diff options
Diffstat (limited to 'Eigen/src/Core/arch/AVX/PacketMath.h')
| -rw-r--r-- | Eigen/src/Core/arch/AVX/PacketMath.h | 15 |
1 files changed, 12 insertions, 3 deletions
diff --git a/Eigen/src/Core/arch/AVX/PacketMath.h b/Eigen/src/Core/arch/AVX/PacketMath.h index f88e36024..3f94f85eb 100644 --- a/Eigen/src/Core/arch/AVX/PacketMath.h +++ b/Eigen/src/Core/arch/AVX/PacketMath.h @@ -117,14 +117,15 @@ template<> struct unpacket_traits<Packet8f> { typedef float type; typedef Packet4f half; typedef Packet8i integer_packet; - enum {size=8, alignment=Aligned32, vectorizable=true}; + typedef uint8_t mask_t; + enum {size=8, alignment=Aligned32, vectorizable=true, masked_load_available=true}; }; template<> struct unpacket_traits<Packet4d> { typedef double type; typedef Packet2d half; - enum {size=4, alignment=Aligned32, vectorizable=true}; + enum {size=4, alignment=Aligned32, vectorizable=true, masked_load_available=false}; }; -template<> struct unpacket_traits<Packet8i> { typedef int type; typedef Packet4i half; enum {size=8, alignment=Aligned32, vectorizable=false}; }; +template<> struct unpacket_traits<Packet8i> { typedef int type; typedef Packet4i half; enum {size=8, alignment=Aligned32, vectorizable=false, masked_load_available=false}; }; template<> EIGEN_STRONG_INLINE Packet8f pset1<Packet8f>(const float& from) { return _mm256_set1_ps(from); } template<> EIGEN_STRONG_INLINE Packet4d pset1<Packet4d>(const double& from) { return _mm256_set1_pd(from); } @@ -385,6 +386,14 @@ template<> EIGEN_STRONG_INLINE Packet8f ploadu<Packet8f>(const float* from) { EI template<> EIGEN_STRONG_INLINE Packet4d ploadu<Packet4d>(const double* from) { EIGEN_DEBUG_UNALIGNED_LOAD return _mm256_loadu_pd(from); } template<> EIGEN_STRONG_INLINE Packet8i ploadu<Packet8i>(const int* from) { EIGEN_DEBUG_UNALIGNED_LOAD return _mm256_loadu_si256(reinterpret_cast<const __m256i*>(from)); } +template<> EIGEN_STRONG_INLINE Packet8f ploadu<Packet8f>(const float* from, uint8_t umask) { + __m256i mask = _mm256_set1_epi8(static_cast<char>(umask)); + const __m256i bit_mask = _mm256_set_epi32(0xffffff7f, 0xffffffbf, 0xffffffdf, 0xffffffef, 0xfffffff7, 0xfffffffb, 0xfffffffd, 0xfffffffe); + mask = _mm256_or_si256(mask, bit_mask); + mask = _mm256_cmpeq_epi32(mask, _mm256_set1_epi32(0xffffffff)); + EIGEN_DEBUG_UNALIGNED_LOAD return _mm256_maskload_ps(from, mask); +} + // Loads 4 floats from memory a returns the packet {a0, a0 a1, a1, a2, a2, a3, a3} template<> EIGEN_STRONG_INLINE Packet8f ploaddup<Packet8f>(const float* from) { |