/external/speex/libspeexdsp/ |
D | resample_sse.h | 50 sum = _mm_add_ps(sum, _mm_movehl_ps(sum, sum)); in inner_product_single() 68 sum = _mm_add_ps(sum, _mm_movehl_ps(sum, sum)); in interpolate_product_single() 88 sum = _mm_add_pd(sum, _mm_cvtps_pd(_mm_movehl_ps(t, t))); in inner_product_double() 92 sum = _mm_add_pd(sum, _mm_cvtps_pd(_mm_movehl_ps(t, t))); in inner_product_double() 108 __m128d f2 = _mm_cvtps_pd(_mm_movehl_ps(f,f)); in interpolate_product_double() 114 sum2 = _mm_add_pd(sum2, _mm_cvtps_pd(_mm_movehl_ps(t, t))); in interpolate_product_double() 118 sum2 = _mm_add_pd(sum2, _mm_cvtps_pd(_mm_movehl_ps(t, t))); in interpolate_product_double()
|
/external/XNNPACK/src/f32-igemm/gen/ |
D | 8x8-minmax-fma3-broadcast.c | 250 vacc7x0123 = _mm_movehl_ps(vacc7x0123, vacc7x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast() 251 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast() 252 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast() 253 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast() 254 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast() 255 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast() 256 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast() 257 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_igemm_minmax_ukernel_8x8__fma3_broadcast()
|
D | 7x8-minmax-fma3-broadcast.c | 228 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_igemm_minmax_ukernel_7x8__fma3_broadcast() 229 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_igemm_minmax_ukernel_7x8__fma3_broadcast() 230 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_igemm_minmax_ukernel_7x8__fma3_broadcast() 231 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_igemm_minmax_ukernel_7x8__fma3_broadcast() 232 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_igemm_minmax_ukernel_7x8__fma3_broadcast() 233 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_igemm_minmax_ukernel_7x8__fma3_broadcast() 234 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_igemm_minmax_ukernel_7x8__fma3_broadcast()
|
D | 7x8-minmax-avx-broadcast.c | 228 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_igemm_minmax_ukernel_7x8__avx_broadcast() 229 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_igemm_minmax_ukernel_7x8__avx_broadcast() 230 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_igemm_minmax_ukernel_7x8__avx_broadcast() 231 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_igemm_minmax_ukernel_7x8__avx_broadcast() 232 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_igemm_minmax_ukernel_7x8__avx_broadcast() 233 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_igemm_minmax_ukernel_7x8__avx_broadcast() 234 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_igemm_minmax_ukernel_7x8__avx_broadcast()
|
D | 6x8-minmax-avx-broadcast.c | 206 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_igemm_minmax_ukernel_6x8__avx_broadcast() 207 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_igemm_minmax_ukernel_6x8__avx_broadcast() 208 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_igemm_minmax_ukernel_6x8__avx_broadcast() 209 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_igemm_minmax_ukernel_6x8__avx_broadcast() 210 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_igemm_minmax_ukernel_6x8__avx_broadcast() 211 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_igemm_minmax_ukernel_6x8__avx_broadcast()
|
D | 6x8-minmax-fma3-broadcast.c | 206 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_igemm_minmax_ukernel_6x8__fma3_broadcast() 207 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_igemm_minmax_ukernel_6x8__fma3_broadcast() 208 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_igemm_minmax_ukernel_6x8__fma3_broadcast() 209 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_igemm_minmax_ukernel_6x8__fma3_broadcast() 210 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_igemm_minmax_ukernel_6x8__fma3_broadcast() 211 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_igemm_minmax_ukernel_6x8__fma3_broadcast()
|
D | 5x8-minmax-fma3-broadcast.c | 184 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_igemm_minmax_ukernel_5x8__fma3_broadcast() 185 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_igemm_minmax_ukernel_5x8__fma3_broadcast() 186 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_igemm_minmax_ukernel_5x8__fma3_broadcast() 187 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_igemm_minmax_ukernel_5x8__fma3_broadcast() 188 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_igemm_minmax_ukernel_5x8__fma3_broadcast()
|
D | 5x8-minmax-avx-broadcast.c | 184 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_igemm_minmax_ukernel_5x8__avx_broadcast() 185 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_igemm_minmax_ukernel_5x8__avx_broadcast() 186 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_igemm_minmax_ukernel_5x8__avx_broadcast() 187 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_igemm_minmax_ukernel_5x8__avx_broadcast() 188 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_igemm_minmax_ukernel_5x8__avx_broadcast()
|
/external/XNNPACK/src/f32-gemm/gen-inc/ |
D | 8x8inc-minmax-fma3-broadcast.c | 225 vacc7x0123 = _mm_movehl_ps(vacc7x0123, vacc7x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast() 226 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast() 227 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast() 228 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast() 229 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast() 230 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast() 231 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast() 232 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_8x8__fma3_broadcast()
|
D | 7x8inc-minmax-avx-broadcast.c | 205 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__avx_broadcast() 206 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__avx_broadcast() 207 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__avx_broadcast() 208 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__avx_broadcast() 209 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__avx_broadcast() 210 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__avx_broadcast() 211 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__avx_broadcast()
|
D | 7x8inc-minmax-fma3-broadcast.c | 205 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__fma3_broadcast() 206 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__fma3_broadcast() 207 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__fma3_broadcast() 208 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__fma3_broadcast() 209 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__fma3_broadcast() 210 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__fma3_broadcast() 211 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_7x8__fma3_broadcast()
|
