Implementation notes: aarch64, hikey960, crypto_kem/kyber90s1024

Computer: hikey960
Architecture: aarch64
CPU ID: 410fd034
SUPERCOP version: 20190816
Operation: crypto_kem
Primitive: kyber90s1024

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
1406925	36306 512 0	52328 1336 1584	`ref`	`clang_-O3_-fwrapv_-mavx2_-fomit-frame-pointer_-Qunused-arguments`	20191209	20190816
1501227	29346 512 0	44547 1440 1568	`ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv`	20191209	20190816
1515150	52178 512 0	70563 1448 1600	`ref`	`gcc_-funroll-loops_-fno-schedule-insns_-O3_-fomit-frame-pointer`	20191209	20190816
1556775	38298 512 0	56483 1440 1568	`ref`	`gcc_-funroll-loops_-fno-schedule-insns_-O2_-fomit-frame-pointer`	20191209	20190816
1686015	36306 512 0	52328 1336 1584	`ref`	`clang_-O3_-fwrapv_-mavx_-maes_-mpclmul_-fomit-frame-pointer_-Qunused-arguments`	20191209	20190816
1789875	28038 512 0	42479 1424 1568	`ref`	`gcc_-fno-schedule-insns_-Os_-fomit-frame-pointer`	20191209	20190816
1789875	28034 512 0	42495 1424 1568	`ref`	`gcc_-funroll-loops_-Os_-fomit-frame-pointer`	20191209	20190816
1823175	29370 512 0	44571 1440 1568	`ref`	`gcc_-O_-fomit-frame-pointer`	20191209	20190816
1823175	29370 512 0	44571 1440 1568	`ref`	`gcc_-fno-schedule-insns_-O_-fomit-frame-pointer`	20191209	20190816
2007558	38678 512 0	57075 1440 1568	`ref`	`gcc_-funroll-loops_-O_-fomit-frame-pointer`	20191209	20190816
2007558	38678 512 0	57075 1440 1568	`ref`	`gcc_-funroll-loops_-fno-schedule-insns_-O_-fomit-frame-pointer`	20191209	20190816
2231100	42526 512 0	59315 1448 1600	`ref`	`gcc_-O3_-fomit-frame-pointer`	20191209	20190816
2389275	52362 512 0	70795 1448 1600	`ref`	`gcc_-funroll-loops_-O3_-fomit-frame-pointer`	20191209	20190816
2439225	42302 512 0	63155 1448 1600	`ref`	`gcc_-fno-schedule-insns_-O3_-fomit-frame-pointer`	20191209	20190816
2644951	28982 512 0	44139 1440 1568	`ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv`	20191209	20190816
2647350	36806 512 0	52808 1336 1584	`ref`	`clang_-O3_-fomit-frame-pointer_-Qunused-arguments`	20191209	20190816
2813850	28038 512 0	42479 1424 1568	`ref`	`gcc_-Os_-fomit-frame-pointer`	20191209	20190816
3118087	38354 512 0	56499 1440 1568	`ref`	`gcc_-funroll-loops_-O2_-fomit-frame-pointer`	20191209	20190816
3435096	36306 512 0	52328 1336 1584	`ref`	`clang_-O3_-fwrapv_-mavx_-fomit-frame-pointer_-Qunused-arguments`	20191209	20190816
3504825	29030 512 0	44179 1440 1568	`ref`	`gcc_-O2_-fomit-frame-pointer`	20191209	20190816
3684255	42186 512 0	58963 1448 1600	`ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv`	20191209	20190816
3771225	28034 512 0	42495 1424 1568	`ref`	`gcc_-funroll-loops_-fno-schedule-insns_-Os_-fomit-frame-pointer`	20191209	20190816
3814473	28062 512 0	42503 1424 1568	`ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv`	20191209	20190816
5285236	29062 512 0	44171 1440 1568	`ref`	`gcc_-fno-schedule-insns_-O2_-fomit-frame-pointer`	20191209	20190816
6893100	72494 512 0	90695 1424 1568	`ref`	`gcc`	20191209	20190816
12987000	72494 512 0	90695 1424 1568	`ref`	`cc`	20191209	20190816
25467354	72494 512 0	90695 1424 1568	`ref`	`gcc_-funroll-loops`	20191209	20190816

Compiler output

Implementation: avx2
Security model: unknown
Compiler: cc

aes256ctr.c: aes256ctr.c:10:23: fatal error: immintrin.h: No such file or directory
aes256ctr.c: #include <immintrin.h>
aes256ctr.c: ^
aes256ctr.c: compilation terminated.

Number of similar (compiler,implementation) pairs: 23, namely:

Compiler	Implementations
cc	avx2
gcc	avx2
gcc -O2 -fomit-frame-pointer	avx2
gcc -O3 -fomit-frame-pointer	avx2
gcc -O -fomit-frame-pointer	avx2
gcc -Os -fomit-frame-pointer	avx2
gcc -fno-schedule-insns -O2 -fomit-frame-pointer	avx2
gcc -fno-schedule-insns -O3 -fomit-frame-pointer	avx2
gcc -fno-schedule-insns -O -fomit-frame-pointer	avx2
gcc -fno-schedule-insns -Os -fomit-frame-pointer	avx2
gcc -funroll-loops	avx2
gcc -funroll-loops -O2 -fomit-frame-pointer	avx2
gcc -funroll-loops -O3 -fomit-frame-pointer	avx2
gcc -funroll-loops -O -fomit-frame-pointer	avx2
gcc -funroll-loops -Os -fomit-frame-pointer	avx2
gcc -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer	avx2
gcc -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer	avx2
gcc -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer	avx2
gcc -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer	avx2
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv	avx2
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv	avx2
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv	avx2
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv	avx2

Compiler output

Implementation: avx2
Security model: unknown
Compiler: clang -O3 -fomit-frame-pointer -Qunused-arguments

aes256ctr.c: In file included from aes256ctr.c:10:
aes256ctr.c: In file included from /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/immintrin.h:27:
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:45:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_vec_init_v2si(__i, 0);
aes256ctr.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:69:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_packsswb((__v4hi)__m1, (__v4hi)__m2);
aes256ctr.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:75:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_packssdw((__v2si)__m1, (__v2si)__m2);
aes256ctr.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:81:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_packuswb((__v4hi)__m1, (__v4hi)__m2);
aes256ctr.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:87:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_punpckhbw((__v8qi)__m1, (__v8qi)__m2);
aes256ctr.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:93:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_punpckhwd((__v4hi)__m1, (__v4hi)__m2);
aes256ctr.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:99:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_punpckhdq((__v2si)__m1, (__v2si)__m2);
aes256ctr.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
aes256ctr.c: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.1/include/mmintrin.h:105:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
aes256ctr.c: return (__m64)__builtin_ia32_punpcklbw((__v8qi)__m1, (__v8qi)__m2);
aes256ctr.c: ...

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

Compiler	Implementations
clang -O3 -fomit-frame-pointer -Qunused-arguments	avx2
clang -O3 -fwrapv -mavx2 -fomit-frame-pointer -Qunused-arguments	avx2
clang -O3 -fwrapv -mavx -fomit-frame-pointer -Qunused-arguments	avx2
clang -O3 -fwrapv -mavx -maes -mpclmul -fomit-frame-pointer -Qunused-arguments	avx2