| /external/zlib/ |
| D | crc32_simd.c | 48 x1 = _mm_xor_si128(x1, _mm_cvtsi32_si128(crc)); in crc32_sse42_simd_()
75 x1 = _mm_xor_si128(x1, x5); in crc32_sse42_simd_()
76 x2 = _mm_xor_si128(x2, x6); in crc32_sse42_simd_()
77 x3 = _mm_xor_si128(x3, x7); in crc32_sse42_simd_()
78 x4 = _mm_xor_si128(x4, x8); in crc32_sse42_simd_()
80 x1 = _mm_xor_si128(x1, y5); in crc32_sse42_simd_()
81 x2 = _mm_xor_si128(x2, y6); in crc32_sse42_simd_()
82 x3 = _mm_xor_si128(x3, y7); in crc32_sse42_simd_()
83 x4 = _mm_xor_si128(x4, y8); in crc32_sse42_simd_()
96 x1 = _mm_xor_si128(x1, x2); in crc32_sse42_simd_()
[all …]
|
| D | crc_folding.c | 239 xmm_shr = _mm_xor_si128(xmm_shr, xmm_mask3); in partial_fold()
312 xmm_crc0 = _mm_xor_si128(xmm_crc0, xmm_t0); in crc_fold_copy()
313 xmm_crc1 = _mm_xor_si128(xmm_crc1, xmm_t1); in crc_fold_copy()
314 xmm_crc2 = _mm_xor_si128(xmm_crc2, xmm_t2); in crc_fold_copy()
315 xmm_crc3 = _mm_xor_si128(xmm_crc3, xmm_t3); in crc_fold_copy()
337 xmm_crc1 = _mm_xor_si128(xmm_crc1, xmm_t0); in crc_fold_copy()
338 xmm_crc2 = _mm_xor_si128(xmm_crc2, xmm_t1); in crc_fold_copy()
339 xmm_crc3 = _mm_xor_si128(xmm_crc3, xmm_t2); in crc_fold_copy()
357 xmm_crc2 = _mm_xor_si128(xmm_crc2, xmm_t0); in crc_fold_copy()
358 xmm_crc3 = _mm_xor_si128(xmm_crc3, xmm_t1); in crc_fold_copy()
[all …]
|
| /external/v8/third_party/zlib/ |
| D | crc32_simd.c | 48 x1 = _mm_xor_si128(x1, _mm_cvtsi32_si128(crc)); in crc32_sse42_simd_()
75 x1 = _mm_xor_si128(x1, x5); in crc32_sse42_simd_()
76 x2 = _mm_xor_si128(x2, x6); in crc32_sse42_simd_()
77 x3 = _mm_xor_si128(x3, x7); in crc32_sse42_simd_()
78 x4 = _mm_xor_si128(x4, x8); in crc32_sse42_simd_()
80 x1 = _mm_xor_si128(x1, y5); in crc32_sse42_simd_()
81 x2 = _mm_xor_si128(x2, y6); in crc32_sse42_simd_()
82 x3 = _mm_xor_si128(x3, y7); in crc32_sse42_simd_()
83 x4 = _mm_xor_si128(x4, y8); in crc32_sse42_simd_()
96 x1 = _mm_xor_si128(x1, x2); in crc32_sse42_simd_()
[all …]
|
| D | crc_folding.c | 241 xmm_shr = _mm_xor_si128(xmm_shr, xmm_mask3); in partial_fold()
314 xmm_crc0 = _mm_xor_si128(xmm_crc0, xmm_t0); in crc_fold_copy()
315 xmm_crc1 = _mm_xor_si128(xmm_crc1, xmm_t1); in crc_fold_copy()
316 xmm_crc2 = _mm_xor_si128(xmm_crc2, xmm_t2); in crc_fold_copy()
317 xmm_crc3 = _mm_xor_si128(xmm_crc3, xmm_t3); in crc_fold_copy()
339 xmm_crc1 = _mm_xor_si128(xmm_crc1, xmm_t0); in crc_fold_copy()
340 xmm_crc2 = _mm_xor_si128(xmm_crc2, xmm_t1); in crc_fold_copy()
341 xmm_crc3 = _mm_xor_si128(xmm_crc3, xmm_t2); in crc_fold_copy()
359 xmm_crc2 = _mm_xor_si128(xmm_crc2, xmm_t0); in crc_fold_copy()
360 xmm_crc3 = _mm_xor_si128(xmm_crc3, xmm_t1); in crc_fold_copy()
[all …]
|
| /external/lzma/C/ |
| D | AesOpt.c | 25 m = _mm_xor_si128(m, *data); in AesCbc_Encode_Intel()
26 m = _mm_xor_si128(m, p[2]); in AesCbc_Encode_Intel()
65 m0 = _mm_xor_si128(t, data[0]); in AesCbc_Decode_Intel()
66 m1 = _mm_xor_si128(t, data[1]); in AesCbc_Decode_Intel()
67 m2 = _mm_xor_si128(t, data[2]); in AesCbc_Decode_Intel()
82 t = _mm_xor_si128(m0, iv); iv = data[0]; data[0] = t; in AesCbc_Decode_Intel()
83 t = _mm_xor_si128(m1, iv); iv = data[1]; data[1] = t; in AesCbc_Decode_Intel()
84 t = _mm_xor_si128(m2, iv); iv = data[2]; data[2] = t; in AesCbc_Decode_Intel()
91 __m128i m = _mm_xor_si128(w[2], *data); in AesCbc_Decode_Intel()
103 m = _mm_xor_si128(m, iv); in AesCbc_Decode_Intel()
[all …]
|
| /external/scrypt/lib/crypto/ |
| D | crypto_scrypt-sse.c | 77 D[i] = _mm_xor_si128(D[i], S[i]); in blkxor()
99 X1 = _mm_xor_si128(X1, _mm_slli_epi32(T, 7)); in salsa20_8()
100 X1 = _mm_xor_si128(X1, _mm_srli_epi32(T, 25)); in salsa20_8()
102 X2 = _mm_xor_si128(X2, _mm_slli_epi32(T, 9)); in salsa20_8()
103 X2 = _mm_xor_si128(X2, _mm_srli_epi32(T, 23)); in salsa20_8()
105 X3 = _mm_xor_si128(X3, _mm_slli_epi32(T, 13)); in salsa20_8()
106 X3 = _mm_xor_si128(X3, _mm_srli_epi32(T, 19)); in salsa20_8()
108 X0 = _mm_xor_si128(X0, _mm_slli_epi32(T, 18)); in salsa20_8()
109 X0 = _mm_xor_si128(X0, _mm_srli_epi32(T, 14)); in salsa20_8()
118 X3 = _mm_xor_si128(X3, _mm_slli_epi32(T, 7)); in salsa20_8()
[all …]
|
| /external/python/cpython3/Modules/_blake2/impl/ |
| D | blake2b-round.h | 36 : (-(c) == 63) ? _mm_xor_si128(_mm_srli_epi64((x), -(c)), _mm_add_epi64((x), (x))) \
37 : _mm_xor_si128(_mm_srli_epi64((x), -(c)), _mm_slli_epi64((x), 64-(-(c))))
39 #define _mm_roti_epi64(r, c) _mm_xor_si128(_mm_srli_epi64( (r), -(c) ),_mm_slli_epi64( (r), 64-(-(c…
51 row4l = _mm_xor_si128(row4l, row1l); \
52 row4h = _mm_xor_si128(row4h, row1h); \
60 row2l = _mm_xor_si128(row2l, row3l); \
61 row2h = _mm_xor_si128(row2h, row3h); \
70 row4l = _mm_xor_si128(row4l, row1l); \
71 row4h = _mm_xor_si128(row4h, row1h); \
79 row2l = _mm_xor_si128(row2l, row3l); \
[all …]
|
| D | blake2s-round.h | 35 : _mm_xor_si128(_mm_srli_epi32( (r), -(c) ),_mm_slli_epi32( (r), 32-(-(c)) )) )
37 #define _mm_roti_epi32(r, c) _mm_xor_si128(_mm_srli_epi32( (r), -(c) ),_mm_slli_epi32( (r), 32-(-(c…
46 row4 = _mm_xor_si128( row4, row1 ); \
49 row2 = _mm_xor_si128( row2, row3 ); \
