reduce code generation and minor speed up

This commit is contained in:
Gael Guennebaud 2010-07-06 19:15:02 +02:00
parent d6454788d9
commit e04c3f2cc0

View File

@ -80,9 +80,9 @@ static void run(Index rows, Index cols, Index depth,
Index mc = std::min(rows,blocking.mc()); // cache block size along the M direction
//Index nc = blocking.nc(); // cache block size along the N direction
ei_gemm_pack_lhs<Scalar, Index, Blocking::mr, LhsStorageOrder, ConjugateLhs> pack_lhs;
ei_gemm_pack_rhs<Scalar, Index, Blocking::nr, RhsStorageOrder, ConjugateRhs> pack_rhs;
ei_gebp_kernel<Scalar, Index, Blocking::mr, Blocking::nr> gebp;
ei_gemm_pack_lhs<Scalar, Index, Blocking::mr, LhsStorageOrder> pack_lhs;
ei_gemm_pack_rhs<Scalar, Index, Blocking::nr, RhsStorageOrder> pack_rhs;
ei_gebp_kernel<Scalar, Index, Blocking::mr, Blocking::nr, ConjugateLhs, ConjugateRhs> gebp;
#ifdef EIGEN_HAS_OPENMP
if(info)
@ -234,7 +234,7 @@ struct ei_gemm_functor
{
if(cols==-1)
cols = m_rhs.cols();
Gemm::run(rows, cols, m_lhs.cols(),
(const Scalar*)&(m_lhs.const_cast_derived().coeffRef(row,0)), m_lhs.outerStride(),
(const Scalar*)&(m_rhs.const_cast_derived().coeffRef(0,col)), m_rhs.outerStride(),