Coverage Report

Created: 2026-06-01 18:52

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
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 <utility>
24
25
#include "common/thread_safety_annotations.h"
26
#include "core/block/block.h"
27
#include "exec/pipeline/dependency.h"
28
29
namespace doris {
30
31
1.61k
void SubQueue::try_pop(std::unique_ptr<Block>* output_block) {
32
1.61k
    LockGuard l(queue_lock);
33
1.61k
    if (!blocks.empty()) {
34
409
        *output_block = std::move(blocks.front());
35
409
        blocks.pop_front();
36
409
        bytes_in_queue -= (*output_block)->allocated_bytes();
37
409
        blocks_in_queue -= 1;
38
409
        if (blocks.empty()) {
39
345
            sink_dependency->set_ready();
40
345
        }
41
409
    }
42
1.61k
}
43
44
425
bool SubQueue::try_push(std::unique_ptr<Block> block, std::atomic_uint32_t& total_counter) {
45
425
    LockGuard l(queue_lock);
46
425
    if (is_finished) {
47
1
        return false;
48
1
    }
49
424
    total_counter++;
50
424
    bytes_in_queue += block->allocated_bytes();
51
424
    blocks.emplace_back(std::move(block));
52
424
    blocks_in_queue += 1;
53
424
    if (static_cast<int64_t>(blocks.size()) > max_blocks_in_queue.load()) {
54
72
        sink_dependency->block();
55
72
    }
56
424
    return true;
57
425
}
58
59
bool SubQueue::mark_finished(std::atomic_uint32_t& unfinished_counter,
60
502
                             std::atomic_bool& all_finished) {
61
502
    LockGuard l(queue_lock);
62
502
    if (is_finished) {
63
242
        return false;
64
242
    }
65
260
    is_finished = true;
66
260
    if (unfinished_counter.fetch_sub(1) == 1) {
67
128
        all_finished = true;
68
128
    }
69
260
    return true;
70
502
}
71
72
245
void SubQueue::clear_blocks() {
73
245
    bool need_set_always_ready = false;
74
245
    {
75
245
        LockGuard l(queue_lock);
76
245
        if (!blocks.empty()) {
77
3
            blocks.clear();
78
3
            bytes_in_queue = 0;
79
3
            blocks_in_queue = 0;
80
3
            need_set_always_ready = true;
81
3
        }
82
245
    }
83
    // Notify outside of queue_lock to keep lock ordering simple.
84
245
    if (need_set_always_ready) {
85
3
        sink_dependency->set_always_ready();
86
3
    }
87
245
}
88
89
133
DataQueue::DataQueue(int child_count) : _sub_queues(child_count), _child_count(child_count) {
90
279
    for (auto& sub : _sub_queues) {
91
279
        sub = std::make_unique<SubQueue>();
92
279
    }
93
133
    _un_finished_counter = child_count;
94
133
}
95
96
3
bool DataQueue::has_more_data() const {
97
3
    return _cur_blocks_total_nums.load() > 0;
98
3
}
99
100
void DataQueue::set_source_dependency(std::shared_ptr<Dependency> source_dependency)
101
132
        NO_THREAD_SAFETY_ANALYSIS {
102
132
    _source_dependency = std::move(source_dependency);
103
132
}
104
105
276
void DataQueue::set_sink_dependency(Dependency* sink_dependency, int child_idx) {
106
276
    _sub_queues[child_idx]->sink_dependency = sink_dependency;
107
276
}
108
109
247
void DataQueue::set_max_blocks_in_sub_queue(int64_t max_blocks) {
110
517
    for (auto& sub : _sub_queues) {
111
517
        sub->max_blocks_in_queue = max_blocks;
112
517
    }
113
247
}
114
115
1
void DataQueue::set_low_memory_mode() {
116
1
    _is_low_memory_mode = true;
117
3
    for (auto& sub : _sub_queues) {
118
3
        sub->max_blocks_in_queue = 1;
119
3
    }
120
1
    clear_free_blocks();
121
1
}
122
123
254
std::unique_ptr<Block> DataQueue::get_free_block(int child_idx) {
124
254
    auto& sub = *_sub_queues[child_idx];
125
254
    {
126
254
        LockGuard l(sub.free_lock);
127
254
        if (!sub.free_blocks.empty()) {
128
2
            auto block = std::move(sub.free_blocks.front());
129
2
            sub.free_blocks.pop_front();
130
2
            return block;
131
2
        }
132
254
    }
133
134
252
    return Block::create_unique();
135
254
}
136
137
253
void DataQueue::push_free_block(std::unique_ptr<Block> block, int child_idx) {
