Developer knowledge network ยท moderated exchange

UnreliableCode Community

Developer Research, Reverse Engineering & Coding Community

Knowledge indexLive
4Categories
919Threads
2.8KPosts
Release

Inverted Ring Buffer Driver: high-throughput memory streaming architecture

dma_ghost
DMA Guru
MEMBER
Rep: 115
Join Date: Jan 2020
Posts: 12
Thanks: 83
3 weeks ago ยท Jul 30, 2026 11:36 PM
#1

High-throughput external memory streaming framework:

  • Zero named devices (IoCreateDevice is never called).
  • Zero IOCTL handlers.
  • Uses atomic memory spinlocks over mapped MDL pages.
  • Sustains 2,500,000 memory read requests per second with < 0.2 microsecond latency.

Tested clean on EasyAntiCheat and BattlEye with 0 detection flags.

ring_buffer_kernel_driver_demo.zip
Size: 84 KB ยท Downloads: 0
SHA256: e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855
kernel_komrade
Driver Dev
MEMBER
Rep: 145
Join Date: Feb 2019
Posts: 26
Thanks: 45
3 weeks ago ยท Jul 31, 2026 3:30 AM
#2

2.5M reads per second is incredible. Shared MDL memory mapping is so much faster than traditional IRP IOCTL dispatch.

ptr_arithmetic
C++ Wizard
MEMBER
Rep: 162
Join Date: May 2018
Posts: 73
Thanks: 42
3 weeks ago ยท Jul 31, 2026 12:59 PM
#3

Cleanest IPC design possible.