Implementation notes: amd64, h8atom, crypto_encode/256x2

Computer: h8atom
Microarchitecture: amd64; Bonnell (30661)
Architecture: amd64
CPU ID: GenuineIntel-00030661-bfebfbff
SUPERCOP version: 20240107
Operation: crypto_encode
Primitive: 256x2

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
2107	133 0 0	10916 816 728	`ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231213	20231212
2114	101 0 0	9134 808 728	`ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231213	20231212
2114	145 0 0	10814 776 800	`ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231213	20231212
2128	145 0 0	12158 776 800	`ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231213	20231212
2142	133 0 0	12092 816 728	`ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231213	20231212
2149	138 0 0	9425 752 768	`ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231213	20231212
2765	132 0 0	12092 816 728	`ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231213	20231212
2765	149 0 0	10469 768 800	`ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231213	20231212
2898	136 0 0	9924 816 728	`ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231213	20231212

Compiler output

Implementation: avx
Security model: constbranchindex
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

encode.c: encode.c:7:16: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_encode_256x2_avx_constbranchindex' that is compiled without support for 'avx'
encode.c: __m256i a0 = _mm256_loadu_si256(0+(__m256i *) v);
encode.c: ^
encode.c: encode.c:7:16: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
encode.c: encode.c:8:16: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_encode_256x2_avx_constbranchindex' that is compiled without support for 'avx'
encode.c: __m256i a1 = _mm256_loadu_si256(1+(__m256i *) v);
encode.c: ^
encode.c: encode.c:8:16: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
encode.c: encode.c:9:16: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_encode_256x2_avx_constbranchindex' that is compiled without support for 'avx'
encode.c: __m256i a2 = _mm256_loadu_si256(2+(__m256i *) v);
encode.c: ^
encode.c: encode.c:9:16: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
encode.c: encode.c:10:16: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_encode_256x2_avx_constbranchindex' that is compiled without support for 'avx'
encode.c: __m256i a3 = _mm256_loadu_si256(3+(__m256i *) v);
encode.c: ^
encode.c: encode.c:10:16: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
encode.c: encode.c:11:16: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_encode_256x2_avx_constbranchindex' that is compiled without support for 'avx'
encode.c: __m256i a4 = _mm256_loadu_si256(4+(__m256i *) v);
encode.c: ^
encode.c: encode.c:11:16: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
encode.c: encode.c:12:16: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_encode_256x2_avx_constbranchindex' that is compiled without support for 'avx'
encode.c: __m256i a5 = _mm256_loadu_si256(5+(__m256i *) v);
encode.c: ^
encode.c: encode.c:12:16: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
encode.c: encode.c:13:16: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_encode_256x2_avx_constbranchindex' that is compiled without support for 'avx'
encode.c: ...

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

Compiler	Implementations
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

encode.c: encode.c: In function 'crypto_encode_256x2_avx_constbranchindex':
encode.c: encode.c:7:11: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
encode.c: 7 | __m256i a0 = _mm256_loadu_si256(0+(__m256i *) v);
encode.c: | ^~
encode.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
encode.c: from encode.c:1:
encode.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:433:1: error: inlining failed in call to 'always_inline' '_mm256_movemask_epi8': target specific option mismatch
encode.c: 433 | _mm256_movemask_epi8 (__m256i __A)
encode.c: | ^~~~~~~~~~~~~~~~~~~~
encode.c: encode.c:32:16: note: called from here
encode.c: 32 | int32_t c7 = _mm256_movemask_epi8(b7);
encode.c: | ^~~~~~~~~~~~~~~~~~~~~~~~
encode.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
encode.c: from encode.c:1:
encode.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:433:1: error: inlining failed in call to 'always_inline' '_mm256_movemask_epi8': target specific option mismatch
encode.c: 433 | _mm256_movemask_epi8 (__m256i __A)
encode.c: | ^~~~~~~~~~~~~~~~~~~~
encode.c: encode.c:31:16: note: called from here
encode.c: 31 | int32_t c6 = _mm256_movemask_epi8(b6);
encode.c: | ^~~~~~~~~~~~~~~~~~~~~~~~
encode.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
encode.c: from encode.c:1:
encode.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:433:1: error: inlining failed in call to 'always_inline' '_mm256_movemask_epi8': target specific option mismatch
encode.c: 433 | _mm256_movemask_epi8 (__m256i __A)
encode.c: | ^~~~~~~~~~~~~~~~~~~~
encode.c: ...

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

Compiler	Implementations
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