diff options
| author |  Benoit Steiner <benoit.steiner.goog@gmail.com> | 2014-03-26 15:58:13 -0700 |
|---|---|---|
| committer | Benoit Steiner <benoit.steiner.goog@gmail.com> | 2014-03-26 15:58:13 -0700 |
| commit | e45a6bed4548fb2821efd3d78471bf1bd2a7d4ef (patch) | |
| tree | 5266760caa12e50808a421ef76bd4181100045ec | |
| parent | cc73164aa84c711c5581d8e163e1bd496c768b8a (diff) | |
Specialized the pload1 packet primitive for Packet8f and Packet4d in order to take advantage of the vbroadcastss and vbroadcastsd instructions whenever possible.
| -rw-r--r-- | Eigen/src/Core/arch/AVX/PacketMath.h | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/Eigen/src/Core/arch/AVX/PacketMath.h b/Eigen/src/Core/arch/AVX/PacketMath.h index 26cc996db..98f8c5416 100644 --- a/Eigen/src/Core/arch/AVX/PacketMath.h +++ b/Eigen/src/Core/arch/AVX/PacketMath.h @@ -87,6 +87,9 @@ template<> EIGEN_STRONG_INLINE Packet8f pset1<Packet8f>(const float& from) { re template<> EIGEN_STRONG_INLINE Packet4d pset1<Packet4d>(const double& from) { return _mm256_set1_pd(from); } template<> EIGEN_STRONG_INLINE Packet8i pset1<Packet8i>(const int& from) { return _mm256_set1_epi32(from); } +template<> EIGEN_STRONG_INLINE Packet8f pload1<Packet8f>(const float* from) { return _mm256_broadcast_ss(from); } +template<> EIGEN_STRONG_INLINE Packet4d pload1<Packet4d>(const double* from) { return _mm256_broadcast_sd(from); } + template<> EIGEN_STRONG_INLINE Packet8f plset<float>(const float& a) { return _mm256_add_ps(_mm256_set1_ps(a), _mm256_set_ps(7.0,6.0,5.0,4.0,3.0,2.0,1.0,0.0)); } template<> EIGEN_STRONG_INLINE Packet4d plset<double>(const double& a) { return _mm256_add_pd(_mm256_set1_pd(a), _mm256_set_pd(3.0,2.0,1.0,0.0)); } |