[CI] add pre-commit integration

.clang-format

@@ -0,0 +1,8 @@
+---
+BasedOnStyle: LLVM
+UseTab: Never
+IndentWidth: 2
+ColumnLimit: 80
+
+Language: Cpp
+Standard: c++17

.editorconfig

@@ -14,7 +14,10 @@ insert_final_newline = true
 indent_size = 4
 
 [*.{cpp,hpp,cxx,cc,c,h,cu,cuh}]
-indent_size = 4
+indent_size = 2
+
+[{*.cmake,CMakeLists.txt}]
+indent_size = 2
 
 [*.{yaml,yml}]
 indent_size = 2

.gitattributes

@@ -1 +1,10 @@
+* text eol=lf
+*.bat eol=crlf
+
+*.svg binary
+*.jpg binary
+*.jpeg binary
+*.png binary
+*.gif binary
+
 *.h linguist-language=C++

.pre-commit-config.yaml

@@ -0,0 +1,59 @@
+# See https://pre-commit.com for more information
+# See https://pre-commit.com/hooks.html for more hooks
+ci:
+  autofix_prs: true
+  autofix_commit_msg: "[Lint]: [pre-commit.ci] auto fixes [...]"
+  autoupdate_commit_msg: "[CI] [pre-commit.ci] autoupdate"
+  autoupdate_schedule: monthly
+default_stages: [pre-commit, pre-push, manual]
+exclude: '^(build|3rdparty)/.*$'  # exclude build and 3rdparty directories
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v6.0.0
+    hooks:
+      - id: check-symlinks
+      - id: destroyed-symlinks
+      # FIXME: enable these hooks
+      # - id: trailing-whitespace
+      # - id: end-of-file-fixer
+      - id: check-added-large-files
+      - id: check-merge-conflict
+        fail_fast: true
+      # FIXME: enable these hooks
+      # - id: check-executables-have-shebangs
+      # - id: check-shebang-scripts-are-executable
+      - id: detect-private-key
+      - id: check-yaml
+      - id: check-toml
+      - id: check-ast
+        fail_fast: true
+      - id: debug-statements
+  - repo: https://github.com/pre-commit/mirrors-clang-format
+    rev: v15.0.7  # sync with requirements-lint.txt
+    hooks:
+      - id: clang-format
+        exclude: |
+          (?ix)(
+            ^.+\.json$
+          )
+  - repo: https://github.com/astral-sh/ruff-pre-commit
+    rev: v0.14.0  # sync with requirements-lint.txt
+    hooks:
+      - id: ruff-check
+        args: [--fix, --exit-non-zero-on-fix]
+  - repo: https://github.com/google/yapf
+    rev: v0.43.0  # sync with requirements-lint.txt
+    hooks:
+      - id: yapf
+        args: [--recursive, --in-place]
+  - repo: https://github.com/codespell-project/codespell
+    rev: v2.4.1  # sync with requirements-lint.txt
+    hooks:
+      - id: codespell
+        additional_dependencies: [".[toml]"]
+        exclude: |
+          (?x)(
+            ^.+\.(cpp|hpp|cxx|cc|c|h|cu|cuh)$|
+            ^.+\.svg$|
+            ^.*\brequirements\b.*\.txt$
+          )

CMakeLists.txt

@@ -56,7 +56,7 @@ else()
 
   # Set default build type to RelWithDebInfo if not provided
   if(NOT CMAKE_BUILD_TYPE)
-  # Set default build type to Release if not provided
+    # Set default build type to Release if not provided
     set(CMAKE_BUILD_TYPE Release CACHE STRING "Build type" FORCE)
     message(STATUS "Setting default build type to ${CMAKE_BUILD_TYPE}")
   endif()
@@ -199,7 +199,7 @@ if(USE_CUDA)
   set(CUDA_MAJOR_VERSION ${CUDAToolkit_VERSION_MAJOR})
   message(STATUS "Setting CUDA_MAJOR_VERSION=${CUDA_MAJOR_VERSION}")
   add_compile_definitions(CUDA_MAJOR_VERSION=${CUDA_MAJOR_VERSION})
-  
+
   list(APPEND TILE_LANG_INCLUDES ${CUDAToolkit_INCLUDE_DIRS})
 endif(USE_CUDA)
 

docs/deeplearning_operators/matmul.md

@@ -8,7 +8,7 @@
 :class: myclass1 myclass2
 :name: a-tip-reference
 
-   This document is still **experimental** and may be incomplete.  
+   This document is still **experimental** and may be incomplete.
    Suggestions and improvements are highly encouraged—please submit a PR!
 :::
 
@@ -256,4 +256,4 @@ For more advanced usage—including partial lowering, explicitly controlling thr
 * [BitBLAS](https://github.com/tile-ai/bitblas)
 * [Triton](https://github.com/openai/triton)
 * [Cutlass](https://github.com/NVIDIA/cutlass)
-* [PyCUDA](https://documen.tician.de/pycuda/)
+* [PyCUDA](https://documen.tician.de/pycuda/)  <!-- codespell:ignore -->

examples/deepseek_v32/fp8_lighting_indexer.py

@@ -258,6 +258,7 @@ def ref_fp8_mqa_logits(q: torch.Tensor, kv: torch.Tensor, weights: torch.Tensor,
     cost = mask.sum()
     return logits, cost
 
+
 def test_fp8_lighting_indexer(S=4096, SKV=8192, H=32, HKV=1, D=64, kv_stride=1):
     q = torch.randn(S, H, D, device="cuda", dtype=torch.bfloat16).to(torch.bfloat16)
     kv = torch.randn(SKV, D, device="cuda", dtype=torch.bfloat16).to(torch.bfloat16)
@@ -302,5 +303,6 @@ def logits_fn():
     print(f"logits_tflops: {logits_tflops}, logits_ms: {logits_ms}")
     print(f"cost_ref: {cost_ref}")
 
+
 if __name__ == "__main__":
     test_fp8_lighting_indexer()