![]() |
alpaka
Abstraction Library for Parallel Kernel Acceleration
|
#include <WorkerGroup.hpp>
Public Member Functions | |
| constexpr | WorkerGroup (T_ThreadIdxOrOrigin threadIdxOrOrigin, T_NumThreadsOrUnit numThreadsOrUnit, T_MultiDimensional=MultiDimensional< true >{}) |
| WorkerGroup constructor. | |
| constexpr auto | idx (auto const &acc) const |
| constexpr auto | size (auto const &acc) const |
Private Member Functions | |
| constexpr auto | getThreadSpace (auto const &acc) const |
| get the thread configuration | |
| constexpr auto | getThreadSpace (auto const &acc) const |
| get the thread configuration | |
Private Attributes | |
| T_NumThreadsOrUnit | m_numThreadsOrUnit |
| T_ThreadIdxOrOrigin | m_threadIdxOrOrigin |
Friends | |
| template<typename T_ThreadGroup, typename T_ThreadIdxOrOriginRange> | |
| struct | DomainSpec |
Definition at line 27 of file WorkerGroup.hpp.
|
inlineconstexpr |
WorkerGroup constructor.
| threadIdxOrOrigin | the index of the thread or onAcc::origin |
| numThreadsOrUnit | the number of threads or the onAcc::unit |
| multiDimensional | keep the dimensionality for both input parameters, if 'linearized' is used the workgroup will be reduced to a one dimensional group. |
Definition at line 36 of file WorkerGroup.hpp.
|
inlineconstexprprivate |
get the thread configuration
Implementation specialization for lazy evaluated acc properties based on an origin and unit.
Definition at line 78 of file WorkerGroup.hpp.
|
inlineconstexprprivate |
get the thread configuration
Implementation specialization for vectors.
Definition at line 63 of file WorkerGroup.hpp.
|
inlineconstexpr |
Definition at line 50 of file WorkerGroup.hpp.
|
inlineconstexpr |
Definition at line 45 of file WorkerGroup.hpp.
|
friend |
Definition at line 57 of file WorkerGroup.hpp.
|
private |
Definition at line 100 of file WorkerGroup.hpp.
|
private |
Definition at line 99 of file WorkerGroup.hpp.