Lines Matching refs:ABC

9 $ABC = "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZ"
39       const __m128i vx${ABC[0:8]} = _mm_cvtepi8_epi16(_mm_loadl_epi64((const __m128i*) input_x));
40       const __m128i vy${ABC[0:8]} = _mm_cvtepi8_epi16(_mm_loadl_epi64((const __m128i*) input_y));
42 …const __m128i vx${ABC[N:N+8]} = _mm_cvtepi8_epi16(_mm_loadl_epi64((const __m128i*) (input_x + ${N}…
43 …const __m128i vy${ABC[N:N+8]} = _mm_cvtepi8_epi16(_mm_loadl_epi64((const __m128i*) (input_y + ${N}…
45       __m128i vx${ABC[0:8]} = _mm_loadl_epi64((const __m128i*) input_x);
46       __m128i vy${ABC[0:8]} = _mm_loadl_epi64((const __m128i*) input_y);
48         __m128i vx${ABC[N:N+8]} = _mm_loadl_epi64((const __m128i*) (input_x + ${N}));
49         __m128i vy${ABC[N:N+8]} = _mm_loadl_epi64((const __m128i*) (input_y + ${N}));
55 …vx${ABC[N:N+8]} = _mm_unpacklo_epi8(vx${ABC[N:N+8]}, _mm_cmpgt_epi8(_mm_setzero_si128(), vx${ABC[N…
56 …vy${ABC[N:N+8]} = _mm_unpacklo_epi8(vy${ABC[N:N+8]}, _mm_cmpgt_epi8(_mm_setzero_si128(), vy${ABC[N…
59       __m128i vxprod${ABC[N:N+8]}hi = _mm_mulhi_epu16(vx${ABC[N:N+8]}, vx_multiplier_lo);
60       __m128i vyprod${ABC[N:N+8]}hi = _mm_mulhi_epu16(vy${ABC[N:N+8]}, vy_multiplier_lo);
61       const __m128i vxprod${ABC[N:N+8]}lo = _mm_mullo_epi16(vx${ABC[N:N+8]}, vx_multiplier_lo);
62       const __m128i vyprod${ABC[N:N+8]}lo = _mm_mullo_epi16(vy${ABC[N:N+8]}, vy_multiplier_lo);
65 …vxprod${ABC[N:N+8]}hi = _mm_add_epi16(vxprod${ABC[N:N+8]}hi, _mm_mullo_epi16(vx${ABC[N:N+8]}, vx_m…
66 …vyprod${ABC[N:N+8]}hi = _mm_add_epi16(vyprod${ABC[N:N+8]}hi, _mm_mullo_epi16(vy${ABC[N:N+8]}, vy_m…
69 …vxprod${ABC[N:N+8]}hi = _mm_sub_epi16(vxprod${ABC[N:N+8]}hi, _mm_and_si128(_mm_srai_epi16(vx${ABC[…
70 …vyprod${ABC[N:N+8]}hi = _mm_sub_epi16(vyprod${ABC[N:N+8]}hi, _mm_and_si128(_mm_srai_epi16(vy${ABC[…
73 …__m128i vacc${ABC[N:N+4]} = _mm_add_epi32(vzero_point_product, _mm_unpacklo_epi16(vxprod${ABC[N:N+…
74 …__m128i vacc${ABC[N+4:N+8]} = _mm_add_epi32(vzero_point_product, _mm_unpackhi_epi16(vxprod${ABC[N:…
77 …vacc${ABC[N:N+4]} = _mm_add_epi32(vacc${ABC[N:N+4]}, _mm_unpacklo_epi16(vyprod${ABC[N:N+8]}lo, vyp…
78 …vacc${ABC[N+4:N+8]} = _mm_add_epi32(vacc${ABC[N+4:N+8]}, _mm_unpackhi_epi16(vyprod${ABC[N:N+8]}lo,…
81 …__m128i vrem${ABC[N:N+4]} = _mm_add_epi32(_mm_and_si128(vacc${ABC[N:N+4]}, vremainder_mask), _mm_c…
