg preview then run !analyze -v in the command line you should be just able to scroll down and click on the text

after you can just copy paste the result as this would tell you what is responsible for the whea

It was on the SSD that died. I lost all the data.

[Phénomènes Mystiques]

WHEA has been popping up lately here also.

I narrowed it down using some techniques and pinned my issue with Windows Hardware Error Architecture and a strange location pertaining OneDrive. After removing said nuisance from said strange location, no errors, critical errors, BSoD, or ghost restarts have happened, although I am still in hour 8 of testing to see if it is resolved. So far so good.

I hope this helps set a path or what not - at least I tried to share my experience.

Oh, OneDrive apparently was making dual security certificates (or something as I cannot remember the specific phrasing) causing WHEA and such.

I blame Windows updates for my issue which seems to hold water.

[Phénomènes Mystiques]

Oprindeligt skrevet af Midnight Aurais:

right i had a brainfart reading the problem guess the 4 hours sleep each night is catching up to me XD

wel if anything it is a good program to diagnose memory dump files so if anyone ever needs it keep it in the back of your mind they also apparantly now got it out of preview stage so now you should even be able to let your memory dump file play out like a recording like the original one could usefull for seeing what things do during crashes

You got that right.

[Illusion of Progress]

Without seeing whatever appropriate event logs, WHEA logs, Watch Dog logs, or memory dumps it may have created, there's nothing anyone can answer as to what caused it. It's a guessing game. There's a lot of WHEA errors and a lot of causes.

If it showed up before the SSD died and hasn't showed up since, I wouldn't worry about it unless you start having further issues.

Oprindeligt skrevet af Midnight Aurais:

go on the microsoft store and download this "WinDbg Preview"

I'm going to echo this recommendation.

I'm dealing with issues with my PC spontaneously restarting and as it was leaving no BSODs or memory dumps, it was a bit more limited in clues to track as I thought my only source was whatever event viewer was logging (which similarly just shows a WHEA error even log, ID 18 in my case which itself just means an MCE was identified [but not necessarily caused by] the CPU, and that too can have a multitude of possible causes).

Then I found Windows keeps WHEA and Watch Dog logs in the "Windows/LiveKernelReports" folder, and I used WinDbg Preview to analyze my WHEA logs (all "124" errors, which just means the CPU itself identified an MCE so it was sort of a mirror of the event viewer logs) and Watch Dog logs (these were more of a clue, as they were all "117", "141", "1a8", and "1b8" errors, all which pointed at the new GPU which was the part I changed before the behavior arrived, and matched symptoms others are having with my same GPU, so I'm RMAing that now and hoping it resolves it).

Lesson is, I echo that this would be a good tool to have around. BSODs aren't always created, and even when they are, stuff like BlueScreenView or WhoCrashed tend to simply list the faulting process, which itself isn't always the cause of the issue. Namely, it may make a hardware problem look like it might be a software problem. Not saying those programs are bad, but there's things they won't cover. A lot of issues (especially hardware issues) cascade into other things failing as well, so something like (for example) smss.exe, a Windows user environment process, might fault and then get listed as the faulting process by those applications, even though it was actually something else that led to it. So expanding your tool set is good.