Implementation notes: amd64, hydra4, crypto_aead/grain128aeadv2
Computer: hydra4
Microarchitecture: amd64; K10 32nm (300f10)
Architecture: amd64
CPU ID: AuthenticAMD-00300f10-178bfbff
SUPERCOP version: 20240625
Operation: crypto_aead
Primitive: grain128aeadv2
| Time | Object size | Test size | Implementation | Compiler | Benchmark date | SUPERCOP version |
|---|
| 462801 | 17675 0 0 | 34077 844 1016 | T:x64 | clang++_-march=native_-O3_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240629 | 20240625 |
| 463433 | 21202 0 0 | 36733 844 1016 | T:x64 | clang++_-march=native_-O2_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240629 | 20240625 |
| 486420 | 7221 0 0 | 21984 780 1112 | T:x64 | g++_-march=native_-mtune=native_-O2_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240629 | 20240625 |
| 488028 | 6823 0 0 | 20903 772 1112 | T:x64 | g++_-march=native_-mtune=native_-O_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240629 | 20240625 |
| 563345 | 2883 0 0 | 16907 836 1016 | T:x64 | clang++_-march=native_-O_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240629 | 20240625 |
| 642379 | 2514 0 0 | 15491 756 1080 | T:x64 | g++_-march=native_-mtune=native_-Os_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240629 | 20240625 |
Checksum failure
7d0b0c2792c151f313da2a4c4f1e9f6ce9e0620579deaee3411fe08dabcaf63d
Number of similar (implementation,compiler) pairs: 1, namely:
| Implementation | Compiler |
| T:x64 | clang++ -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
Checksum failure
d0dd6234d89dbd95039a8c1c4e9e8f6fa0d58228afa4fda263447f76c3102a36
Number of similar (implementation,compiler) pairs: 1, namely:
| Implementation | Compiler |
| T:x64 | g++ -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
Compiler output
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:236:6: error: '__builtin_ia32_pternlogq128_mask' needs target feature avx512vl
grain128aead-v2_opt.cpp: T = _xorand3(_mm_srli_epi16(X, 1), X, _mm_set1_epi16(0x2222));
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:125:27: note: expanded from macro '_xorand3'
grain128aead-v2_opt.cpp: #define _xorand3(a, b, c) _mm_ternarylogic_epi64(a, b, c, 0x28)
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx512vlintrin.h:6565:12: note: expanded from macro '_mm_ternarylogic_epi64'
grain128aead-v2_opt.cpp: (__m128i)__builtin_ia32_pternlogq128_mask((__v2di)(__m128i)(A), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:237:6: error: '__builtin_ia32_pternlogq128_mask' needs target feature avx512vl
grain128aead-v2_opt.cpp: X = _xor3(_mm_slli_epi16(T, 1), X, T);
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:122:25: note: expanded from macro '_xor3'
grain128aead-v2_opt.cpp: #define _xor3(a, b, c) _mm_ternarylogic_epi64(a, b, c, 0x96)
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx512vlintrin.h:6565:12: note: expanded from macro '_mm_ternarylogic_epi64'
grain128aead-v2_opt.cpp: (__m128i)__builtin_ia32_pternlogq128_mask((__v2di)(__m128i)(A), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:238:6: error: '__builtin_ia32_pternlogq128_mask' needs target feature avx512vl
grain128aead-v2_opt.cpp: T = _xorand3(_mm_srli_epi16(X, 2), X, _mm_set1_epi16(0x0c0c));
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:125:27: note: expanded from macro '_xorand3'
grain128aead-v2_opt.cpp: #define _xorand3(a, b, c) _mm_ternarylogic_epi64(a, b, c, 0x28)
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx512vlintrin.h:6565:12: note: expanded from macro '_mm_ternarylogic_epi64'
grain128aead-v2_opt.cpp: ...
