Coverage Report

Created: 2025-03-12 11:23

/root/doris/be/src/vec/runtime/partitioner.h
Line
Count
Source (jump to first uncovered line)
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
#pragma once
19
20
#include "util/runtime_profile.h"
21
#include "vec/exprs/vexpr.h"
22
#include "vec/exprs/vexpr_context.h"
23
24
namespace doris::vectorized {
25
#include "common/compile_check_begin.h"
26
struct ChannelField {
27
    const void* channel_id;
28
    const uint32_t len;
29
30
    template <typename T>
31
27
    const T* get() const {
32
27
        CHECK_EQ(sizeof(T), len) << " sizeof(T): " << sizeof(T) << " len: " << len;
33
27
        return reinterpret_cast<const T*>(channel_id);
34
27
    }
_ZNK5doris10vectorized12ChannelField3getIjEEPKT_v
Line
Count
Source
31
27
    const T* get() const {
32
27
        CHECK_EQ(sizeof(T), len) << " sizeof(T): " << sizeof(T) << " len: " << len;
33
27
        return reinterpret_cast<const T*>(channel_id);
34
27
    }
Unexecuted instantiation: _ZNK5doris10vectorized12ChannelField3getIlEEPKT_v
35
};
36
37
class PartitionerBase {
38
public:
39
20
    PartitionerBase(size_t partition_count) : _partition_count(partition_count) {}
40
20
    virtual ~PartitionerBase() = default;
41
42
    virtual Status init(const std::vector<TExpr>& texprs) = 0;
43
44
    virtual Status prepare(RuntimeState* state, const RowDescriptor& row_desc) = 0;
45
46
    virtual Status open(RuntimeState* state) = 0;
47
48
    virtual Status close(RuntimeState* state) = 0;
49
50
    virtual Status do_partitioning(RuntimeState* state, Block* block, bool eos = false,
51
                                   bool* already_sent = nullptr) const = 0;
52
53
    virtual ChannelField get_channel_ids() const = 0;
54
55
    virtual Status clone(RuntimeState* state, std::unique_ptr<PartitionerBase>& partitioner) = 0;
56
57
0
    size_t partition_count() const { return _partition_count; }
58
59
protected:
60
    const size_t _partition_count;
61
};
62
63
template <typename ChannelIds>
64
class Crc32HashPartitioner : public PartitionerBase {
65
public:
66
19
    Crc32HashPartitioner(int partition_count) : PartitionerBase(partition_count) {}
_ZN5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEEC2Ei
Line
Count
Source
66
5
    Crc32HashPartitioner(int partition_count) : PartitionerBase(partition_count) {}
_ZN5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEEC2Ei
Line
Count
Source
66
14
    Crc32HashPartitioner(int partition_count) : PartitionerBase(partition_count) {}
67
19
    ~Crc32HashPartitioner() override = default;
_ZN5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEED2Ev
Line
Count
Source
67
5
    ~Crc32HashPartitioner() override = default;
_ZN5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEED2Ev
Line
Count
Source
67
14
    ~Crc32HashPartitioner() override = default;
68
69
11
    Status init(const std::vector<TExpr>& texprs) override {
70
11
        return VExpr::create_expr_trees(texprs, _partition_expr_ctxs);
71
11
    }
Unexecuted instantiation: _ZN5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEE4initERKSt6vectorINS_5TExprESaIS5_EE
_ZN5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEE4initERKSt6vectorINS_5TExprESaIS5_EE
Line
Count
Source
69
11
    Status init(const std::vector<TExpr>& texprs) override {
70
11
        return VExpr::create_expr_trees(texprs, _partition_expr_ctxs);
71
11
    }
72
73
9
    Status prepare(RuntimeState* state, const RowDescriptor& row_desc) override {
74
9
        return VExpr::prepare(_partition_expr_ctxs, state, row_desc);
75
9
    }
Unexecuted instantiation: _ZN5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEE7prepareEPNS_12RuntimeStateERKNS_13RowDescriptorE
