Implementation notes: amd64, margaux, crypto_verify/1167

Computer: margaux
Microarchitecture: amd64; Core 2 65nm (6fb)
Architecture: amd64
CPU ID: GenuineIntel-000006fb-bfebfbff
SUPERCOP version: 20240107
Operation: crypto_verify
Primitive: 1167
TimeObject sizeTest sizeImplementationCompilerBenchmark dateSUPERCOP version
1181476 0 012949 804 928refgcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2023121320231212
1311354 0 011883 844 896refclang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2023121320231212
1311354 0 013003 844 896refclang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2023121320231212
1311356 0 012795 844 896refclang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2023121320231212
2761122 0 09968 780 896refgcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2023121320231212
2772134 0 011437 804 928refgcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2023121320231212
294299 0 010331 844 896refclang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2023121320231212
294286 0 09533 836 896refclang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2023121320231212
2966132 0 011012 796 928refgcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2023121320231212

Compiler output

Implementation: avx
Security model: constbranchindex
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
verify.c: verify.c:6:18: error: always_inline function '_mm256_set1_epi8' requires target feature 'avx', but would be inlined into function 'crypto_verify_1167_avx_constbranchindex' that is compiled without support for 'avx'
verify.c: __m256i diff = _mm256_set1_epi8(0);
verify.c: ^
verify.c: verify.c:6:18: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
verify.c: verify.c:13:20: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_verify_1167_avx_constbranchindex' that is compiled without support for 'avx'
verify.c: __m256i x0 = _mm256_loadu_si256((__m256i *) x);
verify.c: ^
verify.c: verify.c:13:20: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
verify.c: verify.c:14:20: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_verify_1167_avx_constbranchindex' that is compiled without support for 'avx'
verify.c: __m256i y0 = _mm256_loadu_si256((__m256i *) y);
verify.c: ^
verify.c: verify.c:14:20: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
verify.c: verify.c:25:11: error: always_inline function '_mm256_srli_epi16' requires target feature 'avx2', but would be inlined into function 'crypto_verify_1167_avx_constbranchindex' that is compiled without support for 'avx2'
verify.c: diff |= _mm256_srli_epi16(diff,8);
verify.c: ^
verify.c: verify.c:25:11: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
verify.c: verify.c:26:11: error: always_inline function '_mm256_srli_epi32' requires target feature 'avx2', but would be inlined into function 'crypto_verify_1167_avx_constbranchindex' that is compiled without support for 'avx2'
verify.c: diff |= _mm256_srli_epi32(diff,16);
verify.c: ^
verify.c: verify.c:26:11: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
verify.c: verify.c:27:11: error: always_inline function '_mm256_srli_epi64' requires target feature 'avx2', but would be inlined into function 'crypto_verify_1167_avx_constbranchindex' that is compiled without support for 'avx2'
verify.c: diff |= _mm256_srli_epi64(diff,32);
verify.c: ^
verify.c: verify.c:27:11: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
verify.c: verify.c:29:19: error: '__builtin_ia32_vec_ext_v32qi' needs target feature avx
verify.c: ...

Number of similar (compiler,implementation) pairs: 5, namely:
CompilerImplementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE avx
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE avx
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE avx
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE avx
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE avx

Compiler output

Implementation: avx
Security model: constbranchindex
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE
verify.c: verify.c: In function 'crypto_verify_1167_avx_constbranchindex':
verify.c: verify.c:6:11: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
verify.c: 6 | __m256i diff = _mm256_set1_epi8(0);
verify.c: | ^~~~
verify.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:43,
verify.c: from verify.c:1:
verify.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avxintrin.h:1340:1: error: inlining failed in call to 'always_inline' '_mm256_set1_epi8': target specific option mismatch
verify.c: 1340 | _mm256_set1_epi8 (char __A)
verify.c: | ^~~~~~~~~~~~~~~~
verify.c: verify.c:6:18: note: called from here
verify.c: 6 | __m256i diff = _mm256_set1_epi8(0);
verify.c: | ^~~~~~~~~~~~~~~~~~~
verify.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:43,
verify.c: from verify.c:1:
verify.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avxintrin.h:927:1: error: inlining failed in call to 'always_inline' '_mm256_loadu_si256': target specific option mismatch
verify.c: 927 | _mm256_loadu_si256 (__m256i_u const *__P)
verify.c: | ^~~~~~~~~~~~~~~~~~
verify.c: verify.c:14:20: note: called from here
verify.c: 14 | __m256i y0 = _mm256_loadu_si256((__m256i *) y);
verify.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
verify.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:43,
verify.c: from verify.c:1:
verify.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avxintrin.h:927:1: error: inlining failed in call to 'always_inline' '_mm256_loadu_si256': target specific option mismatch
verify.c: 927 | _mm256_loadu_si256 (__m256i_u const *__P)
verify.c: | ^~~~~~~~~~~~~~~~~~
verify.c: ...

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE avx
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE avx
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE avx
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE avx