[Misc] Replace /*! with /**. (#4823)

* replace

* blabla

* balbla

* blabla
Co-authored-by: Steve <ubuntu@ip-172-31-34-29.ap-northeast-1.compute.internal>

include/dgl/runtime/tensordispatch.h

-/*!
+/**
  *  Copyright (c) 2020-2022 by Contributors
  * @file array/tensordispatch.h
  * @brief This file defines the dispatcher of tensor operators to
@@ -38,7 +38,7 @@
 #endif  // DGL_USE_CUDA
 #include "ndarray.h"
 
-/*!
+/**
  * @brief Casts a pointer \c entry to a function pointer with signature of \c
  * func.
  */
@@ -47,7 +47,7 @@
 namespace dgl {
 namespace runtime {
 
-/*!
+/**
  * @brief Dispatcher that delegates the function calls to framework-specific C++
  * APIs.
  *
@@ -55,19 +55,19 @@ namespace runtime {
  */
 class TensorDispatcher {
  public:
-  /*! @brief Get the singleton instance. */
+  /** @brief Get the singleton instance. */
   static TensorDispatcher* Global() {
     static TensorDispatcher inst;
     return &inst;
   }
 
-  /*! @brief Whether an adapter library is available. */
+  /** @brief Whether an adapter library is available. */
   inline bool IsAvailable() { return available_; }
 
-  /*! @brief Load symbols from the given tensor adapter library path. */
+  /** @brief Load symbols from the given tensor adapter library path. */
   bool Load(const char* path_cstr);
 
-  /*!
+  /**
    * @brief Allocate a piece of CPU memory via PyTorch's CPUAllocator.
    * Used in CPUDeviceAPI::AllocWorkspace().
    *
@@ -79,7 +79,7 @@ class TensorDispatcher {
     return FUNCCAST(tensoradapter::CPURawAlloc, entry)(nbytes);
   }
 
-  /*!
+  /**
    * @brief Free the CPU memory.
    * Used in CPUDeviceAPI::FreeWorkspace().
    *
@@ -91,7 +91,7 @@ class TensorDispatcher {
   }
 
 #ifdef DGL_USE_CUDA
-  /*!
+  /**
    * @brief Allocate a piece of GPU memory via
    * PyTorch's THCCachingAllocator.
    * Used in CUDADeviceAPI::AllocWorkspace().
@@ -109,7 +109,7 @@ class TensorDispatcher {
     return FUNCCAST(tensoradapter::CUDARawAlloc, entry)(nbytes, stream);
   }
 
-  /*!
+  /**
    * @brief Free the GPU memory.
    * Used in CUDADeviceAPI::FreeWorkspace().
    *
@@ -120,7 +120,7 @@ class TensorDispatcher {
     FUNCCAST(tensoradapter::CUDARawDelete, entry)(ptr);
   }
 
-  /*!
+  /**
    * @brief Find the current PyTorch CUDA stream
    * Used in runtime::getCurrentCUDAStream().
    *
@@ -136,7 +136,7 @@ class TensorDispatcher {
   }
 #endif  // DGL_USE_CUDA
 
-  /*!
+  /**
    * @brief Record streams that are using this tensor.
    * Used in NDArray::RecordStream().
    *
@@ -153,12 +153,12 @@ class TensorDispatcher {
   }
 
  private:
-  /*! @brief ctor */
+  /** @brief ctor */
   TensorDispatcher() = default;
-  /*! @brief dtor */
+  /** @brief dtor */
   ~TensorDispatcher();
 
-  /*!
+  /**
    * @brief List of symbols in the adapter library.
    *
    * Must match the functions in tensoradapter/include/tensoradapter.h.
@@ -170,7 +170,7 @@ class TensorDispatcher {
 #endif  // DGL_USE_CUDA
   };
 
-  /*! @brief Index of each function to the symbol list */
+  /** @brief Index of each function to the symbol list */
   class Op {
    public:
     static constexpr int kCPURawAlloc = 0;
@@ -183,10 +183,10 @@ class TensorDispatcher {
 #endif  // DGL_USE_CUDA
   };
 
-  /*! @brief Number of functions */
+  /** @brief Number of functions */
   static constexpr int num_entries_ = sizeof(names_) / sizeof(names_[0]);
 