D | 6x8inc-minmax-avx-broadcast.c | 185 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__avx_broadcast() 186 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__avx_broadcast() 187 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__avx_broadcast() 188 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__avx_broadcast() 189 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__avx_broadcast() 190 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__avx_broadcast()
|
D | 6x8inc-minmax-fma3-broadcast.c | 185 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__fma3_broadcast() 186 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__fma3_broadcast() 187 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__fma3_broadcast() 188 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__fma3_broadcast() 189 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__fma3_broadcast() 190 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_6x8__fma3_broadcast()
|
D | 5x8inc-minmax-fma3-broadcast.c | 165 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__fma3_broadcast() 166 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__fma3_broadcast() 167 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__fma3_broadcast() 168 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__fma3_broadcast() 169 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__fma3_broadcast()
|
D | 5x8inc-minmax-avx-broadcast.c | 165 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__avx_broadcast() 166 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__avx_broadcast() 167 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__avx_broadcast() 168 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__avx_broadcast() 169 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__avx_broadcast()
|
D | 5x8inc-minmax-sse-load1.c | 186 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__sse_load1() 187 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__sse_load1() 188 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__sse_load1() 189 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__sse_load1() 190 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemminc_minmax_ukernel_5x8__sse_load1()
|
/external/XNNPACK/src/f32-gemm/gen/ |
D | 8x8-minmax-fma3-broadcast.c | 223 vacc7x0123 = _mm_movehl_ps(vacc7x0123, vacc7x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast() 224 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast() 225 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast() 226 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast() 227 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast() 228 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast() 229 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast() 230 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_8x8__fma3_broadcast()
|
D | 7x8-minmax-fma3-broadcast.c | 203 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_gemm_minmax_ukernel_7x8__fma3_broadcast() 204 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemm_minmax_ukernel_7x8__fma3_broadcast() 205 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemm_minmax_ukernel_7x8__fma3_broadcast() 206 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_7x8__fma3_broadcast() 207 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_7x8__fma3_broadcast() 208 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_7x8__fma3_broadcast() 209 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_7x8__fma3_broadcast()
|
D | 7x8-minmax-avx-broadcast.c | 203 vacc6x0123 = _mm_movehl_ps(vacc6x0123, vacc6x0123); in xnn_f32_gemm_minmax_ukernel_7x8__avx_broadcast() 204 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemm_minmax_ukernel_7x8__avx_broadcast() 205 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemm_minmax_ukernel_7x8__avx_broadcast() 206 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_7x8__avx_broadcast() 207 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_7x8__avx_broadcast() 208 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_7x8__avx_broadcast() 209 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_7x8__avx_broadcast()
|
D | 6x8-minmax-avx-broadcast.c | 183 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemm_minmax_ukernel_6x8__avx_broadcast() 184 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemm_minmax_ukernel_6x8__avx_broadcast() 185 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_6x8__avx_broadcast() 186 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_6x8__avx_broadcast() 187 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_6x8__avx_broadcast() 188 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_6x8__avx_broadcast()
|
D | 6x8-minmax-fma3-broadcast.c | 183 vacc5x0123 = _mm_movehl_ps(vacc5x0123, vacc5x0123); in xnn_f32_gemm_minmax_ukernel_6x8__fma3_broadcast() 184 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemm_minmax_ukernel_6x8__fma3_broadcast() 185 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_6x8__fma3_broadcast() 186 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_6x8__fma3_broadcast() 187 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_6x8__fma3_broadcast() 188 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_6x8__fma3_broadcast()
|
D | 5x8-minmax-fma3-broadcast.c | 163 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemm_minmax_ukernel_5x8__fma3_broadcast() 164 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_5x8__fma3_broadcast() 165 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_5x8__fma3_broadcast() 166 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_5x8__fma3_broadcast() 167 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_5x8__fma3_broadcast()
|
D | 5x8-minmax-avx-broadcast.c | 163 vacc4x0123 = _mm_movehl_ps(vacc4x0123, vacc4x0123); in xnn_f32_gemm_minmax_ukernel_5x8__avx_broadcast() 164 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_5x8__avx_broadcast() 165 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_5x8__avx_broadcast() 166 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_5x8__avx_broadcast() 167 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_5x8__avx_broadcast()
|
D | 4x8-minmax-avx-broadcast.c | 143 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_4x8__avx_broadcast() 144 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_4x8__avx_broadcast() 145 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_4x8__avx_broadcast() 146 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_4x8__avx_broadcast()
|
D | 4x8-minmax-fma3-broadcast.c | 143 vacc3x0123 = _mm_movehl_ps(vacc3x0123, vacc3x0123); in xnn_f32_gemm_minmax_ukernel_4x8__fma3_broadcast() 144 vacc2x0123 = _mm_movehl_ps(vacc2x0123, vacc2x0123); in xnn_f32_gemm_minmax_ukernel_4x8__fma3_broadcast() 145 vacc1x0123 = _mm_movehl_ps(vacc1x0123, vacc1x0123); in xnn_f32_gemm_minmax_ukernel_4x8__fma3_broadcast() 146 vacc0x0123 = _mm_movehl_ps(vacc0x0123, vacc0x0123); in xnn_f32_gemm_minmax_ukernel_4x8__fma3_broadcast()
|