GCC Code Coverage Report

Directory: src/
Coverage: low: ≥ 0% medium: ≥ 75.0% high: ≥ 90.0%
Coverage Exec / Excl / Total
Lines: 65.6% 63 / 0 / 96
Functions: 100.0% 22 / 0 / 22
Branches: 89.3% 25 / 4 / 32
tensor.hpp
Line Branch Exec Source
1 #pragma once
2
3 #include <array>
4
5 namespace wala {
6
7 template <typename T, int NDIMS> struct tensor_view {
8 static_assert(NDIMS >= 0, "NDIMS must be nonnegative");
9
10 protected:
11 std::array<int, NDIMS> shape;
12 std::array<int, NDIMS> strides;
13 T* data;
14
15 54 tensor_view(std::array<int, NDIMS> shape_, std::array<int, NDIMS> strides_, T* data_) : shape(shape_), strides(strides_), data(data_) {}
16
17 public:
18 ✗ tensor_view() : shape{0}, strides{0}, data(nullptr) {}
19
20 protected:
21 ✗ int flatten_index(std::array<int, NDIMS> idx) const {
22 ✗ int res = 0;
23 ✗ for (int i = 0; i < NDIMS; i++) { res += idx[i] * strides[i]; }
24 ✗ return res;
25 }
26 18 int flatten_index_checked(std::array<int, NDIMS> idx) const {
27 18 int res = 0;
28
2/2
✓ Branch 30 → 3 taken 36 times.
✓ Branch 30 → 31 taken 18 times.
 54 for (int i = 0; i < NDIMS; i++) {
29 36 assert(0 <= idx[i] && idx[i] < shape[i]);
30 36 res += idx[i] * strides[i];
31 }
32 18 return res;
33 }
34
35 public:
36 12 T& operator[] (std::array<int, NDIMS> idx) const {
37 #ifdef _GLIBCXX_DEBUG
38 12 return data[flatten_index_checked(idx)];
39 #else
40 return data[flatten_index(idx)];
41 #endif
42 }
43 6 T& at(std::array<int, NDIMS> idx) const {
44 6 return data[flatten_index_checked(idx)];
45 }
46
47 template <int D = NDIMS>
48 24 typename std::enable_if<(0 < D), tensor_view<T, NDIMS-1>>::type operator[] (int idx) const {
49
4/4
std::enable_if<(0)<(1), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 0> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 1>::operator[]<1>(int) const:
✓ Branch 15 → 16 taken 6 times.
std::enable_if<(0)<(2), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 1> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 2>::operator[]<2>(int) const:
✓ Branch 17 → 18 taken 6 times.
std::enable_if<(0)<(1), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 0> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 1>::operator[]<1>(int) const:
✓ Branch 15 → 16 taken 6 times.
std::enable_if<(0)<(2), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 1> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 2>::operator[]<2>(int) const:
✓ Branch 17 → 18 taken 6 times.
 84 std::array<int, NDIMS-1> nshape; std::copy(shape.begin()+1, shape.end(), nshape.begin());
50
4/4
std::enable_if<(0)<(1), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 0> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 1>::operator[]<1>(int) const:
✓ Branch 33 → 34 taken 6 times.
std::enable_if<(0)<(2), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 1> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, 2>::operator[]<2>(int) const:
✓ Branch 37 → 38 taken 6 times.
std::enable_if<(0)<(1), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 0> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 1>::operator[]<1>(int) const:
✓ Branch 33 → 34 taken 6 times.
std::enable_if<(0)<(2), wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 1> >::type wala::tensor_view<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, 2>::operator[]<2>(int) const:
✓ Branch 37 → 38 taken 6 times.
 84 std::array<int, NDIMS-1> nstrides; std::copy(strides.begin()+1, strides.end(), nstrides.begin());
51 48 T* ndata = data + (strides[0] * idx);
52 60 return tensor_view<T, NDIMS-1>(nshape, nstrides, ndata);
53 }
54 template <int D = NDIMS>
55 ✗ typename std::enable_if<(0 < D), tensor_view<T, NDIMS-1>>::type at(int idx) const {
56 ✗ assert(0 <= idx && idx < shape[0]);
57 ✗ return operator[](idx);
58 }
59
60 template <int D = NDIMS>
61 12 typename std::enable_if<(0 == D), T&>::type operator * () const {
62 12 return *data;
63 }
64
65 template <typename U, int D> friend struct tensor_view;
66 template <typename U, int D> friend struct tensor;
67 };
68
69 template <typename T, int NDIMS> struct tensor {
70 static_assert(NDIMS >= 0, "NDIMS must be nonnegative");
71
72 protected:
73 std::array<int, NDIMS> shape;
74 std::array<int, NDIMS> strides;
75 int len;
76 T* data;
77
78 public:
79 ✗ tensor() : shape{0}, strides{0}, len(0), data(nullptr) {}
80
81 1 explicit tensor(std::array<int, NDIMS> shape_, const T& t = T()) {
82 1 shape = shape_;
83 1 len = 1;
84
2/2
✓ Branch 29 → 8 taken 2 times.
