BSOD Memory_management

[IceScreamer]

So from about a month ago I have been getting random blue screens with memory_management warnings. In about 5 times out of 7 it happens when using Firefox, a tab would crash, then Firefox would freeze and a BSOD happens.
Out of the possible fixes suggested I reset the BIOS, swapped memory sticks, updated my drivers, ran sfc /scannow, no problems reported, nothing changed, I still got a BSOD.
Last night I ran memtest86 for about 6 hours. It did 5 passes with no errors reported. There is no overheating present and everything works perfectly, apart from the said issue.
It doesn't allow me to upload *.dmp files here for some reason so I linked them to a google drive folder. If there is a more secure way to upload these files do let me know.
https://drive.google.com/drive/folders/1DTtx9OHDWX3kubHgMv3ZguTsk0Uia-jh?usp=sharing
All of them show a problem in the kernel, so I'm stumped. Can anyone with more knowledge analyze this and point me in the right direction?

 

[er557]

clean install a stable windows 10 version(LTSB?) on another partition in dual boot mode, and test that install for issues. Than you'll know if it's an issue with the computer or just a windows glitch.

 

[IceScreamer]

clean install a stable windows 10 version(LTSB?) on another partition in dual boot mode, and test that install for issues. Than you'll know if it's an issue with the computer or just a windows glitch.

I was trying to avoid this, I'm strapped for time but I guess I'll have to.

 

[biffzinker]

Could try a small bump to voltage, I've had memory stress tests pass but Firefox crash or BSOD with memory management.

 

[IceScreamer]

Could try a small bump to voltage, I've had memory stress tests pass but Firefox crash or BSOD with memory management.

Hmm, will check that too thanks, the problem is I can't really force the BSOD on demand.

 

[Regeneration]

Memory management BSOD is mostly related to memory modules/memory controller/memory configuration but sometimes to the entire bus.

MemTest86 is outdated, intended for single-core processors with discrete memory controller and shouldn't be used in 2019.

Run MemTest64 or HCI MemTest in safe mode to stress test the memory modules and configuration.

Prime95 custom torture test of 512K to 4096K with 90% of the RAM should stress the memory controller.

Check out the memory configuration in the BIOS. If its not overclocked, make sure the voltage and timing are set correctly according to the SPD information.

 

[IceScreamer]

Memory management BSOD is mostly related to memory modules/memory controller/memory configuration but sometimes to the entire bus.

MemTest86 is outdated, intended for single-core processors with discrete memory controller and shouldn't be used in 2019.

Run MemTest64 or HCI MemTest in safe mode to stress test the memory modules and configuration.

Prime95 custom torture test of 512K to 4096K with 90% of the RAM should stress the memory controller.

Check out the memory configuration in the BIOS. If its not overclocked, make sure the voltage and timing are set correctly according to the SPD information.

Will do that, thanks.

EDIT: Just did 11 loops of MemTest64 in safe mode, no errors reported.
EDIT2: Prime95 set up as mentioned, 9 tests, 33 mins, 0 warnings 0 errors.

 

[Regeneration]

You need to run each for a few hours.

According to your dump files, it is 0x1A, 0x41790 BSOD (corrupted page table). Bad mapping of physical memory to virtual memory.

Check the OS drive for errors, check the SMART data, re-create the pagefile, if you installed some driver recently, try to uninstall it.

 

[IceScreamer]

You need to run each for a few hours.

According to your dump files, it is 0x1A, 0x41790 BSOD (corrupted page table). Bad mapping of physical memory to virtual memory.

Check the OS drive for errors, check the SMART data, re-create the pagefile, if you installed some driver recently, try to uninstall it.

Currently unable to do the tests for more than an hour, projects coming up. The hard drive is a month old, SMART data is showing no errors. Only driver that was installed was the NVIDIA one and I installed and reinstalled that one a couple of times. I will try the pagefile rebuild.

 

[Regeneration]

So from about a month ago I have been getting random blue screens with memory_management warnings.

The hard drive is a month old

Did you make any other alterations to the PC?

 

[IceScreamer]

Did you make any other alterations to the PC?

Nope, tho I did clone my old hard drive to this one. I mean I doubted that could be issue but there were no errors and everything booted and worked fine.

 

[Regeneration]

Sounds like the pagefile is corrupt, or the clone went bad, or the new drive is defective.

 

[IceScreamer]

Sounds like the pagefile is corrupt, or the clone went bad, or the new drive is defective.

The drive is definitely not defective, I tested it a couple of time. The clone could be bad but I thought I'd be having other issues too.

Is there a way to repair the pagefile, or does just disabling/rebooting/enabling fixes it?

 

[Regeneration]

The drive is definitely not defective, I tested it a couple of time. The clone could be bad but I thought I'd be having other issues too.

Is there a way to repair the pagefile, or does just disabling/rebooting/enabling fixes it?

Disable, reboot, make sure the file is gone, and enable again.

 

[er557]

I would disable the page file altogether, and work like that to see if issue persists, also try enable the large system cache tweak, or change priority in memory management from applications to background processes. The setting is somewhere in advanced "this pc" properties, performance section.
After some time of stability, re-enable the page file to automatically assigned size, to another drive.

