aboutsummaryrefslogtreecommitdiffhomepage
diff options
context:
space:
mode:
-rw-r--r--Eigen/src/Core/arch/AVX512/PacketMath.h10
1 files changed, 4 insertions, 6 deletions
diff --git a/Eigen/src/Core/arch/AVX512/PacketMath.h b/Eigen/src/Core/arch/AVX512/PacketMath.h
index 3842f576b..c111fd7f0 100644
--- a/Eigen/src/Core/arch/AVX512/PacketMath.h
+++ b/Eigen/src/Core/arch/AVX512/PacketMath.h
@@ -526,13 +526,11 @@ EIGEN_STRONG_INLINE Packet8d ploaddup<Packet8d>(const double* from) {
// {a0, a0 a0, a0, a1, a1, a1, a1, a2, a2, a2, a2, a3, a3, a3, a3}
template <>
EIGEN_STRONG_INLINE Packet16f ploadquad<Packet16f>(const float* from) {
- Packet16f tmp = _mm512_undefined_ps();
- tmp = _mm512_insertf32x4(tmp, _mm_load_ps1(from), 0);
- tmp = _mm512_insertf32x4(tmp, _mm_load_ps1(from + 1), 1);
- tmp = _mm512_insertf32x4(tmp, _mm_load_ps1(from + 2), 2);
- tmp = _mm512_insertf32x4(tmp, _mm_load_ps1(from + 3), 3);
- return tmp;
+ Packet16f tmp = _mm512_castps128_ps512(pload<Packet4f>(from));
+ const Packet16i scatter_mask = _mm512_set_epi32(3,3,3,3, 2,2,2,2, 1,1,1,1, 0,0,0,0);
+ return _mm512_permutexvar_ps(scatter_mask, tmp);
}
+
// Loads 2 doubles from memory a returns the packet
// {a0, a0 a0, a0, a1, a1, a1, a1}
template <>