Fix build issues with documentation

cub/cub/agent/agent_radix_sort_onesweep.cuh

@@ -58,7 +58,7 @@ enum RadixSortStoreAlgorithm
   RADIX_SORT_STORE_ALIGNED
 };
 
-#if !_CCCL_COMPILER(NVRTC)
+#if !_CCCL_COMPILER(NVRTC) && !defined(_CCCL_DOXYGEN_INVOKED)
 inline ::std::ostream& operator<<(::std::ostream& os, RadixSortStoreAlgorithm algo)
 {
   switch (algo)
@@ -71,7 +71,7 @@ inline ::std::ostream& operator<<(::std::ostream& os, RadixSortStoreAlgorithm al
       return os << "<unknown RadixSortStoreAlgorithm: " << static_cast<int>(algo) << ">";
   }
 }
-#endif // !_CCCL_COMPILER(NVRTC)
+#endif // !_CCCL_COMPILER(NVRTC) && !_CCCL_DOXYGEN_INVOKED
 
 template <int NominalBlockThreads4B,
           int NominalItemsPerThread4B,

cub/cub/block/block_load.cuh

@@ -696,7 +696,7 @@ enum BlockLoadAlgorithm
   BLOCK_LOAD_WARP_TRANSPOSE_TIMESLICED,
 };
 
-#if !_CCCL_COMPILER(NVRTC)
+#if !_CCCL_COMPILER(NVRTC) && !defined(_CCCL_DOXYGEN_INVOKED)
 inline ::std::ostream& operator<<(::std::ostream& os, BlockLoadAlgorithm algo)
 {
   switch (algo)
@@ -717,7 +717,7 @@ inline ::std::ostream& operator<<(::std::ostream& os, BlockLoadAlgorithm algo)
       return os << "<unknown BlockLoadAlgorithm: " << static_cast<int>(algo) << ">";
   }
 }
-#endif // !_CCCL_COMPILER(NVRTC)
+#endif // !_CCCL_COMPILER(NVRTC) && !_CCCL_DOXYGEN_INVOKED
 
 //! @rst
 //! The BlockLoad class provides :ref:`collective <collective-primitives>` data movement methods for loading a linear

cub/cub/block/block_radix_rank.cuh

@@ -74,7 +74,7 @@ enum RadixRankAlgorithm
   RADIX_RANK_MATCH_EARLY_COUNTS_ATOMIC_OR
 };
 
-#if !_CCCL_COMPILER(NVRTC)
+#if !_CCCL_COMPILER(NVRTC) && !defined(_CCCL_DOXYGEN_INVOKED)
 inline ::std::ostream& operator<<(::std::ostream& os, RadixRankAlgorithm algo)
 {
   switch (algo)
@@ -93,7 +93,7 @@ inline ::std::ostream& operator<<(::std::ostream& os, RadixRankAlgorithm algo)
       return os << "<unknown RadixRankAlgorithm: " << static_cast<int>(algo) << ">";
   }
 }
-#endif // !_CCCL_COMPILER(NVRTC)
+#endif // !_CCCL_COMPILER(NVRTC) && !_CCCL_DOXYGEN_INVOKED
 
 /** Empty callback implementation */
 template <int BINS_PER_THREAD>

cub/cub/block/block_reduce.cuh

@@ -152,7 +152,7 @@ enum BlockReduceAlgorithm
   BLOCK_REDUCE_WARP_REDUCTIONS_NONDETERMINISTIC,
 };
 
-#if !_CCCL_COMPILER(NVRTC)
+#if !_CCCL_COMPILER(NVRTC) && !defined(_CCCL_DOXYGEN_INVOKED)
 inline ::std::ostream& operator<<(::std::ostream& os, const BlockReduceAlgorithm& alg)
 {
   switch (alg)
@@ -169,7 +169,7 @@ inline ::std::ostream& operator<<(::std::ostream& os, const BlockReduceAlgorithm
       return os << "<unknown BlockReduceAlgorithm: " << static_cast<int>(alg) << ">";
   }
 }
-#endif // !_CCCL_COMPILER(NVRTC)
+#endif // !_CCCL_COMPILER(NVRTC) && !_CCCL_DOXYGEN_INVOKED
 
