Summary:
Client reliably crashes with ERROR #132 after being connected for roughly 10-20 minutes. I've been mostly sitting around SW because that is when it happens most consistantly, but have had it happen in other areas as well. Most consistently while idling in the world or while alt-tabbed doing something else on the PC (but not exclusively at
login/char-select or during any specific action I've been able to pin down).
Crash details:
- ERROR #132 (0x85100084) Fatal Exception
- Exception alternates between 0xC0000096 (PRIV_INSTRUCTION) and 0xC0000005 (ACCESS_VIOLATION) across repeated crashes
- Faulting address consistently in the range 0x006AAB00-0x006AAE00 in WoW.exe
- Stack trace is consistent across every crash, always terminating in the same three frames regardless of the path taken to reach them:
006AABAD/006AADAD (varies)
006AA911
0067233E
- The frames above 0067233E vary between crashes (e.g. sometimes via 00633B4B → 00634152, other times via 006320B6 → 00632D2E → 0063524A → 00635739 → 00635BE0 → 00634204), all converging on the same three
leaf frames. This suggests several different code paths funnel into one shared subroutine that's crashing.
What I've already ruled out (crash reproduces identically after each of these changes, so please don't suggest them first):
1. CPU affinity restricted to P-cores only — no change
2. Third-party overlays (OBS game-capture hook, NVIDIA overlay/ShadowPlay) — confirmed loaded in crash dumps but disabling them in-app made no difference
3. Framerate capped (maxFPS 60 / maxFPSBk 30) + VSync — no change
4. Large Address Aware flag disabled on WoW.exe — no change
5. Bypassing VanillaFixes entirely (raw WoW.exe) — no change
6. NVIDIA driver downgraded from 610.88.0 to 572.83 (to rule out the known NVIDIA 576.02+ 32-bit DXVK shader-compiler bug, nvgpucomp32.dll) — no change
7. WDB cache cleared/empty — no change