diff options
Diffstat (limited to 'Eigen/src/Core/arch/SSE/PacketMath.h')
-rwxr-xr-x | Eigen/src/Core/arch/SSE/PacketMath.h | 15 |
1 files changed, 15 insertions, 0 deletions
diff --git a/Eigen/src/Core/arch/SSE/PacketMath.h b/Eigen/src/Core/arch/SSE/PacketMath.h index 70839d68d..baad692e3 100755 --- a/Eigen/src/Core/arch/SSE/PacketMath.h +++ b/Eigen/src/Core/arch/SSE/PacketMath.h @@ -162,6 +162,11 @@ template<> struct unpacket_traits<Packet4f> { typedef float type; enum {size=4, template<> struct unpacket_traits<Packet2d> { typedef double type; enum {size=2, alignment=Aligned16}; typedef Packet2d half; }; template<> struct unpacket_traits<Packet4i> { typedef int type; enum {size=4, alignment=Aligned16}; typedef Packet4i half; }; +#ifndef EIGEN_VECTORIZE_AVX +template<> struct scalar_div_cost<float,true> { enum { value = 7 }; }; +template<> struct scalar_div_cost<double,true> { enum { value = 8 }; }; +#endif + #if EIGEN_COMP_MSVC==1500 // Workaround MSVC 9 internal compiler error. // TODO: It has been detected with win64 builds (amd64), so let's check whether it also happens in 32bits+SSE mode @@ -813,6 +818,16 @@ template<> EIGEN_STRONG_INLINE Packet2d pblend(const Selector<2>& ifPacket, cons #endif } +// Scalar path for pmadd with FMA to ensure consistency with vectorized path. +#ifdef __FMA__ +template<> EIGEN_STRONG_INLINE float pmadd(const float& a, const float& b, const float& c) { + return ::fmaf(a,b,c); +} +template<> EIGEN_STRONG_INLINE double pmadd(const double& a, const double& b, const double& c) { + return ::fma(a,b,c); +} +#endif + } // end namespace internal } // end namespace Eigen |