| /external/rust/crates/quiche/deps/boringssl/src/crypto/poly1305/ |
| D | poly1305_vec.c | 206 p->S21.v = _mm_mul_epu32(p->R21.v, FIVE); in poly1305_first_block()
207 p->S22.v = _mm_mul_epu32(p->R22.v, FIVE); in poly1305_first_block()
208 p->S23.v = _mm_mul_epu32(p->R23.v, FIVE); in poly1305_first_block()
209 p->S24.v = _mm_mul_epu32(p->R24.v, FIVE); in poly1305_first_block()
260 T0 = _mm_mul_epu32(H0, p->R20.v); in poly1305_blocks()
261 T1 = _mm_mul_epu32(H0, p->R21.v); in poly1305_blocks()
262 T2 = _mm_mul_epu32(H0, p->R22.v); in poly1305_blocks()
263 T3 = _mm_mul_epu32(H0, p->R23.v); in poly1305_blocks()
264 T4 = _mm_mul_epu32(H0, p->R24.v); in poly1305_blocks()
265 T5 = _mm_mul_epu32(H1, p->S24.v); in poly1305_blocks()
[all …]
|
| /external/boringssl/src/crypto/poly1305/ |
| D | poly1305_vec.c | 210 p->S21.v = _mm_mul_epu32(p->R21.v, FIVE); in poly1305_first_block()
211 p->S22.v = _mm_mul_epu32(p->R22.v, FIVE); in poly1305_first_block()
212 p->S23.v = _mm_mul_epu32(p->R23.v, FIVE); in poly1305_first_block()
213 p->S24.v = _mm_mul_epu32(p->R24.v, FIVE); in poly1305_first_block()
264 T0 = _mm_mul_epu32(H0, p->R20.v); in poly1305_blocks()
265 T1 = _mm_mul_epu32(H0, p->R21.v); in poly1305_blocks()
266 T2 = _mm_mul_epu32(H0, p->R22.v); in poly1305_blocks()
267 T3 = _mm_mul_epu32(H0, p->R23.v); in poly1305_blocks()
268 T4 = _mm_mul_epu32(H0, p->R24.v); in poly1305_blocks()
269 T5 = _mm_mul_epu32(H1, p->S24.v); in poly1305_blocks()
[all …]
|
| /external/rust/crates/ring/crypto/poly1305/ |
| D | poly1305_vec.c | 214 p->S21.v = _mm_mul_epu32(p->R21.v, FIVE); in poly1305_first_block()
215 p->S22.v = _mm_mul_epu32(p->R22.v, FIVE); in poly1305_first_block()
216 p->S23.v = _mm_mul_epu32(p->R23.v, FIVE); in poly1305_first_block()
217 p->S24.v = _mm_mul_epu32(p->R24.v, FIVE); in poly1305_first_block()
268 T0 = _mm_mul_epu32(H0, p->R20.v); in poly1305_blocks()
269 T1 = _mm_mul_epu32(H0, p->R21.v); in poly1305_blocks()
270 T2 = _mm_mul_epu32(H0, p->R22.v); in poly1305_blocks()
271 T3 = _mm_mul_epu32(H0, p->R23.v); in poly1305_blocks()
272 T4 = _mm_mul_epu32(H0, p->R24.v); in poly1305_blocks()
273 T5 = _mm_mul_epu32(H1, p->S24.v); in poly1305_blocks()
[all …]
|
| /external/libvpx/libvpx/vpx_dsp/x86/ |
| D | highbd_idct4x4_add_sse2.c | 43 temp1[0] = _mm_mul_epu32(temp1[0], cospi_p16_p16); // ([0] + [2])*cospi_16_64 in highbd_idct4_small_sse2()
44 temp1[1] = _mm_mul_epu32(temp1[1], cospi_p16_p16); // ([0] + [2])*cospi_16_64 in highbd_idct4_small_sse2()
45 temp2[0] = _mm_mul_epu32(temp2[0], cospi_p16_p16); // ([0] - [2])*cospi_16_64 in highbd_idct4_small_sse2()
46 temp2[1] = _mm_mul_epu32(temp2[1], cospi_p16_p16); // ([0] - [2])*cospi_16_64 in highbd_idct4_small_sse2()
52 temp1[0] = _mm_mul_epu32(io[1], cospi_p24_p24); // input[1] * cospi_24_64 in highbd_idct4_small_sse2()
53 temp1[1] = _mm_mul_epu32(temp1[3], cospi_p24_p24); // input[1] * cospi_24_64 in highbd_idct4_small_sse2()
54 temp2[0] = _mm_mul_epu32(io[1], cospi_p08_p08); // input[1] * cospi_8_64 in highbd_idct4_small_sse2()
55 temp2[1] = _mm_mul_epu32(temp1[3], cospi_p08_p08); // input[1] * cospi_8_64 in highbd_idct4_small_sse2()
56 temp1[2] = _mm_mul_epu32(io[3], cospi_p08_p08); // input[3] * cospi_8_64 in highbd_idct4_small_sse2()
57 temp1[3] = _mm_mul_epu32(temp2[3], cospi_p08_p08); // input[3] * cospi_8_64 in highbd_idct4_small_sse2()
[all …]
|
| /external/XNNPACK/src/qs8-requantization/ |
| D | precise-sse4.c | 64 const __m128i x_absmul02 = _mm_mul_epu32(x_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse4()
65 const __m128i y_absmul02 = _mm_mul_epu32(y_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse4()
66 const __m128i z_absmul02 = _mm_mul_epu32(z_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse4()
67 const __m128i w_absmul02 = _mm_mul_epu32(w_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse4()
69 const __m128i x_absmul13 = _mm_mul_epu32(x_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse4()
70 const __m128i y_absmul13 = _mm_mul_epu32(y_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse4()
71 const __m128i z_absmul13 = _mm_mul_epu32(z_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse4()
72 const __m128i w_absmul13 = _mm_mul_epu32(w_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse4()
