| Line | Branch | Exec | Source |
|---|---|---|---|
| 1 | #include "levonychev_i_radix_batcher_sort/tbb/include/ops_tbb.hpp" | ||
| 2 | |||
| 3 | #include <tbb/blocked_range.h> | ||
| 4 | #include <tbb/global_control.h> | ||
| 5 | #include <tbb/parallel_for.h> | ||
| 6 | #include <tbb/parallel_invoke.h> | ||
| 7 | |||
| 8 | #include <algorithm> | ||
| 9 | #include <cstddef> | ||
| 10 | #include <ranges> | ||
| 11 | #include <vector> | ||
| 12 | |||
| 13 | #include "levonychev_i_radix_batcher_sort/common/include/common.hpp" | ||
| 14 | #include "util/include/util.hpp" | ||
| 15 | |||
| 16 | namespace levonychev_i_radix_batcher_sort { | ||
| 17 | |||
| 18 |
1/2✓ Branch 1 taken 16 times.
✗ Branch 2 not taken.
|
16 | LevonychevIRadixBatcherSortTBB::LevonychevIRadixBatcherSortTBB(const InType &in) { |
| 19 | SetTypeOfTask(GetStaticTypeOfTask()); | ||
| 20 |
1/2✓ Branch 1 taken 16 times.
✗ Branch 2 not taken.
|
16 | GetInput() = in; |
| 21 | 16 | } | |
| 22 | |||
| 23 | 132 | void LevonychevIRadixBatcherSortTBB::CountingSort(InType &arr, size_t byte_index) { | |
| 24 | const size_t byte = 256; | ||
| 25 |
1/2✓ Branch 2 taken 132 times.
✗ Branch 3 not taken.
|
132 | std::vector<int> count(byte, 0); |
| 26 |
1/4✓ Branch 1 taken 132 times.
✗ Branch 2 not taken.
✗ Branch 3 not taken.
✗ Branch 4 not taken.
|
132 | OutType result(arr.size()); |
| 27 | |||
| 28 | bool is_last_byte = (byte_index == (sizeof(int) - 1ULL)); | ||
| 29 |
2/2✓ Branch 0 taken 384 times.
✓ Branch 1 taken 132 times.
|
516 | for (auto number : arr) { |
| 30 | 384 | int value_of_byte = (number >> (byte_index * 8ULL)) & 0xFF; | |
| 31 | |||
| 32 |
2/2✓ Branch 0 taken 96 times.
✓ Branch 1 taken 288 times.
|
384 | if (is_last_byte) { |
| 33 | 96 | value_of_byte ^= 0x80; | |
| 34 | } | ||
| 35 | |||
| 36 | 384 | ++count[value_of_byte]; | |
| 37 | } | ||
| 38 | |||
| 39 |
2/2✓ Branch 0 taken 33660 times.
✓ Branch 1 taken 132 times.
|
33792 | for (size_t i = 1ULL; i < byte; ++i) { |
| 40 | 33660 | count[i] += count[i - 1]; | |
| 41 | } | ||
| 42 | |||
| 43 |
2/2✓ Branch 0 taken 384 times.
✓ Branch 1 taken 132 times.
|
516 | for (int &val : std::ranges::reverse_view(arr)) { |
| 44 | 384 | int value_of_byte = (val >> (byte_index * 8ULL)) & 0xFF; | |
| 45 | |||
| 46 |
2/2✓ Branch 0 taken 96 times.
✓ Branch 1 taken 288 times.
|
384 | if (is_last_byte) { |
| 47 | 96 | value_of_byte ^= 0x80; | |
| 48 | } | ||
| 49 | |||
| 50 | 384 | result[--count[value_of_byte]] = val; | |
| 51 | } | ||
| 52 |
1/2✓ Branch 1 taken 132 times.
✗ Branch 2 not taken.
|
132 | arr = result; |
| 53 | 132 | } | |
| 54 | |||
| 55 | 16 | bool LevonychevIRadixBatcherSortTBB::ValidationImpl() { | |
| 56 | 16 | return !GetInput().empty(); | |
| 57 | } | ||
| 58 | |||
| 59 | 16 | bool LevonychevIRadixBatcherSortTBB::PreProcessingImpl() { | |
| 60 | 16 | return true; | |
| 61 | } | ||
| 62 | |||
| 63 | 75 | void LevonychevIRadixBatcherSortTBB::BatcherCompareRange(std::vector<int> &arr, int j, int k, int p2) { | |
| 64 | 75 | int range = std::min(k, static_cast<int>(arr.size()) - j - k); | |
| 65 |
2/2✓ Branch 0 taken 123 times.
