# Recurring BSOD 0x1A (MEMORY_MANAGEMENT) 0x61941 PAGE_TABLE_RESERVED_BITS_SET — only in VRChat with EasyAntiCheat EOS
## Summary
Since July 2026 I get a blue screen **only when running VRChat** (which uses EasyAntiCheat EOS). 5 crashes so far: 2 on 2026-07-20, 3 on 2026-08-02. Every dump is the same bugcheck:
- **BugCheck:** `0x1A MEMORY_MANAGEMENT`
- **Arg1:** `0x61941` → *“The paging hierarchy is corrupted”* (per Microsoft docs)
- **Arg2:** pointer to the virtual address that caused the fault
In every single dump, `EasyAntiCheat_EOS.sys` appears at the exact point where the kernel’s **perf-interrupt stack walk** hits EAC code, and the faulting PTE has reserved bits set.
## System Configuration
| Component | Detail |
|—|—|
| OS | Windows 11 **25H2** (build 26200.8875, kernel 10.0.26100.8875) |
| CPU | AMD Ryzen 9 9950X (16C / 32T, AM5) |
| Motherboard | ASUS TUF GAMING X870E-PLUS WIFI7 (BIOS 2306, 2026-06-15) |
| Memory | 2× 32GB DDR5-6400 (CL32-40-40, **EXPO enabled**, running 6400 MT/s) |
| GPU | NVIDIA GeForce RTX 5070 Ti |
| Virtualization | VBS enabled (Credential Guard running, Hyper-V root partition / WSL2). HVCI (Memory Integrity) is **disabled** (a driver is flagged incompatible, so it cannot be enabled) |
## Crash Analysis (identical signature in all 5 dumps)
```
BUGCHECK_CODE: 1a
BUGCHECK_P1: 61941 (paging hierarchy corrupted / PTE reserved bits set)
BUGCHECK_P2:
MODULE_NAME: hardware (MS heuristic bucket)
IMAGE_NAME: hardware_ram
FAILURE_BUCKET_ID: 0x1a_61941_PAGE_TABLE_RESERVED_BITS_SET_IMAGE_hardware_ram
```
Representative stack (latest crash, 2026-08-02 18:04):
```
nt!KeBugCheckEx
nt!MmAccessFault+0x4f7
nt!KiPageFault+0x38b
nt!RtlpxLookupFunctionTable+0x35f
nt!RtlpLookupFunctionEntryForStackWalks+0x211
nt!RtlpWalkFrameChain+0x2b8
nt!RtlWalkFrameChain+0x94
nt!EtwpTraceStackWalk+0x1f9
nt!EtwpLogKernelEvent+0x6ff
nt!EtwTraceSiloKernelEvent+0xa2
nt!EtwpProfileInterrupt+0x11a
nt!KiProcessProfileList+0x78
nt!KeProfileInterruptWithSource+0x39
nt!HalpPerfInterrupt+0x89
nt!KiCallInterruptServiceRoutine+0x5b0
nt!KiInterruptSubDispatchNoLockNoEtw+0x4e
nt!KiInterruptDispatchNoLockNoEtw+0x3c
EasyAntiCheat_EOS+0x28e3e2 ← EAC code is where the walk trips
0xffffdb83`31cdf7a0
0x00000004`7ac4af26
0x282
nt!_guard_retpoline_exit_indirect_rax+0x9
nt!KiIpiGenericCallTarget+0x89
nt!KiIpiProcessRequests+0x4a8
nt!KiIpiInterruptSubDispatch+0x40
nt!KiIpiInterrupt+0x39a
```
Notes:
- The crash path is **profile/perf interrupt → kernel stack walk → reads function table of the sampled address → page fault on a PTE with reserved bits set**.
- `EasyAntiCheat_EOS.sys` addresses are hit at this exact step in **all 5 dumps** (offsets `+0x28e3e2`, `+0x28e3e5`, `+0x28f27b`, `+0x5d58d`).
- kd also reports `WARNING: Unable to verify timestamp for EasyAntiCheat_EOS.sys`.
- The crashing process context varies (VRChat.exe, VRCX.exe, ctfmon.exe, System) — the trigger is the EAC driver being sampled, independent of which process runs.
## Troubleshooting already performed (no effect)
- **Replaced RAM** with a new DDR5-6400 kit → still crashes
- **Replaced motherboard** → still crashes
- **Upgraded GPU** → still crashes
- **memtest86 / memtest**: ran overnight, ≥140% coverage, **0 errors** at the current EXPO 6400 config
- **No WHEA / MCE hardware errors** in the Event Log (rules out a classic RAM/CPU fault path)
- **Reinstalled the EAC driver** → was stable for ~2 weeks, then crashes returned (again only in VRChat)
## Why I believe this is an EAC driver issue, not hardware
1. Crashes happen **exclusively in VRChat / with EAC** — no other game or load crashes.
2. All 5 dumps have the same EAC-specific signature; the fault is a page-table state violation during a stack walk of EAC code, not a physical-bit-flip pattern.
3. memtest at the exact running config passes, and there are zero WHEA events.
4. Replacing RAM / motherboard / GPU changed nothing.
5. MS’s “hardware_ram” bucket is a heuristic label; the concrete evidence points at the driver.
## Request
Is there a known issue between EasyAntiCheat EOS and the kernel perf-interrupt / ETW stack-walk path (Windows 11 24H2/25H2, build 26100/26200)? Any fixed EAC version or workaround to test?
Minidumps and full `!analyze -v` output for all 5 crashes are included in the report package.
## Attachments (in the report package)
- `minidumps/` — all 5 kernel minidump files
- `analysis/` — full `!analyze -v` output for each dump
- `system_info.txt` — hardware/OS inventory