Coverage Report

Created: 2026-06-09 07:29

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
be/src/exec/pipeline/pipeline.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/pipeline/pipeline.h"
19
20
#include <memory>
21
#include <string>
22
#include <utility>
23
24
#include "exec/exchange/local_exchange_source_operator.h"
25
#include "exec/operator/operator.h"
26
#include "exec/pipeline/pipeline_fragment_context.h"
27
#include "exec/pipeline/pipeline_task.h"
28
29
namespace doris {
30
31
738k
void Pipeline::_init_profile() {
32
738k
    auto s = fmt::format("Pipeline (pipeline id={})", _pipeline_id);
33
738k
    _pipeline_profile = std::make_unique<RuntimeProfile>(std::move(s));
34
738k
}
35
36
bool Pipeline::need_to_local_exchange(const DataDistribution target_data_distribution,
37
132k
                                      const int idx) const {
38
132k
    if (!target_data_distribution.need_local_exchange()) {
39
2
        return false;
40
2
    }
41
    // If serial operator exists after `idx`-th operator, we should not improve parallelism.
42
132k
    if (std::any_of(_operators.begin() + idx, _operators.end(),
43
132k
                    [&](OperatorPtr op) -> bool { return op->is_serial_operator(); })) {
44
122
        return false;
45
122
    }
46
    // If all operators are serial and sink is not serial, we should improve parallelism for sink.
47
132k
    if (std::all_of(_operators.begin(), _operators.end(),
48
173k
                    [&](OperatorPtr op) -> bool { return op->is_serial_operator(); })) {
49
29.2k
        if (!_sink->is_serial_operator()) {
50
24.8k
            return true;
51
24.8k
        }
52
103k
    } else if (std::any_of(_operators.begin(), _operators.end(),
53
150k
                           [&](OperatorPtr op) -> bool { return op->is_serial_operator(); })) {
54
        // If non-serial operators exist, we should improve parallelism for those.
55
39.5k
        return true;
56
39.5k
    }
57
68.1k
    if (auto local_exchange_source =
58
68.1k
                std::dynamic_pointer_cast<LocalExchangeSourceOperatorX>(_operators.front());
59
68.1k
        local_exchange_source && is_hash_exchange(target_data_distribution.distribution_type)) {
60
511
        const auto source_exchange_type = local_exchange_source->exchange_type();
61
511
        if (source_exchange_type != ExchangeType::NOOP && !is_hash_exchange(source_exchange_type)) {
62
399
            return true;
63
399
        }
64
511
    }
65
67.7k
    if (target_data_distribution.distribution_type != ExchangeType::BUCKET_HASH_SHUFFLE &&
66
67.7k
        target_data_distribution.distribution_type != ExchangeType::HASH_SHUFFLE) {
67
        // Always do local exchange if non-hash-partition exchanger is required.
68
        // For example, `PASSTHROUGH` exchanger is always required to distribute data evenly.
69
22.3k
        return true;
70
45.4k
    } else if (_operators.front()->is_serial_operator()) {
71
0
        DCHECK(std::all_of(_operators.begin(), _operators.end(),
72
0
                           [&](OperatorPtr op) -> bool { return op->is_serial_operator(); }) &&
73
0
               _sink->is_serial_operator())
74
0
                << debug_string();
75
        // All operators and sink are serial in this path.
76
0
        return false;
77
45.4k
    } else {
78
45.4k
        return _data_distribution.distribution_type != target_data_distribution.distribution_type &&
79
45.4k
               !(is_hash_exchange(_data_distribution.distribution_type) &&
80
1.93k
                 is_hash_exchange(target_data_distribution.distribution_type));
81
45.4k
    }
82
67.7k
}
83
84
740k
Status Pipeline::add_operator(OperatorPtr& op, const int parallelism) {
85
740k
    if (parallelism > 0 && op->is_serial_operator()) {
86
81.6k
        set_num_tasks(parallelism);
87
81.6k
    }
88
740k
    op->set_parallel_tasks(num_tasks());
89
740k
    _operators.emplace_back(op);
90
740k
    if (op->is_source()) {
91
564k
        std::reverse(_operators.begin(), _operators.end());
92
564k
    }
93
740k
    return Status::OK();
94
740k
}
95
96
595k
Status Pipeline::prepare(RuntimeState* state) {
97
595k
    RETURN_IF_ERROR(_operators.back()->prepare(state));
98
594k
    RETURN_IF_ERROR(_sink->prepare(state));
99
594k
    _name.append(std::to_string(id()));
100
594k
    _name.push_back('-');
101
699k
    for (auto& op : _operators) {
102
699k
        _name.append(std::to_string(op->node_id()));
103
699k
        _name.append(op->get_name());
104
699k
    }
105
594k
    _name.push_back('-');
106
594k
    _name.append(std::to_string(_sink->node_id()));
107
594k
    _name.append(_sink->get_name());
108
594k
    return Status::OK();
109
594k
}
110
111
738k
Status Pipeline::set_sink(DataSinkOperatorPtr& sink) {
112
738k
    if (_sink) {
113
0
        return Status::InternalError("set sink twice");
114
0
    }
115
738k
    if (!sink->is_sink()) {
116
0
        return Status::InternalError("should set a sink operator but {}", typeid(sink).name());
117
0
    }
118
738k
    _sink = sink;
119
738k
    return Status::OK();
120
738k
}
121
122
308k
void Pipeline::make_all_runnable(PipelineId wake_by) {
123
308k
    DBUG_EXECUTE_IF("Pipeline::make_all_runnable.sleep", {
124
308k
        auto pipeline_id = DebugPoints::instance()->get_debug_param_or_default<int32_t>(
125
308k
                "Pipeline::make_all_runnable.sleep", "pipeline_id", -1);
126
308k
        if (pipeline_id == id()) {
127
308k
            LOG(WARNING) << "Pipeline::make_all_runnable.sleep sleep 10s";
128
308k
            sleep(10);
129
308k
        }
130
308k
    });
131
132
308k
    if (_sink->count_down_destination()) {
133
898k
        for (auto* task : _tasks) {
134
898k
            if (task) {
135
898k
                task->set_wake_up_early(wake_by);
136
898k
                task->unblock_all_dependencies();
137
898k
            }
138
898k
        }
139
305k
    }
140
308k
}
141
142
} // namespace doris