Implementation notes: amd64, comet, crypto_aead/isapk128av20

Computer: comet
Microarchitecture: amd64; Comet Lake (806ec)
Architecture: amd64
CPU ID: GenuineIntel-000806ec-bfebfbff
SUPERCOP version: 20240107
Operation: crypto_aead
Primitive: isapk128av20

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
398955	152472 0 0	168628 780 1088	`T:opt_64`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222
400519	152204 0 0	166468 780 1088	`T:opt_64`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222
411053	157492 0 0	171291 772 1088	`T:opt_64`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222
413248	154930 0 0	172969 852 1088	`T:opt_64`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
413353	154930 0 0	172673 852 1056	`T:opt_64`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
418712	156017 0 0	170735 844 1088	`T:opt_64`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
445992	168936 0 0	182865 852 1024	`T:opt_64`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
559253	173201 0 0	189625 852 1024	`T:opt_64`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
586246	144243 0 0	156815 756 1056	`T:opt_64`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222
813651	5838 0 0	22445 860 1024	`T:ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
1250330	5088 0 0	21404 788 1088	`T:ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222
1367873	6367 0 0	24277 860 1056	`T:ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
1460667	6415 0 0	24621 860 1088	`T:ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
4171243	4419 0 0	18517 860 1024	`T:ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
4734197	4127 0 0	18540 788 1088	`T:ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222
6752301	4204 0 0	18171 780 1088	`T:ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222
7105747	4388 0 0	19275 852 1088	`T:ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240106	20231222
7766748	3396 0 0	16135 764 1056	`T:ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240106	20231222

Compiler output

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

isap.c: isap.c:105:5: error: always_inline function '_mm512_set_epi32' requires target feature 'avx512f', but would be inlined into function 'isap_rk' that is compiled without support for 'avx512f'
isap.c: declareABCDEX
isap.c: ^
isap.c: ./KeccakP-400-64.macros:42:21: note: expanded from macro 'declareABCDEX'
isap.c: __m512i pchi1 = _mm512_set_epi32(0, 0, 0, 0x000014, 0x180017, 0x160015, 0x0f0013, 0x120011, 0x10000a, 0xe000d, 0xc000b, 0x50009, 0x80007, 0x60000, 0x40003, 0x20001); \
isap.c: ^
isap.c: isap.c:105:5: error: AVX vector return of type '__m512i' (vector of 8 'long long' values) without 'avx512f' enabled changes the ABI
isap.c: ./KeccakP-400-64.macros:42:21: note: expanded from macro 'declareABCDEX'
isap.c: __m512i pchi1 = _mm512_set_epi32(0, 0, 0, 0x000014, 0x180017, 0x160015, 0x0f0013, 0x120011, 0x10000a, 0xe000d, 0xc000b, 0x50009, 0x80007, 0x60000, 0x40003, 0x20001); \
isap.c: ^
isap.c: isap.c:105:5: error: always_inline function '_mm512_set_epi32' requires target feature 'avx512f', but would be inlined into function 'isap_rk' that is compiled without support for 'avx512f'
isap.c: ./KeccakP-400-64.macros:43:21: note: expanded from macro 'declareABCDEX'
isap.c: __m512i pchi2 = _mm512_set_epi32(0, 0, 0, 0x000015, 0x140018, 0x170016, 0x10000f, 0x130012, 0x11000b, 0xa000e, 0xd000c, 0x60005, 0x90008, 0x70001, 0x00004, 0x30002); \
isap.c: ^
isap.c: isap.c:105:5: error: AVX vector return of type '__m512i' (vector of 8 'long long' values) without 'avx512f' enabled changes the ABI
isap.c: ./KeccakP-400-64.macros:43:21: note: expanded from macro 'declareABCDEX'
isap.c: __m512i pchi2 = _mm512_set_epi32(0, 0, 0, 0x000015, 0x140018, 0x170016, 0x10000f, 0x130012, 0x11000b, 0xa000e, 0xd000c, 0x60005, 0x90008, 0x70001, 0x00004, 0x30002); \
isap.c: ^
isap.c: isap.c:105:5: error: always_inline function '_mm512_set_epi32' requires target feature 'avx512f', but would be inlined into function 'isap_rk' that is compiled without support for 'avx512f'
isap.c: ./KeccakP-400-64.macros:44:19: note: expanded from macro 'declareABCDEX'
isap.c: __m512i ppi = _mm512_set_epi32(0, 0, 0, 0x0015, 0xf000e, 0x80002, 0x170011, 0xb0005, 0x40014, 0x13000d, 0x70001, 0x160010, 0xa0009, 0x30018, 0x12000c, 0x60000); \
isap.c: ^
isap.c: isap.c:105:5: error: AVX vector return of type '__m512i' (vector of 8 'long long' values) without 'avx512f' enabled changes the ABI
isap.c: ./KeccakP-400-64.macros:44:19: note: expanded from macro 'declareABCDEX'
isap.c: __m512i ppi = _mm512_set_epi32(0, 0, 0, 0x0015, 0xf000e, 0x80002, 0x170011, 0xb0005, 0x40014, 0x13000d, 0x70001, 0x160010, 0xa0009, 0x30018, 0x12000c, 0x60000); \
isap.c: ...

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

Compiler	Implementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx512
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx512
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx512
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx512
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	T:avx512

Compiler output

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

isap.c: In file included from isap.c:6:
isap.c: isap.c: In function 'isap_rk':
isap.c: KeccakP-400-64.macros:42:13: warning: AVX512F vector return without AVX512F enabled changes the ABI [-Wpsabi]
isap.c: 42 | __m512i pchi1 = _mm512_set_epi32(0, 0, 0, 0x000014, 0x180017, 0x160015, 0x0f0013, 0x120011, 0x10000a, 0xe000d, 0xc000b, 0x50009, 0x80007, 0x60000, 0x40003, 0x20001); \
isap.c: | ^~~~~
isap.c: isap.c:105:5: note: in expansion of macro 'declareABCDEX'
isap.c: 105 | declareABCDEX
isap.c: | ^~~~~~~~~~~~~
isap.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/12/include/immintrin.h:49,
isap.c: from isap.c:3:
isap.c: /usr/lib/gcc/x86_64-linux-gnu/12/include/avx512fintrin.h:7572:1: error: inlining failed in call to 'always_inline' '_mm512_xor_epi64': target specific option mismatch
isap.c: 7572 | _mm512_xor_epi64 (__m512i __A, __m512i __B)
isap.c: | ^~~~~~~~~~~~~~~~
isap.c: KeccakP-400-64.macros:76:11: note: called from here
isap.c: 76 | E.S = _mm512_xor_epi64 (E.S, _mm512_set_epi64(0,0,0,0,0,0,0,KeccakF400RoundConstants[i])); \
isap.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
isap.c: KeccakP-400-unrolling.macros:63:5: note: in expansion of macro 'thetaRhoPiChiIota'
isap.c: 63 | thetaRhoPiChiIota(19, E, S) \
isap.c: | ^~~~~~~~~~~~~~~~~
isap.c: isap.c:56:25: note: in expansion of macro 'rounds8X'
isap.c: 56 | #define PermuteRoundsKX rounds8X
isap.c: | ^~~~~~~~
isap.c: isap.c:131:5: note: in expansion of macro 'PermuteRoundsKX'
isap.c: 131 | PermuteRoundsKX;
isap.c: | ^~~~~~~~~~~~~~~
isap.c: ...

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

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