Skip to content

Fix incorrect results when probing multiple bloom filters #24

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jianhongshi123 wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Fix incorrect results when probing multiple bloom filters #24

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
6cd54a1
fix incorrect results when probing multiple bloom filters
jianhongshi123 Aug 6, 2026
b02b621
move selection vectors into state
jianhongshi123 Aug 28, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 5 additions & 9 deletions src/operators/physical_probe_filter.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -104,6 +104,7 @@ OperatorResultType PhysicalProbeFilter::ExecuteInternal(ExecutionContext &contex
(unsigned long long)build_col.table_index, (unsigned long long)build_col.column_index,
build_table.c_str());
state.bloom_filters.push_back(bf);
state.sel_vector.emplace_back(STANDARD_VECTOR_SIZE);
break; // found the filter for this column
}
}
Expand Down Expand Up @@ -131,7 +132,6 @@ OperatorResultType PhysicalProbeFilter::ExecuteInternal(ExecutionContext &contex

// apply bloom filters
idx_t result_count = row_num;
auto &sel = state.sel;

unique_ptr<ScopedTimer> probe_timer;
if (profiling_stats) {
Expand Down Expand Up @@ -171,7 +171,8 @@ OperatorResultType PhysicalProbeFilter::ExecuteInternal(ExecutionContext &contex
// }

// lookup directly into selection vector
result_count = bf->LookupSel(input, sel, {bound_column_indices[i]}, state.bit_vector.data());

result_count = bf->LookupSel(input, state.sel_vector[i], {bound_column_indices[i]}, state.bit_vector.data());

// early exit if no rows passed
if (result_count == 0) {
Expand All @@ -185,20 +186,15 @@ OperatorResultType PhysicalProbeFilter::ExecuteInternal(ExecutionContext &contex

// apply filter if we filtered rows
if (result_count < row_num) {
input.Slice(sel, result_count);
input.Slice(state.sel_vector[i], result_count);
row_num = result_count;
}
}

// stop probe timer before output work
probe_timer.reset();

// optimization: if all rows passed, just reference input (zero-copy)
if (result_count == row_num) {
chunk.Reference(input);
} else {
chunk.Slice(input, sel, result_count);
}
chunk.Reference(input);

if (profiling_stats) {
profiling_stats->rows_in.fetch_add(original_row_num, std::memory_order_relaxed);
Expand Down
5 changes: 2 additions & 3 deletions src/operators/physical_probe_filter.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,15 +12,14 @@ class PhysicalCreateFilter;

class PhysicalProbeFilterState : public CachingOperatorState {
public:
PhysicalProbeFilterState()
: bloom_filters_initialized(false), sel(STANDARD_VECTOR_SIZE), bit_vector((STANDARD_VECTOR_SIZE + 7) / 8) {
PhysicalProbeFilterState() : bloom_filters_initialized(false), bit_vector((STANDARD_VECTOR_SIZE + 7) / 8) {
}

vector<shared_ptr<PTBloomFilter>> bloom_filters;
bool bloom_filters_initialized;

// reusable buffers to avoid per-chunk heap allocations
SelectionVector sel;
vector<SelectionVector> sel_vector;
vector<uint8_t> bit_vector;
};

Expand Down
Loading