Implementation notes: amd64, h3neo, crypto_sign/rainbow1acyclicc363232

Computer: h3neo
Microarchitecture: amd64; K10 45nm (100f63)
Architecture: amd64
CPU ID: AuthenticAMD-00100f63-078bfbff
SUPERCOP version: 20230530
Operation: crypto_sign
Primitive: rainbow1acyclicc363232

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
26837491	37095 0 52	79547 924 1792	`T:amd64`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27143839	45613 0 52	88937 932 1792	`T:ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27335805	60150 0 52	103985 932 1792	`T:ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27337502	82283 0 52	125897 932 1792	`T:amd64`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27442108	37910 0 52	80849 932 1792	`T:ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27527160	32563 0 52	74979 924 1792	`T:ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27598662	62026 0 52	106081 932 1792	`T:ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27667616	41698 0 52	84617 932 1792	`T:amd64`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
27733513	54169 0 52	97073 932 1792	`T:amd64`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
28181851	79812 0 52	123217 932 1792	`T:amd64`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230704	20230530
28247626	68427 0 52	112617 916 1824	`T:amd64`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530
28270186	148894 0 52	194241 916 1824	`T:ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530
28639793	201126 0 52	246465 916 1824	`T:amd64`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530
28654228	57269 0 52	100641 916 1824	`T:amd64`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530
28684794	32146 0 52	74705 908 1824	`T:amd64`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530
28965835	47295 0 52	91457 916 1824	`T:ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530
29054055	29338 0 52	71825 908 1824	`T:ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530
29608248	46083 0 52	89377 916 1824	`T:ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230704	20230530

Compiler output

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

blas_comm.c: In file included from blas_comm.c:6:
blas_comm.c: In file included from ./blas.h:25:
blas_comm.c: ./blas_avx2.h:88:17: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'gf256v_add_avx2' that is compiled without support for 'avx'
blas_comm.c: __m256i inp = _mm256_loadu_si256( (__m256i*) (a+i*32) );
blas_comm.c: ^
blas_comm.c: ./blas_avx2.h:88:17: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
blas_comm.c: ./blas_avx2.h:89:17: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'gf256v_add_avx2' that is compiled without support for 'avx'
blas_comm.c: __m256i out = _mm256_loadu_si256( (__m256i*) (accu_b+i*32) );
blas_comm.c: ^
blas_comm.c: ./blas_avx2.h:89:17: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
blas_comm.c: ./blas_avx2.h:91:3: error: always_inline function '_mm256_storeu_si256' requires target feature 'avx', but would be inlined into function 'gf256v_add_avx2' that is compiled without support for 'avx'
blas_comm.c: _mm256_storeu_si256( (__m256i*) (accu_b+i*32) , out );
blas_comm.c: ^
blas_comm.c: ./blas_avx2.h:91:3: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
blas_comm.c: 6 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

blas_comm.c: In file included from blas_avx2.h:15,
blas_comm.c: from blas.h:25,
blas_comm.c: from blas_comm.c:6:
blas_comm.c: gf16_avx2.h: In function 'linear_transform_8x8_256b':
blas_comm.c: gf16_avx2.h:28:1: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
blas_comm.c: 28 | {
blas_comm.c: | ^
blas_comm.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:43,
blas_comm.c: from blas_avx2.h:10,
blas_comm.c: from blas.h:25,
blas_comm.c: from blas_comm.c:6:
blas_comm.c: blas_avx2.h: In function 'gf256v_add_avx2':
blas_comm.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avxintrin.h:933:1: error: inlining failed in call to 'always_inline' '_mm256_storeu_si256': target specific option mismatch
blas_comm.c: 933 | _mm256_storeu_si256 (__m256i_u *__P, __m256i __A)
blas_comm.c: | ^~~~~~~~~~~~~~~~~~~
blas_comm.c: In file included from blas.h:25,
blas_comm.c: from blas_comm.c:6:
blas_comm.c: blas_avx2.h:91:17: note: called from here
blas_comm.c: 91 | _mm256_storeu_si256( (__m256i*) (accu_b+i*32) , out );
blas_comm.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blas_comm.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:43,
blas_comm.c: from blas_avx2.h:10,
blas_comm.c: from blas.h:25,
blas_comm.c: from blas_comm.c:6:
blas_comm.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
blas_comm.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

