Implementation notes: amd64, mangetsu, crypto_sign/3icp

Computer: mangetsu
Architecture: amd64
CPU ID: GenuineIntel-000206a7-bfebfbff
SUPERCOP version: 20141014
Operation: crypto_sign
Primitive: 3icp
TimeImplementationCompilerBenchmark dateSUPERCOP version
386276refg++ -m64 -march=corei7-avx -O3 -fomit-frame-pointer2014092620140907
395808refg++ -m64 -march=corei7-avx -O2 -fomit-frame-pointer2014092620140907
409752refg++ -m64 -march=native -mtune=native -O3 -fomit-frame-pointer2014092620140907
412316refclang++ -O3 -fomit-frame-pointer2014092620140907
443080refg++ -m64 -march=native -mtune=native -O2 -fomit-frame-pointer2014092620140907
455736refg++ -m64 -march=corei7-avx -O -fomit-frame-pointer2014092620140907
471932refg++ -march=barcelona -O -fomit-frame-pointer2014092620140907
480380refg++ -m64 -march=barcelona -O3 -fomit-frame-pointer2014092620140907
502924refg++ -m64 -march=core2 -O3 -fomit-frame-pointer2014092620140907
503620refg++ -m64 -march=core2 -msse4.1 -O3 -fomit-frame-pointer2014092620140907
508524refg++ -m64 -march=core-avx-i -O2 -fomit-frame-pointer2014092620140907
540868refg++ -m64 -march=corei7 -O -fomit-frame-pointer2014092620140907
552920refg++ -m64 -march=native -mtune=native -O -fomit-frame-pointer2014092620140907
554004refg++ -m64 -march=core2 -O2 -fomit-frame-pointer2014092620140907
560292refg++ -m64 -march=core2 -msse4.1 -O2 -fomit-frame-pointer2014092620140907
561332refg++ -m64 -march=nocona -O3 -fomit-frame-pointer2014092620140907
580224refg++ -march=k8 -O3 -fomit-frame-pointer2014092620140907
592896refg++ -march=barcelona -O3 -fomit-frame-pointer2014092620140907
598828refg++ -m64 -march=core2 -O -fomit-frame-pointer2014092620140907
604964refg++ -march=nocona -O3 -fomit-frame-pointer2014092620140907
618524refg++ -m64 -march=core-avx-i -O3 -fomit-frame-pointer2014092620140907
621632refg++ -m64 -march=core2 -msse4 -O -fomit-frame-pointer2014092620140907
628428refg++ -m64 -march=corei7 -O3 -fomit-frame-pointer2014092620140907
628656refg++ -m64 -march=k8 -O -fomit-frame-pointer2014092620140907
646520refg++ -march=k8 -O2 -fomit-frame-pointer2014092620140907
651252refg++ -m64 -march=k8 -O2 -fomit-frame-pointer2014092620140907
655000refg++ -m64 -march=core2 -msse4 -O3 -fomit-frame-pointer2014092620140907
676812refg++ -m64 -march=k8 -O3 -fomit-frame-pointer2014092620140907
678860refg++ -m64 -march=core2 -msse4 -O2 -fomit-frame-pointer2014092620140907
679108refg++ -m64 -march=core2 -msse4.1 -O -fomit-frame-pointer2014092620140907
683928refg++ -m64 -march=nocona -O2 -fomit-frame-pointer2014092620140907
695348refg++ -m64 -march=core-avx-i -O -fomit-frame-pointer2014092620140907
710324refg++ -m64 -march=barcelona -O -fomit-frame-pointer2014092620140907
712980refg++ -march=k8 -O -fomit-frame-pointer2014092620140907
736592refg++ -march=barcelona -O2 -fomit-frame-pointer2014092620140907
740600refg++ -m64 -march=barcelona -O2 -fomit-frame-pointer2014092620140907
748784refg++ -m64 -march=corei7 -O2 -fomit-frame-pointer2014092620140907
761072refg++ -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer2014092620140907
763172refg++ -march=nocona -O2 -fomit-frame-pointer2014092620140907
775248refg++ -march=nocona -O -fomit-frame-pointer2014092620140907
797492refg++ -m64 -march=nocona -O -fomit-frame-pointer2014092620140907
886396refg++ -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer2014092620140907
951244refg++ -m64 -march=core2 -Os -fomit-frame-pointer2014092620140907
977832refg++ -m64 -march=native -mtune=native -Os -fomit-frame-pointer2014092620140907
1006876refg++ -Os -fomit-frame-pointer2014092620140907
1024076refg++ -fno-schedule-insns -O2 -fomit-frame-pointer2014092620140907
1057388refg++ -m64 -march=k8 -Os -fomit-frame-pointer2014092620140907
1061192refg++ -O2 -fomit-frame-pointer2014092620140907
1147488refg++ -O3 -fomit-frame-pointer2014092620140907
1172848refg++ -m64 -O3 -fomit-frame-pointer2014092620140907
1208984refg++ -fno-schedule-insns -O -fomit-frame-pointer2014092620140907
1214340refg++ -m64 -march=corei7-avx -Os -fomit-frame-pointer2014092620140907
1221384refg++ -m64 -march=corei7 -Os -fomit-frame-pointer2014092620140907
1227004refg++ -fno-schedule-insns -Os -fomit-frame-pointer2014092620140907
1242960refg++ -m64 -O2 -fomit-frame-pointer2014092620140907
1257468refg++ -O -fomit-frame-pointer2014092620140907
1272212refg++ -march=barcelona -Os -fomit-frame-pointer2014092620140907
1288596refg++ -march=k8 -Os -fomit-frame-pointer2014092620140907
1303880refg++ -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer2014092620140907
1335208refg++ -m64 -march=core-avx-i -Os -fomit-frame-pointer2014092620140907
1369844refg++ -m64 -march=core2 -msse4.1 -Os -fomit-frame-pointer2014092620140907
1417716refg++ -fno-schedule-insns -O3 -fomit-frame-pointer2014092620140907
1421396refg++ -m64 -march=core2 -msse4 -Os -fomit-frame-pointer2014092620140907
1429272refg++ -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer2014092620140907
1536152refg++ -m64 -Os -fomit-frame-pointer2014092620140907
1559008refg++ -march=nocona -Os -fomit-frame-pointer2014092620140907
1589884refg++ -m64 -march=barcelona -Os -fomit-frame-pointer2014092620140907
1737520refg++ -m64 -O -fomit-frame-pointer2014092620140907
1771480refg++ -m64 -march=nocona -Os -fomit-frame-pointer2014092620140907
6807152refg++2014092620140907

