diff options
author | Gael Guennebaud <g.gael@free.fr> | 2014-04-25 10:56:18 +0200 |
---|---|---|
committer | Gael Guennebaud <g.gael@free.fr> | 2014-04-25 10:56:18 +0200 |
commit | 3d8d0f6269478a06f4fcbd4b838c8e9b9d7e9d62 (patch) | |
tree | 9f55e81498c1acc61ea0c10bce0ead69a71875d1 /Eigen/src/Core/arch/NEON/PacketMath.h | |
parent | b0e19db1cf462a07e25429d4f04f7d8e858f670f (diff) |
Enable vectorization of pack_rhs with a column-major RHS.
Rename and generalize Kernel<*> to PacketBlock<*,N>.
Diffstat (limited to 'Eigen/src/Core/arch/NEON/PacketMath.h')
-rw-r--r-- | Eigen/src/Core/arch/NEON/PacketMath.h | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/Eigen/src/Core/arch/NEON/PacketMath.h b/Eigen/src/Core/arch/NEON/PacketMath.h index 6426623cf..37fb840d4 100644 --- a/Eigen/src/Core/arch/NEON/PacketMath.h +++ b/Eigen/src/Core/arch/NEON/PacketMath.h @@ -452,7 +452,7 @@ PALIGN_NEON(3,Packet4i,vextq_s32) #undef PALIGN_NEON template<> EIGEN_DEVICE_FUNC inline void -ptranspose(Kernel<Packet4f>& kernel) { +ptranspose(PacketBlock<Packet4f,4>& kernel) { float32x4x2_t tmp1 = vzipq_f32(kernel.packet[0], kernel.packet[1]); float32x4x2_t tmp2 = vzipq_f32(kernel.packet[2], kernel.packet[3]); @@ -463,7 +463,7 @@ ptranspose(Kernel<Packet4f>& kernel) { } template<> EIGEN_DEVICE_FUNC inline void -ptranspose(Kernel<Packet4i>& kernel) { +ptranspose(PacketBlock<Packet4i,4>& kernel) { int32x4x2_t tmp1 = vzipq_s32(kernel.packet[0], kernel.packet[1]); int32x4x2_t tmp2 = vzipq_s32(kernel.packet[2], kernel.packet[3]); kernel.packet[0] = vcombine_s32(vget_low_s32(tmp1.val[0]), vget_low_s32(tmp2.val[0])); |