be/src/exec/operator/data_queue.cpp
Line | Count | Source |
1 | | // Licensed to the Apache Software Foundation (ASF) under one |
2 | | // or more contributor license agreements. See the NOTICE file |
3 | | // distributed with this work for additional information |
4 | | // regarding copyright ownership. The ASF licenses this file |
5 | | // to you under the Apache License, Version 2.0 (the |
6 | | // "License"); you may not use this file except in compliance |
7 | | // with the License. You may obtain a copy of the License at |
8 | | // |
9 | | // http://www.apache.org/licenses/LICENSE-2.0 |
10 | | // |
11 | | // Unless required by applicable law or agreed to in writing, |
12 | | // software distributed under the License is distributed on an |
13 | | // "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY |
14 | | // KIND, either express or implied. See the License for the |
15 | | // specific language governing permissions and limitations |
16 | | // under the License. |
17 | | |
18 | | #include "exec/operator/data_queue.h" |
19 | | |
20 | | #include <glog/logging.h> |
21 | | |
22 | | #include <algorithm> |
23 | | #include <mutex> |
24 | | #include <utility> |
25 | | |
26 | | #include "core/block/block.h" |
27 | | #include "exec/pipeline/dependency.h" |
28 | | |
29 | | namespace doris { |
30 | | DataQueue::DataQueue(int child_count) |
31 | 4.19k | : _queue_blocks_lock(child_count), |
32 | 4.19k | _queue_blocks(child_count), |
33 | 4.19k | _free_blocks_lock(child_count), |
34 | 4.19k | _free_blocks(child_count), |
35 | 4.19k | _child_count(child_count), |
36 | 4.19k | _is_finished(child_count), |
37 | 4.19k | _is_canceled(child_count), |
38 | 4.19k | _cur_bytes_in_queue(child_count), |
39 | 4.19k | _cur_blocks_nums_in_queue(child_count), |
40 | 4.19k | _flag_queue_idx(0) { |
41 | 12.7k | for (int i = 0; i < child_count; ++i) { |
42 | 8.59k | _queue_blocks_lock[i].reset(new std::mutex()); |
43 | 8.59k | _free_blocks_lock[i].reset(new std::mutex()); |
44 | 8.59k | _is_finished[i] = false; |
45 | 8.59k | _is_canceled[i] = false; |
46 | 8.59k | _cur_bytes_in_queue[i] = 0; |
47 | 8.59k | _cur_blocks_nums_in_queue[i] = 0; |
48 | 8.59k | } |
49 | 4.19k | _un_finished_counter = child_count; |
50 | 4.19k | _sink_dependencies.resize(child_count, nullptr); |
51 | 4.19k | } |
52 | | |
53 | 8.66k | std::unique_ptr<Block> DataQueue::get_free_block(int child_idx) { |
54 | 8.66k | { |
55 | 8.66k | INJECT_MOCK_SLEEP(std::lock_guard<std::mutex> l(*_free_blocks_lock[child_idx])); |
56 | 8.66k | if (!_free_blocks[child_idx].empty()) { |
57 | 3 | auto block = std::move(_free_blocks[child_idx].front()); |
58 | 3 | _free_blocks[child_idx].pop_front(); |
59 | 3 | return block; |
60 | 3 | } |
61 | 8.66k | } |
62 | | |
63 | 8.66k | return Block::create_unique(); |
64 | 8.66k | } |
65 | | |
66 | 8.65k | void DataQueue::push_free_block(std::unique_ptr<Block> block, int child_idx) { |
67 | 8.65k | DCHECK(block->rows() == 0); |
68 | | |
69 | 8.65k | if (!_is_low_memory_mode) { |
