avx2intrin.h File Reference

Go to the source code of this file.

Macros
#define	__DEFAULT_FN_ATTRS   __attribute__((__always_inline__, __nodebug__, __target__("avx2")))
#define	_mm256_mpsadbw_epu8(X, Y, M)   __builtin_ia32_mpsadbw256((X), (Y), (M))
#define	_mm256_alignr_epi8(a, b, n)
#define	_mm256_blend_epi16(V1, V2, M)
#define	_mm256_shuffle_epi32(a, imm)
#define	_mm256_shufflehi_epi16(a, imm)
#define	_mm256_shufflelo_epi16(a, imm)
#define	_mm256_slli_si256(a, count)
#define	_mm256_bslli_epi128(a, count)   _mm256_slli_si256((a), (count))
#define	_mm256_srli_si256(a, count)
#define	_mm256_bsrli_epi128(a, count)   _mm256_srli_si256((a), (count))
#define	_mm_blend_epi32(V1, V2, M)
#define	_mm256_blend_epi32(V1, V2, M)
#define	_mm256_permute4x64_pd(V, M)
#define	_mm256_permute4x64_epi64(V, M)
#define	_mm256_permute2x128_si256(V1, V2, M)
#define	_mm256_extracti128_si256(V, M)
#define	_mm256_inserti128_si256(V1, V2, M)
#define	_mm_mask_i32gather_pd(a, m, i, mask, s)
#define	_mm256_mask_i32gather_pd(a, m, i, mask, s)
#define	_mm_mask_i64gather_pd(a, m, i, mask, s)
#define	_mm256_mask_i64gather_pd(a, m, i, mask, s)
#define	_mm_mask_i32gather_ps(a, m, i, mask, s)
#define	_mm256_mask_i32gather_ps(a, m, i, mask, s)
#define	_mm_mask_i64gather_ps(a, m, i, mask, s)
#define	_mm256_mask_i64gather_ps(a, m, i, mask, s)
#define	_mm_mask_i32gather_epi32(a, m, i, mask, s)
#define	_mm256_mask_i32gather_epi32(a, m, i, mask, s)
#define	_mm_mask_i64gather_epi32(a, m, i, mask, s)
#define	_mm256_mask_i64gather_epi32(a, m, i, mask, s)
#define	_mm_mask_i32gather_epi64(a, m, i, mask, s)
#define	_mm256_mask_i32gather_epi64(a, m, i, mask, s)
#define	_mm_mask_i64gather_epi64(a, m, i, mask, s)
#define	_mm256_mask_i64gather_epi64(a, m, i, mask, s)
#define	_mm_i32gather_pd(m, i, s)
#define	_mm256_i32gather_pd(m, i, s)
#define	_mm_i64gather_pd(m, i, s)
#define	_mm256_i64gather_pd(m, i, s)
#define	_mm_i32gather_ps(m, i, s)
#define	_mm256_i32gather_ps(m, i, s)
#define	_mm_i64gather_ps(m, i, s)
#define	_mm256_i64gather_ps(m, i, s)
#define	_mm_i32gather_epi32(m, i, s)
#define	_mm256_i32gather_epi32(m, i, s)
#define	_mm_i64gather_epi32(m, i, s)
#define	_mm256_i64gather_epi32(m, i, s)
#define	_mm_i32gather_epi64(m, i, s)
#define	_mm256_i32gather_epi64(m, i, s)
#define	_mm_i64gather_epi64(m, i, s)
#define	_mm256_i64gather_epi64(m, i, s)

Macro Definition Documentation

#define __DEFAULT_FN_ATTRS   __attribute__((__always_inline__, __nodebug__, __target__("avx2")))

#define _mm256_alignr_epi8	(		a,
			b,
			n
	)

Value:

__extension__ ({        \
  (__m256i)__builtin_ia32_palignr256((__v32qi)(__m256i)(a), \
                                     (__v32qi)(__m256i)(b), (n)); })

#define _mm256_blend_epi16	(		V1,
			V2,
			M
	)

Value:

