Implementation notes: amd64, jasper, crypto_sign/mqdss48

Computer: jasper
Microarchitecture: amd64; Tremont (906c0)
Architecture: amd64
CPU ID: GenuineIntel-000906c0-20-bfebfbff
SUPERCOP version: 20240425
Operation: crypto_sign
Primitive: mqdss48

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
59829060	13218 0 0	32906 836 1720	`T:ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240430	20240425
61566690	9985 0 0	27832 828 1720	`T:ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240430	20240425
62018844	16674 0 0	38682 836 1720	`T:ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240430	20240425
84695215	34596 0 0	54946 804 1784	`T:ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240430	20240425
112791920	13951 0 0	34754 836 1720	`T:ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240430	20240425
119833582	10289 0 0	28680 828 1720	`T:ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240430	20240425
124086744	8555 0 0	26090 796 1752	`T:ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240430	20240425
163631203	11813 0 0	31090 804 1784	`T:ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240430	20240425
169997587	10095 0 0	28890 804 1784	`T:ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240430	20240425

Compiler output

Implementation: T:avx2
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

gf31.c: gf31.c:14:20: error: always_inline function '_mm256_set1_epi16' requires target feature 'avx', but would be inlined into function 'vgf31_unique' that is compiled without support for 'avx'
gf31.c: __m256i _w31 = _mm256_set1_epi16(31);
gf31.c: ^
gf31.c: gf31.c:14:20: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
gf31.c: gf31.c:18:13: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'vgf31_unique' that is compiled without support for 'avx'
gf31.c: x = _mm256_loadu_si256((__m256i const *) (in + 16*i));
gf31.c: ^
gf31.c: gf31.c:18:13: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
gf31.c: gf31.c:19:56: error: always_inline function '_mm256_cmpeq_epi16' requires target feature 'avx2', but would be inlined into function 'vgf31_unique' that is compiled without support for 'avx2'
gf31.c: x = _mm256_xor_si256(x, _mm256_and_si256(_w31, _mm256_cmpeq_epi16(x, _w31)));
gf31.c: ^
gf31.c: gf31.c:19:56: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
gf31.c: gf31.c:19:33: error: always_inline function '_mm256_and_si256' requires target feature 'avx2', but would be inlined into function 'vgf31_unique' that is compiled without support for 'avx2'
gf31.c: x = _mm256_xor_si256(x, _mm256_and_si256(_w31, _mm256_cmpeq_epi16(x, _w31)));
gf31.c: ^
gf31.c: gf31.c:19:33: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
gf31.c: gf31.c:19:13: error: always_inline function '_mm256_xor_si256' requires target feature 'avx2', but would be inlined into function 'vgf31_unique' that is compiled without support for 'avx2'
gf31.c: x = _mm256_xor_si256(x, _mm256_and_si256(_w31, _mm256_cmpeq_epi16(x, _w31)));
gf31.c: ^
gf31.c: gf31.c:19:13: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
gf31.c: gf31.c:20:9: error: always_inline function '_mm256_storeu_si256' requires target feature 'avx', but would be inlined into function 'vgf31_unique' that is compiled without support for 'avx'
gf31.c: _mm256_storeu_si256((__m256i*)(out + i*16), x);
gf31.c: ^
gf31.c: gf31.c:20:9: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
gf31.c: 12 errors generated.

Number of similar (compiler,implementation) pairs: 5, namely:

Compiler	Implementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx2
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx2
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx2
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx2
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx2

Compiler output

Implementation: T:avx2
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

gf31.c: gf31.c: In function 'vgf31_unique':
gf31.c: gf31.c:14:13: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
gf31.c: 14 | __m256i _w31 = _mm256_set1_epi16(31);
gf31.c: | ^~~~
gf31.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:51,
gf31.c: from gf31.c:2:
gf31.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avxintrin.h:1326:1: error: inlining failed in call to 'always_inline' '_mm256_set1_epi16': target specific option mismatch
gf31.c: 1326 | _mm256_set1_epi16 (short __A)
gf31.c: | ^~~~~~~~~~~~~~~~~
gf31.c: gf31.c:14:20: note: called from here
gf31.c: 14 | __m256i _w31 = _mm256_set1_epi16(31);
gf31.c: | ^~~~~~~~~~~~~~~~~~~~~
gf31.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:51,
gf31.c: from gf31.c:2:
gf31.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avxintrin.h:926:1: error: inlining failed in call to 'always_inline' '_mm256_storeu_si256': target specific option mismatch
gf31.c: 926 | _mm256_storeu_si256 (__m256i_u *__P, __m256i __A)
gf31.c: | ^~~~~~~~~~~~~~~~~~~
gf31.c: gf31.c:20:9: note: called from here
gf31.c: 20 | _mm256_storeu_si256((__m256i*)(out + i*16), x);
gf31.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
gf31.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:53,
gf31.c: from gf31.c:2:
gf31.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx2intrin.h:913:1: error: inlining failed in call to 'always_inline' '_mm256_xor_si256': target specific option mismatch
gf31.c: 913 | _mm256_xor_si256 (__m256i __A, __m256i __B)
gf31.c: | ^~~~~~~~~~~~~~~~
gf31.c: ...

Number of similar (compiler,implementation) pairs: 4, namely:

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:avx2
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:avx2
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:avx2
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:avx2