✓ Branch 29 → 30 taken 1 time.
 3 for (int i = NDIMS-1; i >= 0; i--) {
85 2 strides[i] = len;
86 2 len *= shape[i];
87 }
88
3/4
✓ Branch 33 → 34 taken 1 time.
✗ Branch 33 → 35 not taken.
✓ Branch 44 → 40 taken 6 times.
✓ Branch 44 → 45 taken 1 time.
 7 data = new T[len];
89 1 std::fill(data, data + len, t);
90 1 }
91
92
3/4
✓ Branch 21 → 22 taken 2 times.
✗ Branch 21 → 23 not taken.
✓ Branch 32 → 28 taken 12 times.
✓ Branch 32 → 33 taken 2 times.
 14 tensor(const tensor& o) : shape(o.shape), strides(o.strides), len(o.len), data(new T[len]) {
93
2/2
✓ Branch 56 → 38 taken 12 times.
✓ Branch 56 → 57 taken 2 times.
 14 for (int i = 0; i < len; i++) {
94 24 data[i] = o.data[i];
95 }
96 2 }
97
98 ✗ tensor& operator=(tensor&& o) noexcept {
99 using std::swap;
100 ✗ swap(shape, o.shape);
101 ✗ swap(strides, o.strides);
102 ✗ swap(len, o.len);
103 ✗ swap(data, o.data);
104 ✗ return *this;
105 }
106 ✗ tensor(tensor&& o) : tensor() {
107 ✗ *this = std::move(o);
108 }
109 ✗ tensor& operator=(const tensor& o) {
110 ✗ return *this = tensor(o);
111 }
112
3/4
✓ Branch 5 → 6 taken 3 times.
✗ Branch 5 → 40 not taken.
✓ Branch 20 → 21 taken 18 times.
✓ Branch 20 → 29 taken 3 times.
 21 ~tensor() { delete[] data; }
113
114 using view_t = tensor_view<T, NDIMS>;
115 24 view_t view() {
116 24 return tensor_view<T, NDIMS>(shape, strides, data);
117 }
118 ✗ operator view_t() {
119 ✗ return view();
120 }
121
122 using const_view_t = tensor_view<const T, NDIMS>;
123 6 const_view_t view() const {
124 6 return tensor_view<const T, NDIMS>(shape, strides, data);
125 }
126 ✗ operator const_view_t() const {
127 ✗ return view();
128 }
129
130 12 T& operator[] (std::array<int, NDIMS> idx) { return view()[idx]; }
131 6 T& at(std::array<int, NDIMS> idx) { return view().at(idx); }
132 ✗ const T& operator[] (std::array<int, NDIMS> idx) const { return view()[idx]; }
133 ✗ const T& at(std::array<int, NDIMS> idx) const { return view().at(idx); }
134
135 template <int D = NDIMS>
136 6 typename std::enable_if<(0 < D), tensor_view<T, NDIMS-1>>::type operator[] (int idx) {
137
1/1
✓ Branch 6 → 7 taken 6 times.
 6 return view()[idx];
138 }
139 template <int D = NDIMS>
140 ✗ typename std::enable_if<(0 < D), tensor_view<T, NDIMS-1>>::type at(int idx) {
141 ✗ return view().at(idx);
142 }
143
144 template <int D = NDIMS>
145 6 typename std::enable_if<(0 < D), tensor_view<const T, NDIMS-1>>::type operator[] (int idx) const {
146
1/1
✓ Branch 6 → 7 taken 6 times.
 6 return view()[idx];
147 }
148 template <int D = NDIMS>
149 ✗ typename std::enable_if<(0 < D), tensor_view<const T, NDIMS-1>>::type at(int idx) const {
150 ✗ return view().at(idx);
151 }
152
153 template <int D = NDIMS>
154 ✗ typename std::enable_if<(0 == D), T&>::type operator * () {
155 ✗ return *view();
156 }
157 template <int D = NDIMS>
158 ✗ typename std::enable_if<(0 == D), const T&>::type operator * () const {
159 ✗ return *view();
160 }
161 };
162
163 } // namespace wala
164