 //! @rst
 //! The BlockReduce class provides :ref:`collective <collective-primitives>` methods for computing a

cub/cub/block/block_scan.cuh

@@ -103,7 +103,7 @@ enum BlockScanAlgorithm
   BLOCK_SCAN_WARP_SCANS,
 };
 
-#if !_CCCL_COMPILER(NVRTC)
+#if !_CCCL_COMPILER(NVRTC) && !defined(_CCCL_DOXYGEN_INVOKED)
 inline ::std::ostream& operator<<(::std::ostream& os, BlockScanAlgorithm algo)
 {
   switch (algo)
@@ -118,7 +118,7 @@ inline ::std::ostream& operator<<(::std::ostream& os, BlockScanAlgorithm algo)
       return os << "<unknown BlockScanAlgorithm: " << static_cast<int>(algo) << ">";
   }
 }
-#endif // !_CCCL_COMPILER(NVRTC)
+#endif // !_CCCL_COMPILER(NVRTC) && !_CCCL_DOXYGEN_INVOKED
 
 //! @rst
 //! The BlockScan class provides :ref:`collective <collective-primitives>` methods for computing a parallel prefix

cub/cub/block/block_store.cuh

@@ -530,7 +530,7 @@ enum BlockStoreAlgorithm
   BLOCK_STORE_WARP_TRANSPOSE_TIMESLICED,
 };
 
-#if !_CCCL_COMPILER(NVRTC)
+#if !_CCCL_COMPILER(NVRTC) && !defined(_CCCL_DOXYGEN_INVOKED)
 inline ::std::ostream& operator<<(::std::ostream& os, BlockStoreAlgorithm algo)
 {
   switch (algo)
@@ -551,7 +551,7 @@ inline ::std::ostream& operator<<(::std::ostream& os, BlockStoreAlgorithm algo)
       return os << "<unknown BlockStoreAlgorithm: " << static_cast<int>(algo) << ">";
   }
 }
-#endif // !_CCCL_COMPILER(NVRTC)
+#endif // !_CCCL_COMPILER(NVRTC) && !_CCCL_DOXYGEN_INVOKED
 
 //! @rst
 //! The BlockStore class provides :ref:`collective <collective-primitives>` data movement

cub/cub/device/dispatch/dispatch_common.cuh

@@ -15,32 +15,32 @@
 
 CUB_NAMESPACE_BEGIN
 
-// Options for specifying memory aliasing
+//! @brief Options for specifying memory aliasing
 enum class MayAlias
 {
   Yes,
   No
 };
 
-// Options for specifying sorting order.
+//! @brief Options for specifying sorting order.
 enum class SortOrder
 {
   Ascending,
   Descending
 };
 
-// Options for specifying the behavior of the stream compaction algorithm.
+//! @brief Options for specifying the behavior of the stream compaction algorithm.
 enum class SelectImpl
 {
-  // Stream compaction, discarding rejected items. It's required that memory of input and output are disjoint.
+  //! Stream compaction, discarding rejected items. It's required that memory of input and output are disjoint.
   Select,
-  // Stream compaction, discarding rejected items. Memory of the input may be identical to the memory of the output.
+  //! Stream compaction, discarding rejected items. Memory of the input may be identical to the memory of the output.
   SelectPotentiallyInPlace,
-  // Partition, keeping rejected items. It's required that memory of input and output are disjoint.
+  //! Partition, keeping rejected items. It's required that memory of input and output are disjoint.
   Partition
 };
 
-// Options for forcing inclusive prefix-scan even when initial value has been provided
+//! @brief Options for forcing inclusive prefix-scan even when initial value has been provided
 enum class ForceInclusive
 {
   Yes,

cub/cub/thread/thread_load.cuh

@@ -50,7 +50,7 @@ enum CacheLoadModifier
   LOAD_VOLATILE, ///< Volatile (any memory space)
 };
 