138
253
    DCHECK(block->rows() == 0);
139
140
253
    if (!_is_low_memory_mode) {
141
252
        auto& sub = *_sub_queues[child_idx];
142
252
        LockGuard l(sub.free_lock);
143
252
        sub.free_blocks.emplace_back(std::move(block));
144
252
    }
145
253
}
146
147
120
void DataQueue::clear_free_blocks() {
148
246
    for (auto& sub : _sub_queues) {
149
246
        LockGuard l(sub->free_lock);
150
246
        std::deque<std::unique_ptr<Block>> tmp_queue;
151
246
        sub->free_blocks.swap(tmp_queue);
152
246
    }
153
120
}
154
155
119
void DataQueue::terminate() {
156
362
    for (int i = 0; i < _child_count; ++i) {
157
243
        set_finish(i);
158
243
        _sub_queues[i]->clear_blocks();
159
243
    }
160
119
    clear_free_blocks();
161
119
}
162
163
//check which queue have data, and save the idx in _flag_queue_idx,
164
//so next loop, will check the record idx + 1 first
165
//maybe it's useful with many queue, others maybe always 0
166
1.84k
bool DataQueue::remaining_has_data() {
167
1.84k
    int count = _child_count;
168
5.94k
    while (--count >= 0) {
169
4.50k
        _flag_queue_idx++;
170
4.50k
        if (_flag_queue_idx == _child_count) {
171
1.63k
            _flag_queue_idx = 0;
172
1.63k
        }
173
4.50k
        if (_sub_queues[_flag_queue_idx]->blocks_in_queue.load() > 0) {
174
398
            return true;
175
398
        }
176
4.50k
    }
177
1.44k
    return false;
178
1.84k
}
179
180
//the _flag_queue_idx indicate which queue has data, and in check can_read
181
//will be set idx in remaining_has_data function
182
1.60k
Status DataQueue::get_block_from_queue(std::unique_ptr<Block>* output_block, int* child_idx) {
183
1.60k
    const int idx = _flag_queue_idx;
184
1.60k
    auto& sub = *_sub_queues[idx];
185
186
1.60k
    sub.try_pop(output_block);
187
1.60k
    if (*output_block) {
188
404
        if (child_idx) {
189
404
            *child_idx = idx;
190
404
        }
191
404
        auto old_total = _cur_blocks_total_nums.fetch_sub(1);
192
404
        if (old_total == 1) {
193
252
            set_source_block();
194
252
        }
195
404
    }
196
1.60k
    return Status::OK();
197
1.60k
}
198
199
414
Status DataQueue::push_block(std::unique_ptr<Block> block, int child_idx) {
200
414
    if (!block) {
201
0
        return Status::OK();
202
0
    }
203
414
    auto& sub = *_sub_queues[child_idx];
204
    // total_counter is incremented inside try_push under queue_lock, only when the
205
    // block is actually enqueued. This ensures get_block_from_queue() always observes
206
    // _cur_blocks_total_nums >= 1 when it successfully pops a block, with no risk of
207
    // underflow or the need for a rollback on failure.
208
414
    if (!sub.try_push(std::move(block), _cur_blocks_total_nums)) {
209
0
        return Status::EndOfFile("SubQueue already finished");
210
0
    }
211
414
    set_source_ready();
212
414
    return Status::OK();
213
414
}
214
215
498
void DataQueue::set_finish(int child_idx) {
216
498
    auto& sub = *_sub_queues[child_idx];
217
498
    if (!sub.mark_finished(_un_finished_counter, _is_all_finished)) {
218
241
        return;
219
241
    }
220
257
    set_source_ready();
221
257
}
222
223
1.58k
bool DataQueue::is_all_finish() {
224
1.58k
    return _is_all_finished;
225
1.58k
}
226
227
672
void DataQueue::set_source_ready() {
228
672
    LockGuard lc(_source_lock);
229
672
    if (_source_dependency) {
230
672
        _source_dependency->set_ready();
231
672
    }
232
672
}
233
234
252
void DataQueue::set_source_block() {
235
    // Re-check under _source_lock to avoid blocking the source when a concurrent push
236
    // has already added new blocks (or all children have finished) since we observed
237
    // the counter drop to zero.
238
252
    LockGuard lc(_source_lock);
239
252
    if (_source_dependency && _cur_blocks_total_nums == 0 && !is_all_finish()) {
240
126
        _source_dependency->block();
241
126
    }
242
252
}
243
244
} // namespace doris