Home
last modified time | relevance | path

Searched refs:vtwo (Results 1 – 15 of 15) sorted by relevance

/external/XNNPACK/src/f32-sigmoid/gen/
Davx-rr2-p5-nr2-x80.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80() local
229 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
230 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
231 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
232 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
233 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
234 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
235 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
236 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
237 vr4 = _mm256_mul_ps(vr4, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr4, vd4))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x80()
[all …]
Davx-rr2-p5-nr2-x72.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72() local
212 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
213 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
214 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
215 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
216 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
217 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
218 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
219 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
220 vr4 = _mm256_mul_ps(vr4, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr4, vd4))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x72()
[all …]
Davx-rr2-p5-nr2-x64.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64() local
195 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
196 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
197 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
198 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
199 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
200 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
201 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
202 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
203 vr4 = _mm256_mul_ps(vr4, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr4, vd4))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x64()
[all …]
Davx-rr2-p5-nr2-x56.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56() local
178 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
179 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
180 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
181 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
182 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
183 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
184 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
185 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
186 vr4 = _mm256_mul_ps(vr4, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr4, vd4))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x56()
[all …]
Davx-rr2-p5-nr2-x40.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40() local
144 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
145 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
146 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
147 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
148 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
149 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
150 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
151 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
152 vr4 = _mm256_mul_ps(vr4, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr4, vd4))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x40()
[all …]
Davx-rr2-p5-nr2-x48.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48() local
161 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
162 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
163 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
164 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
165 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
166 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
167 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
168 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
169 vr4 = _mm256_mul_ps(vr4, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr4, vd4))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x48()
[all …]
Davx-rr2-p5-nr2-x32.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32() local
127 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
128 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
129 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
130 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
131 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
132 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
133 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
134 vr3 = _mm256_mul_ps(vr3, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr3, vd3))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
184 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x32()
[all …]
Davx-rr2-p5-nr2-x24.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24() local
110 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
111 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
112 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
113 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
114 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
115 vr2 = _mm256_mul_ps(vr2, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr2, vd2))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
161 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
162 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
200 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x24()
[all …]
Davx-rr2-p5-nr2-x16.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16() local
93 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
94 vr0 = _mm256_mul_ps(vr0, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr0, vd0))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
95 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
96 vr1 = _mm256_mul_ps(vr1, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr1, vd1))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
138 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
139 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
177 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
178 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x16()
Davx-rr2-p5-nr2-x8.c39 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x8() local
69 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x8()
70 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x8()
108 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x8()
109 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_f32_sigmoid_ukernel__avx_rr2_p5_nr2_x8()
/external/XNNPACK/src/math/
Dsigmoid-avx-rr2-p5-nr2.c37 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_math_f32_sigmoid__avx_rr2_p5_nr2() local
97 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_math_f32_sigmoid__avx_rr2_p5_nr2()
98 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_math_f32_sigmoid__avx_rr2_p5_nr2()
Dsigmoid-avx-rr2-p5-nr1.c37 const __m256 vtwo = _mm256_set1_ps(2.0f); in xnn_math_f32_sigmoid__avx_rr2_p5_nr1() local
97 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd))); in xnn_math_f32_sigmoid__avx_rr2_p5_nr1()
Dsigmoid-sse2-rr2-p5-nr1.c37 const __m128 vtwo = _mm_set1_ps(2.0f); in xnn_math_f32_sigmoid__sse2_rr2_p5_nr1() local
95 vr = _mm_mul_ps(vr, _mm_sub_ps(vtwo, _mm_mul_ps(vr, vd))); in xnn_math_f32_sigmoid__sse2_rr2_p5_nr1()
Dsigmoid-sse2-rr2-lut64-p2-nr1.c38 const __m128 vtwo = _mm_set1_ps(2.0f); in xnn_math_f32_sigmoid__sse2_rr2_lut64_p2_nr1() local
125 vr = _mm_mul_ps(vr, _mm_sub_ps(vtwo, _mm_mul_ps(vr, vd))); in xnn_math_f32_sigmoid__sse2_rr2_lut64_p2_nr1()
/external/XNNPACK/src/f32-sigmoid/
Davx-p5.c.in45 const __m256 vtwo = _mm256_set1_ps(2.0f);
108 …vr${ABC[N]} = _mm256_mul_ps(vr${ABC[N]}, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr${ABC[N]}, vd${ABC[N]…
109 …vr${ABC[N]} = _mm256_mul_ps(vr${ABC[N]}, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr${ABC[N]}, vd${ABC[N]…
158 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd)));
159 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd)));
203 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd)));
204 vr = _mm256_mul_ps(vr, _mm256_sub_ps(vtwo, _mm256_mul_ps(vr, vd)));