Implementation notes: amd64, h3neo, crypto_core/invsntrup953

Computer: h3neo
Microarchitecture: amd64; K10 45nm (100f63)
Architecture: amd64
CPU ID: AuthenticAMD-00100f63-078bfbff
SUPERCOP version: 20231107
Operation: crypto_core
Primitive: invsntrup953

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
91669218	3889 0 0	18599 812 992	`ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231110	20231107
104137082	2896 0 0	16509 852 928	`ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231111	20231107
104356344	2870 0 0	16317 852 928	`ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231110	20231107
104369509	2870 0 0	15397 852 928	`ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231111	20231107
108002755	1067 0 0	12127 844 928	`ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231110	20231107
108670136	1170 0 0	12845 852 928	`ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231111	20231107
111909683	947 0 0	12362 788 960	`ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231110	20231107
115082652	1079 0 0	14447 812 992	`ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231110	20231107
117132639	1159 0 0	13700 796 992	`ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231110	20231107

Compiler output

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

recip.c: recip.c:94:19: error: always_inline function '_mm256_set1_epi16' requires target feature 'avx', but would be inlined into function 'vectormodq_swapeliminate' that is compiled without support for 'avx'
recip.c: __m256i f0vec = _mm256_set1_epi16(f0);
recip.c: ^
recip.c: recip.c:94:19: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
recip.c: recip.c:95:19: error: always_inline function '_mm256_set1_epi16' requires target feature 'avx', but would be inlined into function 'vectormodq_swapeliminate' that is compiled without support for 'avx'
recip.c: __m256i g0vec = _mm256_set1_epi16(g0);
recip.c: ^
recip.c: recip.c:95:19: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
recip.c: recip.c:96:48: error: always_inline function '_mm256_set1_epi16' requires target feature 'avx', but would be inlined into function 'vectormodq_swapeliminate' that is compiled without support for 'avx'
recip.c: __m256i f0vecqinv = _mm256_mullo_epi16(f0vec,qinvvec);
recip.c: ^
recip.c: recip.c:80:17: note: expanded from macro 'qinvvec'
recip.c: #define qinvvec _mm256_set1_epi16(qinv)
recip.c: ^
recip.c: recip.c:96:48: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
recip.c: recip.c:80:17: note: expanded from macro 'qinvvec'
recip.c: #define qinvvec _mm256_set1_epi16(qinv)
recip.c: ^
recip.c: recip.c:96:23: error: always_inline function '_mm256_mullo_epi16' requires target feature 'avx2', but would be inlined into function 'vectormodq_swapeliminate' that is compiled without support for 'avx2'
recip.c: __m256i f0vecqinv = _mm256_mullo_epi16(f0vec,qinvvec);
recip.c: ^
recip.c: recip.c:96:23: error: AVX vector argument of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
recip.c: recip.c:97:48: error: always_inline function '_mm256_set1_epi16' requires target feature 'avx', but would be inlined into function 'vectormodq_swapeliminate' that is compiled without support for 'avx'
recip.c: __m256i g0vecqinv = _mm256_mullo_epi16(g0vec,qinvvec);
recip.c: ^
recip.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

recip.c: recip.c: In function 'montproduct':
recip.c: recip.c:83:1: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
recip.c: 83 | {
recip.c: | ^
recip.c: recip.c:82:23: note: the ABI for passing parameters with 32-byte alignment has changed in GCC 4.6
recip.c: 82 | static inline __m256i montproduct(__m256i x,__m256i y,__m256i yqinv)
recip.c: | ^~~~~~~~~~~
recip.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
recip.c: from recip.c:1:
recip.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:808:1: error: inlining failed in call to 'always_inline' '_mm256_sub_epi16': target specific option mismatch
recip.c: 808 | _mm256_sub_epi16 (__m256i __A, __m256i __B)
recip.c: | ^~~~~~~~~~~~~~~~
recip.c: recip.c:89:10: note: called from here
recip.c: 89 | return _mm256_sub_epi16(hi,e);
recip.c: | ^~~~~~~~~~~~~~~~~~~~~~
recip.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
recip.c: from recip.c:1:
recip.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:546:1: error: inlining failed in call to 'always_inline' '_mm256_mulhi_epi16': target specific option mismatch
recip.c: 546 | _mm256_mulhi_epi16 (__m256i __A, __m256i __B)
recip.c: | ^~~~~~~~~~~~~~~~~~
recip.c: recip.c:88:7: note: called from here
recip.c: 88 | e = _mm256_mulhi_epi16(d,qvec);
recip.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~
recip.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:43,
recip.c: from recip.c:1:
recip.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