threading_serial Namespace Reference

Classes
class	task_group

Functions
template<typename Fn , typename... Args, typename Result = std::result_of_t<Fn && (Args && ...)>>
future< Result >	async (Fn &&fn, Args &&...args)
template<typename Int , typename Body , typename Partitioner = auto_partitioner>
void	parallel_for (const blocked_range< Int > &range, const Body &body, const Partitioner &p=Partitioner())
template<typename Index , typename Function , typename Partitioner = auto_partitioner>
void	parallel_for (Index first, Index last, const Function &f, const Partitioner &p=Partitioner())
template<typename Int , typename Value , typename RealBody , typename Reduction , typename Partitioner = auto_partitioner>
Value	parallel_reduce (const blocked_range< Int > &range, const Value &identity, const RealBody &real_body, const Reduction &reduction, const Partitioner &p=Partitioner())
	Parallel iteration with reduction. More...

Function Documentation

template<typename Fn , typename... Args, typename Result = std::result_of_t<Fn && (Args && ...)>>

future<Result> threading_serial::async	(	Fn &&	fn,
		Args &&...	args
	)

Definition at line 11 of file threading_serial.h.

References run_benchmark_import::args.

Referenced by TableFunctions_Namespace::OneHotEncoder_Namespace::allocate_one_hot_cols(), approximate_distinct_tuples(), approximate_distinct_tuples_bbox_intersect(), approximate_distinct_tuples_range(), RangeJoinHashTable::approximateTupleCount(), BoundingBoxIntersectJoinHashTable::approximateTupleCount(), BaselineJoinHashTable::approximateTupleCount(), threading_std::async(), ColumnarResults::compactAndCopyEntriesWithoutTargetSkipping(), ColumnarResults::compactAndCopyEntriesWithTargetSkipping(), Fragmenter_Namespace::InsertOrderFragmenter::compactRows(), compute_bucket_sizes_on_cpu(), foreign_storage::TextFileBufferParser::convertImportBuffersToDataBlocks(), ColumnarResults::copyAllNonLazyColumns(), Fragmenter_Namespace::copyDataOfShard(), StringDictionary::copyStrings(), foreign_storage::create_futures_for_workers(), CodeGenerator::createInValuesBitmap(), fill_one_to_many_baseline_hash_table(), fill_one_to_many_hash_table_impl(), fill_one_to_many_hash_table_sharded_impl(), import_export::TypedImportBuffer::get_data_block_pointers(), anonymous_namespace{RelAlgTranslator.cpp}::get_in_values_expr(), ArrowResultSetConverter::getArrowBatch(), RelAlgTranslator::getInIntegerSetExpr(), Fragmenter_Namespace::InsertOrderFragmenter::getVacuumOffsets(), anonymous_namespace{ForeignDataImporter.cpp}::import_foreign_data(), import_export::Importer::importDelimited(), import_export::Importer::importGDALGeo(), import_export::Importer::importGDALRaster(), import_export::ForeignDataImporter::importGeneralS3(), inclusive_scan(), File_Namespace::FileMgr::init(), BaselineJoinHashTableBuilder::initHashTableOnCpu(), PerfectJoinHashTableBuilder::initOneToManyHashTableOnCpu(), PerfectJoinHashTableBuilder::initOneToOneHashTableOnCpu(), Fragmenter_Namespace::InsertDataLoader::insertChunks(), Fragmenter_Namespace::InsertDataLoader::insertData(), foreign_storage::AbstractTextFileDataWrapper::iterativeFileScan(), ColumnFetcher::linearizeVarLenArrayColFrags(), foreign_storage::LazyParquetChunkLoader::loadRowGroups(), ColumnarResults::locateAndCountEntries(), ColumnarResults::materializeAllColumnsThroughIteration(), ColumnarResults::materializeAllLazyColumns(), foreign_storage::LazyParquetChunkLoader::metadataScan(), ResultSetStorage::moveEntriesToBuffer(), File_Namespace::FileMgr::openFiles(), threading_std::parallel_for(), threading_std::parallel_reduce(), foreign_storage::populate_chunks_using_data_blocks(), StringDictionary::populate_string_array_ids(), foreign_storage::AbstractTextFileDataWrapper::populateChunkMetadata(), foreign_storage::AbstractTextFileDataWrapper::populateChunks(), Parser::InsertIntoTableAsSelectStmt::populateData(), File_Namespace::FileBuffer::read(), ResultSetStorage::reduce(), PerfectJoinHashTable::reify(), BoundingBoxIntersectJoinHashTable::reifyImpl(), RangeJoinHashTable::reifyWithLayout(), BaselineJoinHashTable::reifyWithLayout(), ThreadController_NS::SimpleThreadController< FutureReturnType >::startThread(), StringDictionary::StringDictionary(), Fragmenter_Namespace::InsertOrderFragmenter::updateColumn(), Fragmenter_Namespace::InsertOrderFragmenter::updateColumns(), StorageIOFacility::yieldDeleteCallback(), and StorageIOFacility::yieldUpdateCallback().

                                              {
  std::promise<Result> pr;
  if constexpr (std::is_same<void, Result>::value) {
    fn(std::forward<Args>(args)...);
    pr.set_value();
  } else {
    pr.set_value(fn(std::forward<Args>(args)...));
  }
  return pr.get_future();
}

template<typename Int , typename Body , typename Partitioner = auto_partitioner>

void threading_serial::parallel_for	(	const blocked_range< Int > &	range,
		const Body &	body,
		const Partitioner &	p = Partitioner()
	)

Definition at line 34 of file threading_serial.h.