|
| D | precise-ssse3.c | 63 const __m128i x_absmul02 = _mm_mul_epu32(x_abs0123, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
64 const __m128i y_absmul02 = _mm_mul_epu32(y_abs0123, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
65 const __m128i z_absmul02 = _mm_mul_epu32(z_abs0123, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
66 const __m128i w_absmul02 = _mm_mul_epu32(w_abs0123, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
68 const __m128i x_absmul13 = _mm_mul_epu32(x_abs1032, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
69 const __m128i y_absmul13 = _mm_mul_epu32(y_abs1032, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
70 const __m128i z_absmul13 = _mm_mul_epu32(z_abs1032, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
71 const __m128i w_absmul13 = _mm_mul_epu32(w_abs1032, vmultiplier); in xnn_qs8_requantize_precise__ssse3()
|
| D | precise-sse2.c | 68 const __m128i x_absmul02 = _mm_mul_epu32(x_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse2()
69 const __m128i y_absmul02 = _mm_mul_epu32(y_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse2()
70 const __m128i z_absmul02 = _mm_mul_epu32(z_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse2()
71 const __m128i w_absmul02 = _mm_mul_epu32(w_abs0123, vmultiplier); in xnn_qs8_requantize_precise__sse2()
73 const __m128i x_absmul13 = _mm_mul_epu32(x_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse2()
74 const __m128i y_absmul13 = _mm_mul_epu32(y_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse2()
75 const __m128i z_absmul13 = _mm_mul_epu32(z_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse2()
76 const __m128i w_absmul13 = _mm_mul_epu32(w_abs1032, vmultiplier); in xnn_qs8_requantize_precise__sse2()
|
| D | q31-ssse3.c | 77 const __m128i x_abs_product_even = _mm_mul_epu32(x_abs, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
78 const __m128i y_abs_product_even = _mm_mul_epu32(y_abs, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
79 const __m128i z_abs_product_even = _mm_mul_epu32(z_abs, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
80 const __m128i w_abs_product_even = _mm_mul_epu32(w_abs, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
97 const __m128i x_abs_product_odd = _mm_mul_epu32(x_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
98 const __m128i y_abs_product_odd = _mm_mul_epu32(y_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
99 const __m128i z_abs_product_odd = _mm_mul_epu32(z_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
100 const __m128i w_abs_product_odd = _mm_mul_epu32(w_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__ssse3()
|
| D | q31-sse2.c | 77 const __m128i x_abs_product_even = _mm_mul_epu32(x_abs, vmultiplier); in xnn_qs8_requantize_q31__sse2()
78 const __m128i y_abs_product_even = _mm_mul_epu32(y_abs, vmultiplier); in xnn_qs8_requantize_q31__sse2()
79 const __m128i z_abs_product_even = _mm_mul_epu32(z_abs, vmultiplier); in xnn_qs8_requantize_q31__sse2()
80 const __m128i w_abs_product_even = _mm_mul_epu32(w_abs, vmultiplier); in xnn_qs8_requantize_q31__sse2()
97 const __m128i x_abs_product_odd = _mm_mul_epu32(x_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__sse2()
98 const __m128i y_abs_product_odd = _mm_mul_epu32(y_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__sse2()
99 const __m128i z_abs_product_odd = _mm_mul_epu32(z_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__sse2()
100 const __m128i w_abs_product_odd = _mm_mul_epu32(w_abs_rev, vmultiplier); in xnn_qs8_requantize_q31__sse2()
|
| /external/XNNPACK/src/qu8-requantization/ |
| D | precise-sse4.c | 64 const __m128i x_absmul02 = _mm_mul_epu32(x_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse4()
65 const __m128i y_absmul02 = _mm_mul_epu32(y_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse4()
66 const __m128i z_absmul02 = _mm_mul_epu32(z_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse4()
67 const __m128i w_absmul02 = _mm_mul_epu32(w_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse4()
69 const __m128i x_absmul13 = _mm_mul_epu32(x_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse4()