__extension__ ({       \
  (__m256i)__builtin_shufflevector((__v16hi)(__m256i)(V1),   \
                                   (__v16hi)(__m256i)(V2),   \
                                   (((M) & 0x01) ? 16 : 0),  \
                                   (((M) & 0x02) ? 17 : 1),  \
                                   (((M) & 0x04) ? 18 : 2),  \
                                   (((M) & 0x08) ? 19 : 3),  \
                                   (((M) & 0x10) ? 20 : 4),  \
                                   (((M) & 0x20) ? 21 : 5),  \
                                   (((M) & 0x40) ? 22 : 6),  \
                                   (((M) & 0x80) ? 23 : 7),  \
                                   (((M) & 0x01) ? 24 : 8),  \
                                   (((M) & 0x02) ? 25 : 9),  \
                                   (((M) & 0x04) ? 26 : 10), \
                                   (((M) & 0x08) ? 27 : 11), \
                                   (((M) & 0x10) ? 28 : 12), \
                                   (((M) & 0x20) ? 29 : 13), \
                                   (((M) & 0x40) ? 30 : 14), \
                                   (((M) & 0x80) ? 31 : 15)); })

#define _mm256_blend_epi32	(		V1,
			V2,
			M
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_shufflevector((__v8si)(__m256i)(V1),   \
                                   (__v8si)(__m256i)(V2),   \
                                   (((M) & 0x01) ?  8 : 0), \
                                   (((M) & 0x02) ?  9 : 1), \
                                   (((M) & 0x04) ? 10 : 2), \
                                   (((M) & 0x08) ? 11 : 3), \
                                   (((M) & 0x10) ? 12 : 4), \
                                   (((M) & 0x20) ? 13 : 5), \
                                   (((M) & 0x40) ? 14 : 6), \
                                   (((M) & 0x80) ? 15 : 7)); })

#define _mm256_bslli_epi128	(		a,
			count
	)		_mm256_slli_si256((a), (count))

#define _mm256_bsrli_epi128	(		a,
			count
	)		_mm256_srli_si256((a), (count))

#define _mm256_extracti128_si256	(		V,
			M
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_shufflevector((__v4di)(__m256i)(V), \
                                   (__v4di)_mm256_setzero_si256(), \
                                   (((M) & 1) ? 2 : 0), \
                                   (((M) & 1) ? 3 : 1) ); })

#define _mm256_i32gather_epi32	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_gatherd_d256((__v8si)_mm256_undefined_si256(), \
                                       (int const *)(m), (__v8si)(__m256i)(i), \
                                       (__v8si)_mm256_set1_epi32(-1), (s)); })

#define _mm256_i32gather_epi64	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_gatherd_q256((__v4di)_mm256_undefined_si256(), \
                                       (long long const *)(m), \
                                       (__v4si)(__m128i)(i), \
                                       (__v4di)_mm256_set1_epi64x(-1), (s)); })

#define _mm256_i32gather_pd	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m256d)__builtin_ia32_gatherd_pd256((__v4df)_mm256_undefined_pd(), \
                                        (double const *)(m), \
                                        (__v4si)(__m128i)(i), \
                                        (__v4df)_mm256_cmp_pd(_mm256_setzero_pd(), \
                                                              _mm256_setzero_pd(), \
                                                              _CMP_EQ_OQ), \
                                        (s)); })

#define _mm256_i32gather_ps	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m256)__builtin_ia32_gatherd_ps256((__v8sf)_mm256_undefined_ps(), \
                                       (float const *)(m), \
                                       (__v8si)(__m256i)(i), \
                                       (__v8sf)_mm256_cmp_ps(_mm256_setzero_ps(), \
                                                             _mm256_setzero_ps(), \
                                                             _CMP_EQ_OQ), \
                                       (s)); })

#define _mm256_i64gather_epi32	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherq_d256((__v4si)_mm_undefined_si128(), \
                                       (int const *)(m), (__v4di)(__m256i)(i), \
                                       (__v4si)_mm_set1_epi32(-1), (s)); })

#define _mm256_i64gather_epi64	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_gatherq_q256((__v4di)_mm256_undefined_si256(), \
                                       (long long const *)(m), \
                                       (__v4di)(__m256i)(i), \
                                       (__v4di)_mm256_set1_epi64x(-1), (s)); })

#define _mm256_i64gather_pd	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m256d)__builtin_ia32_gatherq_pd256((__v4df)_mm256_undefined_pd(), \
                                        (double const *)(m), \
                                        (__v4di)(__m256i)(i), \
                                        (__v4df)_mm256_cmp_pd(_mm256_setzero_pd(), \
                                                              _mm256_setzero_pd(), \
                                                              _CMP_EQ_OQ), \
                                        (s)); })

#define _mm256_i64gather_ps	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128)__builtin_ia32_gatherq_ps256((__v4sf)_mm_undefined_ps(), \
                                       (float const *)(m), \
                                       (__v4di)(__m256i)(i), \
                                       (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
                                                            _mm_setzero_ps()), \
                                       (s)); })

#define _mm256_inserti128_si256	(		V1,
			V2,
			M
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_shufflevector((__v4di)(__m256i)(V1), \
                                   (__v4di)_mm256_castsi128_si256((__m128i)(V2)), \
                                   (((M) & 1) ? 0 : 4), \
                                   (((M) & 1) ? 1 : 5), \
                                   (((M) & 1) ? 4 : 2), \
                                   (((M) & 1) ? 5 : 3) ); })

#define _mm256_mask_i32gather_epi32	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_gatherd_d256((__v8si)(__m256i)(a), \
                                       (int const *)(m), \
                                       (__v8si)(__m256i)(i), \
                                       (__v8si)(__m256i)(mask), (s)); })

#define _mm256_mask_i32gather_epi64	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_gatherd_q256((__v4di)(__m256i)(a), \
                                       (long long const *)(m), \
                                       (__v4si)(__m128i)(i), \
                                       (__v4di)(__m256i)(mask), (s)); })

#define _mm256_mask_i32gather_pd	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m256d)__builtin_ia32_gatherd_pd256((__v4df)(__m256d)(a), \
                                        (double const *)(m), \
                                        (__v4si)(__m128i)(i), \
                                        (__v4df)(__m256d)(mask), (s)); })

#define _mm256_mask_i32gather_ps	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m256)__builtin_ia32_gatherd_ps256((__v8sf)(__m256)(a), \
                                       (float const *)(m), \
                                       (__v8si)(__m256i)(i), \
                                       (__v8sf)(__m256)(mask), (s)); })

#define _mm256_mask_i64gather_epi32	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherq_d256((__v4si)(__m128i)(a), \
                                       (int const *)(m), \
                                       (__v4di)(__m256i)(i), \
                                       (__v4si)(__m128i)(mask), (s)); })

#define _mm256_mask_i64gather_epi64	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_gatherq_q256((__v4di)(__m256i)(a), \
                                       (long long const *)(m), \
                                       (__v4di)(__m256i)(i), \
                                       (__v4di)(__m256i)(mask), (s)); })

#define _mm256_mask_i64gather_pd	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m256d)__builtin_ia32_gatherq_pd256((__v4df)(__m256d)(a), \
                                        (double const *)(m), \
                                        (__v4di)(__m256i)(i), \
                                        (__v4df)(__m256d)(mask), (s)); })

#define _mm256_mask_i64gather_ps	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128)__builtin_ia32_gatherq_ps256((__v4sf)(__m128)(a), \
                                       (float const *)(m), \
                                       (__v4di)(__m256i)(i), \
                                       (__v4sf)(__m128)(mask), (s)); })

#define _mm256_mpsadbw_epu8	(		X,
			Y,
			M
	)		__builtin_ia32_mpsadbw256((X), (Y), (M))

#define _mm256_permute2x128_si256	(		V1,
			V2,
			M
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_permti256((__m256i)(V1), (__m256i)(V2), (M)); })

#define _mm256_permute4x64_epi64	(		V,
			M
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_shufflevector((__v4di)(__m256i)(V), \
                                   (__v4di)_mm256_setzero_si256(), \
                                   (M) & 0x3, ((M) & 0xc) >> 2, \
                                   ((M) & 0x30) >> 4, ((M) & 0xc0) >> 6); })

