#include <sdfg/targets/gpu/gpu_offload_reduce_dispatcher.h>

Inheritance diagram for sdfg::gpu::GPUOffloadReduceDispatcher:

Public Member Functions
	GPUOffloadReduceDispatcher (codegen::LanguageExtension &language_extension, StructuredSDFG &sdfg, analysis::AnalysisManager &analysis_manager, structured_control_flow::Reduce &node, codegen::InstrumentationPlan &instrumentation_plan, codegen::ArgCapturePlan &arg_capture_plan, std::unique_ptr< GPUOffloadDispatcherStrategy > strategy)
Public Member Functions inherited from sdfg::gpu::GPUOffloadBaseDispatcher
void	dispatch_node (codegen::PrettyPrinter &main_stream, codegen::PrettyPrinter &globals_stream, codegen::CodeSnippetFactory &library_snippet_factory) override
codegen::InstrumentationInfo	instrumentation_info () const override
	Bad design.
Public Member Functions inherited from sdfg::codegen::NodeDispatcher
	NodeDispatcher (LanguageExtension &language_extension, StructuredSDFG &sdfg, analysis::AnalysisManager &analysis_manager, structured_control_flow::ControlFlowNode &node, InstrumentationPlan &instrumentation_plan, ArgCapturePlan &arg_capture_plan)
virtual	~NodeDispatcher ()=default
InstrumentationInfo	instrumentation_info_for_group () const
virtual void	dispatch (PrettyPrinter &main_stream, PrettyPrinter &globals_stream, CodeSnippetFactory &library_snippet_factory)

Protected Member Functions
void	dispatch_kernel_body (codegen::NestedCodeSnippetFactory &kernel_snippet_factory, codegen::PrettyPrinter &kernel_source_stream, codegen::PrettyPrinter &kernel_header_stream, codegen::LanguageExtension &kernel_lang_ext, symbolic::Symbol indvar, std::vector< std::string > &scope_variables, symbolic::Expression &num_iterations) override
void	validate_before_dispatch (analysis::AnalysisManager &analysis_manager) override
	Node-specific preconditions checked before any code is emitted (default: none).
bool	has_nested_warp_reduction (const std::string &container)
bool	has_nested_block_reduction (const std::string &container)
bool	has_enclosing_block_reduction (const std::string &container)
bool	has_enclosing_grid_reduction (const std::string &container)
bool	block_result_collides_across_grid (const symbolic::Expression &index)
std::string	block_reduce_leader_condition (codegen::LanguageExtension &language_extension, const std::string &container)
std::string	reduce_linear_thread_index (codegen::LanguageExtension &language_extension)
std::string	reduce_axis_stride (codegen::LanguageExtension &language_extension, TargetLevel target_level)
symbolic::Expression	reduce_block_dim (TargetLevel block_level)
symbolic::Expression	reduce_block_size_product ()
std::string	partials_buffer_name (const std::string &container)
bool	is_scalar_accumulator (const std::string &container)
std::string	reduce_base_slot (codegen::LanguageExtension &language_extension, const std::string &container)
void	dispatch_reduction_declarations (codegen::LanguageExtension &language_extension, codegen::PrettyPrinter &stream, codegen::CodeSnippetFactory &library_snippet_factory, TargetLevel target_level)
std::string	reduction_target (codegen::LanguageExtension &language_extension, const std::string &container, symbolic::Expression index)
void	dispatch_reduction_combine (codegen::LanguageExtension &language_extension, codegen::PrettyPrinter &stream, codegen::CodeSnippetFactory &library_snippet_factory, TargetLevel target_level)
bool	uses_register_partial (TargetLevel target_level, const std::string &container)
int64_t	warp_combine_width (TargetLevel target_level, const std::string &container)
int64_t	lane_group_width (TargetLevel target_level)
bool	in_rescale_group (const std::string &container)
void	dispatch_rescale_group_combine (codegen::LanguageExtension &language_extension, codegen::PrettyPrinter &stream, const structured_control_flow::ReductionInfo &anchor, TargetLevel target_level)
void	dispatch_reduction_publish (codegen::LanguageExtension &language_extension, codegen::PrettyPrinter &stream, TargetLevel target_level)
Protected Member Functions inherited from sdfg::gpu::GPUOffloadBaseDispatcher
	GPUOffloadBaseDispatcher (codegen::LanguageExtension &language_extension, StructuredSDFG &sdfg, analysis::AnalysisManager &analysis_manager, structured_control_flow::StructuredLoop &node, codegen::InstrumentationPlan &instrumentation_plan, codegen::ArgCapturePlan &arg_capture_plan, std::unique_ptr< GPUOffloadDispatcherStrategy > strategy)
bool	is_outermost_map (analysis::AnalysisManager &analysis_manager)
void	dispatch_header (codegen::PrettyPrinter &globals_stream, const std::string &kernel_name, std::vector< std::string > &arguments_declaration)
void	dispatch_kernel_preamble (codegen::PrettyPrinter &library_stream, const std::string &kernel_name, std::vector< std::string > &arguments_declaration)
void	emit_lib_dependency_includes (codegen::PrettyPrinter &kernel_header_stream, codegen::NestedCodeSnippetFactory &nested_snippet_factory)
Protected Member Functions inherited from sdfg::codegen::NodeDispatcher
virtual bool	begin_node (PrettyPrinter &stream)
virtual void	end_node (PrettyPrinter &stream, bool has_declaration)