70 const __m128i y_absmul13 = _mm_mul_epu32(y_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse4()
71 const __m128i z_absmul13 = _mm_mul_epu32(z_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse4()
72 const __m128i w_absmul13 = _mm_mul_epu32(w_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse4()
|
| D | precise-ssse3.c | 63 const __m128i x_absmul02 = _mm_mul_epu32(x_abs0123, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
64 const __m128i y_absmul02 = _mm_mul_epu32(y_abs0123, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
65 const __m128i z_absmul02 = _mm_mul_epu32(z_abs0123, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
66 const __m128i w_absmul02 = _mm_mul_epu32(w_abs0123, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
68 const __m128i x_absmul13 = _mm_mul_epu32(x_abs1032, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
69 const __m128i y_absmul13 = _mm_mul_epu32(y_abs1032, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
70 const __m128i z_absmul13 = _mm_mul_epu32(z_abs1032, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
71 const __m128i w_absmul13 = _mm_mul_epu32(w_abs1032, vmultiplier); in xnn_qu8_requantize_precise__ssse3()
|
| D | precise-sse2.c | 68 const __m128i x_absmul02 = _mm_mul_epu32(x_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse2()
69 const __m128i y_absmul02 = _mm_mul_epu32(y_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse2()
70 const __m128i z_absmul02 = _mm_mul_epu32(z_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse2()
71 const __m128i w_absmul02 = _mm_mul_epu32(w_abs0123, vmultiplier); in xnn_qu8_requantize_precise__sse2()
73 const __m128i x_absmul13 = _mm_mul_epu32(x_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse2()
74 const __m128i y_absmul13 = _mm_mul_epu32(y_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse2()
75 const __m128i z_absmul13 = _mm_mul_epu32(z_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse2()
76 const __m128i w_absmul13 = _mm_mul_epu32(w_abs1032, vmultiplier); in xnn_qu8_requantize_precise__sse2()
|
| D | q31-ssse3.c | 77 const __m128i x_abs_product_even = _mm_mul_epu32(x_abs, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
78 const __m128i y_abs_product_even = _mm_mul_epu32(y_abs, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
79 const __m128i z_abs_product_even = _mm_mul_epu32(z_abs, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
80 const __m128i w_abs_product_even = _mm_mul_epu32(w_abs, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
97 const __m128i x_abs_product_odd = _mm_mul_epu32(x_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
98 const __m128i y_abs_product_odd = _mm_mul_epu32(y_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
99 const __m128i z_abs_product_odd = _mm_mul_epu32(z_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
100 const __m128i w_abs_product_odd = _mm_mul_epu32(w_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__ssse3()
|
| D | q31-sse2.c | 77 const __m128i x_abs_product_even = _mm_mul_epu32(x_abs, vmultiplier); in xnn_qu8_requantize_q31__sse2()
78 const __m128i y_abs_product_even = _mm_mul_epu32(y_abs, vmultiplier); in xnn_qu8_requantize_q31__sse2()
79 const __m128i z_abs_product_even = _mm_mul_epu32(z_abs, vmultiplier); in xnn_qu8_requantize_q31__sse2()
80 const __m128i w_abs_product_even = _mm_mul_epu32(w_abs, vmultiplier); in xnn_qu8_requantize_q31__sse2()
97 const __m128i x_abs_product_odd = _mm_mul_epu32(x_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__sse2()
98 const __m128i y_abs_product_odd = _mm_mul_epu32(y_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__sse2()
99 const __m128i z_abs_product_odd = _mm_mul_epu32(z_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__sse2()
100 const __m128i w_abs_product_odd = _mm_mul_epu32(w_abs_rev, vmultiplier); in xnn_qu8_requantize_q31__sse2()
|
| /external/XNNPACK/src/qs8-gavgpool/gen/ |
| D | 7x-minmax-sse41-c24-acc2.c | 134 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
135 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
136 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
137 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
138 const __m128i vabsprod8A = _mm_mul_epu32(vabsacc89AB, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
