Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
68 lines (46 loc) · 1.87 KB

CONTRIBUTING.md

File metadata and controls

68 lines (46 loc) · 1.87 KB
Raw

Contributing to FSP.DMRCrack

Prerequisites

Windows:

Linux:

  • GCC or Clang
  • CMake >= 3.18
  • CUDA Toolkit 12.x (or ROCm 5.7+ for AMD, with -DUSE_HIP=ON)

Python 3.8+ is optional; only needed for the conversion and verification scripts in tools/.

Building

Linux

cmake -S . -B build -DNO_GUI=ON -DCMAKE_BUILD_TYPE=Release
cmake --build build --parallel

Windows (GUI)

Open an x64 Native Tools Command Prompt for VS in the project root:

build.bat

Output: bin\dmrcrack.exe

Test tools (no CUDA required)

build_test.bat       # builds bin\test_strict_score.exe

Code style

  • C99, no C++ in .c files
  • CUDA/HIP kernel code in .cu files
  • 4-space indentation, snake_case for all identifiers
  • Platform-specific code goes through include/platform.h. Don't call Win32 APIs directly outside src/gui.c and src/main.c
  • GPU kernel logic and host-side scoring helpers both live in src/bruteforce.cu so the strict pipeline stays in sync between the two

Submitting changes

  1. Fork and create a feature branch
  2. Make your changes; build.bat (Windows) or CMake (Linux) should succeed without new warnings
  3. If you touched scoring logic, run bin\test_strict_score.exe and tools\verify_decrypt.py against the test data in test/aaaaa/
  4. Open a PR with a description of what changed and why

Reporting issues

Open an issue and include:

  • OS and GPU model
  • CUDA Toolkit version (nvcc --version)
  • Exact error message or unexpected behavior
  • Steps to reproduce (anonymized .bin file if relevant)

License

By contributing you agree that your work will be licensed under GPL v3.0 (see LICENSE).