Update to clang 20

.pre-commit-config.yaml

@@ -42,7 +42,7 @@ repos:
                "python/dask_cudf/dask_cudf"]
         pass_filenames: false
   - repo: https://github.com/pre-commit/mirrors-clang-format
-    rev: v16.0.6
+    rev: v20.1.4
     hooks:
       - id: clang-format
         types_or: [c, c++, cuda]

conda/environments/all_cuda-118_arch-aarch64.yaml

@@ -14,8 +14,9 @@ dependencies:
 - c-compiler
 - cachetools
 - certifi
-- clang-tools=16.0.6
-- clang==16.0.6
+- clang-tools=20.1.4
+- clang-tools==20.1.4
+- clang==20.1.4
 - cmake>=3.30.4
 - cramjam
 - cubinlinker
@@ -36,6 +37,7 @@ dependencies:
 - gcc_linux-aarch64=11.*
 - hypothesis>=6.131.7
 - identify>=2.5.20
+- include-what-you-use==0.24.0
 - ipython
 - jupyter_client
 - libcurand-dev=10.3.0.86

conda/environments/all_cuda-118_arch-x86_64.yaml

@@ -14,8 +14,9 @@ dependencies:
 - c-compiler
 - cachetools
 - certifi
-- clang-tools=16.0.6
-- clang==16.0.6
+- clang-tools=20.1.4
+- clang-tools==20.1.4
+- clang==20.1.4
 - cmake>=3.30.4
 - cramjam
 - cubinlinker
@@ -36,6 +37,7 @@ dependencies:
 - gcc_linux-64=11.*
 - hypothesis>=6.131.7
 - identify>=2.5.20
+- include-what-you-use==0.24.0
 - ipython
 - jupyter_client
 - libcufile-dev=1.4.0.31

conda/environments/all_cuda-128_arch-aarch64.yaml

@@ -14,8 +14,9 @@ dependencies:
 - c-compiler
 - cachetools
 - certifi
-- clang-tools=16.0.6
-- clang==16.0.6
+- clang-tools=20.1.4
+- clang-tools==20.1.4
+- clang==20.1.4
 - cmake>=3.30.4
 - cramjam
 - cuda-cudart-dev
@@ -37,6 +38,7 @@ dependencies:
 - gcc_linux-aarch64=13.*
 - hypothesis>=6.131.7
 - identify>=2.5.20
+- include-what-you-use==0.24.0
 - ipython
 - jupyter_client
 - libcurand-dev

conda/environments/all_cuda-128_arch-x86_64.yaml

@@ -14,8 +14,9 @@ dependencies:
 - c-compiler
 - cachetools
 - certifi
-- clang-tools=16.0.6
-- clang==16.0.6
+- clang-tools=20.1.4
+- clang-tools==20.1.4
+- clang==20.1.4
 - cmake>=3.30.4
 - cramjam
 - cuda-cudart-dev
@@ -37,6 +38,7 @@ dependencies:
 - gcc_linux-64=13.*
 - hypothesis>=6.131.7
 - identify>=2.5.20
+- include-what-you-use==0.24.0
 - ipython
 - jupyter_client
 - libcufile-dev

cpp/CMakeLists.txt

@@ -187,7 +187,7 @@ if(CUDF_CLANG_TIDY)
   # versions are skipped so we don't want to constrain to a patch version that the user can't
   # install.
   set(LLVM_VERSION "${CMAKE_MATCH_1}")
-  set(expected_clang_tidy_version 19.1)
+  set(expected_clang_tidy_version 20.1)
 
   if(NOT expected_clang_tidy_version VERSION_EQUAL LLVM_VERSION)
     message(

cpp/benchmarks/ndsh/utilities.hpp

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2024, NVIDIA CORPORATION.
+ * Copyright (c) 2024-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -28,7 +28,7 @@
 class table_with_names {
  public:
   table_with_names(std::unique_ptr<cudf::table> tbl, std::vector<std::string> col_names)
-    : tbl(std::move(tbl)), col_names(col_names){};
+    : tbl(std::move(tbl)), col_names(col_names) {};
   /**
    * @brief Return the table view
    */

