Unusual random crashes occurring without apparent connection
Unusual random crashes occurring without apparent connection
Hello! I've been assembling my own PCs for roughly two decades now. I’m someone who believes a PC should be built from the ground up. Anything else feels like it doesn’t quite meet the mark. But lately, I’ve encountered some odd problems with my current setup and I’m struggling to figure out the next steps for troubleshooting. Since gaming isn’t a daily habit anymore, it hasn’t been too critical. Most of my regular tasks are handled on my phone or laptop. However, I’ve recently picked up gaming again and now feel compelled to get things running smoothly. I’m hoping someone might suggest something helpful.
Machine: CPU – AMD Ryzen 5 3600X | RAM – Gigabyte X570 AORUS PRO (Bios v F37f) | GPU – Gigabyte RTX 2080 Ti (with custom fan) | Memory – Corsair Vengeance LPX 32 GB DDR4 3600MHz (CMK32GX4M2Z3600C18) | SSD – Samsung MZVLB256HAHQ, Intel Optane 900p 480GB | OS – Samsung 960 Pro 2TB | Version – Windows 10 22H2 19045.4239 (current, but issues persist before)
Issue: I experience random crashes, often appearing as a BSOD. Most commonly, it’s just an unexpected crash or error within the app I’m using—like Firefox freezing when I open a tab, or YouTube playing without warning. I also face random crashes in games, especially Halo: The Master Chief Collection, where the game freezes with a small error window. I can close the game and restart it, but sometimes it runs for 5 minutes or even 2–3 hours. I managed to play Halo 1 and 2 about 20–25 times before, but similar “soft” crashes have appeared in other titles too.
Old troubles: The problems began after rebuilding my machine using the current motherboard, CPU, and RAM. My previous setup was an Intel 6800K with the same GPU, PSU, SSD, sound card, case, and mouse. I upgraded to a new board with the 3600X and 4x4GB RAM. Despite that, I had issues from the start, so I upgraded to 2x16GB DDR4 at 2800MHz or 3600MHz with XMP enabled.
What I’ve tried: I’m not sure if load is the cause. CPU temperatures are normal at idle (around 35°C). I don’t remember what was running under load, but it might have been around 70°C. I ran Memtest86+ for about 12 hours without issues, Prime95 on all cores for 10 hours without problems, and FurMark for a few hours with no crashes. However, when I tried enabling REBAR in UEFI, the system became extremely unstable and would sometimes fail to boot. So I left it off. I updated my BIOS multiple times but didn’t see any changes. Reinstalled Windows, but nothing changed.
Anyone have suggestions on what I should try next?
Visit C:\Windows\Minidump to see if any minidump files exist. If present, return to the Windows directory and transfer the Minidump folder to the Downloads folder (use your desktop if needed). Compress the copied folder and include it in a post. Please adhere strictly to instructions since Windows doesn't allow file changes there.
Haha FF crashes appear to be happening just because I attempted to share this message. I noticed several issues in the event log, most matching what you saw here: https://answers.microsoft.com/en-us/wind...3941745bd6. Some errors seem linked to the last application that failed. For instance, yesterday when Halo crashed I received event 1000. At one point it showed halo2.dll with name: MCC-Win64-Shipping.exe, version 1.3385.0.0, timestamp 0x65959e09. There was also a nvwgf2umx.dll fault, possibly part of the NVIDIA driver. Another time I saw firefox.exe with version 124.0.1.8846, no clear module name, and it reported an unknown fault. All details look similar to what others posted.
I performed a brief WinDBG analysis on the files inside the minidump directory. Here are the key findings:
- A lock managed by AutoBoost was dropped by a thread that didn’t own it.
- This usually happens when another thread releases a lock it shouldn’t, or when a thread attempts to release a lock it no longer controls.
- Sample values:
- Arg1: ffffe58a126dd080
- Thread address: ffffe58a0a37ded0
- Lock address: 00000000ffffffff
- Session ID: 0000000000000000
- Process name: MoUsoCoreWorker.exe
- Additional debug info:
- CPU usage: 2405 msec
- Elapsed time: 5829 msec
- IO memory usage: 8 MB
- Init CPU: 609 msec
- Memory peak: 84 MB
- Code reference: 162
- Target model: 0x162
- Release cache awareness: not enabled
- Error code: 0x162
No further details provided.