Protected Attributes
structured_control_flow::Reduce &	node_
std::map< std::string, tiles::ReductionBufferInfo >	reduction_buffers_
Protected Attributes inherited from sdfg::gpu::GPUOffloadBaseDispatcher
structured_control_flow::StructuredLoop &	node_
std::unique_ptr< GPUOffloadDispatcherStrategy >	strategy_
long	launch_bounds_ = 0
Protected Attributes inherited from sdfg::codegen::NodeDispatcher
LanguageExtension &	language_extension_
StructuredSDFG &	sdfg_
InstrumentationPlan &	instrumentation_plan_
ArgCapturePlan &	arg_capture_plan_
analysis::AnalysisManager &	analysis_manager_

Constructor & Destructor Documentation

◆ GPUOffloadReduceDispatcher()

sdfg::gpu::GPUOffloadReduceDispatcher::GPUOffloadReduceDispatcher	(	codegen::LanguageExtension &	language_extension,
		StructuredSDFG &	sdfg,
		analysis::AnalysisManager &	analysis_manager,
		structured_control_flow::Reduce &	node,
		codegen::InstrumentationPlan &	instrumentation_plan,
		codegen::ArgCapturePlan &	arg_capture_plan,
		std::unique_ptr< GPUOffloadDispatcherStrategy >	strategy )

Member Function Documentation

◆ block_reduce_leader_condition()

std::string sdfg::gpu::GPUOffloadReduceDispatcher::block_reduce_leader_condition	(	codegen::LanguageExtension &	language_extension,
		const std::string &	container )

protected

◆ block_result_collides_across_grid()

bool sdfg::gpu::GPUOffloadReduceDispatcher::block_result_collides_across_grid ( const symbolic::Expression & index )

protected

◆ dispatch_kernel_body()

void sdfg::gpu::GPUOffloadReduceDispatcher::dispatch_kernel_body	(	codegen::NestedCodeSnippetFactory &	kernel_snippet_factory,
		codegen::PrettyPrinter &	kernel_source_stream,
		codegen::PrettyPrinter &	kernel_header_stream,
		codegen::LanguageExtension &	kernel_lang_ext,
		symbolic::Symbol	indvar,
		std::vector< std::string > &	scope_variables,
		symbolic::Expression &	num_iterations )

overrideprotectedvirtual

Implements sdfg::gpu::GPUOffloadBaseDispatcher.

◆ dispatch_reduction_combine()

void sdfg::gpu::GPUOffloadReduceDispatcher::dispatch_reduction_combine	(	codegen::LanguageExtension &	language_extension,
		codegen::PrettyPrinter &	stream,
		codegen::CodeSnippetFactory &	library_snippet_factory,
		TargetLevel	target_level )

protected

◆ dispatch_reduction_declarations()

void sdfg::gpu::GPUOffloadReduceDispatcher::dispatch_reduction_declarations	(	codegen::LanguageExtension &	language_extension,
		codegen::PrettyPrinter &	stream,
		codegen::CodeSnippetFactory &	library_snippet_factory,
		TargetLevel	target_level )

protected

◆ dispatch_reduction_publish()

void sdfg::gpu::GPUOffloadReduceDispatcher::dispatch_reduction_publish	(	codegen::LanguageExtension &	language_extension,
		codegen::PrettyPrinter &	stream,
		TargetLevel	target_level )

protected

◆ dispatch_rescale_group_combine()

void sdfg::gpu::GPUOffloadReduceDispatcher::dispatch_rescale_group_combine	(	codegen::LanguageExtension &	language_extension,
		codegen::PrettyPrinter &	stream,
		const structured_control_flow::ReductionInfo &	anchor,
		TargetLevel	target_level )

protected

◆ has_enclosing_block_reduction()

bool sdfg::gpu::GPUOffloadReduceDispatcher::has_enclosing_block_reduction ( const std::string & container )

