Implementation notes: amd64, shoe, crypto_aead/hs1sivhiv2

Computer: shoe
Microarchitecture: amd64; Broadwell+AES (306d4)
Architecture: amd64
CPU ID: GenuineIntel-000306d4-bfebfbff
SUPERCOP version: 20240107
Operation: crypto_aead
Primitive: hs1sivhiv2

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
11271	19817 0 0	42456 796 1080	`T:dolbeau/amd64-avx2`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
11430	13493 0 0	34416 796 1080	`T:dolbeau/amd64-avx2`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
11635	13976 0 0	34487 788 1080	`T:dolbeau/amd64-avx2`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
11990	13304 0 0	36096 796 1080	`T:faster`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
12284	12012 0 0	30955 772 1048	`T:dolbeau/amd64-avx2`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
12379	9855 0 0	30864 796 1080	`T:faster`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
12433	17884 0 0	40928 828 1048	`T:faster`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
12461	8763 0 0	27827 772 1048	`T:faster`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
12495	18252 0 0	41360 828 1048	`T:faster`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
12892	13213 0 0	36024 828 1048	`T:dolbeau/amd64-avx2`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
12964	13213 0 0	36104 828 1048	`T:dolbeau/amd64-avx2`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
13334	9292 0 0	29863 788 1080	`T:faster`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
13373	12202 0 0	31686 820 1016	`T:dolbeau/amd64-avx2`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
19247	12749 0 0	33672 796 1080	`T:dolbeau/amd64-sse`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
19342	21034 0 0	43688 796 1080	`T:dolbeau/amd64-sse`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
19535	12387 0 0	35360 828 1048	`T:dolbeau/amd64-sse`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
19546	12387 0 0	35248 828 1048	`T:dolbeau/amd64-sse`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
19604	13162 0 0	33671 788 1080	`T:dolbeau/amd64-sse`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
19709	11247 0 0	30195 772 1048	`T:dolbeau/amd64-sse`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
19884	11203 0 0	30734 820 1016	`T:dolbeau/amd64-sse`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
20424	11892 0 0	31838 820 1016	`T:dolbeau/amd64-sse`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
38066	8029 0 0	30496 828 1016	`T:ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
41545	7277 0 0	30400 828 1048	`T:ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
41728	8765 0 0	31968 828 1048	`T:ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
44908	5743 0 0	26696 796 1080	`T:ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
45086	6047 0 0	28792 796 1080	`T:ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
48903	4332 0 0	23363 772 1048	`T:ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
55710	4221 0 0	23942 820 1016	`T:ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212
71041	5876 0 0	26504 796 1080	`T:ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231214	20231212
133754	6216 0 0	26230 820 1016	`T:ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231214	20231212

Test failure

Implementation: T:dolbeau/amd64-avx2
Security model: timingleaks
Compiler: clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

error 111

Number of similar (compiler,implementation) pairs: 2, namely:

Compiler	Implementations
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:dolbeau/amd64-avx2
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:faster

Compiler output

Implementation: T:dolbeau/amd64-avx2
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

encrypt.c: encrypt.c:90:2: error: "This code requires AVX2 to work"
encrypt.c: #error "This code requires AVX2 to work"
encrypt.c: ^
encrypt.c: 1 error generated.

Number of similar (compiler,implementation) pairs: 1, namely:

Compiler	Implementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:dolbeau/amd64-avx2

Compiler output

Implementation: T:dolbeau/amd64-avx512
Security model: timingleaks
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

encrypt.c: encrypt.c:90:2: error: "This code requires AVX512F to work"
encrypt.c: #error "This code requires AVX512F to work"
encrypt.c: ^
encrypt.c: encrypt.c:330:20: error: conflicting types for '_mm512_reduce_add_epi64'
encrypt.c: unsigned long long _mm512_reduce_add_epi64 (__m512i a) {
encrypt.c: ^
encrypt.c: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx512fintrin.h:9319:51: note: previous definition is here
encrypt.c: static __inline__ long long __DEFAULT_FN_ATTRS512 _mm512_reduce_add_epi64(__m512i __W) {
encrypt.c: ^
encrypt.c: encrypt.c:343:21: error: invalid input size for constraint 'Yz'
encrypt.c: : [a] "Yz" (a)
encrypt.c: ^
encrypt.c: 3 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:dolbeau/amd64-avx512
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:dolbeau/amd64-avx512
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:dolbeau/amd64-avx512
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:dolbeau/amd64-avx512
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:dolbeau/amd64-avx512