cpp/benchmarks/string/translate.cpp

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2021-2024, NVIDIA CORPORATION.
+ * Copyright (c) 2021-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -45,9 +45,7 @@ static void bench_translate(nvbench::state& state)
   std::transform(thrust::counting_iterator<int>(0),
                  thrust::counting_iterator<int>(entry_count),
                  entries.begin(),
-                 [](auto idx) -> entry_type {
-                   return entry_type{'!' + idx, '~' - idx};
-                 });
+                 [](auto idx) -> entry_type { return entry_type{'!' + idx, '~' - idx}; });
 
   auto stream = cudf::get_default_stream();
   state.set_cuda_stream(nvbench::make_cuda_stream_view(stream.value()));

cpp/examples/parquet_io/parquet_io_multithreaded.cpp

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2024, NVIDIA CORPORATION.
+ * Copyright (c) 2024-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -321,12 +321,11 @@ std::vector<io_source> extract_input_sources(std::string const& paths,
   std::vector<io_source> input_sources;
   input_sources.reserve(parquet_files.size());
   // Transform input files to the specified io sources
-  std::transform(parquet_files.begin(),
-                 parquet_files.end(),
-                 std::back_inserter(input_sources),
-                 [&](auto const& file_name) {
-                   return io_source{file_name, io_source_type, stream};
-                 });
+  std::transform(
+    parquet_files.begin(),
+    parquet_files.end(),
+    std::back_inserter(input_sources),
+    [&](auto const& file_name) { return io_source{file_name, io_source_type, stream}; });
   stream.synchronize();
   return input_sources;
 }

cpp/include/cudf/column/column_view.hpp

@@ -478,7 +478,7 @@ class column_view : public detail::column_view_base {
 
   std::vector<column_view> _children{};  ///< Based on element type, children
                                          ///< may contain additional data
-};                                       // namespace cudf
+};  // namespace cudf
 
 /**
  * @brief A non-owning, mutable view of device data as a column of elements,
@@ -504,7 +504,7 @@ class mutable_column_view : public detail::column_view_base {
  public:
   mutable_column_view() = default;
 
-  ~mutable_column_view() override{
+  ~mutable_column_view() override {
     // Needed so that the first instance of the implicit destructor for any TU isn't 'constructed'
     // from a host+device function marking the implicit version also as host+device
   };

cpp/include/cudf/hashing/detail/murmurhash3_x64_128.cuh

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2023-2024, NVIDIA CORPORATION.
+ * Copyright (c) 2023-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -76,33 +76,33 @@ MurmurHash3_x64_128<double>::result_type __device__ inline MurmurHash3_x64_128<d
 
 template <>
 MurmurHash3_x64_128<cudf::string_view>::result_type
-  __device__ inline MurmurHash3_x64_128<cudf::string_view>::operator()(
-    cudf::string_view const& key) const
+  __device__ inline MurmurHash3_x64_128<cudf::string_view>::
+  operator()(cudf::string_view const& key) const
 {
   return this->compute_bytes(reinterpret_cast<cuda::std::byte const*>(key.data()),
                              key.size_bytes());
 }
 