139 const __m128i vabsprod9B = _mm_mul_epu32(vabsacc9B, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
140 const __m128i vabsprodCE = _mm_mul_epu32(vabsaccCDEF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
141 const __m128i vabsprodDF = _mm_mul_epu32(vabsaccDF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
142 const __m128i vabsprodGI = _mm_mul_epu32(vabsaccGHIJ, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
143 const __m128i vabsprodHJ = _mm_mul_epu32(vabsaccHJ, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c24_acc2()
[all …]
|
| D | 7x-minmax-sse41-c16-acc2.c | 115 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
116 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
117 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
118 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
119 const __m128i vabsprod8A = _mm_mul_epu32(vabsacc89AB, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
120 const __m128i vabsprod9B = _mm_mul_epu32(vabsacc9B, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
121 const __m128i vabsprodCE = _mm_mul_epu32(vabsaccCDEF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
122 const __m128i vabsprodDF = _mm_mul_epu32(vabsaccDF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
196 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
197 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c16_acc2()
[all …]
|
| D | 7x-minmax-ssse3-c24-acc2.c | 158 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
159 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
160 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
161 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
162 const __m128i vabsprod8A = _mm_mul_epu32(vabsacc89AB, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
163 const __m128i vabsprod9B = _mm_mul_epu32(vabsacc9B, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
164 const __m128i vabsprodCE = _mm_mul_epu32(vabsaccCDEF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
165 const __m128i vabsprodDF = _mm_mul_epu32(vabsaccDF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
166 const __m128i vabsprodGI = _mm_mul_epu32(vabsaccGHIJ, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
167 const __m128i vabsprodHJ = _mm_mul_epu32(vabsaccHJ, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c24_acc2()
[all …]
|
| D | 7x-minmax-sse2-c24-acc2.c | 165 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
166 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
167 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
168 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
169 const __m128i vabsprod8A = _mm_mul_epu32(vabsacc89AB, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
170 const __m128i vabsprod9B = _mm_mul_epu32(vabsacc9B, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
171 const __m128i vabsprodCE = _mm_mul_epu32(vabsaccCDEF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
172 const __m128i vabsprodDF = _mm_mul_epu32(vabsaccDF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
173 const __m128i vabsprodGI = _mm_mul_epu32(vabsaccGHIJ, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
174 const __m128i vabsprodHJ = _mm_mul_epu32(vabsaccHJ, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c24_acc2()
[all …]
|
| D | 7x-minmax-sse41-c8-acc2.c | 96 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
97 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
98 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
99 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
163 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
164 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
165 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
166 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse41_c8_acc2()
|
| D | 7x-minmax-ssse3-c16-acc2.c | 131 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
132 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
133 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
134 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
135 const __m128i vabsprod8A = _mm_mul_epu32(vabsacc89AB, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