Compiler output

Implementation: T:ssse3
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

blas_matrix_ref.c: In file included from blas_matrix_ref.c:6:
blas_matrix_ref.c: In file included from ./blas.h:25:
blas_matrix_ref.c: In file included from ./blas_sse.h:16:
blas_matrix_ref.c: ./gf16_sse.h:34:9: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'linear_transform_8x8_128b' that is compiled without support for 'ssse3'
blas_matrix_ref.c: return _mm_shuffle_epi8(tab_l,v&mask_f)^_mm_shuffle_epi8(tab_h,_mm_srli_epi16(v,4)&mask_f);
blas_matrix_ref.c: ^
blas_matrix_ref.c: ./gf16_sse.h:34:42: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'linear_transform_8x8_128b' that is compiled without support for 'ssse3'
blas_matrix_ref.c: return _mm_shuffle_epi8(tab_l,v&mask_f)^_mm_shuffle_epi8(tab_h,_mm_srli_epi16(v,4)&mask_f);
blas_matrix_ref.c: ^
blas_matrix_ref.c: 2 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:ssse3
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:ssse3
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:ssse3
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:ssse3
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:ssse3

Compiler output

Implementation: T:ssse3
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

blas_matrix_ref.c: In file included from blas_sse.h:12,
blas_matrix_ref.c: from blas.h:25,
blas_matrix_ref.c: from blas_matrix_ref.c:6:
blas_matrix_ref.c: gf16_sse.h: In function 'linear_transform_8x8_128b':
blas_matrix_ref.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/tmmintrin.h:136:1: error: inlining failed in call to 'always_inline' '_mm_shuffle_epi8': target specific option mismatch
blas_matrix_ref.c: 136 | _mm_shuffle_epi8 (__m128i __X, __m128i __Y)
blas_matrix_ref.c: | ^~~~~~~~~~~~~~~~
blas_matrix_ref.c: In file included from blas_sse.h:16,
blas_matrix_ref.c: from blas.h:25,
blas_matrix_ref.c: from blas_matrix_ref.c:6:
blas_matrix_ref.c: gf16_sse.h:34:49: note: called from here
blas_matrix_ref.c: 34 | return _mm_shuffle_epi8(tab_l,v&mask_f)^_mm_shuffle_epi8(tab_h,_mm_srli_epi16(v,4)&mask_f);
blas_matrix_ref.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blas_matrix_ref.c: In file included from blas_sse.h:12,
blas_matrix_ref.c: from blas.h:25,
blas_matrix_ref.c: from blas_matrix_ref.c:6:
blas_matrix_ref.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/tmmintrin.h:136:1: error: inlining failed in call to 'always_inline' '_mm_shuffle_epi8': target specific option mismatch
blas_matrix_ref.c: 136 | _mm_shuffle_epi8 (__m128i __X, __m128i __Y)
blas_matrix_ref.c: | ^~~~~~~~~~~~~~~~
blas_matrix_ref.c: In file included from blas_sse.h:16,
blas_matrix_ref.c: from blas.h:25,
blas_matrix_ref.c: from blas_matrix_ref.c:6:
blas_matrix_ref.c: gf16_sse.h:34:16: note: called from here
blas_matrix_ref.c: 34 | return _mm_shuffle_epi8(tab_l,v&mask_f)^_mm_shuffle_epi8(tab_h,_mm_srli_epi16(v,4)&mask_f);
blas_matrix_ref.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blas_matrix_ref.c: ...

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

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:ssse3
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:ssse3
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:ssse3
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:ssse3