✓ Branch 1 taken 75 times.
|
198 | for (int i = 0; i < range; ++i) { |
| 66 | 123 | int idx1 = j + i; | |
| 67 | 123 | int idx2 = j + i + k; | |
| 68 |
4/4✓ Branch 0 taken 117 times.
✓ Branch 1 taken 6 times.
✓ Branch 2 taken 75 times.
✓ Branch 3 taken 42 times.
|
123 | if ((idx1 & p2) == (idx2 & p2) && (arr[idx1] > arr[idx2])) { |
| 69 | std::swap(arr[idx1], arr[idx2]); | ||
| 70 | } | ||
| 71 | } | ||
| 72 | 75 | } | |
| 73 | |||
| 74 | 12 | void LevonychevIRadixBatcherSortTBB::BatcherMergeIterative(std::vector<int> &arr, int start_p) { | |
| 75 | 12 | int n = static_cast<int>(arr.size()); | |
| 76 |
2/2✓ Branch 0 taken 15 times.
✓ Branch 1 taken 12 times.
|
27 | for (int pv = start_p; pv < n; pv <<= 1) { |
| 77 | 15 | int p2 = pv << 1; | |
| 78 |
2/2✓ Branch 0 taken 39 times.
✓ Branch 1 taken 15 times.
|
54 | for (int k = pv; k > 0; k >>= 1) { |
| 79 | 39 | int num_iters = (n - k - (k % pv) + 2 * k - 1) / (2 * k); | |
| 80 | |||
| 81 | 39 | tbb::parallel_for(0, num_iters, [&](int i) { | |
| 82 | 75 | int j = (k % pv) + (i * (2 * k)); | |
| 83 | 75 | BatcherCompareRange(arr, j, k, p2); | |
| 84 | 75 | }); | |
| 85 | } | ||
| 86 | } | ||
| 87 | 12 | } | |
| 88 | |||
| 89 | 16 | bool LevonychevIRadixBatcherSortTBB::RunImpl() { | |
| 90 | 16 | GetOutput() = GetInput(); | |
| 91 | 16 | int n = static_cast<int>(GetOutput().size()); | |
| 92 |
2/2✓ Branch 0 taken 12 times.
✓ Branch 1 taken 4 times.
|
16 | if (n <= 1) { |
| 93 | return true; | ||
| 94 | } | ||
| 95 | 12 | int num_threads = ppc::util::GetNumThreads(); | |
| 96 | 12 | int grain = std::max(1, n / num_threads); | |
| 97 | |||
| 98 | 45 | tbb::parallel_for(tbb::blocked_range<int>(0, n, grain), [&](const tbb::blocked_range<int> &r) { | |
| 99 | int left = r.begin(); | ||
| 100 | int right = r.end(); | ||
| 101 | |||
| 102 | 33 | std::vector<int> local_block(GetOutput().begin() + left, GetOutput().begin() + right); | |
| 103 |
2/2✓ Branch 0 taken 132 times.
✓ Branch 1 taken 33 times.
|
165 | for (size_t i = 0; i < sizeof(int); ++i) { |
| 104 |
1/2✓ Branch 1 taken 132 times.
✗ Branch 2 not taken.
|
132 | CountingSort(local_block, i); |
| 105 | } | ||
| 106 |
1/2✓ Branch 0 taken 33 times.
✗ Branch 1 not taken.
|
33 | std::ranges::copy(local_block, GetOutput().begin() + left); |
| 107 | 33 | }); | |
| 108 | |||
| 109 | int block_size = grain; | ||
| 110 | int start_p = 1; | ||
| 111 |
2/2✓ Branch 0 taken 21 times.
✓ Branch 1 taken 12 times.
|
33 | while (start_p < block_size) { |
| 112 | 21 | start_p <<= 1; | |
| 113 | } | ||
| 114 | |||
| 115 | 12 | BatcherMergeIterative(GetOutput(), start_p); | |
| 116 | 12 | return true; | |
| 117 | } | ||
| 118 | |||
| 119 | 16 | bool LevonychevIRadixBatcherSortTBB::PostProcessingImpl() { | |
| 120 | 16 | return true; | |
| 121 | } | ||
| 122 | |||
| 123 | } // namespace levonychev_i_radix_batcher_sort | ||
| 124 |