Remove packet ops pinsertfirst and pinsertlast that are only used in a single place, and can be replaced by other ops when constructing the first/final packet in linspaced_op_impl::packetOp.

I cannot measure any performance changes for SSE, AVX, or AVX512.

name                                 old time/op             new time/op             delta
BM_LinSpace<float>/1                 1.63ns ± 0%             1.63ns ± 0%   ~             (p=0.762 n=5+5)
BM_LinSpace<float>/8                 4.92ns ± 3%             4.89ns ± 3%   ~             (p=0.421 n=5+5)
BM_LinSpace<float>/64                34.6ns ± 0%             34.6ns ± 0%   ~             (p=0.841 n=5+5)
BM_LinSpace<float>/512                217ns ± 0%              217ns ± 0%   ~             (p=0.421 n=5+5)
BM_LinSpace<float>/4k                1.68µs ± 0%             1.68µs ± 0%   ~             (p=1.000 n=5+5)
BM_LinSpace<float>/32k               13.3µs ± 0%             13.3µs ± 0%   ~             (p=0.905 n=5+4)
BM_LinSpace<float>/256k               107µs ± 0%              107µs ± 0%   ~             (p=0.841 n=5+5)
BM_LinSpace<float>/1M                 427µs ± 0%              427µs ± 0%   ~             (p=0.690 n=5+5)

2 files changed, 0 insertions, 54 deletions

diff --git a/Eigen/src/Core/arch/AVX512/Complex.h b/Eigen/src/Core/arch/AVX512/Complex.h
index 75bdf57f1..dc2ae0a35 100644
--- a/Eigen/src/Core/arch/AVX512/Complex.h
+++ b/Eigen/src/Core/arch/AVX512/Complex.h
@@ -440,30 +440,6 @@ ptranspose(PacketBlock<Packet4cd,4>& kernel) {
   kernel.packet[0] = Packet4cd(_mm512_shuffle_f64x2(T0, T2, EIGEN_SSE_SHUFFLE_MASK(0,2,0,2))); // [a0 b0 c0 d0]
 }
 
-template<> EIGEN_STRONG_INLINE Packet8cf pinsertfirst(const Packet8cf& a, std::complex<float> b)
-{
-  Packet2cf tmp = Packet2cf(_mm512_extractf32x4_ps(a.v,0));
-  tmp = pinsertfirst(tmp, b);
-  return Packet8cf( _mm512_insertf32x4(a.v, tmp.v, 0) );
-}
-
-template<> EIGEN_STRONG_INLINE Packet4cd pinsertfirst(const Packet4cd& a, std::complex<double> b)
-{
-  return Packet4cd(_mm512_castsi512_pd( _mm512_inserti32x4(_mm512_castpd_si512(a.v), _mm_castpd_si128(pset1<Packet1cd>(b).v), 0) ));
-}
-
-template<> EIGEN_STRONG_INLINE Packet8cf pinsertlast(const Packet8cf& a, std::complex<float> b)
-{
-  Packet2cf tmp = Packet2cf(_mm512_extractf32x4_ps(a.v,3) );
-  tmp = pinsertlast(tmp, b);
-  return Packet8cf( _mm512_insertf32x4(a.v, tmp.v, 3) );
-}
-
-template<> EIGEN_STRONG_INLINE Packet4cd pinsertlast(const Packet4cd& a, std::complex<double> b)
-{
-  return Packet4cd(_mm512_castsi512_pd( _mm512_inserti32x4(_mm512_castpd_si512(a.v), _mm_castpd_si128(pset1<Packet1cd>(b).v), 3) ));
-}
-
 } // end namespace internal
 
 } // end namespace Eigen
diff --git a/Eigen/src/Core/arch/AVX512/PacketMath.h b/Eigen/src/Core/arch/AVX512/PacketMath.h
index 346d1f06e..10a1d4adb 100644
--- a/Eigen/src/Core/arch/AVX512/PacketMath.h
+++ b/Eigen/src/Core/arch/AVX512/PacketMath.h
@@ -1194,26 +1194,6 @@ EIGEN_STRONG_INLINE Packet8d pblend(const Selector<8>& ifPacket,
   return _mm512_mask_blend_pd(m, elsePacket, thenPacket);
 }
 
-template<> EIGEN_STRONG_INLINE Packet16f pinsertfirst(const Packet16f& a, float b)
-{
-  return _mm512_mask_broadcastss_ps(a, (1), _mm_load_ss(&b));
-}
-
-template<> EIGEN_STRONG_INLINE Packet8d pinsertfirst(const Packet8d& a, double b)
-{
-  return _mm512_mask_broadcastsd_pd(a, (1), _mm_load_sd(&b));
-}
-
-template<> EIGEN_STRONG_INLINE Packet16f pinsertlast(const Packet16f& a, float b)
-{
-  return _mm512_mask_broadcastss_ps(a, (1<<15), _mm_load_ss(&b));
-}
-
-template<> EIGEN_STRONG_INLINE Packet8d pinsertlast(const Packet8d& a, double b)
-{
-  return _mm512_mask_broadcastsd_pd(a, (1<<7), _mm_load_sd(&b));
-}
-
 template<> EIGEN_STRONG_INLINE Packet16i pcast<Packet16f, Packet16i>(const Packet16f& a) {
   return _mm512_cvttps_epi32(a);
 }
@@ -1432,16 +1412,6 @@ template<> EIGEN_STRONG_INLINE Packet16h preverse(const Packet16h& a)
                                            _mm_shuffle_epi8(_mm256_extractf128_si256(a,0),m), 1);
 }
 
-template<> EIGEN_STRONG_INLINE Packet16h pinsertfirst(const Packet16h& a, Eigen::half b)
-{
-  return _mm256_insert_epi16(a,b.x,0);
-}
-
-template<> EIGEN_STRONG_INLINE Packet16h pinsertlast(const Packet16h& a, Eigen::half b)
-{
-  return _mm256_insert_epi16(a,b.x,15);
-}
-
 template<> EIGEN_STRONG_INLINE Packet16h pgather<Eigen::half, Packet16h>(const Eigen::half* from, Index stride)
 {
   return _mm256_set_epi16(