84 …vacc${ABC[N:N+4]} = _mm_sub_epi32(_mm_sra_epi32(vacc${ABC[N:N+4]}, vshift), _mm_cmpgt_epi32(vrem${…
87 …__m128i vout${ABC[N:N+8]} = _mm_adds_epi16(_mm_packs_epi32(vacc${ABC[N:N+4]}, vacc${ABC[N+4:N+8]})…
90       vout${ABC[N:N+8]} = _mm_max_epi16(vout${ABC[N:N+8]}, voutput_min);
93       vout${ABC[N:N+8]} = _mm_min_epi16(vout${ABC[N:N+8]}, voutput_max);
97         const __m128i vout${ABC[N:N+16]} = _mm_packs_epi16(vout${ABC[N:N+8]}, vout${ABC[N+8:N+16]});
99 …const __m128i vout${ABC[N:N+8]}${ABC[N:N+8]} = _mm_packs_epi16(vout${ABC[N:N+8]}, vout${ABC[N:N+8]…
102       _mm_storeu_si128((__m128i*) output, vout${ABC[0:16]});
104       _mm_storel_epi64((__m128i*) output, vout${ABC[0:8]}${ABC[0:8]});
107         _mm_storeu_si128((__m128i*) (output + ${N}), vout${ABC[N:N+16]});
109         _mm_storel_epi64((__m128i*) (output + ${N}), vout${ABC[N:N+8]}${ABC[N:N+8]});
115         const __m128i vx${ABC[0:8]} = _mm_cvtepi8_epi16(_mm_loadl_epi64((const __m128i*) input_x));
116         const __m128i vy${ABC[0:8]} = _mm_cvtepi8_epi16(_mm_loadl_epi64((const __m128i*) input_y));
118         __m128i vx${ABC[0:8]} = _mm_loadl_epi64((const __m128i*) input_x);
119         __m128i vy${ABC[0:8]} = _mm_loadl_epi64((const __m128i*) input_y);
125 …vx${ABC[0:8]} = _mm_unpacklo_epi8(vx${ABC[0:8]}, _mm_cmpgt_epi8(_mm_setzero_si128(), vx${ABC[0:8]}…
126 …vy${ABC[0:8]} = _mm_unpacklo_epi8(vy${ABC[0:8]}, _mm_cmpgt_epi8(_mm_setzero_si128(), vy${ABC[0:8]}…
128       __m128i vxprod${ABC[0:8]}hi = _mm_mulhi_epu16(vx${ABC[0:8]}, vx_multiplier_lo);
129       __m128i vyprod${ABC[0:8]}hi = _mm_mulhi_epu16(vy${ABC[0:8]}, vy_multiplier_lo);
130       const __m128i vxprod${ABC[0:8]}lo = _mm_mullo_epi16(vx${ABC[0:8]}, vx_multiplier_lo);
131       const __m128i vyprod${ABC[0:8]}lo = _mm_mullo_epi16(vy${ABC[0:8]}, vy_multiplier_lo);
133 …vxprod${ABC[0:8]}hi = _mm_add_epi16(vxprod${ABC[0:8]}hi, _mm_mullo_epi16(vx${ABC[0:8]}, vx_multipl…
134 …vyprod${ABC[0:8]}hi = _mm_add_epi16(vyprod${ABC[0:8]}hi, _mm_mullo_epi16(vy${ABC[0:8]}, vy_multipl…
136 …vxprod${ABC[0:8]}hi = _mm_sub_epi16(vxprod${ABC[0:8]}hi, _mm_and_si128(_mm_srai_epi16(vx${ABC[0:8]…
137 …vyprod${ABC[0:8]}hi = _mm_sub_epi16(vyprod${ABC[0:8]}hi, _mm_and_si128(_mm_srai_epi16(vy${ABC[0:8]…
