database/doris
2
0
Fork 0
Code Issues Pull Requests Actions 3 Packages Projects Releases Wiki Activity

[Improvement](hash) refactor of hash map context (#24966)

Browse Source 
refactor of hash map context
This commit is contained in:
Pxl
2023-10-12 18:10:21 +08:00
committed by GitHub
parent 04bda138d6
commit 1a0344df16
57 changed files with 1225 additions and 2767 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
be/src/vec/exec/distinct_vaggregation_node.cpp
View File

@ -86,43 +86,27 @@ void DistinctAggregationNode::_emplace_into_hash_table_to_distinct(IColumn::Sele
[&](auto&& agg_method) -> void {
SCOPED_TIMER(_hash_table_compute_timer);
using HashMethodType = std::decay_t<decltype(agg_method)>;
using HashTableType = std::decay_t<decltype(agg_method.data)>;
using AggState = typename HashMethodType::State;
AggState state(key_columns, _probe_key_sz, nullptr);
_pre_serialize_key_if_need(state, agg_method, key_columns, num_rows);
AggState state(key_columns, _probe_key_sz);
agg_method.init_serialized_keys(key_columns, _probe_key_sz, num_rows);
if constexpr (HashTableTraits<HashTableType>::is_phmap) {
const auto& keys = state.get_keys();
if (_hash_values.size() < num_rows) {
_hash_values.resize(num_rows);
}
size_t row = 0;
auto creator = [&](const auto& ctor, auto& key, auto& origin) {
HashMethodType::try_presis_key(key, origin, *_agg_arena_pool);
ctor(key, dummy_mapped_data);
distinct_row.push_back(row);
};
for (size_t i = 0; i < num_rows; ++i) {
_hash_values[i] = agg_method.data.hash(keys[i]);
}
SCOPED_TIMER(_hash_table_emplace_timer);
for (size_t i = 0; i < num_rows; ++i) {
if (LIKELY(i + HASH_MAP_PREFETCH_DIST < num_rows)) {
agg_method.data.prefetch_by_hash(
_hash_values[i + HASH_MAP_PREFETCH_DIST]);
}
auto result = state.emplace_with_key(
agg_method.data, state.pack_key_holder(keys[i], *_agg_arena_pool),
_hash_values[i], i);
if (result.is_inserted()) {
distinct_row.push_back(i);
}
}
} else {
SCOPED_TIMER(_hash_table_emplace_timer);
for (size_t i = 0; i < num_rows; ++i) {
auto result = state.emplace_key(agg_method.data, i, *_agg_arena_pool);
if (result.is_inserted()) {
result.set_mapped(dummy_mapped_data);
distinct_row.push_back(i);
}
}
auto creator_for_null_key = [&](auto& mapped) {
mapped = dummy_mapped_data;
distinct_row.push_back(row);
};
SCOPED_TIMER(_hash_table_emplace_timer);
for (; row < num_rows; ++row) {
agg_method.lazy_emplace(state, row, creator, creator_for_null_key);
}
COUNTER_UPDATE(_hash_table_input_counter, num_rows);
},
_agg_data->method_variant);