Question Persistent Kernel Mode Exception on Legacy Environment - Debugging, PLS HELP!!!!!!

[celestial_]

Hello community,

I’m currently running a specialized setup on a machine I’ve built using some of my favorite nostalgic hardware. I’m a student of Computer Architecture and Cybersecurity from Brazil, and I’ve been trying to maintain a stable environment for FoxPro development and legacy stress-testing.

The Problem:I am hitting a Kernel-Mode Exception (BSOD 0x0000008E) during high-load I/O operations. Given my background, I’ve already ruled out simple driver mismatches. I suspect there’s a conflict in the memory paging or a failure in how the NT Kernel is handling legacy interrupts on this specific architecture.

Hardware Specs :

• CPU: Intel Core 2 Quad Q9650 (The gold standard of the LGA 775 era).
• GPU: NVIDIA GeForce GTX 980 Ti (A modern classic for high-end legacy builds).
• RAM: 8GB Corsair Dominator DDR2 (With those iconic tall heat sinks).
• Storage: Western Digital VelociRaptor 10k RPM (For that authentic mechanical "crunch" sound).
• OS: Windows 10 (Heavily modified for XP aesthetics and legacy kernel compatibility).

Debug Logs & Progress:

1. Memory Dump: Analyzed the .dmp file; the faulting module seems to be ntoskrnl.exe.
2. Instruction Set: Verified if the application was trying to execute non-NX memory, but the DEP (Data Execution Prevention) settings seem correct.
3. Low-Level Check: Ran a full MemTest86+ cycle to ensure the Northbridge isn’t overheating under load.

I’m trying to determine if this is a Race Condition within the kernel's scheduler or if I should look into patching the HAL (Hardware Abstraction Layer) for better stability with this specific CPU/Northbridge combo.

Has anyone encountered similar behavior when pushing FoxPro or similar database engines to their limit on quad-core legacy architectures? I'm looking for a low-level perspective before I start disassembling the binaries.

Thanks in advance!