Number of similar (implementation,compiler) pairs: 4, namely:
| Implementation | Compiler |
| T:avx512 | clang++ -march=native -O2 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
| T:avx512 | clang++ -march=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
| T:avx512 | clang++ -march=native -O -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall |
| T:avx512 | clang++ -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
Compiler output
grain128aead-v2_opt.cpp: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:63,
grain128aead-v2_opt.cpp: from /usr/lib/gcc/x86_64-linux-gnu/10/include/x86intrin.h:32,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.h:41,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.cpp:10:
grain128aead-v2_opt.cpp: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx512vlintrin.h: In function 'u64 grain_keystream64(grain_ctx*)':
grain128aead-v2_opt.cpp: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx512vlintrin.h:10568:1: error: inlining failed in call to 'always_inline' '__m128i _mm_ternarylogic_epi64(__m128i, __m128i, __m128i, int)': target specific option mismatch
grain128aead-v2_opt.cpp: 10568 | _mm_ternarylogic_epi64 (__m128i __A, __m128i __B, __m128i __C,
grain128aead-v2_opt.cpp: | ^~~~~~~~~~~~~~~~~~~~~~
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:207:32: note: called from here
grain128aead-v2_opt.cpp: 207 | u64 y = ys ^ _mm_cvtsi128_si64(_xor2(shr8(_xor3(_andxor3(s4, s7, s5), _and3(ts, b7, s6), b1), 7), shr8(_andxor3(shr8(b7, 2), s2, b1), 5)));
grain128aead-v2_opt.cpp: | ~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
grain128aead-v2_opt.cpp: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:63,
grain128aead-v2_opt.cpp: from /usr/lib/gcc/x86_64-linux-gnu/10/include/x86intrin.h:32,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.h:41,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.cpp:10:
grain128aead-v2_opt.cpp: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx512vlintrin.h:10568:1: error: inlining failed in call to 'always_inline' '__m128i _mm_ternarylogic_epi64(__m128i, __m128i, __m128i, int)': target specific option mismatch
grain128aead-v2_opt.cpp: 10568 | _mm_ternarylogic_epi64 (__m128i __A, __m128i __B, __m128i __C,
grain128aead-v2_opt.cpp: | ^~~~~~~~~~~~~~~~~~~~~~
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:207:32: note: called from here
grain128aead-v2_opt.cpp: 207 | u64 y = ys ^ _mm_cvtsi128_si64(_xor2(shr8(_xor3(_andxor3(s4, s7, s5), _and3(ts, b7, s6), b1), 7), shr8(_andxor3(shr8(b7, 2), s2, b1), 5)));
grain128aead-v2_opt.cpp: | ~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
grain128aead-v2_opt.cpp: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:63,
grain128aead-v2_opt.cpp: from /usr/lib/gcc/x86_64-linux-gnu/10/include/x86intrin.h:32,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.h:41,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.cpp:10:
grain128aead-v2_opt.cpp: ...
Number of similar (implementation,compiler) pairs: 8, namely:
| Implementation | Compiler |
| T:avx512 | g++ -march=native -mtune=native -O2 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:avx512 | g++ -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:avx512 | g++ -march=native -mtune=native -O -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:avx512 | g++ -march=native -mtune=native -Os -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:gf2 | g++ -march=native -mtune=native -O2 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:gf2 | g++ -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:gf2 | g++ -march=native -mtune=native -O -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:gf2 | g++ -march=native -mtune=native -Os -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
Compiler output
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:234:6: error: '__builtin_ia32_vgf2p8affineqb_v16qi' needs target feature gfni
grain128aead-v2_opt.cpp: X = _mm_gf2p8affine_epi64_epi8(X, _mm_set1_epi64x(0x0104104002082080ULL), 0);
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/gfniintrin.h:64:12: note: expanded from macro '_mm_gf2p8affine_epi64_epi8'
grain128aead-v2_opt.cpp: (__m128i)__builtin_ia32_vgf2p8affineqb_v16qi((__v16qi)(__m128i)(A), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:241:6: error: '__builtin_ia32_pternlogq128_mask' needs target feature avx512vl
grain128aead-v2_opt.cpp: T = _xorand3(_mm_srli_epi16(X, 4), X, _mm_set1_epi16(0x00f0));
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:125:27: note: expanded from macro '_xorand3'
grain128aead-v2_opt.cpp: #define _xorand3(a, b, c) _mm_ternarylogic_epi64(a, b, c, 0x28)
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx512vlintrin.h:6565:12: note: expanded from macro '_mm_ternarylogic_epi64'
grain128aead-v2_opt.cpp: (__m128i)__builtin_ia32_pternlogq128_mask((__v2di)(__m128i)(A), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:242:6: error: '__builtin_ia32_pternlogq128_mask' needs target feature avx512vl
grain128aead-v2_opt.cpp: X = _xor3(_mm_slli_epi16(T, 4), X, T);
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:122:25: note: expanded from macro '_xor3'
grain128aead-v2_opt.cpp: #define _xor3(a, b, c) _mm_ternarylogic_epi64(a, b, c, 0x96)
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx512vlintrin.h:6565:12: note: expanded from macro '_mm_ternarylogic_epi64'
grain128aead-v2_opt.cpp: (__m128i)__builtin_ia32_pternlogq128_mask((__v2di)(__m128i)(A), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:243:6: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'grain_getz' that is compiled without support for 'ssse3'
grain128aead-v2_opt.cpp: ...