54 row4 = _mm_xor_si128( row4, row1 ); \
57 row2 = _mm_xor_si128( row2, row3 ); \
|
| D | blake2b.c | 325 row4l = _mm_xor_si128( LOADU( &blake2b_IV[4] ), LOADU( &S->t[0] ) ); in blake2b_compress()
326 row4h = _mm_xor_si128( LOADU( &blake2b_IV[6] ), LOADU( &S->f[0] ) ); in blake2b_compress()
339 row1l = _mm_xor_si128( row3l, row1l ); in blake2b_compress()
340 row1h = _mm_xor_si128( row3h, row1h ); in blake2b_compress()
341 STOREU( &S->h[0], _mm_xor_si128( LOADU( &S->h[0] ), row1l ) ); in blake2b_compress()
342 STOREU( &S->h[2], _mm_xor_si128( LOADU( &S->h[2] ), row1h ) ); in blake2b_compress()
343 row2l = _mm_xor_si128( row4l, row2l ); in blake2b_compress()
344 row2h = _mm_xor_si128( row4h, row2h ); in blake2b_compress()
345 STOREU( &S->h[4], _mm_xor_si128( LOADU( &S->h[4] ), row2l ) ); in blake2b_compress()
346 STOREU( &S->h[6], _mm_xor_si128( LOADU( &S->h[6] ), row2h ) ); in blake2b_compress()
|
| /external/XNNPACK/src/requantization/ |
| D | q31-sse2.c | 66 const __m128i x_abs = _mm_sub_epi32(_mm_xor_si128(x, x_neg_mask), x_neg_mask); in xnn_requantize_q31__sse2()
67 const __m128i y_abs = _mm_sub_epi32(_mm_xor_si128(y, y_neg_mask), y_neg_mask); in xnn_requantize_q31__sse2()
68 const __m128i z_abs = _mm_sub_epi32(_mm_xor_si128(z, z_neg_mask), z_neg_mask); in xnn_requantize_q31__sse2()
69 const __m128i w_abs = _mm_sub_epi32(_mm_xor_si128(w, w_neg_mask), w_neg_mask); in xnn_requantize_q31__sse2()
86 …const __m128i x_product_even = _mm_sub_epi64(_mm_xor_si128(x_abs_product_even, x_neg_mask_even), x… in xnn_requantize_q31__sse2()
87 …const __m128i y_product_even = _mm_sub_epi64(_mm_xor_si128(y_abs_product_even, y_neg_mask_even), y… in xnn_requantize_q31__sse2()
88 …const __m128i z_product_even = _mm_sub_epi64(_mm_xor_si128(z_abs_product_even, z_neg_mask_even), z… in xnn_requantize_q31__sse2()
89 …const __m128i w_product_even = _mm_sub_epi64(_mm_xor_si128(w_abs_product_even, w_neg_mask_even), w… in xnn_requantize_q31__sse2()
106 …const __m128i x_product_odd = _mm_sub_epi64(_mm_xor_si128(x_abs_product_odd, x_neg_mask_odd), x_ne… in xnn_requantize_q31__sse2()
107 …const __m128i y_product_odd = _mm_sub_epi64(_mm_xor_si128(y_abs_product_odd, y_neg_mask_odd), y_ne… in xnn_requantize_q31__sse2()
[all …]
|
| D | precise-sse2.c | 57 const __m128i x_abs0123 = _mm_sub_epi32(_mm_xor_si128(x, x_neg_mask), x_neg_mask); in xnn_requantize_precise__sse2()
58 const __m128i y_abs0123 = _mm_sub_epi32(_mm_xor_si128(y, y_neg_mask), y_neg_mask); in xnn_requantize_precise__sse2()
59 const __m128i z_abs0123 = _mm_sub_epi32(_mm_xor_si128(z, z_neg_mask), z_neg_mask); in xnn_requantize_precise__sse2()
60 const __m128i w_abs0123 = _mm_sub_epi32(_mm_xor_si128(w, w_neg_mask), w_neg_mask); in xnn_requantize_precise__sse2()
100 const __m128i x_scaled = _mm_sub_epi32(_mm_xor_si128(x_abs_scaled, x_neg_mask), x_neg_mask); in xnn_requantize_precise__sse2()
101 const __m128i y_scaled = _mm_sub_epi32(_mm_xor_si128(y_abs_scaled, y_neg_mask), y_neg_mask); in xnn_requantize_precise__sse2()
102 const __m128i z_scaled = _mm_sub_epi32(_mm_xor_si128(z_abs_scaled, z_neg_mask), z_neg_mask); in xnn_requantize_precise__sse2()
103 const __m128i w_scaled = _mm_sub_epi32(_mm_xor_si128(w_abs_scaled, w_neg_mask), w_neg_mask); in xnn_requantize_precise__sse2()
|
| D | q31-ssse3.c | 86 …const __m128i x_product_even = _mm_sub_epi64(_mm_xor_si128(x_abs_product_even, x_neg_mask_even), x… in xnn_requantize_q31__ssse3()
87 …const __m128i y_product_even = _mm_sub_epi64(_mm_xor_si128(y_abs_product_even, y_neg_mask_even), y… in xnn_requantize_q31__ssse3()
88 …const __m128i z_product_even = _mm_sub_epi64(_mm_xor_si128(z_abs_product_even, z_neg_mask_even), z… in xnn_requantize_q31__ssse3()
89 …const __m128i w_product_even = _mm_sub_epi64(_mm_xor_si128(w_abs_product_even, w_neg_mask_even), w… in xnn_requantize_q31__ssse3()
106 …const __m128i x_product_odd = _mm_sub_epi64(_mm_xor_si128(x_abs_product_odd, x_neg_mask_odd), x_ne… in xnn_requantize_q31__ssse3()
107 …const __m128i y_product_odd = _mm_sub_epi64(_mm_xor_si128(y_abs_product_odd, y_neg_mask_odd), y_ne… in xnn_requantize_q31__ssse3()
108 …const __m128i z_product_odd = _mm_sub_epi64(_mm_xor_si128(z_abs_product_odd, z_neg_mask_odd), z_ne… in xnn_requantize_q31__ssse3()
109 …const __m128i w_product_odd = _mm_sub_epi64(_mm_xor_si128(w_abs_product_odd, w_neg_mask_odd), w_ne… in xnn_requantize_q31__ssse3()
|
| /external/libvpx/libvpx/vp8/encoder/x86/ |
| D | vp8_quantize_sse2.c | 66 x0 = _mm_xor_si128(z0, sz0); in vp8_regular_quantize_b_sse2()
67 x1 = _mm_xor_si128(z1, sz1); in vp8_regular_quantize_b_sse2()
101 y0 = _mm_xor_si128(y0, sz0); in vp8_regular_quantize_b_sse2()
102 y1 = _mm_xor_si128(y1, sz1); in vp8_regular_quantize_b_sse2()
163 x0 = _mm_xor_si128(z0, sz0); in vp8_fast_quantize_b_sse2()
164 x1 = _mm_xor_si128(z1, sz1); in vp8_fast_quantize_b_sse2()
177 y0 = _mm_xor_si128(y0, sz0); in vp8_fast_quantize_b_sse2()
178 y1 = _mm_xor_si128(y1, sz1); in vp8_fast_quantize_b_sse2()
202 x0 = _mm_xor_si128(x0, ones); in vp8_fast_quantize_b_sse2()
