Implementation notes: amd64, h3neo, crypto_kem/ntruhps4096821

Computer: h3neo
Microarchitecture: amd64; K10 45nm (100f63)
Architecture: amd64
CPU ID: AuthenticAMD-00100f63-078bfbff
SUPERCOP version: 20240107
Operation: crypto_kem
Primitive: ntruhps4096821

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
3799832	37491 0 0	114813 860 1728	`T:compact`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
3807578	26145 0 0	102533 860 1728	`T:compact`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
3835748	37491 0 0	121357 860 1728	`compact`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
3839040	26145 0 0	109077 860 1728	`compact`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
3915550	40471 0 0	118133 860 1728	`T:compact`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
3970709	36323 0 0	114887 820 1760	`T:compact`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107
3973468	40471 0 0	124677 860 1728	`compact`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
4000974	36323 0 0	121431 820 1760	`compact`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107
8331376	7232 0 0	82885 860 1728	`T:compact`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
8347824	7232 0 0	89429 860 1728	`compact`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
9312616	12582 0 0	89021 860 1728	`T:ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
9316148	17657 0 0	95013 860 1728	`T:ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
9346575	12582 0 0	95565 860 1728	`ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
9347866	17657 0 0	101557 860 1728	`ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
9879970	5803 0 0	80887 852 1728	`T:compact`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
9913907	5803 0 0	87431 852 1728	`compact`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20240111	20240107
10036247	6808 0 0	81655 852 1728	`T:ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
10069305	6808 0 0	88199 852 1728	`ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
10119732	19163 0 0	96949 860 1728	`T:ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
10159974	19163 0 0	103493 860 1728	`ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
11780822	5520 0 0	82791 820 1760	`T:compact`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107
11820447	5520 0 0	89335 820 1760	`compact`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107
13054509	17977 0 0	96143 820 1760	`T:ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
13084588	17977 0 0	102687 820 1760	`ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
13089795	7577 0 0	84559 820 1760	`T:ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
13124385	7577 0 0	91103 820 1760	`ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
18000873	6266 0 0	81877 860 1728	`T:ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
18033964	6266 0 0	88421 860 1728	`ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231229	20231212
18097264	6352 0 0	80895 812 1728	`T:ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
18135812	6352 0 0	87439 812 1728	`ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
18700613	6829 0 0	82573 812 1760	`T:ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
18736836	6829 0 0	89117 812 1760	`ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231229	20231212
19424426	4683 0 0	87501 812 1760	`compact`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107
19499697	4683 0 0	80957 812 1760	`T:compact`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107
20098329	4054 0 0	79135 812 1728	`T:compact`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107
20128231	4054 0 0	85679 812 1728	`compact`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20240111	20240107

Compiler output

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

poly_s3_inv.c: poly_s3_inv.c:416:11: error: always_inline function '_mm256_set_epi32' requires target feature 'avx', but would be inlined into function '__poly_S3_inv' that is compiled without support for 'avx'
poly_s3_inv.c: F0[0] = _mm256_set_epi32(-1,-1,-1,-1,-1,-1,-1,-1);
poly_s3_inv.c: ^
poly_s3_inv.c: poly_s3_inv.c:416:11: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
poly_s3_inv.c: poly_s3_inv.c:417:11: error: always_inline function '_mm256_set_epi32' requires target feature 'avx', but would be inlined into function '__poly_S3_inv' that is compiled without support for 'avx'
poly_s3_inv.c: F0[1] = _mm256_set_epi32(-1,-1,-1,-1,-1,-1,-1,-1);
poly_s3_inv.c: ^
poly_s3_inv.c: poly_s3_inv.c:417:11: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
poly_s3_inv.c: poly_s3_inv.c:418:11: error: always_inline function '_mm256_set_epi32' requires target feature 'avx', but would be inlined into function '__poly_S3_inv' that is compiled without support for 'avx'
poly_s3_inv.c: F0[2] = _mm256_set_epi32(-1,-1,-1,-1,-1,-1,-1,-1);
poly_s3_inv.c: ^
poly_s3_inv.c: poly_s3_inv.c:418:11: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
poly_s3_inv.c: poly_s3_inv.c:419:11: error: always_inline function '_mm256_set_epi32' requires target feature 'avx', but would be inlined into function '__poly_S3_inv' that is compiled without support for 'avx'
poly_s3_inv.c: F0[3] = _mm256_set_epi32(0,8191,0,8191,0,8191,0,16383);
poly_s3_inv.c: ^
poly_s3_inv.c: poly_s3_inv.c:419:11: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
poly_s3_inv.c: poly_s3_inv.c:420:11: error: always_inline function '_mm256_set1_epi32' requires target feature 'avx', but would be inlined into function '__poly_S3_inv' that is compiled without support for 'avx'
poly_s3_inv.c: F1[0] = _mm256_set1_epi32(0);
poly_s3_inv.c: ^
poly_s3_inv.c: poly_s3_inv.c:420:11: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
poly_s3_inv.c: poly_s3_inv.c:421:11: error: always_inline function '_mm256_set1_epi32' requires target feature 'avx', but would be inlined into function '__poly_S3_inv' that is compiled without support for 'avx'
poly_s3_inv.c: F1[1] = _mm256_set1_epi32(0);
poly_s3_inv.c: ^
poly_s3_inv.c: poly_s3_inv.c:421:11: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
poly_s3_inv.c: poly_s3_inv.c:422:11: error: always_inline function '_mm256_set1_epi32' requires target feature 'avx', but would be inlined into function '__poly_S3_inv' that is compiled without support for 'avx'
poly_s3_inv.c: ...

Number of similar (compiler,implementation) pairs: 10, namely:

Compiler	Implementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	avx2 T:avx2
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	avx2 T:avx2
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	avx2 T:avx2
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	avx2 T:avx2
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	avx2 T:avx2

Compiler output

Implementation: avx2
Security model: constbranchindex
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

poly_s3_inv.c: poly_s3_inv.c: In function '__poly_S3_inv':
poly_s3_inv.c: poly_s3_inv.c:416:9: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
poly_s3_inv.c: 416 | F0[0] = _mm256_set_epi32(-1,-1,-1,-1,-1,-1,-1,-1);
poly_s3_inv.c: | ~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
poly_s3_inv.c: poly_s3_inv.c: In function 'vec256_swap':
poly_s3_inv.c: poly_s3_inv.c:176:20: note: the ABI for passing parameters with 32-byte alignment has changed in GCC 4.6
poly_s3_inv.c: 176 | static inline void vec256_swap(vec256 *f,vec256 *g,int len,vec256 mask)
poly_s3_inv.c: | ^~~~~~~~~~~
poly_s3_inv.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
poly_s3_inv.c: from poly.h:4,
poly_s3_inv.c: from poly_s3_inv.c:1:
poly_s3_inv.c: poly_s3_inv.c: In function 'vec256_frombits':
poly_s3_inv.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:597:1: error: inlining failed in call to 'always_inline' '_mm256_shuffle_epi32': target specific option mismatch
poly_s3_inv.c: 597 | _mm256_shuffle_epi32 (__m256i __A, const int __mask)
poly_s3_inv.c: | ^~~~~~~~~~~~~~~~~~~~
poly_s3_inv.c: poly_s3_inv.c:67:9: note: called from here
poly_s3_inv.c: 67 | h = _mm256_shuffle_epi32(h,0xd8);
poly_s3_inv.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
poly_s3_inv.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
poly_s3_inv.c: from poly.h:4,
poly_s3_inv.c: from poly_s3_inv.c:1:
poly_s3_inv.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:1071:1: error: inlining failed in call to 'always_inline' '_mm256_permute4x64_epi64': target specific option mismatch
poly_s3_inv.c: 1071 | _mm256_permute4x64_epi64 (__m256i __X, const int __M)
poly_s3_inv.c: | ^~~~~~~~~~~~~~~~~~~~~~~~
poly_s3_inv.c: poly_s3_inv.c:66:9: note: called from here
poly_s3_inv.c: ...

Number of similar (compiler,implementation) pairs: 8, namely:

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	avx2 T:avx2
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	avx2 T:avx2
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	avx2 T:avx2
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	avx2 T:avx2