139 …__m128i vacc${ABC[0:4]} = _mm_add_epi32(vzero_point_product, _mm_unpacklo_epi16(vxprod${ABC[0:8]}l…
140 …__m128i vacc${ABC[4:8]} = _mm_add_epi32(vzero_point_product, _mm_unpackhi_epi16(vxprod${ABC[0:8]}l…
142 …vacc${ABC[0:4]} = _mm_add_epi32(vacc${ABC[0:4]}, _mm_unpacklo_epi16(vyprod${ABC[0:8]}lo, vyprod${A…
143 …vacc${ABC[4:8]} = _mm_add_epi32(vacc${ABC[4:8]}, _mm_unpackhi_epi16(vyprod${ABC[0:8]}lo, vyprod${A…
145 …t __m128i vrem${ABC[0:4]} = _mm_add_epi32(_mm_and_si128(vacc${ABC[0:4]}, vremainder_mask), _mm_cmp…
146 …t __m128i vrem${ABC[4:8]} = _mm_add_epi32(_mm_and_si128(vacc${ABC[4:8]}, vremainder_mask), _mm_cmp…
148 …vacc${ABC[0:4]} = _mm_sub_epi32(_mm_sra_epi32(vacc${ABC[0:4]}, vshift), _mm_cmpgt_epi32(vrem${ABC[…
149 …vacc${ABC[4:8]} = _mm_sub_epi32(_mm_sra_epi32(vacc${ABC[4:8]}, vshift), _mm_cmpgt_epi32(vrem${ABC[…
151 …__m128i vout${ABC[0:8]} = _mm_adds_epi16(_mm_packs_epi32(vacc${ABC[0:4]}, vacc${ABC[4:8]}), voutpu…
152       vout${ABC[0:8]} = _mm_max_epi16(vout${ABC[0:8]}, voutput_min);
153       vout${ABC[0:8]} = _mm_min_epi16(vout${ABC[0:8]}, voutput_max);
155       __m128i vout${ABC[0:8]}${ABC[0:8]} = _mm_packs_epi16(vout${ABC[0:8]}, vout${ABC[0:8]});
159           _mm_storel_epi64((__m128i*) output, vout${ABC[0:8]}${ABC[0:8]});
164             *((uint32_t*) output) = (uint32_t) _mm_cvtsi128_si32(vout${ABC[0:8]}${ABC[0:8]});
165             vout${ABC[0:8]}${ABC[0:8]} = _mm_srli_epi64(vout${ABC[0:8]}${ABC[0:8]}, 32);
169             *((uint16_t*) output) = (uint16_t) _mm_extract_epi16(vout${ABC[0:8]}${ABC[0:8]}, 0);
170             vout${ABC[0:8]}${ABC[0:8]} = _mm_srli_epi32(vout${ABC[0:8]}${ABC[0:8]}, 16);
175               *output = (int8_t) _mm_extract_epi8(vout${ABC[0:8]}${ABC[0:8]}, 0);
177               *output = (int32_t) _mm_cvtsi128_si32(vout${ABC[0:8]}${ABC[0:8]});
183           *((uint32_t*) output) = (uint32_t) _mm_cvtsi128_si32(vout${ABC[0:8]}${ABC[0:8]});
184           vout${ABC[0:8]}${ABC[0:8]} = _mm_srli_epi64(vout${ABC[0:8]}${ABC[0:8]}, 32);
188           *((uint16_t*) output) = (uint16_t) _mm_extract_epi16(vout${ABC[0:8]}${ABC[0:8]}, 0);
189           vout${ABC[0:8]}${ABC[0:8]} = _mm_srli_epi32(vout${ABC[0:8]}${ABC[0:8]}, 16);
194             *output = (int8_t) _mm_extract_epi8(vout${ABC[0:8]}${ABC[0:8]}, 0);
196             *output = (int32_t) _mm_cvtsi128_si32(vout${ABC[0:8]}${ABC[0:8]});