bsod KERNEL_MODE_HEAP_CORRUPTION (13a) help

[Junieteru]

Could someone help me analyze this dump file? I got a random BSOD while playing around with Discord. When I shared my screen, I noticed some black visual glitches when I moved my mouse cursor around, and then it crashed. This is what I saw on my screen right before it happened. 

 

081025-10140-01.dmp memory.csv sysdata.xml sysinfo.txt WERInternalMetadata.xml WERInternalRequest.xml

[Junieteru]

I tried to analyze it myself, but honestly, I don't understand anything that was shown to me.

My system specs:

• Ryzen 7 9800X3D PBO -15 offset 
• RTX 5080 MSI Gaming Trio OC White undervolted 925Mv GPU clock 2800MHz Mem clock 2000Mhz+ 
• 32GB DDR5 6000MTs EXPO enabled 
• X870e Aorus Pro ICE 
• XPG 1300w PSU 
• 2TB Samsung 990 Pro with Heatsink

[Bjoolz]

4 hours ago, Junieteru said:

Could someone help me analyze this dump file? I got a random BSOD while playing around with Discord. When I shared my screen, I noticed some black visual glitches when I moved my mouse cursor around, and then it crashed. This is what I saw on my screen right before it happened. 

 

081025-10140-01.dmp 5.6 MB · 1 download memory.csv 87.78 kB · 0 downloads sysdata.xml 534.3 kB · 0 downloads sysinfo.txt 13.03 kB · 0 downloads WERInternalMetadata.xml 6.05 kB · 0 downloads WERInternalRequest.xml 7.3 kB · 0 downloads

That type of glitch on the BSOD screen is fairly rare, but there are lots of bugs with how it's rendered if it's a high resolution monitor (Higher than 1080p) so I'm not going to make any assumptions based on that (This one is fairly common). 

 

The dump file just shows a corruption in heap memory (Heap memory is just something you can assign a memory region to be, heap memory has less strict rules so it's easier to work with for some applications). It's not pointing to any driver, Windows is just going through it and finds an error. A few weeks ago I saw a lot of people getting very similar crashes to this (If not identical, too long ago to remember exactly). We suspected some software bug or bug in Windows, but at least one of the guys came back and said it stopped after he replaced the RAM. So it could be a bug in something, but it could also be hardware.

 

Memory doesn't have to mean RAM, but it's usually the main suspect. Windows puts low priority data from RAM into the page file and loads it back in when needed so storage can look like memory (And memory can look like storage). The memory controller is in the CPU and if this fails it will just look like memory.

 

Your CPU's memory controller is rated for 5600MT/s. While 6000 isn't a big overclock, it's still an overclock and I prefer stock settings when troubleshooting. There is one newer BIOS version which says in the patch notes that it improves memory compatibility. So it's worth a shot to update the BIOS. 

 

If you have more dump files, the more the merrier. A large part of debugging is looking at patterns across multiple crashes. 

[Junieteru]

8 minutes ago, Bjoolz said:

That type of glitch on the BSOD screen is fairly rare, but there are lots of bugs with how it's rendered if it's a high resolution monitor (Higher than 1080p) so I'm not going to make any assumptions based on that (This one is fairly common). 

 

The dump file just shows a corruption in heap memory (Heap memory is just something you can assign a memory region to be, heap memory has less strict rules so it's easier to work with for some applications). It's not pointing to any driver, Windows is just going through it and finds an error. A few weeks ago I saw a lot of people getting very similar crashes to this (If not identical, too long ago to remember exactly). We suspected some software bug or bug in Windows, but at least one of the guys came back and said it stopped after he replaced the RAM. So it could be a bug in something, but it could also be hardware.

 

Memory doesn't have to mean RAM, but it's usually the main suspect. Windows puts low priority data from RAM into the page file and loads it back in when needed so storage can look like memory (And memory can look like storage). The memory controller is in the CPU and if this fails it will just look like memory.

 

Your CPU's memory controller is rated for 5600MT/s. While 6000 isn't a big overclock, it's still an overclock and I prefer stock settings when troubleshooting. There is one newer BIOS version which says in the patch notes that it improves memory compatibility. So it's worth a shot to update the BIOS. 

 

If you have more dump files, the more the merrier. A large part of debugging is looking at patterns across multiple crashes. 

Thanks for replying dude, I love you. I don't have any other dump files besides this one. It was really random and only happened once. I tried to replicate what I did before the BSOD, but everything worked normally.

I'll update my BIOS version now and see what happens. 

WATCHDOG4401-20250810-2144.dmp

[Bjoolz]

14 minutes ago, Junieteru said:

Thanks for replying dude, I love you. I don't have any other dump files besides this one. It was really random and only happened once. I tried to replicate what I did before the BSOD, but everything worked normally.

I'll update my BIOS version now and see what happens. 

WATCHDOG4401-20250810-2144.dmp 1.24 MB · 0 downloads

This one is from a Live Kernel Event. They are usually from program crashes, though they can also be from drivers/hardware. If what Windows does to remedy the situation works, you get one of these instead of a BSOD if it's a driver/hardware issue (You get a BSOD if it doesn't work). Programs can't really cause BSODs because they write to a different memory region which isn't critical so programs can just crash without it causing a BSOD (They can tell a driver to do something stupid which results in a BSOD though). DirectX noticed a black screen, did a check on the display state and then it must have been resolved because it didn't BSOD. The next command after checking the display state is to create a live dump.