-  /*! @brief Entrypoints of each function */
+  /** @brief Entrypoints of each function */
   void* entrypoints_[num_entries_] = {
       nullptr, nullptr,
 #ifdef DGL_USE_CUDA

include/dgl/runtime/threading_backend.h

-/*!
+/**
  *  Copyright (c) 2018 by Contributors
  * @file dgl/runtime/threading_backend.h
  * @brief Utilities for manipulating thread pool threads.
@@ -14,7 +14,7 @@ namespace dgl {
 namespace runtime {
 namespace threading {
 
-/*!
+/**
  * @brief A platform-agnostic abstraction for managing a collection of
  *        thread pool threads.
  */
@@ -22,7 +22,7 @@ class ThreadGroup {
  public:
   class Impl;
 
-  /*!
+  /**
    * @brief Creates a collection of threads which run a provided function.
    *
    * @param num_workers The total number of worker threads in this group.
@@ -39,7 +39,7 @@ class ThreadGroup {
       bool exclude_worker0 = false);
   ~ThreadGroup();
 
-  /*!
+  /**
    * @brief Blocks until all non-main threads in the pool finish.
    */
   void Join();
@@ -49,7 +49,7 @@ class ThreadGroup {
     kLittle = -1,
   };
 
-  /*!
+  /**
    * @brief configure the CPU id affinity
    *
    * @param mode The preferred CPU type (1 = big, -1 = little).
@@ -67,14 +67,14 @@ class ThreadGroup {
   Impl* impl_;
 };
 
-/*!
+/**
  * @brief Platform-agnostic no-op.
  */
 // This used to be Yield(), renaming to YieldThread() because windows.h defined
 // it as a macro in later SDKs.
 void YieldThread();
 
-/*!
+/**
  * @return the maximum number of effective workers for this system.
  */
 int MaxConcurrency();

include/dgl/runtime/util.h

-/*!
+/**
  *  Copyright (c) 2017 by Contributors
  * @file dgl/runtime/util.h
  * @brief Useful runtime util.
@@ -11,7 +11,7 @@
 namespace dgl {
 namespace runtime {
 
-/*!
+/**
  * @brief Check whether type matches the given spec.
  * @param t The type
  * @param code The type code.
@@ -28,7 +28,7 @@ inline bool TypeMatch(DGLDataType t, int code, int bits, int lanes = 1) {
 namespace dgl {
 namespace ir {
 namespace intrinsic {
-/*! @brief The kind of structure field info used in intrinsic */
+/** @brief The kind of structure field info used in intrinsic */
 enum DGLStructFieldKind : int {
   // array head address
   kArrAddr,

include/dgl/sampler.h

-/*!
+/**
  *  Copyright (c) 2018 by Contributors
  * @file dgl/sampler.h
  * @brief DGL sampler header.
@@ -20,7 +20,7 @@ class ImmutableGraph;
 
 class SamplerOp {
  public:
-  /*!
+  /**
    * @brief Sample a graph from the seed vertices with neighbor sampling.
    * The neighbors are sampled with a uniform distribution.
    *
@@ -39,7 +39,7 @@ class SamplerOp {
       const std::string &edge_type, int num_hops, int expand_factor,
       const bool add_self_loop, const ValueType *probability);
 
-  /*!
+  /**
    * @brief Sample a graph from the seed vertices with layer sampling.
    * The layers are sampled with a uniform distribution.
    *

include/dgl/sampling/negative.h

-/*!
+/**
  *  Copyright (c) 2020 by Contributors
  * @file dgl/sampling/negative.h
  * @brief Negative sampling.
@@ -14,7 +14,7 @@
 namespace dgl {
 namespace sampling {
 
-/*!
+/**
  * @brief Given an edge type, uniformly sample source-destination pairs that do
  * not have an edge in between using rejection sampling.
  *

include/dgl/sampling/neighbor.h

-/*!
+/**
  *  Copyright (c) 2020 by Contributors
  * @file dgl/sampling/neighbor.h
  * @brief Neighborhood-based sampling.
@@ -14,7 +14,7 @@
 namespace dgl {
 namespace sampling {
 
-/*!
+/**
  * @brief Sample from the neighbors of the given nodes and return the sampled
  * edges as a graph.
  *
@@ -47,7 +47,7 @@ HeteroSubgraph SampleNeighbors(
     const std::vector<FloatArray>& probability,
     const std::vector<IdArray>& exclude_edges, bool replace = true);
 
