Implementation notes: amd64, hydra4, crypto_core/weightsntrup857
Computer: hydra4
Microarchitecture: amd64; K10 32nm (300f10)
Architecture: amd64
CPU ID: AuthenticAMD-00300f10-178bfbff
SUPERCOP version: 20240716
Operation: crypto_core
Primitive: weightsntrup857
Time | Object size | Test size | Implementation | Compiler | Benchmark date | SUPERCOP version |
268 | 379 0 0 | 13392 812 920 | ref | clang_-march=native_-O2_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
268 | 379 0 0 | 14224 812 920 | ref | clang_-march=native_-O3_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
268 | 379 0 0 | 14224 812 920 | ref | clang_-mcpu=native_-O3_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
273 | 322 0 0 | 13880 780 984 | ref | gcc_-march=native_-mtune=native_-O3_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
1763 | 118 0 0 | 12464 780 984 | ref | gcc_-march=native_-mtune=native_-O2_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
1763 | 105 0 0 | 11847 772 984 | ref | gcc_-march=native_-mtune=native_-O_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
1764 | 104 0 0 | 11686 804 920 | ref | clang_-march=native_-O_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
2609 | 97 0 0 | 10739 756 952 | ref | gcc_-march=native_-mtune=native_-Os_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
2614 | 97 0 0 | 11150 804 920 | ref | clang_-march=native_-Os_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall | 20240627 | 20240625 |
Compiler output
weight.c: weight.c:20:9: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_core_weightsntrup857_avx_constbranchindex' that is compiled without support for 'avx'
weight.c: sum = _mm256_loadu_si256((__m256i *) (in+p-32));
weight.c: ^
weight.c: weight.c:20:9: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
weight.c: weight.c:21:10: error: always_inline function '_mm256_set_epi8' requires target feature 'avx', but would be inlined into function 'crypto_core_weightsntrup857_avx_constbranchindex' that is compiled without support for 'avx'
weight.c: sum &= endingmask;
weight.c: ^
weight.c: ./params.h:2:20: note: expanded from macro 'endingmask'
weight.c: #define endingmask _mm256_set_epi8(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0)
weight.c: ^
weight.c: weight.c:21:10: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
weight.c: ./params.h:2:20: note: expanded from macro 'endingmask'
weight.c: #define endingmask _mm256_set_epi8(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0)
weight.c: ^
weight.c: weight.c:24:20: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_core_weightsntrup857_avx_constbranchindex' that is compiled without support for 'avx'
weight.c: __m256i bits = _mm256_loadu_si256((__m256i *) in);
weight.c: ^
weight.c: weight.c:24:20: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
weight.c: weight.c:25:13: error: always_inline function '_mm256_set1_epi8' requires target feature 'avx', but would be inlined into function 'crypto_core_weightsntrup857_avx_constbranchindex' that is compiled without support for 'avx'
weight.c: bits &= _mm256_set1_epi8(1);
weight.c: ^
weight.c: weight.c:25:13: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
weight.c: weight.c:26:11: error: always_inline function '_mm256_add_epi8' requires target feature 'avx2', but would be inlined into function 'crypto_core_weightsntrup857_avx_constbranchindex' that is compiled without support for 'avx2'
weight.c: sum = _mm256_add_epi8(sum,bits);
weight.c: ^
weight.c: ...
Number of similar (implementation,compiler) pairs: 5, namely:
Implementation | Compiler |
avx | clang -march=native -O2 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
avx | clang -march=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
avx | clang -march=native -O -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
avx | clang -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
avx | clang -mcpu=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
Compiler output
weight.c: weight.c: In function 'crypto_core_weightsntrup857_avx_constbranchindex':
weight.c: weight.c:20:7: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
weight.c: 20 | sum = _mm256_loadu_si256((__m256i *) (in+p-32));
weight.c: | ~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
weight.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:51,
weight.c: from weight.c:1:
weight.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avxintrin.h:1278:1: error: inlining failed in call to 'always_inline' '_mm256_set_epi8': target specific option mismatch
weight.c: 1278 | _mm256_set_epi8 (char __q31, char __q30, char __q29, char __q28,
weight.c: | ^~~~~~~~~~~~~~~
weight.c: In file included from weight.c:10:
weight.c: params.h:2:20: note: called from here
weight.c: 2 | #define endingmask _mm256_set_epi8(1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,0,0,0,0,0,0,0)
weight.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
weight.c: weight.c:21:10: note: in expansion of macro 'endingmask'
weight.c: 21 | sum &= endingmask;
weight.c: | ^~~~~~~~~~
weight.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:51,
weight.c: from weight.c:1:
weight.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avxintrin.h:920:1: error: inlining failed in call to 'always_inline' '_mm256_loadu_si256': target specific option mismatch
weight.c: 920 | _mm256_loadu_si256 (__m256i_u const *__P)
weight.c: | ^~~~~~~~~~~~~~~~~~
weight.c: weight.c:20:9: note: called from here
weight.c: 20 | sum = _mm256_loadu_si256((__m256i *) (in+p-32));
weight.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
weight.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:53,
weight.c: ...
Number of similar (implementation,compiler) pairs: 4, namely:
Implementation | Compiler |
avx | gcc -march=native -mtune=native -O2 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
avx | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
avx | gcc -march=native -mtune=native -O -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
avx | gcc -march=native -mtune=native -Os -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
Passed TIMECOP
TIMECOP iterations: 1
Number of similar (implementation,compiler) pairs: 9, namely:
Implementation | Compiler |
ref | clang -march=native -O2 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
ref | clang -march=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
ref | clang -march=native -O -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
ref | clang -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
ref | clang -mcpu=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1) |
ref | gcc -march=native -mtune=native -O2 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
ref | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
ref | gcc -march=native -mtune=native -O -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |
ref | gcc -march=native -mtune=native -Os -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110) |