_ZN5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEE7prepareEPNS_12RuntimeStateERKNS_13RowDescriptorE
Line
Count
Source
73
9
    Status prepare(RuntimeState* state, const RowDescriptor& row_desc) override {
74
9
        return VExpr::prepare(_partition_expr_ctxs, state, row_desc);
75
9
    }
76
77
4
    Status open(RuntimeState* state) override { return VExpr::open(_partition_expr_ctxs, state); }
Unexecuted instantiation: _ZN5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEE4openEPNS_12RuntimeStateE
_ZN5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEE4openEPNS_12RuntimeStateE
Line
Count
Source
77
4
    Status open(RuntimeState* state) override { return VExpr::open(_partition_expr_ctxs, state); }
78
79
0
    Status close(RuntimeState* state) override { return Status::OK(); }
Unexecuted instantiation: _ZN5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEE5closeEPNS_12RuntimeStateE
Unexecuted instantiation: _ZN5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEE5closeEPNS_12RuntimeStateE
80
81
    Status do_partitioning(RuntimeState* state, Block* block, bool eos,
82
                           bool* already_sent) const override;
83
84
27
    ChannelField get_channel_ids() const override { return {_hash_vals.data(), sizeof(uint32_t)}; }
_ZNK5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEE15get_channel_idsEv
Line
Count
Source
84
21
    ChannelField get_channel_ids() const override { return {_hash_vals.data(), sizeof(uint32_t)}; }
_ZNK5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEE15get_channel_idsEv
Line
Count
Source
84
6
    ChannelField get_channel_ids() const override { return {_hash_vals.data(), sizeof(uint32_t)}; }
85
86
    Status clone(RuntimeState* state, std::unique_ptr<PartitionerBase>& partitioner) override;
87
88
protected:
89
27
    Status _get_partition_column_result(Block* block, std::vector<int>& result) const {
90
27
        int counter = 0;
91
27
        for (auto ctx : _partition_expr_ctxs) {
92
27
            RETURN_IF_ERROR(ctx->execute(block, &result[counter++]));
93
27
        }
94
27
        return Status::OK();
95
27
    }
_ZNK5doris10vectorized20Crc32HashPartitionerINS0_17ShuffleChannelIdsEE28_get_partition_column_resultEPNS0_5BlockERSt6vectorIiSaIiEE
Line
Count
Source
89
21
    Status _get_partition_column_result(Block* block, std::vector<int>& result) const {
90
21
        int counter = 0;
91
21
        for (auto ctx : _partition_expr_ctxs) {
92
21
            RETURN_IF_ERROR(ctx->execute(block, &result[counter++]));
93
21
        }
94
21
        return Status::OK();
95
21
    }
_ZNK5doris10vectorized20Crc32HashPartitionerINS0_24SpillPartitionChannelIdsEE28_get_partition_column_resultEPNS0_5BlockERSt6vectorIiSaIiEE
Line
Count
Source
89
6
    Status _get_partition_column_result(Block* block, std::vector<int>& result) const {
90
6
        int counter = 0;
91
6
        for (auto ctx : _partition_expr_ctxs) {
92
6
            RETURN_IF_ERROR(ctx->execute(block, &result[counter++]));
93
6
        }
94
6
        return Status::OK();
95
6
    }
96
97
    void _do_hash(const ColumnPtr& column, uint32_t* __restrict result, int idx) const;
98
99
    VExprContextSPtrs _partition_expr_ctxs;
100
    mutable std::vector<uint32_t> _hash_vals;
101
};
102
103
struct ShuffleChannelIds {
104
    template <typename HashValueType>
105
210
    HashValueType operator()(HashValueType l, size_t r) {
106
210
        return l % r;
107
210
    }
108
};
109
110
struct SpillPartitionChannelIds {
111
    template <typename HashValueType>
112
3.14M
    HashValueType operator()(HashValueType l, size_t r) {
113
3.14M
        return ((l >> 16) | (l << 16)) % r;
114
3.14M
    }
115
};
116
#include "common/compile_check_end.h"
117
} // namespace doris::vectorized