-#if !_CCCL_COMPILER(NVRTC)
+#if !_CCCL_COMPILER(NVRTC) && !defined(_CCCL_DOXYGEN_INVOKED)
 inline ::std::ostream& operator<<(::std::ostream& os, CacheLoadModifier modifier)
 {
   switch (modifier)
@@ -73,7 +73,7 @@ inline ::std::ostream& operator<<(::std::ostream& os, CacheLoadModifier modifier
       return os << "<unknown CacheLoadModifier: " << static_cast<int>(modifier) << ">";
   }
 }
-#endif // !_CCCL_COMPILER(NVRTC)
+#endif // !_CCCL_COMPILER(NVRTC) && !_CCCL_DOXYGEN_INVOKED
 
 /**
  * @name Thread I/O (cache modified)

cudax/include/cuda/experimental/__container/uninitialized_buffer.cuh

@@ -256,7 +256,7 @@ public:
   }
 
   //! @rst
-  //! Returns a \c const reference to the :ref:`any_resource <cudax-memory-resource-any-resource>`
+  //! Returns a \c const reference to the :ref:`any_resource <libcudacxx-memory-resource-any-resource>`
   //! that holds the memory resource used to allocate the buffer
   //! @endrst
   [[nodiscard]] _CCCL_HIDE_FROM_ABI const __resource& memory_resource() const noexcept

docs/cudax/container.rst

@@ -20,12 +20,12 @@ annotations are checked by the type system.
    :widths: 25 45 30
    :header-rows: 0
 
-   * - :ref:`<cuda/experimental/container.cuh> <cudax-containers-uninitialized-buffer>`
+   * - :ref:`<cuda/experimental/container.cuh> <libcudacxx-containers-uninitialized-buffer>`
      - Facilities providing uninitialized *heterogeneous* potentially stream ordered storage satisfying a set of properties
      - cudax 2.7.0 / CCCL 2.7.0
-   * - :ref:`<cuda/experimental/container.cuh> <cudax-containers-heterogeneous-iterator>`
+   * - :ref:`<cuda/experimental/container.cuh> <libcudacxx-containers-heterogeneous-iterator>`
      - Iterator providing type safe iteration and memory access through a set of properties
      - cudax 2.7.0 / CCCL 2.7.0
-   * - :ref:`<cuda/experimental/container.cuh> <cudax-containers-async-vector>`
+   * - :ref:`<cuda/experimental/container.cuh> <libcudacxx-containers-async-vector>`
      - High level containers, that provide *heterogeneous* storage satisfying a set of properties
      - cudax 2.7.0 / CCCL 2.7.0

docs/cudax/index.rst

@@ -18,9 +18,9 @@ CUDA Experimental
 However, any feature within this library has important use cases and we encourage users to experiment with them.
 
 Specifically, ``cudax`` provides:
-   - :ref:`uninitialized storage <cudax-containers-uninitialized-buffer>`
-   - :ref:`an owning type erased memory resource <cudax-memory-resource-any-async-resource>`
-   - :ref:`stream-ordered memory resources <cudax-memory-resource-async>`
+   - :ref:`uninitialized storage <libcudacxx-containers-uninitialized-buffer>`
+   - :ref:`an owning type erased memory resource <libcudacxx-memory-resource-any-async-resource>`
+   - :ref:`stream-ordered memory resources <libcudacxx-memory-resource-async>`
    - :ref:`graph functionality <cudax-graph>`
    - dimensions description functionality
    - :ref:`an implementation of the STF (Sequential Task Flow) programming model <stf>`

docs/cudax/memory_resource.rst

@@ -18,16 +18,16 @@ Memory Resources
 
 The ``<cuda/experimental/memory_resource.cuh>`` header provides:
 
