Fixed issue with operator types used as both lvalue/rvalue no assigning

NVIDIA · Jun 25, 2024 · 9bcfeab · 9bcfeab
1 parent ed09e1c
commit 9bcfeab
Show file tree

Hide file tree

Showing 14 changed files with 137 additions and 7 deletions.
diff --git a/include/matx/operators/collapse.h b/include/matx/operators/collapse.h
@@ -51,6 +51,7 @@ namespace matx
         using scalar_type = typename T1::scalar_type;
         using shape_type = index_t;
         using matxoplvalue = bool;
+        using self_type = LCollapseOp<DIM, T1>;
 
         __MATX_INLINE__ std::string str() const { return "lcollapse<" + std::to_string(DIM) + ">(" + op_.str() + ")"; }
         __MATX_INLINE__ LCollapseOp(const T1 op) : op_(op)
@@ -132,6 +133,12 @@ namespace matx
             return op_.Size(DIM + dim - 1);
         }
 
+        ~LCollapseOp() = default;
+        LCollapseOp(const LCollapseOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }              
+
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 
           if constexpr (is_matx_transform_op<R>()) {
@@ -199,6 +206,7 @@ namespace matx
         using scalar_type = typename T1::scalar_type;
         using shape_type = index_t;
         using matxlvalue = bool;
+        using self_type = RCollapseOp<DIM, T1>;
 
         __MATX_INLINE__ std::string str() const { return "rcollapse<" + std::to_string(DIM) + ">(" + op_.str() + ")"; }
 
@@ -281,6 +289,12 @@ namespace matx
             return op_.Size(dim);
         }
 
+        ~RCollapseOp() = default;
+        RCollapseOp(const RCollapseOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }           
+
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 
           if constexpr (is_matx_transform_op<R>()) {

diff --git a/include/matx/operators/concat.h b/include/matx/operators/concat.h
@@ -51,6 +51,7 @@ namespace matx
     {
       using first_type = cuda::std::tuple_element_t<0, cuda::std::tuple<Ts...>>;
       using first_value_type = typename first_type::scalar_type;
+      using self_type = ConcatOp<Ts...>;
 
       static constexpr int RANK = first_type::Rank();
 
@@ -165,6 +166,12 @@ namespace matx
           return cuda::std::get<0>(ops_).Size(dim);
       }
 
+      ~ConcatOp() = default;
+      ConcatOp(const ConcatOp &rhs) = default;
+      __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+        return set(*this, rhs); 
+      }        
+
       template<typename R> 
       __MATX_INLINE__ auto operator=(const R &rhs) { 
         if constexpr (is_matx_transform_op<R>()) {

diff --git a/include/matx/operators/overlap.h b/include/matx/operators/overlap.h
@@ -48,6 +48,7 @@ namespace matx
       public:
         using scalar_type = typename T::scalar_type;
         using shape_type = index_t;
+        using self_type = OverlapOp<DIM, T>;
 
       private:
         typename base_type<T>::type op_;
@@ -118,7 +119,13 @@ namespace matx
           if constexpr (is_matx_op<T>()) {
             op_.PostRun(std::forward<ShapeType>(shape), std::forward<Executor>(ex));
           }
-        }               
+        } 
+
+        ~OverlapOp() = default;
+        OverlapOp(const OverlapOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }                       
 
         template<typename R>
         __MATX_INLINE__ auto operator=(const R &rhs) {

diff --git a/include/matx/operators/permute.h b/include/matx/operators/permute.h
@@ -47,6 +47,7 @@ namespace matx
     {
       public: 
         using scalar_type = typename T::scalar_type;
+        using self_type = PermuteOp<T>;
 
       private:
         typename base_type<T>::type op_;
@@ -133,7 +134,13 @@ namespace matx
           if constexpr (is_matx_op<T>()) {
             op_.PostRun(std::forward<ShapeType>(shape), std::forward<Executor>(ex));
           }
-        }               
+        }   
+
+        ~PermuteOp() = default;
+        PermuteOp(const PermuteOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }                      
 