Number of similar (implementation,compiler) pairs: 4, namely:
| Implementation | Compiler |
| T:gf2 | clang++ -march=native -O2 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
| T:gf2 | clang++ -march=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
| T:gf2 | clang++ -march=native -O -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall |
| T:gf2 | clang++ -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
Compiler output
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:243:6: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'grain_getz' that is compiled without support for 'ssse3'
grain128aead-v2_opt.cpp: X = _mm_shuffle_epi8(X, _mm_set_epi64x(0x0e0c0a0806040200ULL, 0x0f0d0b0907050301ULL));
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:190:7: error: '__builtin_ia32_palignr128' needs target feature ssse3
grain128aead-v2_opt.cpp: b0 = _mm_alignr_epi8(tb, b0, 4);
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/tmmintrin.h:148:12: note: expanded from macro '_mm_alignr_epi8'
grain128aead-v2_opt.cpp: (__m128i)__builtin_ia32_palignr128((__v16qi)(__m128i)(a), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:317:14: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
grain128aead-v2_opt.cpp: __m128i x = _mm_clmulepi64_si128(reg, imsg, 0x00);
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/__wmmintrin_pclmul.h:45:13: note: expanded from macro '_mm_clmulepi64_si128'
grain128aead-v2_opt.cpp: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:318:14: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
grain128aead-v2_opt.cpp: __m128i y = _mm_clmulepi64_si128(reg, imsg, 0x0f);
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/__wmmintrin_pclmul.h:45:13: note: expanded from macro '_mm_clmulepi64_si128'
grain128aead-v2_opt.cpp: ((__m128i)__builtin_ia32_pclmulqdq128((__v2di)(__m128i)(X), \
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:281:34: error: '__builtin_ia32_pclmulqdq128' needs target feature pclmul
grain128aead-v2_opt.cpp: _mm_storeu_si128((__m128i*)tmp, _mm_clmulepi64_si128(reg, imsg, 0x00));
grain128aead-v2_opt.cpp: ^
grain128aead-v2_opt.cpp: /usr/lib/llvm-11/lib/clang/11.0.1/include/__wmmintrin_pclmul.h:45:13: note: expanded from macro '_mm_clmulepi64_si128'
grain128aead-v2_opt.cpp: ...
Number of similar (implementation,compiler) pairs: 4, namely:
| Implementation | Compiler |
| T:sse | clang++ -march=native -O2 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
| T:sse | clang++ -march=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
| T:sse | clang++ -march=native -O -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall |
| T:sse | clang++ -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
Compiler output
grain128aead-v2_opt.cpp: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:35,
grain128aead-v2_opt.cpp: from /usr/lib/gcc/x86_64-linux-gnu/10/include/x86intrin.h:32,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.h:41,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.cpp:10:
grain128aead-v2_opt.cpp: /usr/lib/gcc/x86_64-linux-gnu/10/include/tmmintrin.h: In function 'u64 grain_keystream64(grain_ctx*)':
grain128aead-v2_opt.cpp: /usr/lib/gcc/x86_64-linux-gnu/10/include/tmmintrin.h:185:1: error: inlining failed in call to 'always_inline' '__m128i _mm_alignr_epi8(__m128i, __m128i, int)': target specific option mismatch
grain128aead-v2_opt.cpp: 185 | _mm_alignr_epi8(__m128i __X, __m128i __Y, const int __N)
grain128aead-v2_opt.cpp: | ^~~~~~~~~~~~~~~
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:190:22: note: called from here
grain128aead-v2_opt.cpp: 190 | b0 = _mm_alignr_epi8(tb, b0, 4);
grain128aead-v2_opt.cpp: | ~~~~~~~~~~~~~~~^~~~~~~~~~~
grain128aead-v2_opt.cpp: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:35,
grain128aead-v2_opt.cpp: from /usr/lib/gcc/x86_64-linux-gnu/10/include/x86intrin.h:32,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.h:41,
grain128aead-v2_opt.cpp: from grain128aead-v2_opt.cpp:10:
grain128aead-v2_opt.cpp: /usr/lib/gcc/x86_64-linux-gnu/10/include/tmmintrin.h:185:1: error: inlining failed in call to 'always_inline' '__m128i _mm_alignr_epi8(__m128i, __m128i, int)': target specific option mismatch
grain128aead-v2_opt.cpp: 185 | _mm_alignr_epi8(__m128i __X, __m128i __Y, const int __N)
grain128aead-v2_opt.cpp: | ^~~~~~~~~~~~~~~
grain128aead-v2_opt.cpp: grain128aead-v2_opt.cpp:190:22: note: called from here
grain128aead-v2_opt.cpp: 190 | b0 = _mm_alignr_epi8(tb, b0, 4);
grain128aead-v2_opt.cpp: | ~~~~~~~~~~~~~~~^~~~~~~~~~~
Number of similar (implementation,compiler) pairs: 4, namely:
| Implementation | Compiler |
| T:sse | g++ -march=native -mtune=native -O2 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:sse | g++ -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:sse | g++ -march=native -mtune=native -O -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
| T:sse | g++ -march=native -mtune=native -Os -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |