# coda-freshman **Repository Path**: crystus/coda-freshman ## Basic Information - **Project Name**: coda-freshman - **Description**: coda freshman sourcecode - **Primary Language**: C - **License**: LGPL-3.0 - **Default Branch**: master - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2025-04-06 - **Last Updated**: 2026-05-01 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README ## HOW TO MAKE ### SM code Look up in this url(https://developer.nvidia.com/cuda-gpus) 120 Blackwell RTX50 100 Blackwell-Server B200 90 Hopper-Server H100 89 Ada RTX40 87 Ampere-IOT Jetson Orin Nano 86 Ampere RTX30 75 Turin RTX20/GTX1650 Older GPUS in this url(https://developer.nvidia.com/cuda-legacy-gpus) 61 Pascal GTX10 53 Jetson Nano ### CMAKE toolkit In case cmake extention cannot find cuda toolkits, add following lines to $HOME/.local/share/CMakeTools/cmake-tools-kits.json, then restart vscode ``` [ ... { "name": "CUDA Compiler", "compilers": { "C": "/usr/local/bin/nvcc", "CXX": "/usr/local/bin/nvcc" }, "isTrusted": true } ... ] ``` ## HOW TO DEBUG ### COMPILER ARGS Remember to add "-G -g" in nvcc commmand line args, or add following line CMakelists.txt ``` set(CMAKE_CUDA_FLAGS_DEBUG "-g -G") ``` ### GPUs Debug on a multi-gpu server to avoid share GPU with X-window, or use follow commands to start in CLI mode: ``` sudo systemctl set-default multi-user.target sudo reboot ``` If want to change to GUI, run: ``` sudo systemctl set-default graphical.target sudo reboot ``` ### Permission User must be in debug group, use id command to check, if not, add current user to the group: ``` sudo usermod -a -G debug $USER ``` ### DEBUG IN WSL Use WSL2.0 in Windows 11, install NVIDIA driver in Windows, and check nvidia-smi works OK: ``` /usr/lib/wsl/lib/nvidia-smi ``` And install WSL version CUDA toolkit in WSL from this URL(https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&Distribution=WSL-Ubuntu&target_version=2.0&target_type=runfile_local) Add register settings in windows(from https://docs.nvidia.com/cuda/cuda-gdb/index.html#supported-platforms), make sure this capability is enabled via the registry key >HKEY_LOCAL_MACHINE\SOFTWARE\NVIDIA Corporation\GPUDebugger\EnableInterface set to (DWORD) 1. If there is no GPUDebugger entry, create it.