Compiler output

Implementation: T:dolbeau/amd64-avx512
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

encrypt.c: encrypt.c:90:2: error: #error "This code requires AVX512F to work"
encrypt.c: 90 | #error "This code requires AVX512F to work"
encrypt.c: | ^~~~~
encrypt.c: encrypt.c:330:20: error: conflicting types for '_mm512_reduce_add_epi64'
encrypt.c: 330 | unsigned long long _mm512_reduce_add_epi64 (__m512i a) {
encrypt.c: | ^~~~~~~~~~~~~~~~~~~~~~~
encrypt.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:55,
encrypt.c: from encrypt.c:54:
encrypt.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx512fintrin.h:16018:1: note: previous definition of '_mm512_reduce_add_epi64' was here
encrypt.c: 16018 | _mm512_reduce_add_epi64 (__m512i __A)
encrypt.c: | ^~~~~~~~~~~~~~~~~~~~~~~
encrypt.c: encrypt.c: In function '_mm512_reduce_add_epi64':
encrypt.c: encrypt.c:330:20: note: the ABI for passing parameters with 64-byte alignment has changed in GCC 4.6
encrypt.c: 330 | unsigned long long _mm512_reduce_add_epi64 (__m512i a) {
encrypt.c: | ^~~~~~~~~~~~~~~~~~~~~~~
encrypt.c: In file included from encrypt.c:195:
encrypt.c: c368.h: In function 'chacha_noxor368':
encrypt.c: c368.h:111:11: warning: AVX512F vector return without AVX512F enabled changes the ABI [-Wpsabi]
encrypt.c: 111 | __m512i x_0 = _mm512_set1_epi32(x[0]);
encrypt.c: | ^~~

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

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:dolbeau/amd64-avx512
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:dolbeau/amd64-avx512
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:dolbeau/amd64-avx512
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	T:dolbeau/amd64-avx512

Compiler output

Implementation: T:dolbeau/amd64-sse
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

encrypt.c: In file included from encrypt.c:192:
encrypt.c: ./c256.h:98:7: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'chacha_noxor368' that is compiled without support for 'ssse3'
encrypt.c: VEC4_QUARTERROUND( 0, 4, 8,12);
encrypt.c: ^
encrypt.c: ./c256.h:17:36: note: expanded from macro 'VEC4_QUARTERROUND'
encrypt.c: #define VEC4_QUARTERROUND(a,b,c,d) VEC4_QUARTERROUND_SHUFFLE(a,b,c,d)
encrypt.c: ^
encrypt.c: ./c256.h:12:86: note: expanded from macro 'VEC4_QUARTERROUND_SHUFFLE'
encrypt.c: x_##a = _mm_add_epi32(x_##a, x_##b); t_##a = _mm_xor_si128(x_##d, x_##a); x_##d = _mm_shuffle_epi8(t_##a, rot16); \
encrypt.c: ^
encrypt.c: ./c256.h:98:7: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'chacha_noxor368' that is compiled without support for 'ssse3'
encrypt.c: ./c256.h:17:36: note: expanded from macro 'VEC4_QUARTERROUND'
encrypt.c: #define VEC4_QUARTERROUND(a,b,c,d) VEC4_QUARTERROUND_SHUFFLE(a,b,c,d)
encrypt.c: ^
encrypt.c: ./c256.h:14:86: note: expanded from macro 'VEC4_QUARTERROUND_SHUFFLE'
encrypt.c: x_##a = _mm_add_epi32(x_##a, x_##b); t_##a = _mm_xor_si128(x_##d, x_##a); x_##d = _mm_shuffle_epi8(t_##a, rot8); \
encrypt.c: ^
encrypt.c: ./c256.h:99:7: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'chacha_noxor368' that is compiled without support for 'ssse3'
encrypt.c: VEC4_QUARTERROUND( 1, 5, 9,13);
encrypt.c: ^
encrypt.c: ./c256.h:17:36: note: expanded from macro 'VEC4_QUARTERROUND'
encrypt.c: #define VEC4_QUARTERROUND(a,b,c,d) VEC4_QUARTERROUND_SHUFFLE(a,b,c,d)
encrypt.c: ^
encrypt.c: ./c256.h:12:86: note: expanded from macro 'VEC4_QUARTERROUND_SHUFFLE'
encrypt.c: x_##a = _mm_add_epi32(x_##a, x_##b); t_##a = _mm_xor_si128(x_##d, x_##a); x_##d = _mm_shuffle_epi8(t_##a, rot16); \
encrypt.c: ...

Number of similar (compiler,implementation) pairs: 1, namely:

Compiler	Implementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:dolbeau/amd64-sse