Implementation notes: amd64, jasper, crypto_sign/dilithium3

Computer: jasper
Microarchitecture: amd64; Tremont (906c0)
Architecture: amd64
CPU ID: GenuineIntel-000906c0-20-bfebfbff
SUPERCOP version: 20231107
Operation: crypto_sign
Primitive: dilithium3

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
1732254	36940 0 0	58296 820 1720	`ref`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230605	20230530
1798500	35816 0 0	55864 820 1720	`ref`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230605	20230530
1854047	24969 0 0	43992 820 1720	`ref`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230605	20230530
1967723	19870 0 0	37606 812 1720	`ref`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230605	20230530
2024158	18647 0 0	35718 812 1720	`ref`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230605	20230530
2136532	47033 0 0	67072 788 1784	`ref`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230605	20230530
2150007	20932 0 0	39864 788 1784	`ref`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230605	20230530
2282853	17214 0 0	34112 780 1752	`ref`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230605	20230530
2392805	18429 0 0	36968 788 1784	`ref`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20230605	20230530

Compiler output

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

fips202x4.c: fips202x4.c:22:12: error: always_inline function '_mm256_setzero_si256' requires target feature 'avx', but would be inlined into function 'keccakx4_absorb_once' that is compiled without support for 'avx'
fips202x4.c: s[i] = _mm256_setzero_si256();
fips202x4.c: ^
fips202x4.c: fips202x4.c:22:12: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
fips202x4.c: fips202x4.c:24:9: error: always_inline function '_mm256_set_epi64x' requires target feature 'avx', but would be inlined into function 'keccakx4_absorb_once' that is compiled without support for 'avx'
fips202x4.c: idx = _mm256_set_epi64x((long long)in3, (long long)in2, (long long)in1, (long long)in0);
fips202x4.c: ^
fips202x4.c: fips202x4.c:24:9: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
fips202x4.c: fips202x4.c:27:11: error: '__builtin_ia32_gatherq_q256' needs target feature avx2
fips202x4.c: t = _mm256_i64gather_epi64((long long *)pos, idx, 1);
fips202x4.c: ^
fips202x4.c: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx2intrin.h:1140:12: note: expanded from macro '_mm256_i64gather_epi64'
fips202x4.c: (__m256i)__builtin_ia32_gatherq_q256((__v4di)_mm256_undefined_si256(), \
fips202x4.c: ^
fips202x4.c: fips202x4.c:27:11: error: always_inline function '_mm256_undefined_si256' requires target feature 'avx', but would be inlined into function 'keccakx4_absorb_once' that is compiled without support for 'avx'
fips202x4.c: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx2intrin.h:1140:48: note: expanded from macro '_mm256_i64gather_epi64'
fips202x4.c: (__m256i)__builtin_ia32_gatherq_q256((__v4di)_mm256_undefined_si256(), \
fips202x4.c: ^
fips202x4.c: fips202x4.c:27:11: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
fips202x4.c: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx2intrin.h:1140:48: note: expanded from macro '_mm256_i64gather_epi64'
fips202x4.c: (__m256i)__builtin_ia32_gatherq_q256((__v4di)_mm256_undefined_si256(), \
fips202x4.c: ^
fips202x4.c: fips202x4.c:27:11: error: always_inline function '_mm256_set1_epi64x' requires target feature 'avx', but would be inlined into function 'keccakx4_absorb_once' that is compiled without support for 'avx'
fips202x4.c: /usr/lib/llvm-11/lib/clang/11.0.1/include/avx2intrin.h:1143:48: note: expanded from macro '_mm256_i64gather_epi64'
fips202x4.c: (__v4di)_mm256_set1_epi64x(-1), (s))
fips202x4.c: ...

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

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

Compiler output

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

fips202x4.c: fips202x4.c: In function 'keccakx4_absorb_once':
fips202x4.c: fips202x4.c:22:10: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
fips202x4.c: 22 | s[i] = _mm256_setzero_si256();
fips202x4.c: | ~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
fips202x4.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:51,
fips202x4.c: from fips202x4.c:3:
fips202x4.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avxintrin.h:1235:1: error: inlining failed in call to 'always_inline' '_mm256_setzero_si256': target specific option mismatch
fips202x4.c: 1235 | _mm256_setzero_si256 (void)
fips202x4.c: | ^~~~~~~~~~~~~~~~~~~~
fips202x4.c: fips202x4.c:22:12: note: called from here
fips202x4.c: 22 | s[i] = _mm256_setzero_si256();
fips202x4.c: | ^~~~~~~~~~~~~~~~~~~~~~
fips202x4.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:51,
fips202x4.c: from fips202x4.c:3:
fips202x4.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avxintrin.h:1296:1: error: inlining failed in call to 'always_inline' '_mm256_set_epi64x': target specific option mismatch
fips202x4.c: 1296 | _mm256_set_epi64x (long long __A, long long __B, long long __C,
fips202x4.c: | ^~~~~~~~~~~~~~~~~
fips202x4.c: fips202x4.c:24:9: note: called from here
fips202x4.c: 24 | idx = _mm256_set_epi64x((long long)in3, (long long)in2, (long long)in1, (long long)in0);
fips202x4.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
fips202x4.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:53,
fips202x4.c: from fips202x4.c:3:
fips202x4.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx2intrin.h:913:1: error: inlining failed in call to 'always_inline' '_mm256_xor_si256': target specific option mismatch
fips202x4.c: 913 | _mm256_xor_si256 (__m256i __A, __m256i __B)
fips202x4.c: | ^~~~~~~~~~~~~~~~
fips202x4.c: ...

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

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