mirror of
https://gitlab.com/libeigen/eigen.git
synced 2026-04-10 11:34:33 +08:00
re-enable fast pset1-pstore by introducing a new higher level pstore1 function
This commit is contained in:
@@ -305,6 +305,19 @@ template<> EIGEN_STRONG_INLINE void pstoreu<double>(double* to, const Packet2d&
|
||||
template<> EIGEN_STRONG_INLINE void pstoreu<float>(float* to, const Packet4f& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu((double*)to, _mm_castps_pd(from)); }
|
||||
template<> EIGEN_STRONG_INLINE void pstoreu<int>(int* to, const Packet4i& from) { EIGEN_DEBUG_UNALIGNED_STORE pstoreu((double*)to, _mm_castsi128_pd(from)); }
|
||||
|
||||
// some compilers might be tempted to perform multiple moves instead of using a vector path.
|
||||
template<> EIGEN_STRONG_INLINE void pstore1<Packet4f>(float* to, const float& a)
|
||||
{
|
||||
Packet4f pa = _mm_set_ss(a);
|
||||
pstore(to, vec4f_swizzle1(pa,0,0,0,0));
|
||||
}
|
||||
// some compilers might be tempted to perform multiple moves instead of using a vector path.
|
||||
template<> EIGEN_STRONG_INLINE void pstore1<Packet2d>(double* to, const double& a)
|
||||
{
|
||||
Packet2d pa = _mm_set_sd(a);
|
||||
pstore(to, vec2d_swizzle1(pa,0,0));
|
||||
}
|
||||
|
||||
template<> EIGEN_STRONG_INLINE void prefetch<float>(const float* addr) { _mm_prefetch((const char*)(addr), _MM_HINT_T0); }
|
||||
template<> EIGEN_STRONG_INLINE void prefetch<double>(const double* addr) { _mm_prefetch((const char*)(addr), _MM_HINT_T0); }
|
||||
template<> EIGEN_STRONG_INLINE void prefetch<int>(const int* addr) { _mm_prefetch((const char*)(addr), _MM_HINT_T0); }
|
||||
|
||||
Reference in New Issue
Block a user