-   -  :ref:`any_synchronous_resource <cudax-memory-resource-any-resource>` and
-      :ref:`any_resource <cudax-memory-resource-any-async-resource>` type erased memory resources similar to
+   -  :ref:`any_synchronous_resource <libcudacxx-memory-resource-any-resource>` and
+      :ref:`any_resource <libcudacxx-memory-resource-any-async-resource>` type erased memory resources similar to
       ``std::any``. In contrast to :ref:`resource_ref <libcudacxx-extended-api-memory-resources-resource-ref>` they
       own the contained resource.
-   -  :ref:`device_memory_resource <cudax-memory-resource-async>` A standard C++ interface for *heterogeneous*,
+   -  :ref:`device_memory_resource <libcudacxx-memory-resource-async>` A standard C++ interface for *heterogeneous*,
       *stream-ordered* memory allocation tailored to the needs of CUDA C++ developers. This design builds off of the
       success of the `RAPIDS Memory Manager (RMM) <https://github.com/rapidsai/rmm>`__ project and evolves the design
       based on lessons learned.
-   -  :ref:`shared_resource <cudax-memory-resource-shared-resource>` a type erased reference counted memory resource.
-      In contrast to :ref:`any_resource <cudax-memory-resource-any-resource>` it additionally provides shared ownership
+   -  :ref:`shared_resource <libcudacxx-memory-resource-shared-resource>` a type erased reference counted memory resource.
+      In contrast to :ref:`any_resource <libcudacxx-memory-resource-any-resource>` it additionally provides shared ownership
       semantics.
 
 ``<cuda/experimental/memory_resource.cuh>`` is not intended to replace RMM, but instead moves the definition of the

docs/libcudacxx/Doxyfile

@@ -9,13 +9,14 @@ GENERATE_XML           = YES
 XML_OUTPUT             = xml
 XML_PROGRAMLISTING     = YES
 
-INPUT                  = ../../libcudacxx/include/cuda/__iterator \
-                         ../../libcudacxx/include/cuda/__stream \
+INPUT                  = ../../libcudacxx/include/cuda/__algorithm \
+                         ../../libcudacxx/include/cuda/__container \
                          ../../libcudacxx/include/cuda/__device \
                          ../../libcudacxx/include/cuda/__event \
-                         ../../libcudacxx/include/cuda/__algorithm \
-                         ../../libcudacxx/include/cuda/__memory_resource \
+                         ../../libcudacxx/include/cuda/__iterator \
                          ../../libcudacxx/include/cuda/__memory_pool \
+                         ../../libcudacxx/include/cuda/__memory_resource \
+                         ../../libcudacxx/include/cuda/__stream \
                          ../../libcudacxx/include/nv
 
 RECURSIVE              = YES

docs/libcudacxx/extended_api/math.rst

@@ -18,6 +18,7 @@ Math
    math/uabs
    math/fast_mod_div
    math/mul_hi
+   math/sincos
 
 .. list-table::
    :widths: 25 45 30 30

docs/libcudacxx/extended_api/memory_resource.rst

@@ -44,5 +44,5 @@ With the current design it ranges from cumbersome to impossible to verify whethe
 that are e.g. accessible on device, or whether it can utilize other allocation mechanisms.
 
 To better support asynchronous CUDA `stream-ordered allocations <https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#stream-ordered-memory-allocator>`__
-libcu++ provides :ref:`cuda::stream_ref <libcudacxx-extended-api-streams-stream-ref>` as a wrapper around
+libcu++ provides :cpp:class:`cuda::stream_ref <cuda::stream_ref>` as a wrapper around
 ``cudaStream_t``. The definition of ``cuda::stream_ref`` can be found in the ``<cuda/stream>`` header.

docs/libcudacxx/ptx/pragmas.rst

@@ -6,7 +6,7 @@ PTX Pragmas
 .. toctree::
    :maxdepth: 1
 
-   pragma/enable_smem_spilling
+   pragmas/enable_smem_spilling
 
 .. list-table:: `.pragma Strings <https://docs.nvidia.com/cuda/parallel-thread-execution/#descriptions-pragma-strings>`__
    :widths: 50 50

docs/libcudacxx/runtime/event.rst

@@ -53,6 +53,7 @@ Availability: CCCL 3.1.0 / CUDA 13.1
    }
 
 .. _cccl-runtime-event-timed-event:
+
 ``cuda::timed_event``
 -----------------------------------------------------
 

libcudacxx/include/cuda/__container/buffer.h

@@ -62,7 +62,7 @@ template <class _Env>
 inline constexpr bool __buffer_compatible_env = ::cuda::std::is_same_v<_Env, ::cuda::std::execution::env<>>;
 