#define _mm256_permute4x64_pd	(		V,
			M
	)

Value:

__extension__ ({ \
  (__m256d)__builtin_shufflevector((__v4df)(__m256d)(V), \
                                   (__v4df)_mm256_setzero_pd(), \
                                   (M) & 0x3, ((M) & 0xc) >> 2, \
                                   ((M) & 0x30) >> 4, ((M) & 0xc0) >> 6); })

#define _mm256_shuffle_epi32	(		a,
			imm
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_shufflevector((__v8si)(__m256i)(a), \
                                   (__v8si)_mm256_setzero_si256(), \
                                   (imm) & 0x3, ((imm) & 0xc) >> 2, \
                                   ((imm) & 0x30) >> 4, ((imm) & 0xc0) >> 6, \
                                   4 + (((imm) & 0x03) >> 0), \
                                   4 + (((imm) & 0x0c) >> 2), \
                                   4 + (((imm) & 0x30) >> 4), \
                                   4 + (((imm) & 0xc0) >> 6)); })

#define _mm256_shufflehi_epi16	(		a,
			imm
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_shufflevector((__v16hi)(__m256i)(a), \
                                   (__v16hi)_mm256_setzero_si256(), \
                                   0, 1, 2, 3, \
                                   4 + (((imm) & 0x03) >> 0), \
                                   4 + (((imm) & 0x0c) >> 2), \
                                   4 + (((imm) & 0x30) >> 4), \
                                   4 + (((imm) & 0xc0) >> 6), \
                                   8, 9, 10, 11, \
                                   12 + (((imm) & 0x03) >> 0), \
                                   12 + (((imm) & 0x0c) >> 2), \
                                   12 + (((imm) & 0x30) >> 4), \
                                   12 + (((imm) & 0xc0) >> 6)); })

#define _mm256_shufflelo_epi16	(		a,
			imm
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_shufflevector((__v16hi)(__m256i)(a), \
                                   (__v16hi)_mm256_setzero_si256(), \
                                   (imm) & 0x3,((imm) & 0xc) >> 2, \
                                   ((imm) & 0x30) >> 4, ((imm) & 0xc0) >> 6, \
                                   4, 5, 6, 7, \
                                   8 + (((imm) & 0x03) >> 0), \
                                   8 + (((imm) & 0x0c) >> 2), \
                                   8 + (((imm) & 0x30) >> 4), \
                                   8 + (((imm) & 0xc0) >> 6), \
                                   12, 13, 14, 15); })

#define _mm256_slli_si256	(		a,
			count
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_pslldqi256((__m256i)(a), (count)*8); })

#define _mm256_srli_si256	(		a,
			count
	)

Value:

__extension__ ({ \
  (__m256i)__builtin_ia32_psrldqi256((__m256i)(a), (count)*8); })

#define _mm_blend_epi32	(		V1,
			V2,
			M
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_shufflevector((__v4si)(__m128i)(V1),  \
                                   (__v4si)(__m128i)(V2),  \
                                   (((M) & 0x01) ? 4 : 0), \
                                   (((M) & 0x02) ? 5 : 1), \
                                   (((M) & 0x04) ? 6 : 2), \
                                   (((M) & 0x08) ? 7 : 3)); })

#define _mm_i32gather_epi32	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherd_d((__v4si)_mm_undefined_si128(), \
                                    (int const *)(m), (__v4si)(__m128i)(i), \
                                    (__v4si)_mm_set1_epi32(-1), (s)); })

#define _mm_i32gather_epi64	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherd_q((__v2di)_mm_undefined_si128(), \
                                    (long long const *)(m), \
                                    (__v4si)(__m128i)(i), \
                                    (__v2di)_mm_set1_epi64x(-1), (s)); })

#define _mm_i32gather_pd	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128d)__builtin_ia32_gatherd_pd((__v2df)_mm_undefined_pd(), \
                                     (double const *)(m), \
                                     (__v4si)(__m128i)(i), \
                                     (__v2df)_mm_cmpeq_pd(_mm_setzero_pd(), \
                                                          _mm_setzero_pd()), \
                                     (s)); })

