| Damjan Marion | c576622 | 2018-04-16 00:18:34 +0200 | [diff] [blame] | 1 | /* |
| 2 | * Copyright (c) 2018 Cisco and/or its affiliates. |
| 3 | * Licensed under the Apache License, Version 2.0 (the "License"); |
| 4 | * you may not use this file except in compliance with the License. |
| 5 | * You may obtain a copy of the License at: |
| 6 | * |
| 7 | * http://www.apache.org/licenses/LICENSE-2.0 |
| 8 | * |
| 9 | * Unless required by applicable law or agreed to in writing, software |
| 10 | * distributed under the License is distributed on an "AS IS" BASIS, |
| 11 | * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 12 | * See the License for the specific language governing permissions and |
| 13 | * limitations under the License. |
| 14 | */ |
| 15 | |
| 16 | #ifndef included_vector_avx2_h |
| 17 | #define included_vector_avx2_h |
| 18 | |
| 19 | #include <vppinfra/clib.h> |
| 20 | #include <x86intrin.h> |
| 21 | |
| 22 | #define foreach_avx2_vec256i \ |
| 23 | _(i,8,32,epi8) _(i,16,16,epi16) _(i,32,8,epi32) _(i,64,4,epi64x) |
| 24 | #define foreach_avx2_vec256u \ |
| 25 | _(u,8,32,epi8) _(u,16,16,epi16) _(u,32,8,epi32) _(u,64,4,epi64x) |
| 26 | #define foreach_avx2_vec256f \ |
| 27 | _(f,32,8,ps) _(f,64,4,pd) |
| 28 | |
| 29 | /* splat, load_unaligned, store_unaligned, is_all_zero, is_all_equal */ |
| 30 | #define _(t, s, c, i) \ |
| 31 | static_always_inline t##s##x##c \ |
| 32 | t##s##x##c##_splat (t##s x) \ |
| 33 | { return (t##s##x##c) _mm256_set1_##i (x); } \ |
| 34 | \ |
| 35 | static_always_inline t##s##x##c \ |
| 36 | t##s##x##c##_load_unaligned (void *p) \ |
| 37 | { return (t##s##x##c) _mm256_loadu_si256 (p); } \ |
| 38 | \ |
| 39 | static_always_inline void \ |
| 40 | t##s##x##c##_store_unaligned (t##s##x##c v, void *p) \ |
| 41 | { _mm256_storeu_si256 ((__m256i *) p, (__m256i) v); } \ |
| 42 | \ |
| 43 | static_always_inline int \ |
| 44 | t##s##x##c##_is_all_zero (t##s##x##c x) \ |
| 45 | { return _mm256_testz_si256 ((__m256i) x, (__m256i) x); } \ |
| 46 | \ |
| 47 | static_always_inline int \ |
| 48 | t##s##x##c##_is_all_equal (t##s##x##c v, t##s x) \ |
| 49 | { return t##s##x##c##_is_all_zero (v != t##s##x##c##_splat (x)); }; \ |
| 50 | \ |
| 51 | |
| 52 | foreach_avx2_vec256i foreach_avx2_vec256u |
| 53 | #undef _ |
| 54 | always_inline u32x8 |
| 55 | u32x8_permute (u32x8 v, u32x8 idx) |
| 56 | { |
| 57 | return (u32x8) _mm256_permutevar8x32_epi32 ((__m256i) v, (__m256i) idx); |
| 58 | } |
| 59 | |
| 60 | always_inline u32x4 |
| 61 | u32x8_extract_lo (u32x8 v) |
| 62 | { |
| 63 | return (u32x4) _mm256_extracti128_si256 ((__m256i) v, 0); |
| 64 | } |
| 65 | |
| 66 | always_inline u32x4 |
| 67 | u32x8_extract_hi (u32x8 v) |
| 68 | { |
| 69 | return (u32x4) _mm256_extracti128_si256 ((__m256i) v, 1); |
| 70 | } |
| 71 | |
| 72 | #endif /* included_vector_avx2_h */ |
| 73 | |
| 74 | /* |
| 75 | * fd.io coding-style-patch-verification: ON |
| 76 | * |
| 77 | * Local Variables: |
| 78 | * eval: (c-set-style "gnu") |
| 79 | * End: |
| 80 | */ |