lib/Headers/avxvnniint8intrin.h

25 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
33 /// _mm_dpbssd_epi32(__m128i __W, __m128i __A, __m128i __B);
38 /// \param __A
47 /// 	tmp1.word := SignExtend16(__A.byte[4*j]) * SignExtend16(__B.byte[4*j])
48 /// 	tmp2.word := SignExtend16(__A.byte[4*j+1]) * SignExtend16(__B.byte[4*j+1])
49 /// 	tmp3.word := SignExtend16(__A.byte[4*j+2]) * SignExtend16(__B.byte[4*j+2])
50 /// 	tmp4.word := SignExtend16(__A.byte[4*j+3]) * SignExtend16(__B.byte[4*j+3])
56                                                                  __m128i __A,
58   return (__m128i)__builtin_ia32_vpdpbssd128((__v4si)__W, (__v4si)__A,
62 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
70 /// _mm256_dpbssd_epi32(__m256i __W, __m256i __A, __m256i __B);
75 /// \param __A
84 /// 	tmp1.word := SignExtend16(__A.byte[4*j]) * SignExtend16(__B.byte[4*j])
85 /// 	tmp2.word := SignExtend16(__A.byte[4*j+1]) * SignExtend16(__B.byte[4*j+1])
86 /// 	tmp3.word := SignExtend16(__A.byte[4*j+2]) * SignExtend16(__B.byte[4*j+2])
87 /// 	tmp4.word := SignExtend16(__A.byte[4*j+3]) * SignExtend16(__B.byte[4*j+3])
93 _mm256_dpbssd_epi32(__m256i __W, __m256i __A, __m256i __B) {
94   return (__m256i)__builtin_ia32_vpdpbssd256((__v8si)__W, (__v8si)__A,
98 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
107 /// _mm_dpbssds_epi32( __m128i __W, __m128i __A, __m128i __B);
112 /// \param __A
121 /// 	tmp1.word := SignExtend16(__A.byte[4*j]) * SignExtend16(__B.byte[4*j])
122 /// 	tmp2.word := SignExtend16(__A.byte[4*j+1]) * SignExtend16(__B.byte[4*j+1])
123 /// 	tmp3.word := SignExtend16(__A.byte[4*j+2]) * SignExtend16(__B.byte[4*j+2])
124 /// 	tmp4.word := SignExtend16(__A.byte[4*j+3]) * SignExtend16(__B.byte[4*j+3])
130                                                                   __m128i __A,
132   return (__m128i)__builtin_ia32_vpdpbssds128((__v4si)__W, (__v4si)__A,
136 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
145 /// _mm256_dpbssds_epi32(__m256i __W, __m256i __A, __m256i __B);
150 /// \param __A
159 /// 	tmp1.word := SignExtend16(__A.byte[4*j]) * SignExtend16(__B.byte[4*j])
160 /// 	tmp2.word := SignExtend16(__A.byte[4*j+1]) * SignExtend16(__B.byte[4*j+1])
161 /// 	tmp3.word := SignExtend16(__A.byte[4*j+2]) * SignExtend16(__B.byte[4*j+2])
162 /// 	tmp4.word := SignExtend16(__A.byte[4*j+3]) * SignExtend16(__B.byte[4*j+3])
168 _mm256_dpbssds_epi32(__m256i __W, __m256i __A, __m256i __B) {
169   return (__m256i)__builtin_ia32_vpdpbssds256((__v8si)__W, (__v8si)__A,
173 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
181 /// _mm_dpbsud_epi32(__m128i __W, __m128i __A, __m128i __B);
186 /// \param __A
195 /// 	tmp1.word := Signed(SignExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j]))
196 /// 	tmp2.word := Signed(SignExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1]))
197 /// 	tmp3.word := Signed(SignExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2]))
198 /// 	tmp4.word := Signed(SignExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3]))
204                                                                  __m128i __A,
206   return (__m128i)__builtin_ia32_vpdpbsud128((__v4si)__W, (__v4si)__A,
210 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
218 /// _mm256_dpbsud_epi32(__m256i __W, __m256i __A, __m256i __B);
223 /// \param __A
232 /// 	tmp1.word := Signed(SignExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j]))
233 /// 	tmp2.word := Signed(SignExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1]))
234 /// 	tmp3.word := Signed(SignExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2]))
235 /// 	tmp4.word := Signed(SignExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3]))
241 _mm256_dpbsud_epi32(__m256i __W, __m256i __A, __m256i __B) {
242   return (__m256i)__builtin_ia32_vpdpbsud256((__v8si)__W, (__v8si)__A,
246 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
