There is a single 4D example in the documentation. In the following description the 4 dimensions are referenced as:
column
row
layer
image
Each image is composed of a row of blocks. Each block is a layer.
The buffer is a 4x4x4x4 data set. The data is transferred in a complex way. The Source data organization is as follows:
Utils/GetTiles.py data/Input_1D_256.txt 4D 4 4 4 4 0
0 1 2 3 16 17 18 19 32 33 34 35 48 49 50 51
4 5 6 7 20 21 22 23 36 37 38 39 52 53 54 55
8 9 10 11 24 25 26 27 40 41 42 43 56 57 58 59
12 13 14 15 28 29 30 31 44 45 46 47 60 61 62 63
64 65 66 67 80 81 82 83 96 97 98 99 112 113 114 115
68 69 70 71 84 85 86 87 100 101 102 103 116 117 118 119
72 73 74 75 88 89 90 91 104 105 106 107 120 121 122 123
76 77 78 79 92 93 94 95 108 109 110 111 124 125 126 127
128 129 130 131 144 145 146 147 160 161 162 163 176 177 178 179
132 133 134 135 148 149 150 151 164 165 166 167 180 181 182 183
136 137 138 139 152 153 154 155 168 169 170 171 184 185 186 187
140 141 142 143 156 157 158 159 172 173 174 175 188 189 190 191
192 193 194 195 208 209 210 211 224 225 226 227 240 241 242 243
196 197 198 199 212 213 214 215 228 229 230 231 244 245 246 247
200 201 202 203 216 217 218 219 232 233 234 235 248 249 250 251
204 205 206 207 220 221 222 223 236 237 238 239 252 253 254 255
The read access is declared as this:
adf::read_access(mtxin.out[0]) = adf::tiling({
.buffer_dimension = {4, 4, 4, 4},
.tiling_dimension = {4, 1, 1, 4},
.offset = {0, 0, 0, 0},
.tile_traversal = {
{.dimension = 2, .stride = 1, .wrap = 4},
{.dimension = 1, .stride = 1, .wrap = 4}}});
Considering the tiling dimension we can see that a tile is composed of 16 elements which are selected as the 4 first columns (first row) of the 4 first layers of the image:
0 1 2 3
64 65 66 67
128 129 130 131
192 193 194 195
The first traversal parameter is on dimension 2, the layer dimension. So the second extracted tile will be the 4 first rows of the second layer of each image:
16 17 18 19
80 81 82 83
144 145 146 147
208 209 210 211
The second traversal parameter is on dimension 1, the row dimension. The fifth extracted tile (there is a wrap of 4 on the first traversal parameter) will be the second rows of the first 4 layers of the image:
4 5 6 7
68 69 70 71
132 133 134 135
196 197 198 199
And so on…
Utils/GetTiles.py doc_x86simulator_output/data/Output_40.txt 4D 4 4 4 4 0
Tile: 0
0 1 2 3 16 17 18 19 32 33 34 35 48 49 50 51
64 65 66 67 80 81 82 83 96 97 98 99 112 113 114 115
128 129 130 131 144 145 146 147 160 161 162 163 176 177 178 179
192 193 194 195 208 209 210 211 224 225 226 227 240 241 242 243
4 5 6 7 20 21 22 23 36 37 38 39 52 53 54 55
68 69 70 71 84 85 86 87 100 101 102 103 116 117 118 119
132 133 134 135 148 149 150 151 164 165 166 167 180 181 182 183
196 197 198 199 212 213 214 215 228 229 230 231 244 245 246 247
8 9 10 11 24 25 26 27 40 41 42 43 56 57 58 59
72 73 74 75 88 89 90 91 104 105 106 107 120 121 122 123
136 137 138 139 152 153 154 155 168 169 170 171 184 185 186 187
200 201 202 203 216 217 218 219 232 233 234 235 248 249 250 251
12 13 14 15 28 29 30 31 44 45 46 47 60 61 62 63
76 77 78 79 92 93 94 95 108 109 110 111 124 125 126 127
140 141 142 143 156 157 158 159 172 173 174 175 188 189 190 191
204 205 206 207 220 221 222 223 236 237 238 239 252 253 254 255