 

[IceScreamer]

I rebuilt the pagefile and now I'll see and report if something happens. Thanks for all the suggestions.

 

[silentbogo]

Had the exact same issue at work this fall.
Built a brand-new PC, and every attempt at installing windows would result in memory management BSOD.
At first I thought it was bad motherboard(lots of complaints about that cheap B250 board from Gigabyte and its memory-related issues), but came out to be defective DDR4.
But since you've already tried swapping memory in your PC, I'd assume that motherboard could be a problem.

According to your dump files, it is 0x1A, 0x41790 BSOD (corrupted page table). Bad mapping of physical memory to virtual memory.

That's not the pagefile, that's related to paging of the physical memory.

 

[IceScreamer]

Had the exact same issue at work this fall.
Built a brand-new PC, and every attempt at installing windows would result in memory management BSOD.
At first I thought it was bad motherboard(lots of complaints about that cheap B250 board from Gigabyte and its memory-related issues), but came out to be defective DDR4.
But since you've already tried swapping memory in your PC, I'd assume that motherboard could be a problem.


That's not the pagefile, that's related to paging of the physical memory.

By swapped memory stick I meant swapped their position in the memory slots, not completely different sticks.

 

[eidairaman1]

By swapped memory stick I meant swapped their position in the memory slots, not completely different sticks.

Ok test 1 module at a time now.

 

[Shambles1980]

have you bent any socket pins?

 

[Bill_Bright]

swapped memory sticks

Swapped with what? Totally different sticks?

MemTest86 is outdated

No its not.

intended for single-core processors with discrete memory controller and shouldn't be used in 2019.

Ummm, sorry, but that is totally inaccurate.

Do not confuse MemTest86 with other "outdated" products of similar (and even the same) names. This tester from PassMark is actively being maintained current, and under constant development to ensure it remains current. Note the most recent version, 8.1 was just released less than 2 weeks ago on Jan 4, 2019.

So MemTest86 from Passmark is, by far, the most current, and is fully capable of working with multi-core processors.

That said, no software based memory tester is 100% conclusive. If MemTest86 reports any errors (even just 1) the RAM is bad. But it may not report any errors and the RAM still fails when installed and used with real-world applications, or when paired with other RAM. To conclusively test your RAM, you need to use sophisticated and very expensive test equipment, like this $2,495 Memory Tester (and that's for the cheap model)! So it is usually easier (and cheaper!) to swap in known good RAM and see what happens.

 

[IceScreamer]

Ok test 1 module at a time now.

Will try if it crashes again.

have you bent any socket pins?

Pretty sure I didn't but I might check if it crashes again.

Swapped with what? Totally different sticks?
No its not.
Ummm, sorry, but that is totally inaccurate.

Do not confuse MemTest86 with other "outdated" products of similar (and even the same) names. This tester from PassMark is actively being maintained current, and under constant development to ensure it remains current. Note the most recent version, 8.1 was just released less than 2 weeks ago on Jan 4, 2019.

So MemTest86 from Passmark is, by far, the most current, and is fully capable of working with multi-core processors.

That said, no software based memory tester is 100% conclusive. If MemTest86 reports any errors (even just 1) the RAM is bad. But it may not report any errors and the RAM still fails when installed and used with real-world applications, or when paired with other RAM. To conclusively test your RAM, you need to use sophisticated and very expensive test equipment, like this $2,495 Memory Tester (and that's for the cheap model)! So it is usually easier (and cheaper!) to swap in known good RAM and see what happens.

Swapped places of the two sticks, like stick from slot A to slot B and vice versa, I know that this sometimes fixes these sort of issues.

 

[Mr.Scott]

Had exact same issue only using Google Chrome.
Chrome was actually causing a memory leak. Un-installed Chrome and re-installed, problem gone.

 

[eidairaman1]

Had exact same issue only using Google Chrome.
Chrome was actually causing a memory leak. Un-installed Chrome and re-installed, problem gone.

Sounds like Firefox at times too

 

[Regeneration]

Ummm, sorry, but that is totally inaccurate.

Do not confuse MemTest86 with other "outdated" products of similar (and even the same) names. This tester from PassMark is actively being maintained current, and under constant development to ensure it remains current. Note the most recent version, 8.1 was just released less than 2 weeks ago on Jan 4, 2019.

So MemTest86 from Passmark is, by far, the most current, and is fully capable of working with multi-core processors.

That said, no software based memory tester is 100% conclusive. If MemTest86 reports any errors (even just 1) the RAM is bad. But it may not report any errors and the RAM still fails when installed and used with real-world applications, or when paired with other RAM. To conclusively test your RAM, you need to use sophisticated and very expensive test equipment, like this $2,495 Memory Tester (and that's for the cheap model)! So it is usually easier (and cheaper!) to swap in known good RAM and see what happens.

Despite being maintained by PassMark, MemTest86 is still a single-threaded application at core, incapable of real multi-threading without producing false positives. Insufficient for today's hardware and must be written from scratch.