 //! @rst
-//! .. _cudax-containers-async-vector:
+//! .. _libcudacxx-containers-async-vector:
 //!
 //! buffer
 //! -------------
@@ -520,9 +520,8 @@ class buffer
   //! @}
 
   //! @rst
-  //! Returns a \c const reference to the :ref:`any_resource
-  //! <cuda-memory-resource-any-resource>` that holds the memory resource used
-  //! to allocate the buffer
+  //! Returns a \c const reference to the :ref:`any_resource <libcudacxx-memory-resource-any-resource>` that holds the
+  //! memory resource used to allocate the buffer
   //! @endrst
   [[nodiscard]] _CCCL_HIDE_FROM_ABI const __resource_t& memory_resource() const noexcept
   {

libcudacxx/include/cuda/__container/heterogeneous_iterator.h

@@ -44,7 +44,7 @@ enum class _IsConstIter
   __yes,
 };
 //! @rst
-//! .. _cudax-containers-heterogeneous-iterator:
+//! .. _libcudacxx-containers-heterogeneous-iterator:
 //!
 //! Type safe iterator over heterogeneous memory
 //! ---------------------------------------------

libcudacxx/include/cuda/__container/uninitialized_async_buffer.h

@@ -43,7 +43,7 @@
 _CCCL_BEGIN_NAMESPACE_CUDA
 
 //! @rst
-//! .. _cudax-containers-uninitialized-async-buffer:
+//! .. _libcudacxx-containers-uninitialized-async-buffer:
 //!
 //! Uninitialized stream-ordered type-safe memory storage
 //! ------------------------------------------------------

libcudacxx/include/cuda/__iterator/permutation_iterator.h

@@ -266,6 +266,7 @@ class permutation_iterator
     return __tmp;
   }
 
+#ifndef _CCCL_DOXYGEN_INVOKED // Doxygen has issues with constexpr friend operators
   //! @brief Advances a @c permutation_iterator by a given number of elements
   //! @param __iter The original @c permutation_iterator
   //! @param __n The number of elements to advance
@@ -292,16 +293,6 @@ class permutation_iterator
     return permutation_iterator{__iter.__iter_, __iter.__index_ + __n};
   }
 
-  //! @brief Advances the @c permutation_iterator by a given number of elements
-  //! @param __n The number of elements to advance
-  //! @return Equivalent to ``index + __n``
-  _CCCL_EXEC_CHECK_DISABLE
-  _CCCL_API constexpr permutation_iterator& operator+=(difference_type __n) noexcept(noexcept(__index_ += __n))
-  {
-    __index_ += __n;
-    return *this;
-  }
-
   //! @brief Decrements a @c permutation_iterator by a given number of elements
   //! @param __iter The original @c permutation_iterator
   //! @param __n The number of elements to decrement
@@ -314,6 +305,17 @@ class permutation_iterator
   {
     return permutation_iterator{__iter.__iter_, __iter.__index_ - __n};
   }
+#endif // !_CCCL_DOXYGEN_INVOKED
+
+  //! @brief Advances the @c permutation_iterator by a given number of elements
+  //! @param __n The number of elements to advance
+  //! @return Equivalent to ``index + __n``
+  _CCCL_EXEC_CHECK_DISABLE
+  _CCCL_API constexpr permutation_iterator& operator+=(difference_type __n) noexcept(noexcept(__index_ += __n))
+  {
+    __index_ += __n;
+    return *this;
+  }
 
   //! @brief Decrements the @c permutation_iterator by a given number of elements
   //! @param __n The number of elements to decrement