Eigen::internal::TensorExecutor< Expression, Device, Vectorizable, Tiling > Class Template Reference

#include <TensorExecutor.h>

Public Types
typedef Expression::Index	StorageIndex

Static Public Member Functions
static EIGEN_STRONG_INLINE void	run (const Expression &expr, const Device &device=DefaultDevice())

Detailed Description

template<typename Expression, typename Device, bool Vectorizable, TiledEvaluation Tiling>
class Eigen::internal::TensorExecutor< Expression, Device, Vectorizable, Tiling >

Default strategy: the expression is evaluated sequentially with a single cpu thread, without vectorization and block evaluation.

Member Typedef Documentation

StorageIndex

template<typename Expression , typename Device , bool Vectorizable, TiledEvaluation Tiling>

typedef Expression::Index Eigen::internal::TensorExecutor< Expression, Device, Vectorizable, Tiling >::StorageIndex

Member Function Documentation

run()

template<typename Expression , typename Device , bool Vectorizable, TiledEvaluation Tiling>

static EIGEN_STRONG_INLINE void Eigen::internal::TensorExecutor< Expression, Device, Vectorizable, Tiling >::run ( const Expression &  expr, const Device &  device = DefaultDevice()  )

inlinestatic

                                                                                                      {
    TensorEvaluator<Expression, Device> evaluator(expr, device);
    const bool needs_assign = evaluator.evalSubExprsIfNeeded(NULL);
    if (needs_assign) {
      const StorageIndex size = array_prod(evaluator.dimensions());
      for (StorageIndex i = 0; i < size; ++i) {
        evaluator.evalScalar(i);
      }
    }
    evaluator.cleanup();
  }

References Eigen::internal::array_prod(), i, and size.

Referenced by Eigen::TensorDevice< ExpressionType, DeviceType >::operator+=(), Eigen::TensorDevice< ExpressionType, DeviceType >::operator-=(), Eigen::TensorBase< Derived, AccessLevel >::operator=(), Eigen::TensorDevice< ExpressionType, DeviceType >::operator=(), Eigen::Tensor< Scalar_, NumIndices_, Options_, IndexType_ >::operator=(), Eigen::TensorEvaluator< const TensorConvolutionOp< Indices, InputArgType, KernelArgType >, Device >::preloadKernel(), Eigen::TensorEvaluator< const TensorConvolutionOp< Indices, InputArgType, KernelArgType >, Eigen::SyclDevice >::preloadKernel(), Eigen::Tensor< Scalar_, NumIndices_, Options_, IndexType_ >::Tensor(), and Eigen::TensorFixedSize< Scalar_, Dimensions_, Options_, IndexType >::TensorFixedSize().

---

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

• TensorExecutor.h