be/src/exec/operator/data_queue.h
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 | | #pragma once |
18 | | |
19 | | #include <atomic> |
20 | | #include <cstdint> |
21 | | #include <deque> |
22 | | #include <memory> |
23 | | #include <string> |
24 | | #include <vector> |
25 | | |
26 | | #include "common/status.h" |
27 | | #include "common/thread_safety_annotations.h" |
28 | | #include "core/block/block.h" |
29 | | |
30 | | namespace doris { |
31 | | |
32 | | class Dependency; |
33 | | |
34 | | struct DataQueueBlock { |
35 | | std::unique_ptr<Block> block; |
36 | | bool eos = false; |
37 | | |
38 | | private: |
39 | | int child_idx = 0; |
40 | | friend class DataQueue; |
41 | | }; |
42 | | |
43 | | // Per child sub-queue. Groups all parallel state so that the lock/field |
44 | | // relationship is explicit and can be checked by clang -Wthread-safety. |
45 | | struct SubQueue { |
46 | | // Protects the `blocks` deque and serializes high-level state |
47 | | // transitions (push/pop/finish/cancel) on this sub-queue. |
48 | | AnnotatedMutex queue_lock; |
49 | | std::deque<std::unique_ptr<Block>> blocks GUARDED_BY(queue_lock); |
50 | | |
51 | | // The following fields are only accessed while holding queue_lock. |
52 | | int64_t bytes_in_queue GUARDED_BY(queue_lock) = 0; |
53 | | bool is_finished GUARDED_BY(queue_lock) = false; |
54 | | |
55 | | // Protects the `free_blocks` deque only. |
56 | | AnnotatedMutex free_lock; |
57 | | std::deque<std::unique_ptr<Block>> free_blocks GUARDED_BY(free_lock); |
58 | | |
59 | | // blocks_in_queue is readable from lock-free fast paths (get_block_from_queue), |
60 | | // so it remains atomic and is intentionally not GUARDED_BY. |
61 | | std::atomic_uint32_t blocks_in_queue {0}; |
62 | | |
63 | | // Maximum number of blocks allowed in this sub-queue before the sink is blocked. |
64 | | // Updated by DataQueue::set_max_blocks_in_sub_queue / set_low_memory_mode. |
65 | | std::atomic_int64_t max_blocks_in_queue {1}; |
66 | | |
67 | | // Set once during init via set_sink_dependency, then read-only. |
68 | | Dependency* sink_dependency = nullptr; |
69 | | |
70 | | // Pop a block under queue_lock. |
71 | | // Notifies sink_dependency->set_ready() (outside the lock) if the queue becomes empty. |
72 | | // output_block is null if the queue was empty. |
73 | | void try_pop(std::unique_ptr<Block>* output_block); |
74 | | |
75 | | // Push a block under queue_lock and atomically increment total_counter. |
76 | | // Returns false (without incrementing) if already finished. |
77 | | // Calls sink_dependency->block() (outside the lock) if the queue exceeds max_blocks_in_queue. |
78 | | bool try_push(std::unique_ptr<Block> block, std::atomic_uint32_t& total_counter); |
79 | | |
80 | | // Mark this sub-queue finished. Returns false if already finished (idempotent). |
81 | | // Decrements unfinished_counter and may set all_finished within queue_lock. |
82 | | bool mark_finished(std::atomic_uint32_t& unfinished_counter, std::atomic_bool& all_finished); |
83 | | |
84 | | // Clear all pending blocks under queue_lock. |
85 | | // Calls sink_dependency->set_always_ready() (outside the lock) if any blocks were cleared. |
86 | | void clear_blocks(); |
87 | | }; |
88 | | |
89 | | class DataQueue { |
90 | | public: |
91 | | //always one is enough, but in union node it's has more children |
92 | | DataQueue(int child_count = 1); |
93 | 17 | ~DataQueue() = default; |
94 | | |
95 | | Result<DataQueueBlock> get_block_from_queue(); |
96 | | Status push_block(std::unique_ptr<Block> block, int child_idx, bool eos); |
97 | | |
98 | | std::unique_ptr<Block> get_free_block(int child_idx = 0); |
99 | | void push_free_block(DataQueueBlock&& queue_block); |
100 | | |
101 | | std::string debug_string() const; |
102 | | |
103 | | void set_source_dependency(std::shared_ptr<Dependency> source_dependency) |
104 | | NO_THREAD_SAFETY_ANALYSIS; |
105 | | void set_sink_dependency(Dependency* sink_dependency, int child_idx); |
106 | | void set_max_blocks_in_sub_queue(int64_t max_blocks); |
107 | | void set_low_memory_mode(); |
108 | | |
109 | | void terminate(); |
110 | | |
111 | | private: |
112 | | bool is_all_finish() const; |
113 | | bool has_more_data() const; |
114 | | void clear_free_blocks(); |
115 | | void mark_finish(int child_idx); |
116 | | void set_source_ready(); |
117 | | void set_source_always_ready(); |
118 | | void set_source_block(); |
119 | | |
120 | | std::vector<std::unique_ptr<SubQueue>> _sub_queues; |
121 | | |
122 | | //how many deque will be init, always will be one |
123 | | int _child_count = 0; |
124 | | std::atomic_uint32_t _un_finished_counter = 0; |
125 | | std::atomic_bool _is_all_finished = false; |
126 | | std::atomic_uint32_t _cur_blocks_total_nums = 0; |
127 | | |
128 | | //this will be indicate which queue has data, it's useful when have many queues |
129 | | int _flag_queue_idx = 0; |
130 | | // only used by streaming agg source operator |
131 | | |
132 | | std::atomic_bool _is_low_memory_mode = false; |
133 | | |
134 | | // _source_dependency is written once during initialization (set_source_dependency) |
135 | | // and read/used only while holding _source_lock thereafter. |
136 | | std::shared_ptr<Dependency> _source_dependency GUARDED_BY(_source_lock) = nullptr; |
137 | | AnnotatedMutex _source_lock; |
138 | | }; |
139 | | |
140 | | } // namespace doris |