Multiple BSOD for many years

[RonanPaixao]

OK. That's it! Enough is enough! I have an Asus laptop that is plagued with BSODs almost since I bought it a little over 3 years ago (will never buy Asus again!). I have moderate-to-advanced experience with computers, but have zero experience with kernel hacking.


So, how can I debug where the problem is? I have been unable to find the problem so far, because the problem seems to shift and the dumps point (mostly) to ntoskrnl.exe or ntkrnlmp.exe.


Here's what I have already tried so far:

1. Updated all drivers.
2. Ran the System File Checker tool.
3. Reinstalled Windows (was Windows 10 Home).
4. Removed one RAM module (couldn't remove the other since it is soldered to the motherboard).
5. Ran Windows Memory Diagnostic.
6. Ran MEMTEST86.
7. Ran CheckDisk (AKA chkdsk) in both online and offline modes.
8. Used the Driver Verifier with standard, maximum and custom configurations (some configurations would crash the system so badly that it wouldn't show the BSOD nor create dump files). The time I got it to crash and register is shown as a "SPECIAL_POOL_DETECTED_MEMORY_CORRUPTION" (0xc1) BugCheck.
9. Upgraded Windows 10 from Home edition to Pro.
10. Installed an SSD and moved the system to it, installing from scratch (from an external USB drive with clean Windows install from Microsoft) without Asus' drivers (the only driver I installed was nVidia's).

For most steps above, there was at least one BSOD between it and the next step.


The usual suspects are:

1. VMWare drivers (I also use it for work, but it appears that using it increases the likelihood of a crash, but I can't really quantify that).
2. nVidia drivers or card (too bad I can't replace it on a laptop!).
3. Memory corruption (even though I did run many memory tests).
4. Hard drive (or swap) corruption (unlikely to occur after I changed to the SSD, and unfortunately it would be a big disturbance if I removed the original drive, which has many of the files that I use for work).
5. Network drivers (after I saw the ndis.sys BSOD. I also switched to using cabled network instead of Wi-Fi, but it changed nothing).
6. USB 3.0 drivers: I used the WhoCrashed utility and found that there were some reports in the C:\Windows\LiveKernelReports folder pointing to bugcheck 0x144 (BUGCODE_USB3_DRIVER), but the reports are fewer than the BSOD and the timestamps do not match them.

Note: I use plain Windows Defender Antivirus.

Note 2: The crashes show up even with the computer idle at night (I avoid suspending due to the crashes).

Note 3: During activity, I have experienced crashes in random activities (e.g. using drawing programs, opening a selection box in Firefox, typing in Word).

Note 4: I do have an external monitor attached to the laptop, which may increase the BSOD if the reason is the video driver/card.


I have also accumulated almost a year's worth of Minidumps and Memory dumps. I have scripted and parsed the DumpChk utility (unfortunately, it is much less powerful since Windows 10), and built this spreadsheet and graph:

Note 5: I have ran some dumps with WinDbg with the `!analyze -v` command. Didn't get much farther (output seems almost the same as dumpchk).


Spreadsheet and minidumps uploaded here: https://1drv.ms/u/s!Ah5klMDLKUZ0hhvuy-Bt9CgaMQGu?e=FK0TOd

Continue reading...