Implementation notes: x86, thoth, crypto_aead/minalpherv11

Computer: thoth
Architecture: x86
CPU ID: AuthenticAMD-00000622-0183f9ff
SUPERCOP version: 20160806
Operation: crypto_aead
Primitive: minalpherv11

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
7559112	`ref`	`clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments`	20160727	20160724
7983324	`ref`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160727	20160724
11023299	`ref`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160727	20160724
12295453	`ref`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160727	20160724
12914370	`ref`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160727	20160724

Compiler output

Implementation: crypto_aead/minalpherv11/avx2
Compiler: clang++ -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments

MinalpherCore.cpp: MinalpherCore.cpp:85:27: error: always_inline function '_mm256_set_epi32' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: const __m256i ymmMask = _mm256_set_epi32(0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F);
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:88:17: error: always_inline function '_mm256_setzero_si256' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: __m256i tmp = _mm256_permute4x64_epi64(DATA, 0xD8);
MinalpherCore.cpp: ^
MinalpherCore.cpp: /usr/lib/llvm-3.8/bin/../lib/clang/3.8.0/include/avx2intrin.h:877:44: note: expanded from macro '_mm256_permute4x64_epi64'
MinalpherCore.cpp: (__v4di)_mm256_setzero_si256(), \
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:92:10: error: always_inline function '_mm256_and_si256' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: tmp0 = _mm256_and_si256(ymmMask, tmp);
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:93:10: error: always_inline function '_mm256_andnot_si256' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: tmp1 = _mm256_andnot_si256(ymmMask, tmp);
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:96:10: error: always_inline function '_mm256_srli_epi16' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: tmp1 = _mm256_srli_epi16(tmp1, 4);
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:99:18: error: always_inline function '_mm256_unpacklo_epi8' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: __m256i ymm0 = _mm256_unpacklo_epi8(tmp0, tmp1);
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:100:18: error: always_inline function '_mm256_unpackhi_epi8' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: __m256i ymm1 = _mm256_unpackhi_epi8(tmp0, tmp1);
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:106:27: error: always_inline function '_mm256_set_epi8' requires target feature 'sse3', but would be inlined into function 'minalpher_core_single_block' that is compiled without support for 'sse3'
MinalpherCore.cpp: ...

Number of similar (compiler,implementation) pairs: 1, namely:

Compiler	Implementations
clang++ -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments	avx2

Compiler output

Implementation: crypto_aead/minalpherv11/avx2
Compiler: g++ -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv

MinalpherCore.cpp: MinalpherCore.cpp: In function 'void minalpher_core_single_block(const __m256i&, __m256i&, __m256i&)':
MinalpherCore.cpp: MinalpherCore.cpp:85:138: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
MinalpherCore.cpp: const __m256i ymmMask = _mm256_set_epi32(0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F, 0x0F0F0F0F);
MinalpherCore.cpp: ^
MinalpherCore.cpp: In file included from /usr/lib/gcc/i686-linux-gnu/5/include/immintrin.h:41:0,
MinalpherCore.cpp: from /usr/lib/gcc/i686-linux-gnu/5/include/x86intrin.h:46,
MinalpherCore.cpp: from define.h:47,
MinalpherCore.cpp: from MinalpherCore.cpp:33:
MinalpherCore.cpp: /usr/lib/gcc/i686-linux-gnu/5/include/avxintrin.h: In function 'void __static_initialization_and_destruction_0(int, int)':
MinalpherCore.cpp: /usr/lib/gcc/i686-linux-gnu/5/include/avxintrin.h:1252:1: error: inlining failed in call to always_inline '__m256i _mm256_set_epi8(char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char)': target specific option mismatch
MinalpherCore.cpp: _mm256_set_epi8 (char __q31, char __q30, char __q29, char __q28,
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:705:55: error: called from here
MinalpherCore.cpp: const __m256i ShuffleQuadrupleBlock3 = _mm256_set_epi8(0xc, 0xd, 0x9, 0x8, 0xf, 0xe, 0xb, 0xa, 0x4, 0x5, 0x1, 0x0, 0x7, 0x6, 0x3, 0x2, 0xc, 0xd, 0x9, 0x8, 0xf, 0xe, 0xb, 0xa, 0x4, 0x5, 0x1, 0x0, 0x7, 0x6, 0x3, 0x2);
MinalpherCore.cpp: ^
MinalpherCore.cpp: In file included from /usr/lib/gcc/i686-linux-gnu/5/include/immintrin.h:41:0,
MinalpherCore.cpp: from /usr/lib/gcc/i686-linux-gnu/5/include/x86intrin.h:46,
MinalpherCore.cpp: from define.h:47,
MinalpherCore.cpp: from MinalpherCore.cpp:33:
MinalpherCore.cpp: /usr/lib/gcc/i686-linux-gnu/5/include/avxintrin.h:1252:1: error: inlining failed in call to always_inline '__m256i _mm256_set_epi8(char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char, char)': target specific option mismatch
MinalpherCore.cpp: _mm256_set_epi8 (char __q31, char __q30, char __q29, char __q28,
MinalpherCore.cpp: ^
MinalpherCore.cpp: MinalpherCore.cpp:704:55: error: called from here
MinalpherCore.cpp: const __m256i ShuffleQuadrupleBlock2 = _mm256_set_epi8(0x9, 0x8, 0xf, 0xe, 0xd, 0xc, 0xa, 0xb, 0x1, 0x0, 0x7, 0x6, 0x5, 0x4, 0x2, 0x3, 0x9, 0x8, 0xf, 0xe, 0xd, 0xc, 0xa, 0xb, 0x1, 0x0, 0x7, 0x6, 0x5, 0x4, 0x2, 0x3);
MinalpherCore.cpp: ^
MinalpherCore.cpp: ...

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

Compiler	Implementations
g++ -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv	avx2
g++ -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv	avx2
g++ -march=native -mtune=native -O -fomit-frame-pointer -fwrapv	avx2
g++ -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv	avx2