OmniSciDB  72c90bc290
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Macros Groups Pages
gpu_enabled Namespace Reference

Functions

template<typename... ARGS>
DEVICE auto accumulate (ARGS &&...args)
 
template<typename... ARGS>
DEVICE auto copy (ARGS &&...args)
 
template<typename... ARGS>
DEVICE void fill (ARGS &&...args)
 
template<typename... ARGS>
DEVICE void iota (ARGS &&...args)
 
template<typename... ARGS>
DEVICE auto lower_bound (ARGS &&...args)
 
template<typename... ARGS>
DEVICE void partial_sum (ARGS &&...args)
 
template<typename... ARGS>
DEVICE void reverse (ARGS &&...args)
 
template<typename... ARGS>
DEVICE void sort (ARGS &&...args)
 
template<typename... ARGS>
DEVICE void swap (ARGS &&...args)
 
template<typename... ARGS>
DEVICE auto upper_bound (ARGS &&...args)
 

Function Documentation

template<typename... ARGS>
DEVICE auto gpu_enabled::accumulate ( ARGS &&...  args)

Definition at line 42 of file gpu_enabled.h.

References run_benchmark_import::args.

Referenced by BoundingBoxIntersectJoinHashTable::approximateTupleCount(), QueryFragmentDescriptor::buildFragmentPerKernelMapForUnion(), anonymous_namespace{RelAlgExecutor.cpp}::check_none_encoded_string_cast_tuple_limit(), StringDictionary::copyStrings(), RelAlgExecutor::createUnionWorkUnit(), PerfectJoinHashTable::genChunkKey(), ColSlotContext::getAlignedPaddedSizeForRange(), ColSlotContext::getAllSlotsPaddedSize(), Executor::getChunkRequestInfo(), getExpressionRange(), ColSlotContext::getTotalBytesOfColumnarBuffers(), Executor::getUniqueThreadSharedResultSets(), hide_sensitive_data_from_query(), ColumnarResults::mergeResults(), QueryExecutionContext::prepareKernelParams(), ResultSetManager::reduce(), Executor::reduceMultiDeviceResultSets(), PerfectJoinHashTable::reify(), BoundingBoxIntersectJoinHashTable::reifyWithLayout(), BaselineJoinHashTable::reifyWithLayout(), QueryRewriter::rewriteColumnarDelete(), QueryRewriter::rewriteColumnarUpdate(), ExecutionKernel::runImpl(), and quantile::detail::Centroids< RealType, IndexType >::totalWeight().

42  {
43 #ifdef __CUDACC__
44  return thrust::reduce(thrust::device, std::forward<ARGS>(args)...);
45 #else
46  return std::accumulate(std::forward<ARGS>(args)...);
47 #endif
48 }
DEVICE auto accumulate(ARGS &&...args)
Definition: gpu_enabled.h:42

+ Here is the caller graph for this function:

template<typename... ARGS>
DEVICE auto gpu_enabled::copy ( ARGS &&...  args)

Definition at line 51 of file gpu_enabled.h.

References run_benchmark_import::args.

Referenced by RelLeftDeepTreeIdsCollector::aggregateResult(), quantile::detail::Centroids< RealType, IndexType >::appendAndSortCurrent(), StringNoneEncoder::appendData(), anonymous_namespace{ArrowResultSetConverter.cpp}::appendToListColumnBuilder(), anonymous_namespace{WindowContext.cpp}::apply_lag_to_partition(), anonymous_namespace{WindowContext.cpp}::apply_permutation_to_partition(), CodeGenerator::codegenDictLike(), CodeGenerator::codegenDictRegexp(), CodeGenerator::codegenDictStrCmp(), CodeGenerator::codegenFunctionOper(), WindowFunctionContext::computePartitionBuffer(), ResultSet::copy(), anonymous_namespace{SysCatalog.cpp}::copy_catalog_if_read_only(), StringDictionary::copyStrings(), file_delete(), UdfCompiler::generateAST(), ExtensionFunctionsWhitelist::get_ext_funcs(), DBHandler::get_tables_meta_impl(), QueryPlanDagExtractor::handleLeftDeepJoinTree(), DBHandler::insert_chunks(), ColumnFetcher::linearizeVarLenArrayColFrags(), ResultSet::parallelTop(), pop_n_rows_from_merged_heaps_gpu(), ArrowForeignStorageBase::replaceNullValuesImpl(), ResultSet::setTargetMetaInfo(), quantile::detail::anonymous_namespace{quantile.h}::Skipped< RealType, IndexType >::shiftRange(), and File_Namespace::FileBuffer::writeHeader().

