Memtest reboots before test begins

[Lork]

A friend of mine is trying to diagnose an issue with random shutdowns, but on trying to run MemTest86, he says it displays some white text saying it's preparing to do something, then reboots, looping this way as many times as he allows it to.

He was able to get a log file from the usb stick he used, which I've attached to this post.

[David (PassMark)]

Last few lines in the log before the reboot were,
2017-11-08 18:38:42 - Testing MP support
2017-11-08 18:38:42 - MPSupportTestMPServices - AP dispatch test
2017-11-08 18:38:42 - MPSupportTestMPServices - Starting AP#1

So it has failed while trying to test if the UEFI BIOS supports Multi-threading (MT).
Motherboard is a ASUS SABERTOOTH 990FX.

So this might be a UEFI Firmware bug (which are fairly common). Or it might be the system itself which fails under the load of mutlithreading.

[Lork]

I doubt it's failing simply because of MT - he regularly plays games that I'm sure utilize MT without rebooting like this, at least not immediately anyway.

Do you have any suggestions for a workaround? Like maybe a way to set MemTest86 to run without using MT?

[keith]

To workaround this multiprocessor bug, you need to add the motherboard model to the blacklist.cfg under EFI\BOOT\ of the USB drive.

In your case, you would need to append the following line:

"SABERTOOTH 990FX R2.0","zzzzzzzz",EXACT,RESTRICT_MP

[Lork]

Thanks. Using the blacklist he was able to start the test, but then his computer rebooted a couple minutes into it. It seems like the issue is most likely something other than memory.

[David (PassMark)]

If you have random shutdowns both in Windows and MemTest86, then you surely have a hardware problem. The sudden random shutdown problem is never easy to track down however. Unless you have high end test equipment it normally means you have to do a bunch of trail and error.

[Lork]

We're pretty sure we tracked it down to a CPU overheating problem. He had someone re-apply the thermal paste and hasn't had any issues since.

Anyway, I'm just coming back because he mentioned that the two crashes (on multi-core test and a few minutes into memtest proper) were not the same. The latter caused the computer to completely power down, consistent with the other random shutdowns he experienced, but the former would just reboot as many times as he would allow it. It's most likely a UEFI bug, so you may want to add the motherboard to the blacklist and inform Asus about it.

Thanks again for the help.

[David (PassMark)]

Glad you fixed it. Never easy to track down an problem when it turns out to be multiple different issues.

[digicube]

Looks like I have an identical problem. It reboots when fetching info on multi-threading. My CPU is AMD FX-8350 and mobo is M5A97_LE_R20. I'll reapply thermal paste and see if that's fixes it. If not, then it must be the mobo. 16GB 2134MHz RAM no errors memtest running for 4 hours.

[digicube]

Memtest log file upload attachment, need 10 words, can't update previous post.

[digicube]

I've put my mobo on the black list,. Memtest reboots when I run parallel CPU test. Does this mean I have a bad CPU or just overheating?

[digicube]

I ran OCCT for 4 hours, so most likely a bad mobo.

[digicube]

Oops looks like I have one too many stand-offs below the motherboard and that's what causing the computer freezes.