#define _mm_i32gather_ps	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128)__builtin_ia32_gatherd_ps((__v4sf)_mm_undefined_ps(), \
                                    (float const *)(m), \
                                    (__v4si)(__m128i)(i), \
                                    (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
                                                         _mm_setzero_ps()), \
                                    (s)); })

#define _mm_i64gather_epi32	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherq_d((__v4si)_mm_undefined_si128(), \
                                    (int const *)(m), (__v2di)(__m128i)(i), \
                                    (__v4si)_mm_set1_epi32(-1), (s)); })

#define _mm_i64gather_epi64	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherq_q((__v2di)_mm_undefined_si128(), \
                                    (long long const *)(m), \
                                    (__v2di)(__m128i)(i), \
                                    (__v2di)_mm_set1_epi64x(-1), (s)); })

#define _mm_i64gather_pd	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128d)__builtin_ia32_gatherq_pd((__v2df)_mm_undefined_pd(), \
                                     (double const *)(m), \
                                     (__v2di)(__m128i)(i), \
                                     (__v2df)_mm_cmpeq_pd(_mm_setzero_pd(), \
                                                          _mm_setzero_pd()), \
                                     (s)); })

#define _mm_i64gather_ps	(		m,
			i,
			s
	)

Value:

__extension__ ({ \
  (__m128)__builtin_ia32_gatherq_ps((__v4sf)_mm_undefined_ps(), \
                                    (float const *)(m), \
                                    (__v2di)(__m128i)(i), \
                                    (__v4sf)_mm_cmpeq_ps(_mm_setzero_ps(), \
                                                         _mm_setzero_ps()), \
                                    (s)); })

#define _mm_mask_i32gather_epi32	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherd_d((__v4si)(__m128i)(a), \
                                    (int const *)(m), \
                                    (__v4si)(__m128i)(i), \
                                    (__v4si)(__m128i)(mask), (s)); })

#define _mm_mask_i32gather_epi64	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherd_q((__v2di)(__m128i)(a), \
                                    (long long const *)(m), \
                                    (__v4si)(__m128i)(i), \
                                    (__v2di)(__m128i)(mask), (s)); })

#define _mm_mask_i32gather_pd	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128d)__builtin_ia32_gatherd_pd((__v2df)(__m128i)(a), \
                                     (double const *)(m), \
                                     (__v4si)(__m128i)(i), \
                                     (__v2df)(__m128d)(mask), (s)); })

#define _mm_mask_i32gather_ps	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128)__builtin_ia32_gatherd_ps((__v4sf)(__m128)(a), \
                                    (float const *)(m), \
                                    (__v4si)(__m128i)(i), \
                                    (__v4sf)(__m128)(mask), (s)); })

#define _mm_mask_i64gather_epi32	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherq_d((__v4si)(__m128i)(a), \
                                    (int const *)(m), \
                                    (__v2di)(__m128i)(i), \
                                    (__v4si)(__m128i)(mask), (s)); })

#define _mm_mask_i64gather_epi64	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128i)__builtin_ia32_gatherq_q((__v2di)(__m128i)(a), \
                                    (long long const *)(m), \
                                    (__v2di)(__m128i)(i), \
                                    (__v2di)(__m128i)(mask), (s)); })

#define _mm_mask_i64gather_pd	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128d)__builtin_ia32_gatherq_pd((__v2df)(__m128d)(a), \
                                     (double const *)(m), \
                                     (__v2di)(__m128i)(i), \
                                     (__v2df)(__m128d)(mask), (s)); })

#define _mm_mask_i64gather_ps	(		a,
			m,
			i,
			mask,
			s
	)

Value:

__extension__ ({ \
  (__m128)__builtin_ia32_gatherq_ps((__v4sf)(__m128)(a), \
                                    (float const *)(m), \
                                    (__v2di)(__m128i)(i), \
                                    (__v4sf)(__m128)(mask), (s)); })