51  {
52 #ifdef __CUDACC__
53  return thrust::copy(thrust::device, std::forward<ARGS>(args)...);
54 #else
55  return std::copy(std::forward<ARGS>(args)...);
56 #endif
57 }
DEVICE auto copy(ARGS &&...args)
Definition: gpu_enabled.h:51

+ Here is the caller graph for this function:

template<typename... ARGS>
DEVICE void gpu_enabled::fill ( ARGS &&...  args)

Definition at line 60 of file gpu_enabled.h.

References run_benchmark_import::args.

Referenced by NoneEncoder< T >::appendValidatedOrNonValidatedData(), anonymous_namespace{WindowContext.cpp}::apply_nth_value_to_partition(), quantile::detail::TDigest< RealType, IndexType >::mergeSorted(), ArrowForeignStorageBase::replaceNullValuesImpl(), and ColSlotContext::setAllSlotsSize().

60  {
61 #ifdef __CUDACC__
62  thrust::fill(thrust::device, std::forward<ARGS>(args)...);
63 #else
64  std::fill(std::forward<ARGS>(args)...);
65 #endif
66 }
DEVICE void fill(ARGS &&...args)
Definition: gpu_enabled.h:60

+ Here is the caller graph for this function:

template<typename... ARGS>
DEVICE void gpu_enabled::iota ( ARGS &&...  args)

Definition at line 69 of file gpu_enabled.h.

References run_benchmark_import::args.

Referenced by find_push_down_filters(), import_export::ForeignDataImporter::importGeneralS3(), anonymous_namespace{RelAlgOptimizer.cpp}::mark_live_columns(), foreign_storage::LazyParquetChunkLoader::previewFiles(), Fragmenter_Namespace::SortedOrderFragmenter::sortData(), WindowFunctionContext::sortPartition(), anonymous_namespace{FromTableReordering.cpp}::traverse_join_cost_graph(), and WindowFunctionContext::WindowFunctionContext().

69  {
70 #ifdef __CUDACC__
71  thrust::sequence(thrust::device, std::forward<ARGS>(args)...);
72 #else
73  std::iota(std::forward<ARGS>(args)...);
74 #endif
75 }
DEVICE void iota(ARGS &&...args)
Definition: gpu_enabled.h:69

+ Here is the caller graph for this function:

template<typename... ARGS>
DEVICE auto gpu_enabled::lower_bound ( ARGS &&...  args)
template<typename... ARGS>
DEVICE void gpu_enabled::partial_sum ( ARGS &&...  args)

Definition at line 87 of file gpu_enabled.h.

References run_benchmark_import::args, and inclusive_scan().

Referenced by ColumnarResults::compactAndCopyEntries(), WindowFunctionContext::fillPartitionEnd(), WindowFunctionContext::fillPartitionStart(), find_push_down_filters(), inclusive_scan(), quantile::detail::TDigest< RealType, IndexType >::partialSumOfCounts(), anonymous_namespace{RelLeftDeepInnerJoin.cpp}::RebindRexInputsFromLeftDeepJoin::RebindRexInputsFromLeftDeepJoin(), and WindowFunctionContext::WindowFunctionContext().