203 x1 = _mm_xor_si128(x1, ones); in vp8_fast_quantize_b_sse2()
|
| /external/OpenCL-CTS/test_common/harness/ |
| D | mt19937.cpp | 145 …__m128i vr = _mm_xor_si128( _mm_loadu_si128( (__m128i*)(mt + kk + M)), (__m128i) _mm_srli_epi32( v… in genrand_int32()
146 …vr = _mm_xor_si128( vr, vmag01 ); … in genrand_int32()
170 …__m128i vr = _mm_xor_si128( _mm_loadu_si128( (__m128i*)(mt + kk + M - N)), _mm_srli_epi32( vy, 1 )… in genrand_int32()
171 …vr = _mm_xor_si128( vr, vmag01 ); … in genrand_int32()
188 … vy = _mm_xor_si128( vy, _mm_srli_epi32( vy, 11 ) ); // y ^= (y >> 11); in genrand_int32()
189 …vy = _mm_xor_si128( vy, _mm_and_si128( _mm_slli_epi32( vy, 7 ), c0.v) ); // y ^= (y << 7) &… in genrand_int32()
190 …vy = _mm_xor_si128( vy, _mm_and_si128( _mm_slli_epi32( vy, 15 ), c1.v) ); // y ^= (y << 15) … in genrand_int32()
191 … vy = _mm_xor_si128( vy, _mm_srli_epi32( vy, 18 ) ); // y ^= (y >> 18); in genrand_int32()
|
| /external/libvpx/libvpx/vp9/encoder/x86/ |
| D | vp9_quantize_sse2.c | 64 qcoeff0 = _mm_xor_si128(coeff0, coeff0_sign); in vp9_quantize_fp_sse2()
65 qcoeff1 = _mm_xor_si128(coeff1, coeff1_sign); in vp9_quantize_fp_sse2()
77 qcoeff0 = _mm_xor_si128(qtmp0, coeff0_sign); in vp9_quantize_fp_sse2()
78 qcoeff1 = _mm_xor_si128(qtmp1, coeff1_sign); in vp9_quantize_fp_sse2()
131 qcoeff0 = _mm_xor_si128(coeff0, coeff0_sign); in vp9_quantize_fp_sse2()
132 qcoeff1 = _mm_xor_si128(coeff1, coeff1_sign); in vp9_quantize_fp_sse2()
146 qcoeff0 = _mm_xor_si128(qtmp0, coeff0_sign); in vp9_quantize_fp_sse2()
147 qcoeff1 = _mm_xor_si128(qtmp1, coeff1_sign); in vp9_quantize_fp_sse2()
|
| D | vp9_highbd_block_error_intrin_sse2.c | 36 cmp0 = _mm_xor_si128(_mm_cmpgt_epi32(mm_coeff, max), in vp9_highbd_block_error_sse2()
38 cmp1 = _mm_xor_si128(_mm_cmpgt_epi32(mm_coeff2, max), in vp9_highbd_block_error_sse2()
40 cmp2 = _mm_xor_si128(_mm_cmpgt_epi32(mm_dqcoeff, max), in vp9_highbd_block_error_sse2()
42 cmp3 = _mm_xor_si128(_mm_cmpgt_epi32(mm_dqcoeff2, max), in vp9_highbd_block_error_sse2()
|
| /external/fec/ |
| D | viterbi615_sse2.c | 143 …m0 = _mm_add_epi16(_mm_xor_si128(Branchtab615[0].v[i],sym0v),_mm_xor_si128(Branchtab615[1].v[i],sy… in update_viterbi615_blk_sse2()
144 …m1 = _mm_add_epi16(_mm_xor_si128(Branchtab615[2].v[i],sym2v),_mm_xor_si128(Branchtab615[3].v[i],sy… in update_viterbi615_blk_sse2()
145 …m2 = _mm_add_epi16(_mm_xor_si128(Branchtab615[4].v[i],sym4v),_mm_xor_si128(Branchtab615[5].v[i],sy… in update_viterbi615_blk_sse2()
|