References threading_common::blocked_range< Value >::begin(), cpu_threads(), threading_common::blocked_range< Value >::end(), and threading_common::blocked_range< Value >::size().

Referenced by TreeModelPredictionMgr::allocateAndPopulateHostBuffers(), StringDictionary::buildDictionaryNumericTranslationMap(), StringDictionary::buildDictionaryTranslationMap(), StringDictionaryProxy::buildNumericTranslationMap(), GeoRaster< T, Z >::calculate_slope_and_aspect(), CodeGenerator::codegenTreeRegPredict(), WindowFunctionContext::compute(), GeoRaster< T, Z >::compute(), compute_table_function_col_chunk_stats(), GeoRaster< T, Z >::computeParallelImpl(), GeoRaster< T, Z >::computeParallelReductionAggImpl(), DataBufferCache::copyData(), ArrowForeignStorageBase::createDecimalColumn(), ArrowForeignStorageBase::createDictionaryEncodedColumn(), GeoRaster< T, Z >::fill_bins_from_box_neighborhood(), GeoRaster< T, Z >::fill_bins_from_gaussian_neighborhood(), RasterFormat_Namespace::fill_col_halo(), RasterFormat_Namespace::fill_row_halo(), RasterFormat_Namespace::format_raster_data(), generate_random_strings__cpu_(), generate_series_parallel(), get_column_mean(), get_column_metadata(), get_column_min_max(), get_column_stats(), get_column_std_dev(), TableFunctions_Namespace::OneHotEncoder_Namespace::get_top_k_keys(), StringDictionary::getBulk(), StringDictionary::getStringViews(), StringDictionary::hashStrings(), QueryMemoryInitializer::initRowGroups(), Mandelbrot::mandelbrot_cpu_template(), ResultSet::ResultSetComparator< BUFFER_ITERATOR_TYPE >::materializeModeColumn(), TableFunctions_Namespace::OneHotEncoder_Namespace::one_hot_encode(), GeoRaster< T, Z >::outputDenseColumn(), GeoRaster< T, Z >::outputDenseColumns(), GeoRaster< T, Z >::outputDenseColumnsAndFill(), threading_tbb::parallel_for(), parallel_for(), ArrowForeignStorageBase::parseArrowTable(), r2_score_impl(), TableFunctions_Namespace::remove_null_rows(), ArrowForeignStorageBase::replaceNullValuesImpl(), synthesize_metadata(), CrossSectionTableFunctions::tf_cross_section_2d_impl(), StringDictionaryProxy::transientLookupBulkParallelUnlocked(), TableFunctions_Namespace::unmask_data(), and z_std_normalize_col().

                                                        {
  const Int worker_count = cpu_threads();

  for (Int i = 0,
           start_entry = range.begin(),
           stop_entry = range.end(),
           stride = (range.size() + worker_count - 1) / worker_count;
       i < worker_count && start_entry < stop_entry;
       ++i, start_entry += stride) {
    const auto end_entry = std::min(start_entry + stride, stop_entry);
    body(blocked_range<Int>(start_entry, end_entry));
  }
}

Here is the call graph for this function:

Here is the caller graph for this function:

template<typename Index , typename Function , typename Partitioner = auto_partitioner>

void threading_serial::parallel_for	(	Index	first,
		Index	last,
		const Function &	f,
		const Partitioner &	p = Partitioner()
	)

Parallel iteration over a range of integers with a default step value and default partitioner

Definition at line 53 of file threading_serial.h.

References threading_common::blocked_range< Value >::begin(), threading_common::blocked_range< Value >::end(), f(), and parallel_for().

                                                        {
  parallel_for(
      blocked_range<Index>(first, last),
      [&f](const blocked_range<Index>& r) {
        for (auto i = r.begin(), e = r.end(); i < e; i++) {
          f(i);
        }
      },
      p);
}

Here is the call graph for this function:

template<typename Int , typename Value , typename RealBody , typename Reduction , typename Partitioner = auto_partitioner>

Value threading_serial::parallel_reduce	(	const blocked_range< Int > &	range,
		const Value &	identity,
		const RealBody &	real_body,
		const Reduction &	reduction,
		const Partitioner &	p = Partitioner()
	)

Parallel iteration with reduction.

Definition at line 74 of file threading_serial.h.

References threading_common::blocked_range< Value >::begin(), cpu_threads(), threading_common::blocked_range< Value >::end(), and threading_common::blocked_range< Value >::size().

Referenced by anonymous_namespace{ColumnarResults.cpp}::computeTotalNofValuesForColumnArray(), anonymous_namespace{ColumnarResults.cpp}::computeTotalNofValuesForColumnGeoType(), anonymous_namespace{ColumnarResults.cpp}::computeTotalNofValuesForColumnTextEncodingNone(), threading_tbb::parallel_reduce(), ResultSet::parallelRowCount(), FixedLengthEncoder< T, V >::updateStatsEncoded(), and NoneEncoder< T >::updateStatsEncoded().

                                                            {
  const size_t worker_count = cpu_threads();
  std::vector<Value> worker_threads;
  worker_threads.reserve(worker_count);

  for (Int i = 0,
           start_entry = range.begin(),
           stop_entry = range.end(),
           stride = (range.size() + worker_count - 1) / worker_count;
       i < worker_count && start_entry < stop_entry;
       ++i, start_entry += stride) {
    const auto end_entry = std::min(start_entry + stride, stop_entry);
    // TODO grainsize?
    worker_threads.emplace_back(
        real_body(blocked_range<Int>(start_entry, end_entry), Value{}));
  }
  Value v = identity;
  for (auto& child : worker_threads) {
    v = reduction(v, child);
  }

  return v;
}

Here is the call graph for this function:

Here is the caller graph for this function: