For the last two years I’ve been experiencing system crashes approximately once every 3-4 wakes from suspend. The pc goes to sleep normally, but when it’s woken up it shows a blank black screen for around 10 minutes, after which it reboots and a MEMORY.DMP is generated in the system folder.
I’ve found no interesting entries in the event viewer, just the usual warning and errors from the unexpected restart.
In 2024, I’ve tried examining the dumps with help from Gary Nebbett at learn.microsoft, but debugging a windows memory dump in WinDbg is definitely outside my expertise.. Discussion here. But I am a C programmer, so if someone is willingful to provide further insights I should be able to follow somewhat.
A quote from the discussion:
The lock is in kernel mode and a bug is the kernel-mode code not acquiring locks in a defined order - probably because there was inadequate knowledge of all the locks that might be used in some scenarios.
It is probably some user-mode behaviour that triggers a scenario that reveals the bug. Getting the bug fixed would be difficult and lengthy, but we can try to identify the user-mode behaviour that triggers the bug and avoid it.
Here a onedrive directory with dumps (the latest being from today as this post) and system traces: link
The issue grew in frequency in the last years with every driver update. I’m not sure it correlates with more recent “black screen” reports on newer boards.
All classic troubleshooting steps have been performed. (except system install, but os image was DISM-checked a bunch of times in the last two years). Monitors where replaced in the course of this time (with different technologies: hdmi/displayport, static/vrr, 8bit/10bit ecc).
Microsoft Windows 10 Pro (currently 22H2 19045.6093)
Nvidia GeForce RTX 2080FE driver (currently 576.88)
No third party security products