| /external/libaom/libaom/av1/encoder/x86/ |
| D | highbd_block_error_intrin_sse2.c | 37 cmp0 = _mm_xor_si128(_mm_cmpgt_epi32(mm_coeff, max), in av1_highbd_block_error_sse2()
39 cmp1 = _mm_xor_si128(_mm_cmpgt_epi32(mm_coeff2, max), in av1_highbd_block_error_sse2()
41 cmp2 = _mm_xor_si128(_mm_cmpgt_epi32(mm_dqcoeff, max), in av1_highbd_block_error_sse2()
43 cmp3 = _mm_xor_si128(_mm_cmpgt_epi32(mm_dqcoeff2, max), in av1_highbd_block_error_sse2()
|
| /external/neven/Embedded/common/src/b_BasicEm/ |
| D | MathSSE2.c | 44 m_XMM4 = _mm_xor_si128( m_XMM4, m_XMM4 ); in bbs_dotProduct_64SSE2()
45 m_XMM6 = _mm_xor_si128( m_XMM6, m_XMM6 ); in bbs_dotProduct_64SSE2()
46 m_XMM7 = _mm_xor_si128( m_XMM7, m_XMM7 ); in bbs_dotProduct_64SSE2()
155 m_XMM5 = _mm_xor_si128( m_XMM5, m_XMM5 ); in bbs_dotProduct_128SSE2()
156 m_XMM6 = _mm_xor_si128( m_XMM6, m_XMM6 ); in bbs_dotProduct_128SSE2()
258 m_XMM5 = _mm_xor_si128( m_XMM5, m_XMM5 ); in bbs_dotProduct_u128SSE2()
259 m_XMM6 = _mm_xor_si128( m_XMM6, m_XMM6 ); in bbs_dotProduct_u128SSE2()
|
| /external/XNNPACK/src/q8-gemm/ |
| D | 4x4c2-sse2.c | 211 …const __m128i vabsacc0x0123 = _mm_sub_epi32(_mm_xor_si128(vacc0x0123, vnmask0x0123), vnmask0x0123); in xnn_q8_gemm_ukernel_4x4c2__sse2()
212 …const __m128i vabsacc1x0123 = _mm_sub_epi32(_mm_xor_si128(vacc1x0123, vnmask1x0123), vnmask1x0123); in xnn_q8_gemm_ukernel_4x4c2__sse2()
213 …const __m128i vabsacc2x0123 = _mm_sub_epi32(_mm_xor_si128(vacc2x0123, vnmask2x0123), vnmask2x0123); in xnn_q8_gemm_ukernel_4x4c2__sse2()
214 …const __m128i vabsacc3x0123 = _mm_sub_epi32(_mm_xor_si128(vacc3x0123, vnmask3x0123), vnmask3x0123); in xnn_q8_gemm_ukernel_4x4c2__sse2()
231 const __m128i vprod0x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod0x02, vnmask0x02), vnmask0x02); in xnn_q8_gemm_ukernel_4x4c2__sse2()
232 const __m128i vprod1x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod1x02, vnmask1x02), vnmask1x02); in xnn_q8_gemm_ukernel_4x4c2__sse2()
233 const __m128i vprod2x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod2x02, vnmask2x02), vnmask2x02); in xnn_q8_gemm_ukernel_4x4c2__sse2()
234 const __m128i vprod3x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod3x02, vnmask3x02), vnmask3x02); in xnn_q8_gemm_ukernel_4x4c2__sse2()
251 const __m128i vprod0x13 = _mm_sub_epi64(_mm_xor_si128(vabsprod0x13, vnmask0x13), vnmask0x13); in xnn_q8_gemm_ukernel_4x4c2__sse2()
252 const __m128i vprod1x13 = _mm_sub_epi64(_mm_xor_si128(vabsprod1x13, vnmask1x13), vnmask1x13); in xnn_q8_gemm_ukernel_4x4c2__sse2()
