Class muda::DeviceReduce
ClassList > muda > DeviceReduce
Inherits the following classes: muda::CubWrapper
Public Types inherited from muda::LaunchBase
See muda::LaunchBase
Type | Name |
---|---|
typedef T | derived_type |
Public Attributes inherited from muda::CubWrapper
See muda::CubWrapper
Type | Name |
---|---|
Stream * | m_muda_stream = = nullptr |
Public Functions
Type | Name |
---|---|
DeviceReduce & | ArgMax (InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | ArgMax (void * d_temp_storage, size_t & temp_storage_bytes, InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | ArgMin (InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | ArgMin (void * d_temp_storage, size_t & temp_storage_bytes, InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | Max (InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | Max (void * d_temp_storage, size_t & temp_storage_bytes, InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | Min (InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | Min (void * d_temp_storage, size_t & temp_storage_bytes, InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | Reduce (InputIteratorT d_in, OutputIteratorT d_out, int num_items, ReductionOpT reduction_op, T init) |
DeviceReduce & | Reduce (void * d_temp_storage, size_t & temp_storage_bytes, InputIteratorT d_in, OutputIteratorT d_out, int num_items, ReductionOpT reduction_op, T init) |
DeviceReduce & | ReduceByKey (KeysInputIteratorT d_keys_in, UniqueOutputIteratorT d_unique_out, ValuesInputIteratorT d_values_in, AggregatesOutputIteratorT d_aggregates_out, NumRunsOutputIteratorT d_num_runs_out, ReductionOpT reduction_op, int num_items) |
DeviceReduce & | ReduceByKey (void * d_temp_storage, size_t & temp_storage_bytes, KeysInputIteratorT d_keys_in, UniqueOutputIteratorT d_unique_out, ValuesInputIteratorT d_values_in, AggregatesOutputIteratorT d_aggregates_out, NumRunsOutputIteratorT d_num_runs_out, ReductionOpT reduction_op, int num_items) |
DeviceReduce & | Sum (InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
DeviceReduce & | Sum (void * d_temp_storage, size_t & temp_storage_bytes, InputIteratorT d_in, OutputIteratorT d_out, int num_items) |
Public Functions inherited from muda::CubWrapper
See muda::CubWrapper
Type | Name |
---|---|
CubWrapper (Stream & stream=Stream::Default()) |
|
void | kernel_name (std::string_view) = delete |
Public Functions inherited from muda::LaunchBase
See muda::LaunchBase
Type | Name |
---|---|
MUDA_GENERIC | LaunchBase (::cudaStream_t stream) |
T & | callback (const std::function< void(::cudaStream_t, ::cudaError)> & callback) |
T & | file_line (std::string_view file, int line) |
T & | kernel_name (std::string_view name) |
Next | next (Next n) |
Next | next (Args &&... args) |
T & | pop_range () |
T & | push_range (const std::string & name) |
T & | record (cudaEvent_t e, int flag=cudaEventRecordDefault) |
T & | record (ComputeGraphVar< cudaEvent_t > & e, const std::vector< ComputeGraphVarBase * > & vars) |
T & | record (ComputeGraphVar< cudaEvent_t > & e, ComputeGraphVar< ViewT > &... vars) |
T & | wait (cudaEvent_t e, int flag=cudaEventWaitDefault) |
T & | wait (const ComputeGraphVar< cudaEvent_t > & e, const std::vector< ComputeGraphVarBase * > & vars) |
T & | wait (const ComputeGraphVar< cudaEvent_t > & e, ComputeGraphVar< ViewT > &... vars) |
T & | wait () |
T & | when (cudaEvent_t e, int flag=cudaEventWaitDefault) |
~LaunchBase () |
Public Functions inherited from muda::LaunchCore
See muda::LaunchCore
Type | Name |
---|---|
MUDA_GENERIC | LaunchCore (::cudaStream_t stream) |
void | callback (const std::function< void(::cudaStream_t, ::cudaError)> & callback) |
void | init_stream (::cudaStream_t s) |
void | pop_range () |
void | push_range (const std::string & name) |
void | record (cudaEvent_t e, int flag=cudaEventRecordDefault) |
void | record (ComputeGraphVar< cudaEvent_t > & e, const std::vector< ComputeGraphVarBase * > & vars) |
void | record (ComputeGraphVar< cudaEvent_t > & e, ComputeGraphVar< ViewT > &... vars) |
void | wait (cudaEvent_t e, int flag=cudaEventWaitDefault) |
void | wait (const ComputeGraphVar< cudaEvent_t > & e, const std::vector< ComputeGraphVarBase * > & vars) |
void | wait (const ComputeGraphVar< cudaEvent_t > & e, ComputeGraphVar< ViewT > &... vars) |
void | wait () |
void | when (cudaEvent_t e, int flag=cudaEventWaitDefault) |
~LaunchCore () |