87  {
88 #ifdef __CUDACC__
89  thrust::inclusive_scan(thrust::device, std::forward<ARGS>(args)...);
90 #else
91  std::partial_sum(std::forward<ARGS>(args)...);
92 #endif
93 }
void inclusive_scan(InputIterator first, InputIterator last, OutputIterator out, const size_t thread_count)
DEVICE void partial_sum(ARGS &&...args)
Definition: gpu_enabled.h:87

+ Here is the call graph for this function:

+ Here is the caller graph for this function:

template<typename... ARGS>
DEVICE void gpu_enabled::reverse ( ARGS &&...  args)

Definition at line 96 of file gpu_enabled.h.

References run_benchmark_import::args.

Referenced by quantile::detail::Centroids< RealType, IndexType >::appendAndSortCurrent(), and RaExecutionSequence::RaExecutionSequence().

96  {
97 #ifdef __CUDACC__
98  thrust::reverse(thrust::device, std::forward<ARGS>(args)...);
99 #else
100  std::reverse(std::forward<ARGS>(args)...);
101 #endif
102 }
DEVICE void reverse(ARGS &&...args)
Definition: gpu_enabled.h:96

+ Here is the caller graph for this function:

template<typename... ARGS>
DEVICE void gpu_enabled::sort ( ARGS &&...  args)

Definition at line 105 of file gpu_enabled.h.

References run_benchmark_import::args.

Referenced by anonymous_namespace{RelAlgOptimizer.cpp}::add_new_indices_for(), anonymous_namespace{RelAlgDag.cpp}::add_window_function_pre_project(), Catalog_Namespace::Catalog::addColumnDescriptor(), Catalog_Namespace::Catalog::addTableToMap(), quantile::detail::Centroids< RealType, IndexType >::appendAndSortCurrent(), SpeculativeTopNMap::asRows(), Catalog_Namespace::Catalog::buildColumnsMapUnlocked(), HashJoin::collectFragmentIds(), Parser::AddColumnStmt::execute(), RelAlgExecutor::executeRelAlgQuerySingleStep(), RelAlgExecutor::executeRelAlgStep(), RelAlgExecutor::executeSort(), import_export::fill_missing_columns(), DBHandler::get_completion_hints(), anonymous_namespace{RelAlgExecutor.cpp}::get_input_desc_impl(), anonymous_namespace{RelAlgOptimizer.cpp}::get_live_ins(), Fragmenter_Namespace::InsertOrderFragmenter::getChunkMetadata(), File_Namespace::CachingFileMgr::init(), File_Namespace::FileMgr::init(), ColumnFetcher::linearizeVarLenArrayColFrags(), RelRexDagVisitor::make_handlers(), quantile::detail::TDigest< RealType, IndexType >::mergeBuffer(), RaExecutionSequence::mergeSortWithInput(), DBHandler::parse_to_ra(), anonymous_namespace{RelAlgOptimizer.cpp}::pick_always_live_col_idx(), project_separates_sort(), anonymous_namespace{RelAlgOptimizer.cpp}::propagate_input_renumbering(), Catalog_Namespace::Catalog::removeColumnDescriptor(), Executor::resultsUnion(), sort_column_limit__cpu_template(), File_Namespace::FileMgr::sortAndCopyFilePagesForCompaction(), StringDictionary::sortCache(), DataRecycler< std::optional< HashType >, EMPTY_META_INFO >::sortCacheContainerByQueryMetric(), CacheMetricTracker::sortCacheInfoByQueryMetric(), Fragmenter_Namespace::sortIndexesImpl(), WindowFunctionContext::sortPartition(), RaExecutionSequence::stepsToNextBroadcast(), ResultSet::topPermutation(), QueryPlanDagCache::translateColVarsToInfoHash(), and RelAlgVisitor< std::vector< unsigned > >::visit().

105  {
106 #ifdef __CUDACC__
107  thrust::sort(thrust::device, std::forward<ARGS>(args)...);
108 #else
109  std::sort(std::forward<ARGS>(args)...);
110 #endif
111 }
DEVICE void sort(ARGS &&...args)
Definition: gpu_enabled.h:105

+ Here is the caller graph for this function: