build: migrate CUDA kernel build to CMake

Replace torch CUDAExtension/ParallelBuildExtension with a CMake-based build. Each kernel compiles as an independent pybind11 module in parallel via cmake --build -j, outputting to astrai/extension/lib.

- Add csrc/CMakeLists.txt (5 kernel targets, torch/pybind11 linking)
- setup.py: _CMakeBuildExt invokes cmake; auto-detect CUDA arch via torch
- Remove csrc/build.py (REGISTRY/build flags now in CMakeLists)
- Fix rel-err eps in attn_test.cu (1e-8 -> 1e-4, bf16 scale)
- Update docs/developer/cuda_kernels.md build section
- .gitignore: allow csrc/CMakeLists.txt

This commit is contained in:

ViperEkura

2026-08-04 21:27:22 +08:00

parent cdf9145ecf

commit 49a9c6b3d2

6 changed files with 214 additions and 201 deletions

.gitignore

View File

@@ -9,6 +9,7 @@
 !scripts/**/*.py
 !tests/**/*.py
 !csrc/**/*.py
 !csrc/CMakeLists.txt
 !csrc/**/*.cu
 !csrc/**/*.h