70 | 8.65k | INJECT_MOCK_SLEEP(std::lock_guard<std::mutex> l(*_free_blocks_lock[child_idx])); |
71 | 8.65k | _free_blocks[child_idx].emplace_back(std::move(block)); |
72 | 8.65k | } |
73 | 8.65k | } |
74 | | |
75 | 4.17k | void DataQueue::clear_free_blocks() { |
76 | 12.8k | for (size_t child_idx = 0; child_idx < _free_blocks.size(); ++child_idx) { |
77 | 8.63k | std::lock_guard<std::mutex> l(*_free_blocks_lock[child_idx]); |
78 | 8.63k | std::deque<std::unique_ptr<Block>> tmp_queue; |
79 | 8.63k | _free_blocks[child_idx].swap(tmp_queue); |
80 | 8.63k | } |
81 | 4.17k | } |
82 | | |
83 | 4.17k | void DataQueue::terminate() { |
84 | 12.8k | for (int i = 0; i < _queue_blocks.size(); i++) { |
85 | 8.64k | set_finish(i); |
86 | 8.64k | INJECT_MOCK_SLEEP(std::lock_guard<std::mutex> l(*_queue_blocks_lock[i])); |
87 | 8.64k | if (_cur_blocks_nums_in_queue[i] > 0) { |
88 | 1 | _queue_blocks[i].clear(); |
89 | 1 | _cur_bytes_in_queue[i] = 0; |
90 | 1 | _cur_blocks_nums_in_queue[i] = 0; |
91 | 1 | _sink_dependencies[i]->set_always_ready(); |
92 | 1 | } |
93 | 8.64k | } |
94 | 4.17k | clear_free_blocks(); |
95 | 4.17k | } |
96 | | |
97 | | //check which queue have data, and save the idx in _flag_queue_idx, |
98 | | //so next loop, will check the record idx + 1 first |
99 | | //maybe it's useful with many queue, others maybe always 0 |
100 | 53.5k | bool DataQueue::remaining_has_data() { |
101 | 53.5k | int count = _child_count; |
102 | 147k | while (--count >= 0) { |
103 | 103k | _flag_queue_idx++; |
104 | 103k | if (_flag_queue_idx == _child_count) { |
105 | 49.1k | _flag_queue_idx = 0; |
106 | 49.1k | } |
107 | 103k | if (_cur_blocks_nums_in_queue[_flag_queue_idx] > 0) { |
108 | 8.81k | return true; |
109 | 8.81k | } |
110 | 103k | } |
111 | 44.7k | return false; |
112 | 53.5k | } |
113 | | |
114 | | //the _flag_queue_idx indicate which queue has data, and in check can_read |
115 | | //will be set idx in remaining_has_data function |
116 | 45.1k | Status DataQueue::get_block_from_queue(std::unique_ptr<Block>* output_block, int* child_idx) { |
117 | 45.1k | if (_is_canceled[_flag_queue_idx]) { |
118 | 0 | return Status::InternalError("Current queue of idx {} have beed canceled: ", |
119 | 0 | _flag_queue_idx); |
120 | 0 | } |
121 | | |
122 | 45.1k | { |
123 | 45.1k | INJECT_MOCK_SLEEP(std::lock_guard<std::mutex> l(*_queue_blocks_lock[_flag_queue_idx])); |
124 | 45.1k | if (_cur_blocks_nums_in_queue[_flag_queue_idx] > 0) { |
125 | 8.82k | *output_block = std::move(_queue_blocks[_flag_queue_idx].front()); |
126 | 8.82k | _queue_blocks[_flag_queue_idx].pop_front(); |
127 | 8.82k | if (child_idx) { |
128 | 8.82k | *child_idx = _flag_queue_idx; |
129 | 8.82k | } |
130 | 8.82k | _cur_bytes_in_queue[_flag_queue_idx] -= (*output_block)->allocated_bytes(); |
131 | 8.82k | _cur_blocks_nums_in_queue[_flag_queue_idx] -= 1; |
132 | 8.82k | if (_cur_blocks_nums_in_queue[_flag_queue_idx] == 0) { |
133 | 8.75k | _sink_dependencies[_flag_queue_idx]->set_ready(); |
134 | 8.75k | } |
135 | 8.82k | auto old_value = _cur_blocks_total_nums.fetch_sub(1); |