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 

diff --git a/include/matx/operators/remap.h b/include/matx/operators/remap.h
@@ -57,6 +57,7 @@ namespace matx
         using scalar_type = typename T::scalar_type;
         using shape_type = std::conditional_t<has_shape_type_v<T>, typename T::shape_type, index_t>; 
         using index_type = typename IdxType::scalar_type;
+        using self_type = RemapOp<DIM, T, IdxType>;
         static_assert(std::is_integral<index_type>::value, "RemapOp: Type for index operator must be integral");
         static_assert(IdxType::Rank() <= 1, "RemapOp: Rank of index operator must be 0 or 1");
         static_assert(DIM<T::Rank(), "RemapOp: DIM must be less than Rank of tensor");
@@ -134,7 +135,13 @@ namespace matx
           if constexpr (is_matx_op<T>()) {
             op_.PostRun(std::forward<ShapeType>(shape), std::forward<Executor>(ex));
           }
-        }               
+        } 
+
+        ~RemapOp() = default;
+        RemapOp(const RemapOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }                      
 
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 

diff --git a/include/matx/operators/reshape.h b/include/matx/operators/reshape.h
@@ -56,6 +56,7 @@ namespace matx
       public:
         using matxop = bool;
         using matxoplvalue = bool;
+        using self_type = ReshapeOp<RANK, T, ShapeType>;
 
         __MATX_INLINE__ std::string str() const { return "reshape(" + op_.str() + ")"; }
 
@@ -149,7 +150,13 @@ namespace matx
           if constexpr (is_matx_op<T>()) {
             op_.PostRun(std::forward<S2>(shape), std::forward<Executor>(ex));
           }
-        }               
+        }  
+
+        ~ReshapeOp() = default;
+        ReshapeOp(const ReshapeOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }                      
 
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 

diff --git a/include/matx/operators/reverse.h b/include/matx/operators/reverse.h
@@ -57,6 +57,7 @@ namespace matx
         using matxop = bool;
         using matxoplvalue = bool;
         using scalar_type = typename T1::scalar_type;
+        using self_type = ReverseOp<DIM, T1>;
 
         __MATX_INLINE__ std::string str() const { return "reverse(" + op_.str() + ")"; }
 
@@ -112,7 +113,13 @@ namespace matx
           if constexpr (is_matx_op<T1>()) {
             op_.PostRun(std::forward<ShapeType>(shape), std::forward<Executor>(ex));
           }
-        }               
+        }   
+
+        ~ReverseOp() = default;
+        ReverseOp(const ReverseOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }                       
 
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 

diff --git a/include/matx/operators/set.h b/include/matx/operators/set.h
@@ -68,7 +68,6 @@ class set : public BaseOp<set<T, Op>> {
 public:
   // Type specifier for reflection on class
   using scalar_type = typename T::scalar_type;
-  using shape_type = std::conditional_t<has_shape_type_v<T>, typename T::shape_type, index_t>;
   using tensor_type = T;
   using op_type = Op;
   using matx_setop = bool;
@@ -136,7 +135,9 @@ class set : public BaseOp<set<T, Op>> {
       return r;
     }
   }
-  __MATX_DEVICE__ __MATX_HOST__ inline decltype(auto) operator()(cuda::std::array<shape_type, T::Rank()> idx) const noexcept
+
+  template <typename ShapeType>
+  __MATX_DEVICE__ __MATX_HOST__ inline decltype(auto) operator()(cuda::std::array<ShapeType, T::Rank()> idx) const noexcept
   {
     auto res = cuda::std::apply([&](auto &&...args)  {
         return _internal_mapply(args...);

diff --git a/include/matx/operators/shift.h b/include/matx/operators/shift.h
@@ -61,6 +61,7 @@ namespace matx
         using matxop = bool;
         using matxoplvalue = bool;
         using scalar_type = typename T1::scalar_type;
+        using self_type = ShiftOp<DIM, T1, T2>;
 
         __MATX_INLINE__ std::string str() const { return "shift(" + op_.str() + ")"; }
 
@@ -131,6 +132,12 @@ namespace matx
           return detail::matx_max(size1,size2);
         }
 
+        ~ShiftOp() = default;
+        ShiftOp(const ShiftOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }            
+
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 
           if constexpr (is_matx_transform_op<R>()) {

diff --git a/include/matx/operators/slice.h b/include/matx/operators/slice.h
@@ -48,6 +48,7 @@ namespace matx
       public: 
         using scalar_type = typename T::scalar_type;
         using shape_type = index_t; 
+        using self_type = SliceOp<DIM, T>;
 
       private:
         typename base_type<T>::type op_;
@@ -158,6 +159,12 @@ namespace matx
           return sizes_[dim];
         }
 