136 const __m128i vabsprod9B = _mm_mul_epu32(vabsacc9B, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
137 const __m128i vabsprodCE = _mm_mul_epu32(vabsaccCDEF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
138 const __m128i vabsprodDF = _mm_mul_epu32(vabsaccDF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
229 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
230 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c16_acc2()
[all …]
|
| D | 7x-minmax-sse2-c16-acc2.c | 136 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
137 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
138 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
139 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
140 const __m128i vabsprod8A = _mm_mul_epu32(vabsacc89AB, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
141 const __m128i vabsprod9B = _mm_mul_epu32(vabsacc9B, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
142 const __m128i vabsprodCE = _mm_mul_epu32(vabsaccCDEF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
143 const __m128i vabsprodDF = _mm_mul_epu32(vabsaccDF, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
237 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
238 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__sse2_c16_acc2()
[all …]
|
| D | 7x-minmax-ssse3-c8-acc2.c | 104 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
105 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
106 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
107 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
184 const __m128i vabsprod02 = _mm_mul_epu32(vabsacc0123, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
185 const __m128i vabsprod13 = _mm_mul_epu32(vabsacc13, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
186 const __m128i vabsprod46 = _mm_mul_epu32(vabsacc4567, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
187 const __m128i vabsprod57 = _mm_mul_epu32(vabsacc57, vmultiplier); in xnn_qs8_gavgpool_minmax_ukernel_7x__ssse3_c8_acc2()
|
| /external/flac/src/libFLAC/ |
| D | lpc_intrin_sse2.c | 440 …xmm7 = _mm_mul_epu32(xmm7, xmm5); /* we use _unsigned_ multiplication and discard high dword of th… in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
446 xmm6 = _mm_mul_epu32(xmm6, xmm4); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
453 xmm6 = _mm_mul_epu32(xmm6, xmm3); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
460 xmm6 = _mm_mul_epu32(xmm6, xmm2); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
467 xmm6 = _mm_mul_epu32(xmm6, xmm1); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
474 xmm6 = _mm_mul_epu32(xmm6, xmm0); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
500 xmm7 = _mm_mul_epu32(xmm7, xmm5); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
506 xmm6 = _mm_mul_epu32(xmm6, xmm4); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
513 xmm6 = _mm_mul_epu32(xmm6, xmm3); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
520 xmm6 = _mm_mul_epu32(xmm6, xmm2); in FLAC__lpc_compute_residual_from_qlp_coefficients_intrin_sse2()
[all …]
|
| /external/rust/crates/quiche/deps/boringssl/src/crypto/fipsmodule/modes/ |
| D | gcm_nohw.c | 100 _mm_xor_si128(_mm_mul_epu32(a0a0, b0b1), _mm_mul_epu32(a2a2, b2b3)); in gcm_mul32_nohw()
102 _mm_xor_si128(_mm_mul_epu32(a2a2, b0b1), _mm_mul_epu32(a0a0, b2b3)); in gcm_mul32_nohw()
113 c0c1 = _mm_xor_si128(c0c1, _mm_mul_epu32(a1a1, b3b0)); in gcm_mul32_nohw()
114 c0c1 = _mm_xor_si128(c0c1, _mm_mul_epu32(a3a3, b1b2)); in gcm_mul32_nohw()
115 c2c3 = _mm_xor_si128(c2c3, _mm_mul_epu32(a3a3, b3b0)); in gcm_mul32_nohw()
116 c2c3 = _mm_xor_si128(c2c3, _mm_mul_epu32(a1a1, b1b2)); in gcm_mul32_nohw()
|
| /external/boringssl/src/crypto/fipsmodule/modes/ |
| D | gcm_nohw.c | 100 _mm_xor_si128(_mm_mul_epu32(a0a0, b0b1), _mm_mul_epu32(a2a2, b2b3)); in gcm_mul32_nohw()
102 _mm_xor_si128(_mm_mul_epu32(a2a2, b0b1), _mm_mul_epu32(a0a0, b2b3)); in gcm_mul32_nohw()
113 c0c1 = _mm_xor_si128(c0c1, _mm_mul_epu32(a1a1, b3b0)); in gcm_mul32_nohw()
114 c0c1 = _mm_xor_si128(c0c1, _mm_mul_epu32(a3a3, b1b2)); in gcm_mul32_nohw()
115 c2c3 = _mm_xor_si128(c2c3, _mm_mul_epu32(a3a3, b3b0)); in gcm_mul32_nohw()
116 c2c3 = _mm_xor_si128(c2c3, _mm_mul_epu32(a1a1, b1b2)); in gcm_mul32_nohw()
|