 template <>
 MurmurHash3_x64_128<numeric::decimal32>::result_type
-  __device__ inline MurmurHash3_x64_128<numeric::decimal32>::operator()(
-    numeric::decimal32 const& key) const
+  __device__ inline MurmurHash3_x64_128<numeric::decimal32>::
+  operator()(numeric::decimal32 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
 MurmurHash3_x64_128<numeric::decimal64>::result_type
-  __device__ inline MurmurHash3_x64_128<numeric::decimal64>::operator()(
-    numeric::decimal64 const& key) const
+  __device__ inline MurmurHash3_x64_128<numeric::decimal64>::
+  operator()(numeric::decimal64 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
 MurmurHash3_x64_128<numeric::decimal128>::result_type
-  __device__ inline MurmurHash3_x64_128<numeric::decimal128>::operator()(
-    numeric::decimal128 const& key) const
+  __device__ inline MurmurHash3_x64_128<numeric::decimal128>::
+  operator()(numeric::decimal128 const& key) const
 {
   return this->compute(key.value());
 }

cpp/include/cudf/hashing/detail/murmurhash3_x86_32.cuh

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2024, NVIDIA CORPORATION.
+ * Copyright (c) 2017-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -79,49 +79,49 @@ MurmurHash3_x86_32<double>::result_type __device__ inline MurmurHash3_x86_32<dou
 
 template <>
 MurmurHash3_x86_32<cudf::string_view>::result_type
-  __device__ inline MurmurHash3_x86_32<cudf::string_view>::operator()(
-    cudf::string_view const& key) const
+  __device__ inline MurmurHash3_x86_32<cudf::string_view>::
+  operator()(cudf::string_view const& key) const
 {
   return this->compute_bytes(reinterpret_cast<cuda::std::byte const*>(key.data()),
                              key.size_bytes());
 }
 
 template <>
 MurmurHash3_x86_32<numeric::decimal32>::result_type
-  __device__ inline MurmurHash3_x86_32<numeric::decimal32>::operator()(
-    numeric::decimal32 const& key) const
+  __device__ inline MurmurHash3_x86_32<numeric::decimal32>::
+  operator()(numeric::decimal32 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
 MurmurHash3_x86_32<numeric::decimal64>::result_type
-  __device__ inline MurmurHash3_x86_32<numeric::decimal64>::operator()(
-    numeric::decimal64 const& key) const
+  __device__ inline MurmurHash3_x86_32<numeric::decimal64>::
+  operator()(numeric::decimal64 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
 MurmurHash3_x86_32<numeric::decimal128>::result_type
-  __device__ inline MurmurHash3_x86_32<numeric::decimal128>::operator()(
-    numeric::decimal128 const& key) const
+  __device__ inline MurmurHash3_x86_32<numeric::decimal128>::
+  operator()(numeric::decimal128 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
 MurmurHash3_x86_32<cudf::list_view>::result_type
-  __device__ inline MurmurHash3_x86_32<cudf::list_view>::operator()(
-    cudf::list_view const& key) const
+  __device__ inline MurmurHash3_x86_32<cudf::list_view>::
+  operator()(cudf::list_view const& key) const
 {
   CUDF_UNREACHABLE("List column hashing is not supported");
 }
 
 template <>
 MurmurHash3_x86_32<cudf::struct_view>::result_type
-  __device__ inline MurmurHash3_x86_32<cudf::struct_view>::operator()(
-    cudf::struct_view const& key) const
+  __device__ inline MurmurHash3_x86_32<cudf::struct_view>::
+  operator()(cudf::struct_view const& key) const
 {
   CUDF_UNREACHABLE("Direct hashing of struct_view is not supported");
 }

cpp/include/cudf/hashing/detail/xxhash_32.cuh

@@ -73,30 +73,30 @@ XXHash_32<double>::result_type __device__ inline XXHash_32<double>::operator()(
 }
 
 template <>
-XXHash_32<cudf::string_view>::result_type
-  __device__ inline XXHash_32<cudf::string_view>::operator()(cudf::string_view const& key) const
+XXHash_32<cudf::string_view>::result_type __device__ inline XXHash_32<cudf::string_view>::
+operator()(cudf::string_view const& key) const
 {
   return this->compute_bytes(reinterpret_cast<cuda::std::byte const*>(key.data()),
                              key.size_bytes());
 }
 