[all …]
|
| /external/XNNPACK/src/q8-igemm/ |
| D | 4x4c2-sse2.c | 191 …const __m128i vabsacc0x0123 = _mm_sub_epi32(_mm_xor_si128(vacc0x0123, vnmask0x0123), vnmask0x0123); in xnn_q8_igemm_ukernel_4x4c2__sse2()
192 …const __m128i vabsacc1x0123 = _mm_sub_epi32(_mm_xor_si128(vacc1x0123, vnmask1x0123), vnmask1x0123); in xnn_q8_igemm_ukernel_4x4c2__sse2()
193 …const __m128i vabsacc2x0123 = _mm_sub_epi32(_mm_xor_si128(vacc2x0123, vnmask2x0123), vnmask2x0123); in xnn_q8_igemm_ukernel_4x4c2__sse2()
194 …const __m128i vabsacc3x0123 = _mm_sub_epi32(_mm_xor_si128(vacc3x0123, vnmask3x0123), vnmask3x0123); in xnn_q8_igemm_ukernel_4x4c2__sse2()
211 const __m128i vprod0x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod0x02, vnmask0x02), vnmask0x02); in xnn_q8_igemm_ukernel_4x4c2__sse2()
212 const __m128i vprod1x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod1x02, vnmask1x02), vnmask1x02); in xnn_q8_igemm_ukernel_4x4c2__sse2()
213 const __m128i vprod2x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod2x02, vnmask2x02), vnmask2x02); in xnn_q8_igemm_ukernel_4x4c2__sse2()
214 const __m128i vprod3x02 = _mm_sub_epi64(_mm_xor_si128(vabsprod3x02, vnmask3x02), vnmask3x02); in xnn_q8_igemm_ukernel_4x4c2__sse2()
231 const __m128i vprod0x13 = _mm_sub_epi64(_mm_xor_si128(vabsprod0x13, vnmask0x13), vnmask0x13); in xnn_q8_igemm_ukernel_4x4c2__sse2()
232 const __m128i vprod1x13 = _mm_sub_epi64(_mm_xor_si128(vabsprod1x13, vnmask1x13), vnmask1x13); in xnn_q8_igemm_ukernel_4x4c2__sse2()
[all …]
|
| /external/XNNPACK/src/q8-gavgpool/ |
| D | up7-sse2.c | 92 const __m128i vabs_lo0123 = _mm_sub_epi32(_mm_xor_si128(vacc_lo, vneg_mask_lo), vneg_mask_lo); in xnn_q8_gavgpool_ukernel_up7__sse2()
93 const __m128i vabs_hi0123 = _mm_sub_epi32(_mm_xor_si128(vacc_hi, vneg_mask_hi), vneg_mask_hi); in xnn_q8_gavgpool_ukernel_up7__sse2()
117 …const __m128i vscaled_lo = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_lo, vneg_mask_lo), vneg_mask_lo… in xnn_q8_gavgpool_ukernel_up7__sse2()
118 …const __m128i vscaled_hi = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_hi, vneg_mask_hi), vneg_mask_hi… in xnn_q8_gavgpool_ukernel_up7__sse2()
161 const __m128i vabs_lo0123 = _mm_sub_epi32(_mm_xor_si128(vacc_lo, vneg_mask_lo), vneg_mask_lo); in xnn_q8_gavgpool_ukernel_up7__sse2()
162 const __m128i vabs_hi0123 = _mm_sub_epi32(_mm_xor_si128(vacc_hi, vneg_mask_hi), vneg_mask_hi); in xnn_q8_gavgpool_ukernel_up7__sse2()