[Junieteru]

4 minutes ago, Bjoolz said:

This one is from a Live Kernel Event. They are usually from program crashes, though they can also be from drivers/hardware. If what Windows does to remedy the situation works, you get one of these instead of a BSOD if it's a driver/hardware issue (You get a BSOD if it doesn't work). DirectX noticed a black screen, did a check on the display state and then it must have been resolved because it didn't BSOD. The next command after checking the display state is to create a live dump.  

Thanks! I didn't know that. I was kind of scared because next to it, it says "hardware error," but they're just live kernel events. Is there anything else I could show you?

I also checked the event viewer and saw no WHEA logs 

[Junieteru]

55 minutes ago, Bjoolz said:

That type of glitch on the BSOD screen is fairly rare, but there are lots of bugs with how it's rendered if it's a high resolution monitor (Higher than 1080p) so I'm not going to make any assumptions based on that (This one is fairly common). 

 

The dump file just shows a corruption in heap memory (Heap memory is just something you can assign a memory region to be, heap memory has less strict rules so it's easier to work with for some applications). It's not pointing to any driver, Windows is just going through it and finds an error. A few weeks ago I saw a lot of people getting very similar crashes to this (If not identical, too long ago to remember exactly). We suspected some software bug or bug in Windows, but at least one of the guys came back and said it stopped after he replaced the RAM. So it could be a bug in something, but it could also be hardware.

 

Memory doesn't have to mean RAM, but it's usually the main suspect. Windows puts low priority data from RAM into the page file and loads it back in when needed so storage can look like memory (And memory can look like storage). The memory controller is in the CPU and if this fails it will just look like memory.

 

Your CPU's memory controller is rated for 5600MT/s. While 6000 isn't a big overclock, it's still an overclock and I prefer stock settings when troubleshooting. There is one newer BIOS version which says in the patch notes that it improves memory compatibility. So it's worth a shot to update the BIOS. 

 

If you have more dump files, the more the merrier. A large part of debugging is looking at patterns across multiple crashes. 

Well, I should have done nothing. I tried to update my BIOS to the latest version, F7, and I think my PC just bricked. Now, it just shows a black screen with a bar on the top left. The debug code says 97, and the red light is on the VGA. Damn.

[Junieteru]

23 minutes ago, Junieteru said:

Well, I should have done nothing. I tried to update my BIOS to the latest version, F7, and I think my PC just bricked. Now, it just shows a black screen with a bar on the top left. The debug code says 97, and the red light is on the VGA. Damn.

Restarted CMOS and its back to life!

[Bjoolz]

1 hour ago, Junieteru said:

Restarted CMOS and its back to life!

I just saw you post on reddit, but didn't notice until I opened the dump file. I did spot one new thing though that I missed the first time and that is the pool tag of the driver that owns the memory. The pool tag is "NVRM". Just from the name that sounds like Nvidia, but we can check to make sure. 

 

• Open Command Prompt
• Type in "cd /d C:\Windows\system32\drivers" to change the directory to the driver directory.
• Next copy paste this to search for the pool tag "findstr /m /l /s NVRM *.sys"

It should give you a list of all the drivers that have that piece of text in them. You can sometimes get multiple results, but hopefully it's just one. 

 

If it comes back as the Nvidia driver and because you had those other issues also pointing to GPU issues I would do DDU.

[Junieteru]

5 hours ago, Bjoolz said:

I just saw you post on reddit, but didn't notice until I opened the dump file. I did spot one new thing though that I missed the first time and that is the pool tag of the driver that owns the memory. The pool tag is "NVRM". Just from the name that sounds like Nvidia, but we can check to make sure. 

 

• Open Command Prompt
• Type in "cd /d C:\Windows\system32\drivers" to change the directory to the driver directory.
• Next copy paste this to search for the pool tag "findstr /m /l /s NVRM *.sys"

It should give you a list of all the drivers that have that piece of text in them. You can sometimes get multiple results, but hopefully it's just one. 

 

If it comes back as the Nvidia driver and because you had those other issues also pointing to GPU issues I would do DDU.

What if I don't get any results? Because I got none... did I do something wrong, or is there something missing?

[Bjoolz]

1 hour ago, Junieteru said:

What if I don't get any results? Because I got none... did I do something wrong, or is there something missing?

It might not be in that location. You could try the directory C:\Windows instead, but the scan will take longer (How long depends on the drive speed). It might look like it freezes. You can also get a lot more junk hits. I tried it on my PC and it returned "System32\DriverStore\FileRepository\nvmoi.inf_amd64_d19b99b84b6175a8\nvlddmkm.sys". Nvlddmkm.sys is the Nvidia driver.  So it should be the same on your PC. 

[Junieteru]

2 hours ago, Bjoolz said:

It might not be in that location. You could try the directory C:\Windows instead, but the scan will take longer (How long depends on the drive speed). It might look like it freezes. You can also get a lot more junk hits. I tried it on my PC and it returned "System32\DriverStore\FileRepository\nvmoi.inf_amd64_d19b99b84b6175a8\nvlddmkm.sys". Nvlddmkm.sys is the Nvidia driver.  So it should be the same on your PC. 

Yeah it makes sense to be a failure in the nvidia driver, I tested my GPU using different stress tests and benchmarks including OCCT and 3dmark and I had no problems running VRAM test and 3D adaptative test, including nomad and firestrike from 3dmark, so... yeah, maybe a weird bug from discord overlay/hardware acceleration + nvidia drivers on 50 series. 

I have a second rig with a 4070ti super and 7800x3d, if it happens again I guess I could just switch my GPU and test 

I was also thinking about just doing a fresh install of Windows, might try that as well