protected

◆ has_enclosing_grid_reduction()

bool sdfg::gpu::GPUOffloadReduceDispatcher::has_enclosing_grid_reduction ( const std::string & container )

protected

◆ has_nested_block_reduction()

bool sdfg::gpu::GPUOffloadReduceDispatcher::has_nested_block_reduction ( const std::string & container )

protected

◆ has_nested_warp_reduction()

bool sdfg::gpu::GPUOffloadReduceDispatcher::has_nested_warp_reduction ( const std::string & container )

protected

◆ in_rescale_group()

bool sdfg::gpu::GPUOffloadReduceDispatcher::in_rescale_group ( const std::string & container )

protected

◆ is_scalar_accumulator()

bool sdfg::gpu::GPUOffloadReduceDispatcher::is_scalar_accumulator ( const std::string & container )

protected

◆ lane_group_width()

int64_t sdfg::gpu::GPUOffloadReduceDispatcher::lane_group_width ( TargetLevel target_level )

protected

◆ partials_buffer_name()

std::string sdfg::gpu::GPUOffloadReduceDispatcher::partials_buffer_name ( const std::string & container )

protected

◆ reduce_axis_stride()

std::string sdfg::gpu::GPUOffloadReduceDispatcher::reduce_axis_stride	(	codegen::LanguageExtension &	language_extension,
		TargetLevel	target_level )

protected

◆ reduce_base_slot()

std::string sdfg::gpu::GPUOffloadReduceDispatcher::reduce_base_slot	(	codegen::LanguageExtension &	language_extension,
		const std::string &	container )

protected

◆ reduce_block_dim()

symbolic::Expression sdfg::gpu::GPUOffloadReduceDispatcher::reduce_block_dim ( TargetLevel block_level )

protected

◆ reduce_block_size_product()

symbolic::Expression sdfg::gpu::GPUOffloadReduceDispatcher::reduce_block_size_product ( )

protected

◆ reduce_linear_thread_index()

std::string sdfg::gpu::GPUOffloadReduceDispatcher::reduce_linear_thread_index ( codegen::LanguageExtension & language_extension )

protected

◆ reduction_target()

std::string sdfg::gpu::GPUOffloadReduceDispatcher::reduction_target	(	codegen::LanguageExtension &	language_extension,
		const std::string &	container,
		symbolic::Expression	index )

protected

◆ uses_register_partial()

bool sdfg::gpu::GPUOffloadReduceDispatcher::uses_register_partial	(	TargetLevel	target_level,
		const std::string &	container )

protected

◆ validate_before_dispatch()

void sdfg::gpu::GPUOffloadReduceDispatcher::validate_before_dispatch ( analysis::AnalysisManager & analysis_manager )

overrideprotectedvirtual

Node-specific preconditions checked before any code is emitted (default: none).

Reimplemented from sdfg::gpu::GPUOffloadBaseDispatcher.

◆ warp_combine_width()

int64_t sdfg::gpu::GPUOffloadReduceDispatcher::warp_combine_width	(	TargetLevel	target_level,
		const std::string &	container )

protected

Member Data Documentation

◆ node_

structured_control_flow::Reduce& sdfg::gpu::GPUOffloadReduceDispatcher::node_

protected

◆ reduction_buffers_

std::map<std::string, tiles::ReductionBufferInfo> sdfg::gpu::GPUOffloadReduceDispatcher::reduction_buffers_

protected

The documentation for this class was generated from the following file:

opt/include/sdfg/targets/gpu/gpu_offload_reduce_dispatcher.h

Public Member Functions

Protected Member Functions

Protected Attributes

Constructor & Destructor Documentation

◆ GPUOffloadReduceDispatcher()

Member Function Documentation

◆ block_reduce_leader_condition()

◆ block_result_collides_across_grid()

◆ dispatch_kernel_body()

◆ dispatch_reduction_combine()

◆ dispatch_reduction_declarations()

◆ dispatch_reduction_publish()

◆ dispatch_rescale_group_combine()

◆ has_enclosing_block_reduction()

◆ has_enclosing_grid_reduction()

◆ has_nested_block_reduction()

◆ has_nested_warp_reduction()

◆ in_rescale_group()

◆ is_scalar_accumulator()

◆ lane_group_width()

◆ partials_buffer_name()

◆ reduce_axis_stride()

◆ reduce_base_slot()

◆ reduce_block_dim()

◆ reduce_block_size_product()

◆ reduce_linear_thread_index()

◆ reduction_target()

◆ uses_register_partial()

◆ validate_before_dispatch()

◆ warp_combine_width()

Member Data Documentation

◆ node_

◆ reduction_buffers_