-/*!
+/**
  * Select the neighbors with k-largest weights on the connecting edges for each
  * given node.
  *

include/dgl/sampling/randomwalks.h

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file dgl/samplinig/randomwalks.h
  * @brief Random walk functions.
@@ -17,7 +17,7 @@ namespace dgl {
 
 namespace sampling {
 
-/*!
+/**
  * @brief Metapath-based random walk.
  * @param hg The heterograph.
  * @param seeds A 1D array of seed nodes, with the type the source type of the
@@ -37,7 +37,7 @@ std::tuple<IdArray, IdArray, TypeArray> RandomWalk(
     const HeteroGraphPtr hg, const IdArray seeds, const TypeArray metapath,
     const std::vector<FloatArray> &prob);
 
-/*!
+/**
  * @brief Metapath-based random walk with restart probability.
  * @param hg The heterograph.
  * @param seeds A 1D array of seed nodes, with the type the source type of the
@@ -58,7 +58,7 @@ std::tuple<IdArray, IdArray, TypeArray> RandomWalkWithRestart(
     const HeteroGraphPtr hg, const IdArray seeds, const TypeArray metapath,
     const std::vector<FloatArray> &prob, double restart_prob);
 
-/*!
+/**
  * @brief Metapath-based random walk with stepwise restart probability. Useful
  *        for PinSAGE-like models.
  * @param hg The heterograph.

include/dgl/scheduler.h

-/*!
+/**
  *  Copyright (c) 2018 by Contributors
  * @file dgl/scheduler.h
  * @brief Operations on graph index.
@@ -16,7 +16,7 @@ typedef dgl::runtime::NDArray IdArray;
 
 namespace sched {
 
-/*!
+/**
  * @brief Generate degree bucketing schedule
  * @tparam IdType Graph's index data type, can be int32_t or int64_t
  * @param msg_ids The edge id for each message
@@ -35,7 +35,7 @@ template <class IdType>
 std::vector<IdArray> DegreeBucketing(
     const IdArray& msg_ids, const IdArray& vids, const IdArray& recv_ids);
 
-/*!
+/**
  * @brief Generate degree bucketing schedule for group_apply edge
  * @tparam IdType Graph's index data type, can be int32_t or int64_t
  * @param uids One end vertex of edge by which edges are grouped

include/dgl/transform.h

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file dgl/transform.h
  * @brief DGL graph transformations
@@ -18,7 +18,7 @@ namespace dgl {
 
 namespace transform {
 
-/*!
+/**
  * @brief Given a list of graphs, remove the common nodes that do not have
  * inbound and outbound edges.
  *
@@ -37,7 +37,7 @@ std::pair<std::vector<HeteroGraphPtr>, std::vector<IdArray>> CompactGraphs(
     const std::vector<HeteroGraphPtr> &graphs,
     const std::vector<IdArray> &always_preserve);
 
-/*!
+/**
  * @brief Convert a graph into a bipartite-structured graph for message passing.
  *
  * Specifically, we create one node type \c ntype_l on the "left" side and
@@ -83,7 +83,7 @@ std::tuple<HeteroGraphPtr, std::vector<IdArray>, std::vector<IdArray>> ToBlock(
     HeteroGraphPtr graph, const std::vector<IdArray> &rhs_nodes,
     bool include_rhs_in_lhs);
 
-/*!
+/**
  * @brief Convert a multigraph to a simple graph.
  *
  * @return A triplet of
@@ -116,7 +116,7 @@ std::tuple<HeteroGraphPtr, std::vector<IdArray>, std::vector<IdArray>> ToBlock(
 std::tuple<HeteroGraphPtr, std::vector<IdArray>, std::vector<IdArray>>
 ToSimpleGraph(const HeteroGraphPtr graph);
 
-/*!
+/**
  * @brief Remove edges from a graph.
  *
  * @param graph The graph.

include/dgl/zerocopy_serializer.h

-/*!
+/**
  *  Copyright (c) 2020 by Contributors
  * @file rpc/shared_mem_serializer.h
  * @brief headers for serializer.
@@ -23,7 +23,7 @@
 
 namespace dgl {
 
-/*!
+/**
  *
  * StreamWithBuffer is backed up by dmlc::MemoryFixedSizeStream or
  * dmlc::MemoryStringStream. This class supports serializing and deserializing
@@ -62,7 +62,7 @@ class StreamWithBuffer : public dmlc::SeekStream {
     explicit Buffer(void* data) : data(data) {}
   };
 
-  /*!
+  /**
    * @brief This constructor is for writing scenario or reading from local
    * machine
    * @param strm The backup stream to write/load from
@@ -74,7 +74,7 @@ class StreamWithBuffer : public dmlc::SeekStream {
       : strm_(std::move(strm)),
         buffer_list_(),
         send_to_remote_(send_to_remote) {}
-  /*!
+  /**
    * @brief This constructor is for reading from remote
    * @param strm The stream to write/load from zerocopy write/load
    * @param data_ptr_list list of pointer to reconstruct NDArray
@@ -94,7 +94,7 @@ class StreamWithBuffer : public dmlc::SeekStream {
     }
   }
 
-  /*!
+  /**
    * @brief Construct stream backed up by string
    * @param blob The string to write/load from zerocopy write/load
    * @param send_to_remote Whether this stream will be deserialized at remote
@@ -105,7 +105,7 @@ class StreamWithBuffer : public dmlc::SeekStream {
       : strm_(new dmlc::MemoryStringStream(blob)),
         send_to_remote_(send_to_remote) {}
 
-  /*!
+  /**
    * @brief Construct stream backed up by string
    * @param p_buffer buffer pointer
    * @param size buffer size
@@ -117,7 +117,7 @@ class StreamWithBuffer : public dmlc::SeekStream {
       : strm_(new dmlc::MemoryFixedSizeStream(p_buffer, size)),
         send_to_remote_(send_to_remote) {}
 
-  /*!
+  /**
    * @brief Construct stream backed up by string, and reconstruct NDArray
    * from data_ptr_list
    * @param blob The string to write/load from zerocopy write/load
@@ -130,7 +130,7 @@ class StreamWithBuffer : public dmlc::SeekStream {
     }
   }
 
-  /*!
+  /**
    * @brief Construct stream backed up by string, and reconstruct NDArray
    * from data_ptr_list
    * @param p_buffer buffer pointer
@@ -155,14 +155,14 @@ class StreamWithBuffer : public dmlc::SeekStream {
   using dmlc::Stream::Read;
   using dmlc::Stream::Write;
 
-  /*!
+  /**
    * @brief push NDArray into stream
    * If send_to_remote=true, the NDArray will be saved to the buffer list
    * If send_to_remote=false, the NDArray will be saved to the backedup string
    */
   void PushNDArray(const runtime::NDArray& tensor);
 
-  /*!
+  /**
    * @brief pop NDArray from stream
    * If send_to_remote=true, the NDArray will be reconstructed from buffer list
    * If send_to_remote=false, the NDArray will be reconstructed from shared
@@ -170,12 +170,12 @@ class StreamWithBuffer : public dmlc::SeekStream {
    */
   dgl::runtime::NDArray PopNDArray();
 
-  /*!
+  /**
    * @brief Get whether this stream is for remote usage
    */
   bool send_to_remote() { return send_to_remote_; }
 
-  /*!
+  /**
    * @brief Get underlying buffer list
    */
   const std::deque<Buffer>& buffer_list() const { return buffer_list_; }

include/intel/cpu_support.h

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file intel/cpu_support.h
  * @brief Intel CPU support
@@ -53,7 +53,7 @@ struct IntelKernel {
   }
 };
 
-/*!
+/**
  * @brief Element-wise addition kernel using Intel AVX512 instructions.
  * @note it uses AVX512.
  */

include/intel/meta_utils.h

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file intel/meta_utils.h
  * @brief Meta programming utils

src/api/api_container.cc

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file api/api_container.cc
  * @brief Runtime container APIs. (reference: tvm/src/api/api_lang.cc)

src/api/api_test.cc

-/*!
+/**
  *  Copyright (c) 2022 by Contributors
  * @file api/api_test.cc
  * @brief C APIs for testing FFI

src/array/arith.h

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file array/arith.h
  * @brief Arithmetic functors

src/array/array.cc

-/*!
+/**
  *  Copyright (c) 2019-2021 by Contributors
  * @file array/array.cc
  * @brief DGL array utilities implementation
@@ -956,7 +956,7 @@ COOToSimple(const COOMatrix& coo) {
   const COOMatrix &coalesced_adj = coalesced_result.first;
   const IdArray &count = coalesced_result.second;
 
-  /*
+  /**
    * eids_shuffled actually already contains the mapping from old edge space to the
    * new one:
    *

src/array/array_arith.cc

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file array/array_aritch.cc
  * @brief DGL array arithmetic operations

src/array/array_op.h

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file array/array_op.h
  * @brief Array operator templates

src/array/check.h

-/*!
+/**
  *  Copyright (c) 2019 by Contributors
  * @file array/check.h
  * @brief DGL check utilities

src/array/cpu/array_cumsum.cc

-/*!
+/**
  *  Copyright (c) 2020 by Contributors
  * @file array/cpu/array_cumsum.cc
  * @brief Array cumsum CPU implementation