255 /// _mm_dpbsuds_epi32( __m128i __W, __m128i __A, __m128i __B);
260 /// \param __A
269 /// 	tmp1.word := Signed(SignExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j]))
270 /// 	tmp2.word := Signed(SignExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1]))
271 /// 	tmp3.word := Signed(SignExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2]))
272 /// 	tmp4.word := Signed(SignExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3]))
278                                                                   __m128i __A,
280   return (__m128i)__builtin_ia32_vpdpbsuds128((__v4si)__W, (__v4si)__A,
284 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
293 /// _mm256_dpbsuds_epi32(__m256i __W, __m256i __A, __m256i __B);
298 /// \param __A
307 /// 	tmp1.word := Signed(SignExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j]))
308 /// 	tmp2.word := Signed(SignExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1]))
309 /// 	tmp3.word := Signed(SignExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2]))
310 /// 	tmp4.word := Signed(SignExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3]))
316 _mm256_dpbsuds_epi32(__m256i __W, __m256i __A, __m256i __B) {
317   return (__m256i)__builtin_ia32_vpdpbsuds256((__v8si)__W, (__v8si)__A,
321 /// Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in \a __A with
329 /// _mm_dpbuud_epi32(__m128i __W, __m128i __A, __m128i __B);
334 /// \param __A
343 /// 	tmp1.word := ZeroExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j])
344 /// 	tmp2.word := ZeroExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1])
345 /// 	tmp3.word := ZeroExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2])
346 /// 	tmp4.word := ZeroExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3])
352                                                                  __m128i __A,
354   return (__m128i)__builtin_ia32_vpdpbuud128((__v4si)__W, (__v4si)__A,
358 /// Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in \a __A with
366 /// _mm256_dpbuud_epi32(__m256i __W, __m256i __A, __m256i __B);
371 /// \param __A
380 /// 	tmp1.word := ZeroExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j])
381 /// 	tmp2.word := ZeroExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1])
382 /// 	tmp3.word := ZeroExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2])
383 /// 	tmp4.word := ZeroExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3])
389 _mm256_dpbuud_epi32(__m256i __W, __m256i __A, __m256i __B) {
390   return (__m256i)__builtin_ia32_vpdpbuud256((__v8si)__W, (__v8si)__A,
394 /// Multiply groups of 4 adjacent pairs of unsigned 8-bit integers in \a __A with
403 /// _mm_dpbuuds_epi32( __m128i __W, __m128i __A, __m128i __B);
408 /// \param __A
417 /// 	tmp1.word := ZeroExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j])
418 /// 	tmp2.word := ZeroExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1])
419 /// 	tmp3.word := ZeroExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2])
420 /// 	tmp4.word := ZeroExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3])
426                                                                   __m128i __A,
428   return (__m128i)__builtin_ia32_vpdpbuuds128((__v4si)__W, (__v4si)__A,
432 /// Multiply groups of 4 adjacent pairs of signed 8-bit integers in \a __A with
441 /// _mm256_dpbuuds_epi32(__m256i __W, __m256i __A, __m256i __B);
446 /// \param __A
455 /// 	tmp1.word := ZeroExtend16(__A.byte[4*j]) * ZeroExtend16(__B.byte[4*j])
456 /// 	tmp2.word := ZeroExtend16(__A.byte[4*j+1]) * ZeroExtend16(__B.byte[4*j+1])
457 /// 	tmp3.word := ZeroExtend16(__A.byte[4*j+2]) * ZeroExtend16(__B.byte[4*j+2])
458 /// 	tmp4.word := ZeroExtend16(__A.byte[4*j+3]) * ZeroExtend16(__B.byte[4*j+3])
464 _mm256_dpbuuds_epi32(__m256i __W, __m256i __A, __m256i __B) {
465   return (__m256i)__builtin_ia32_vpdpbuuds256((__v8si)__W, (__v8si)__A,