136 | 8.82k | if (old_value == 1 && _source_dependency) { |
137 | 8.34k | set_source_block(); |
138 | 8.34k | } |
139 | 8.82k | } |
140 | 45.1k | } |
141 | 45.1k | return Status::OK(); |
142 | 45.1k | } |
143 | | |
144 | 8.80k | Status DataQueue::push_block(std::unique_ptr<Block> block, int child_idx) { |
145 | 8.80k | if (!block) { |
146 | 0 | return Status::OK(); |
147 | 0 | } |
148 | 8.80k | { |
149 | 8.80k | INJECT_MOCK_SLEEP(std::lock_guard<std::mutex> l(*_queue_blocks_lock[child_idx])); |
150 | 8.80k | if (_is_finished[child_idx]) { |
151 | 0 | return Status::EndOfFile("Already finish"); |
152 | 0 | } |
153 | 8.80k | _cur_bytes_in_queue[child_idx] += block->allocated_bytes(); |
154 | 8.80k | _queue_blocks[child_idx].emplace_back(std::move(block)); |
155 | 8.80k | _cur_blocks_nums_in_queue[child_idx] += 1; |
156 | | |
157 | 8.80k | if (_cur_blocks_nums_in_queue[child_idx] > _max_blocks_in_sub_queue) { |
158 | 77 | _sink_dependencies[child_idx]->block(); |
159 | 77 | } |
160 | 8.80k | _cur_blocks_total_nums++; |
161 | | |
162 | 8.80k | set_source_ready(); |
163 | 8.80k | } |
164 | 0 | return Status::OK(); |
165 | 8.80k | } |
166 | | |
167 | 17.3k | void DataQueue::set_finish(int child_idx) { |
168 | 17.3k | INJECT_MOCK_SLEEP(std::lock_guard<std::mutex> l(*_queue_blocks_lock[child_idx])); |
169 | 17.3k | if (_is_finished[child_idx]) { |
170 | 8.65k | return; |
171 | 8.65k | } |
172 | 8.66k | _is_finished[child_idx] = true; |
173 | 8.66k | if (_un_finished_counter.fetch_sub(1) == 1) { |
174 | 4.20k | _is_all_finished = true; |
175 | 4.20k | } |
176 | 8.66k | set_source_ready(); |
177 | 8.66k | } |
178 | | |
179 | 0 | void DataQueue::set_canceled(int child_idx) { |
180 | 0 | INJECT_MOCK_SLEEP(std::lock_guard<std::mutex> l(*_queue_blocks_lock[child_idx])); |
181 | 0 | DCHECK(!_is_finished[child_idx]); |
182 | 0 | _is_canceled[child_idx] = true; |
183 | 0 | _is_finished[child_idx] = true; |
184 | 0 | if (_un_finished_counter.fetch_sub(1) == 1) { |
185 | 0 | _is_all_finished = true; |
186 | 0 | } |
187 | 0 | set_source_ready(); |
188 | 0 | } |
189 | | |
190 | 3 | bool DataQueue::is_finish(int child_idx) { |
191 | 3 | return _is_finished[child_idx]; |
192 | 3 | } |
193 | | |
194 | 53.0k | bool DataQueue::is_all_finish() { |
195 | 53.0k | return _is_all_finished; |
196 | 53.0k | } |
197 | | |
198 | 17.5k | void DataQueue::set_source_ready() { |
199 | 17.5k | if (_source_dependency) { |
200 | 17.5k | std::unique_lock lc(_source_lock); |
201 | 17.5k | _source_dependency->set_ready(); |
202 | 17.5k | } |
203 | 17.5k | } |
204 | | |
205 | 8.34k | void DataQueue::set_source_block() { |
206 | 8.34k | if (_cur_blocks_total_nums == 0 && !is_all_finish()) { |
207 | 4.17k | std::unique_lock lc(_source_lock); |
208 | | // Performing the judgment twice, attempting to avoid blocking the source as much as possible. |
209 | 4.18k | if (_cur_blocks_total_nums == 0 && !is_all_finish()) { |
210 | 4.17k | _source_dependency->block(); |
211 | 4.17k | } |
212 | 4.17k | } |
213 | 8.34k | } |
214 | | |
215 | | } // namespace doris |