 template <>
-XXHash_32<numeric::decimal32>::result_type
-  __device__ inline XXHash_32<numeric::decimal32>::operator()(numeric::decimal32 const& key) const
+XXHash_32<numeric::decimal32>::result_type __device__ inline XXHash_32<numeric::decimal32>::
+operator()(numeric::decimal32 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
-XXHash_32<numeric::decimal64>::result_type
-  __device__ inline XXHash_32<numeric::decimal64>::operator()(numeric::decimal64 const& key) const
+XXHash_32<numeric::decimal64>::result_type __device__ inline XXHash_32<numeric::decimal64>::
+operator()(numeric::decimal64 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
-XXHash_32<numeric::decimal128>::result_type
-  __device__ inline XXHash_32<numeric::decimal128>::operator()(numeric::decimal128 const& key) const
+XXHash_32<numeric::decimal128>::result_type __device__ inline XXHash_32<numeric::decimal128>::
+operator()(numeric::decimal128 const& key) const
 {
   return this->compute(key.value());
 }
@@ -109,8 +109,8 @@ XXHash_32<cudf::list_view>::result_type __device__ inline XXHash_32<cudf::list_v
 }
 
 template <>
-XXHash_32<cudf::struct_view>::result_type
-  __device__ inline XXHash_32<cudf::struct_view>::operator()(cudf::struct_view const& key) const
+XXHash_32<cudf::struct_view>::result_type __device__ inline XXHash_32<cudf::struct_view>::
+operator()(cudf::struct_view const& key) const
 {
   CUDF_UNREACHABLE("Direct hashing of struct_view is not supported");
 }

cpp/include/cudf/hashing/detail/xxhash_64.cuh

@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2024, NVIDIA CORPORATION.
+ * Copyright (c) 2024-2025, NVIDIA CORPORATION.
  *
  * Licensed under the Apache License, Version 2.0 (the "License");
  * you may not use this file except in compliance with the License.
@@ -71,30 +71,30 @@ XXHash_64<double>::result_type __device__ inline XXHash_64<double>::operator()(
 }
 
 template <>
-XXHash_64<cudf::string_view>::result_type
-  __device__ inline XXHash_64<cudf::string_view>::operator()(cudf::string_view const& key) const
+XXHash_64<cudf::string_view>::result_type __device__ inline XXHash_64<cudf::string_view>::
+operator()(cudf::string_view const& key) const
 {
   return this->compute_bytes(reinterpret_cast<cuda::std::byte const*>(key.data()),
                              key.size_bytes());
 }
 
 template <>
-XXHash_64<numeric::decimal32>::result_type
-  __device__ inline XXHash_64<numeric::decimal32>::operator()(numeric::decimal32 const& key) const
+XXHash_64<numeric::decimal32>::result_type __device__ inline XXHash_64<numeric::decimal32>::
+operator()(numeric::decimal32 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
-XXHash_64<numeric::decimal64>::result_type
-  __device__ inline XXHash_64<numeric::decimal64>::operator()(numeric::decimal64 const& key) const
+XXHash_64<numeric::decimal64>::result_type __device__ inline XXHash_64<numeric::decimal64>::
+operator()(numeric::decimal64 const& key) const
 {
   return this->compute(key.value());
 }
 
 template <>
-XXHash_64<numeric::decimal128>::result_type
-  __device__ inline XXHash_64<numeric::decimal128>::operator()(numeric::decimal128 const& key) const
+XXHash_64<numeric::decimal128>::result_type __device__ inline XXHash_64<numeric::decimal128>::
+operator()(numeric::decimal128 const& key) const
 {
   return this->compute(key.value());
 }

cpp/include/cudf/io/data_sink.hpp

@@ -101,7 +101,7 @@ class data_sink {
   /**
    * @brief Base class destructor
    */
-  virtual ~data_sink(){};
+  virtual ~data_sink() {};
 
   /**
    * @pure @brief Append the buffer content to the sink