+        ~SliceOp() = default;
+        SliceOp(const SliceOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        }            
+
         template<typename R> 
         __MATX_INLINE__ auto operator=(const R &rhs) { 
           if constexpr (is_matx_transform_op<R>()) {

diff --git a/include/matx/operators/stack.h b/include/matx/operators/stack.h
@@ -50,6 +50,7 @@ namespace matx
     {
       using first_type = cuda::std::tuple_element_t<0, cuda::std::tuple<Ts...>>;
       using first_value_type = typename first_type::scalar_type;
+      using self_type = StackOp<Ts...>;
 
       static constexpr int RANK = first_type::Rank();
 
@@ -181,6 +182,12 @@ namespace matx
         }
       }
 
+      ~StackOp() = default;
+      StackOp(const StackOp &rhs) = default;
+      __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+        return set(*this, rhs); 
+      }       
+
       template<typename R> 
       __MATX_INLINE__ auto operator=(const R &rhs) { 
         if constexpr (is_matx_transform_op<R>()) {

diff --git a/include/matx/operators/transpose.h b/include/matx/operators/transpose.h
@@ -57,6 +57,7 @@ namespace detail {
       using matx_transform_op = bool;
       using matxoplvalue = bool;
       using transpose_xform_op = bool;
+      using self_type = TransposeMatrixOp<OpA>;
 
       __MATX_INLINE__ std::string str() const { return "transpose_matrix(" + get_type_str(a_) + ")"; }
       __MATX_INLINE__ TransposeMatrixOp(OpA a) : a_(a) {
@@ -121,6 +122,12 @@ namespace detail {
         return out_dims_[dim];
       }
 
+      ~TransposeMatrixOp() = default;
+      TransposeMatrixOp(const TransposeMatrixOp &rhs) = default;
+      __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+        return set(*this, rhs); 
+      }       
+
       template<typename R> 
       __MATX_INLINE__ auto operator=(const R &rhs) { 
         if constexpr (is_matx_transform_op<R>()) {

diff --git a/include/matx/operators/updownsample.h b/include/matx/operators/updownsample.h
@@ -54,6 +54,7 @@ namespace matx
         using matxop = bool;
         using matxoplvalue = bool;
         using scalar_type = typename T::scalar_type;
+        using self_type = UpsampleOp<T>;
 
         static __MATX_INLINE__ constexpr __MATX_HOST__ __MATX_DEVICE__ int32_t Rank()
         {
@@ -109,6 +110,12 @@ namespace matx
           }
         }
 
+        ~UpsampleOp() = default;
+        UpsampleOp(const UpsampleOp &rhs) = default;
+        __MATX_INLINE__ auto operator=(const self_type &rhs) { 
+          return set(*this, rhs); 
+        } 
+
         template<typename R> __MATX_INLINE__ auto operator=(const R &rhs) { return set(*this, rhs); }
     };
   }

diff --git a/test/00_operators/OperatorTests.cu b/test/00_operators/OperatorTests.cu
@@ -1589,6 +1589,44 @@ TYPED_TEST(OperatorTestsNumericAllExecs, RemapOp)
     }
   }
 
+  {
+    // Remap as both LHS and RHS
+    auto in = make_tensor<TestType>({4,4,4});
+    auto out = make_tensor<TestType>({4,4,4});
+    TestType c = GenerateData<TestType>();
+    for (int i = 0; i < in.Size(0); i++){
+      for (int j = 0; j < in.Size(1); j++){
+        for (int k = 0; k < in.Size(2); k++){
+          in(i,j,k) = c;
+        }
+      }
+    }
+
+    auto map1 = matx::make_tensor<int>({2});
+    auto map2 = matx::make_tensor<int>({2});
+    map1(0) = 1;
+    map1(1) = 2;
+    map2(0) = 0;
+    map2(1) = 1;
+
+    (out=0).run();
+    (matx::remap<2>(out, map2) = matx::remap<2>(in, map1)).run();
+    exec.sync();
+
+    for (int i = 0; i < in.Size(0); i++){
+      for (int j = 0; j < in.Size(1); j++){
+        for (int k = 0; k < in.Size(2); k++){
+          if (k > 1) {
+            EXPECT_TRUE(out(i,j,k) == (TestType)0);
+          }
+          else {
+            EXPECT_TRUE(out(i,j,k) == in(i,j,k));
+          }
+        }
+      }
+    } 
+  }
+
   MATX_EXIT_HANDLER();
 }