186 …const __m128i vscaled_lo = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_lo, vneg_mask_lo), vneg_mask_lo… in xnn_q8_gavgpool_ukernel_up7__sse2()
187 …const __m128i vscaled_hi = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_hi, vneg_mask_hi), vneg_mask_hi… in xnn_q8_gavgpool_ukernel_up7__sse2()
|
| /external/libvpx/libvpx/vpx_dsp/x86/ |
| D | loopfilter_sse2.c | 67 ps1ps0 = _mm_xor_si128(p1p0, t80); /* ^ 0x80 */ \
68 qs1qs0 = _mm_xor_si128(q1q0, t80); \
103 qs1qs0 = _mm_xor_si128(qs1qs0, t80); /* ^ 0x80 */ \
104 ps1ps0 = _mm_xor_si128(ps1ps0, t80); /* ^ 0x80 */ \
272 hev = _mm_xor_si128(_mm_cmpeq_epi8(hev, zero), ff); in vpx_lpf_horizontal_16_sse2()
277 mask = _mm_xor_si128(_mm_cmpeq_epi8(mask, zero), ff); in vpx_lpf_horizontal_16_sse2()
296 __m128i qs1ps1 = _mm_xor_si128(q1p1, t80); in vpx_lpf_horizontal_16_sse2()
297 __m128i qs0ps0 = _mm_xor_si128(q0p0, t80); in vpx_lpf_horizontal_16_sse2()
298 __m128i qs0 = _mm_xor_si128(p0q0, t80); in vpx_lpf_horizontal_16_sse2()
299 __m128i qs1 = _mm_xor_si128(p1q1, t80); in vpx_lpf_horizontal_16_sse2()
[all …]
|
| /external/XNNPACK/src/q8-avgpool/ |
| D | up9-sse2.c | 111 const __m128i vabs_lo0123 = _mm_sub_epi32(_mm_xor_si128(vacc_lo, vneg_mask_lo), vneg_mask_lo); in xnn_q8_avgpool_ukernel_up9__sse2()
112 const __m128i vabs_hi0123 = _mm_sub_epi32(_mm_xor_si128(vacc_hi, vneg_mask_hi), vneg_mask_hi); in xnn_q8_avgpool_ukernel_up9__sse2()
136 …const __m128i vscaled_lo = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_lo, vneg_mask_lo), vneg_mask_lo… in xnn_q8_avgpool_ukernel_up9__sse2()
137 …const __m128i vscaled_hi = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_hi, vneg_mask_hi), vneg_mask_hi… in xnn_q8_avgpool_ukernel_up9__sse2()
186 const __m128i vabs_lo0123 = _mm_sub_epi32(_mm_xor_si128(vacc_lo, vneg_mask_lo), vneg_mask_lo); in xnn_q8_avgpool_ukernel_up9__sse2()
187 const __m128i vabs_hi0123 = _mm_sub_epi32(_mm_xor_si128(vacc_hi, vneg_mask_hi), vneg_mask_hi); in xnn_q8_avgpool_ukernel_up9__sse2()
211 …const __m128i vscaled_lo = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_lo, vneg_mask_lo), vneg_mask_lo… in xnn_q8_avgpool_ukernel_up9__sse2()
212 …const __m128i vscaled_hi = _mm_sub_epi32(_mm_xor_si128(vabs_scaled_hi, vneg_mask_hi), vneg_mask_hi… in xnn_q8_avgpool_ukernel_up9__sse2()
|
| /external/jemalloc/test/include/test/ |
| D | SFMT-sse2.h | 71 z = _mm_xor_si128(z, x); in mm_recursion()
72 z = _mm_xor_si128(z, v); in mm_recursion()
75 z = _mm_xor_si128(z, x); in mm_recursion()
76 z = _mm_xor_si128(z, y); in mm_recursion()
|