Public Static Functions inherited from muda::LaunchCore
See muda::LaunchCore
Type | Name |
---|---|
void | file_line (std::string_view file, int line) |
void | kernel_name (std::string_view name) |
void | wait_device () |
void | wait_event (cudaEvent_t event) |
void | wait_stream (::cudaStream_t stream) |
Protected Types inherited from muda::LaunchCore
See muda::LaunchCore
Type | Name |
---|---|
typedef std::shared_ptr< T > | S |
Protected Attributes inherited from muda::LaunchCore
See muda::LaunchCore
Type | Name |
---|---|
::cudaStream_t | m_stream |
Protected Functions inherited from muda::CubWrapper
See muda::CubWrapper
Type | Name |
---|---|
std::byte * | prepare_buffer (size_t reqSize) |
Protected Functions inherited from muda::LaunchBase
See muda::LaunchBase
Type | Name |
---|---|
T & | pop_kernel_label () |
Protected Functions inherited from muda::LaunchCore
See muda::LaunchCore
Type | Name |
---|---|
MUDA_HOST void | pop_kernel_label () |
MUDA_GENERIC::cudaStream_t | stream () const |
Public Functions Documentation
function ArgMax [1/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::ArgMax (
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function ArgMax [2/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::ArgMax (
void * d_temp_storage,
size_t & temp_storage_bytes,
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function ArgMin [1/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::ArgMin (
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function ArgMin [2/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::ArgMin (
void * d_temp_storage,
size_t & temp_storage_bytes,
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function Max [1/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::Max (
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function Max [2/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::Max (
void * d_temp_storage,
size_t & temp_storage_bytes,
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function Min [1/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::Min (
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function Min [2/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::Min (
void * d_temp_storage,
size_t & temp_storage_bytes,
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function Reduce [1/2]
template<typename InputIteratorT, typename OutputIteratorT, typename ReductionOpT, typename T>
inline DeviceReduce & muda::DeviceReduce::Reduce (
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items,
ReductionOpT reduction_op,
T init
)
function Reduce [2/2]
template<typename InputIteratorT, typename OutputIteratorT, typename ReductionOpT, typename T>
inline DeviceReduce & muda::DeviceReduce::Reduce (
void * d_temp_storage,
size_t & temp_storage_bytes,
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items,
ReductionOpT reduction_op,
T init
)
function ReduceByKey [1/2]
template<typename KeysInputIteratorT, typename UniqueOutputIteratorT, typename ValuesInputIteratorT, typename AggregatesOutputIteratorT, typename NumRunsOutputIteratorT, typename ReductionOpT>
inline DeviceReduce & muda::DeviceReduce::ReduceByKey (
KeysInputIteratorT d_keys_in,
UniqueOutputIteratorT d_unique_out,
ValuesInputIteratorT d_values_in,
AggregatesOutputIteratorT d_aggregates_out,
NumRunsOutputIteratorT d_num_runs_out,
ReductionOpT reduction_op,
int num_items
)
function ReduceByKey [2/2]
template<typename KeysInputIteratorT, typename UniqueOutputIteratorT, typename ValuesInputIteratorT, typename AggregatesOutputIteratorT, typename NumRunsOutputIteratorT, typename ReductionOpT>
inline DeviceReduce & muda::DeviceReduce::ReduceByKey (
void * d_temp_storage,
size_t & temp_storage_bytes,
KeysInputIteratorT d_keys_in,
UniqueOutputIteratorT d_unique_out,
ValuesInputIteratorT d_values_in,
AggregatesOutputIteratorT d_aggregates_out,
NumRunsOutputIteratorT d_num_runs_out,
ReductionOpT reduction_op,
int num_items
)
function Sum [1/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::Sum (
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
function Sum [2/2]
template<typename InputIteratorT, typename OutputIteratorT>
inline DeviceReduce & muda::DeviceReduce::Sum (
void * d_temp_storage,
size_t & temp_storage_bytes,
InputIteratorT d_in,
OutputIteratorT d_out,
int num_items
)
The documentation for this class was generated from the following file src/muda/cub/device/device_reduce.h