Test failure

Implementation: crypto_sign/3icp/ref
Compiler: g++ -m64 -march=core-avx2 -O2 -fomit-frame-pointer
error 111

Number of similar (compiler,implementation) pairs: 3, namely:
CompilerImplementations
g++ -m64 -march=core-avx2 -O2 -fomit-frame-pointer ref
g++ -m64 -march=core-avx2 -O -fomit-frame-pointer ref
g++ -m64 -march=core-avx2 -Os -fomit-frame-pointer ref

Compiler output

Implementation: crypto_sign/3icp/ref
Compiler: clang++ -O3 -fwrapv -march=native -fomit-frame-pointer
ThreeIC_p-test.cpp: In file included from ThreeIC_p-test.cpp:16:
ThreeIC_p-test.cpp: In file included from ./ThreeIC_p.h:6:
ThreeIC_p-test.cpp: ./tower.h:364:9: error: argument to '__builtin_ia32_palignr128' must be a constant integer
ThreeIC_p-test.cpp: return _mm_alignr_epi8(a1, a0, i*2);
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: /usr/bin/../lib64/clang/3.3/include/tmmintrin.h:72:12: note: expanded from macro '_mm_alignr_epi8'
ThreeIC_p-test.cpp: (__m128i)__builtin_ia32_palignr128((__v16qi)__a, (__v16qi)__b, (n)); })
ThreeIC_p-test.cpp: ^ ~~~
ThreeIC_p-test.cpp: In file included from ThreeIC_p-test.cpp:16:
ThreeIC_p-test.cpp: In file included from ./ThreeIC_p.h:6:
ThreeIC_p-test.cpp: ./tower.h:364:9: error: cannot initialize return object of type '__m128i' with an rvalue of type 'void'
ThreeIC_p-test.cpp: return _mm_alignr_epi8(a1, a0, i*2);
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: /usr/bin/../lib64/clang/3.3/include/tmmintrin.h:69:34: note: expanded from macro '_mm_alignr_epi8'
ThreeIC_p-test.cpp: #define _mm_alignr_epi8(a, b, n) __extension__ ({ \
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: 2 errors generated.

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
clang++ -O3 -fwrapv -march=native -fomit-frame-pointer ref
clang++ -O3 -fwrapv -mavx2 -fomit-frame-pointer ref
clang++ -O3 -fwrapv -mavx -fomit-frame-pointer ref
clang++ -O3 -fwrapv -mavx -maes -mpclmul -fomit-frame-pointer ref

Compiler output

Implementation: crypto_sign/3icp/ref
Compiler: clang++ -O3 -fwrapv -march=native -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer
ThreeIC_p-test.cpp: clang-3.3: warning: argument unused during compilation: '-fpolly'
ThreeIC_p-test.cpp: In file included from ThreeIC_p-test.cpp:16:
ThreeIC_p-test.cpp: In file included from ./ThreeIC_p.h:6:
ThreeIC_p-test.cpp: ./tower.h:364:9: error: argument to '__builtin_ia32_palignr128' must be a constant integer
ThreeIC_p-test.cpp: return _mm_alignr_epi8(a1, a0, i*2);
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: /usr/bin/../lib64/clang/3.3/include/tmmintrin.h:72:12: note: expanded from macro '_mm_alignr_epi8'
ThreeIC_p-test.cpp: (__m128i)__builtin_ia32_palignr128((__v16qi)__a, (__v16qi)__b, (n)); })
ThreeIC_p-test.cpp: ^ ~~~
ThreeIC_p-test.cpp: In file included from ThreeIC_p-test.cpp:16:
ThreeIC_p-test.cpp: In file included from ./ThreeIC_p.h:6:
ThreeIC_p-test.cpp: ./tower.h:364:9: error: cannot initialize return object of type '__m128i' with an rvalue of type 'void'
ThreeIC_p-test.cpp: return _mm_alignr_epi8(a1, a0, i*2);
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: /usr/bin/../lib64/clang/3.3/include/tmmintrin.h:69:34: note: expanded from macro '_mm_alignr_epi8'
ThreeIC_p-test.cpp: #define _mm_alignr_epi8(a, b, n) __extension__ ({ \
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: 2 errors generated.

Number of similar (compiler,implementation) pairs: 3, namely:
CompilerImplementations
clang++ -O3 -fwrapv -march=native -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer ref
clang++ -O3 -fwrapv -mavx -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer ref
clang++ -O3 -fwrapv -mavx -maes -mpclmul -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer ref

Compiler output

Implementation: crypto_sign/3icp/ref
Compiler: clang++ -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer
ThreeIC_p-test.cpp: clang-3.3: warning: argument unused during compilation: '-mcpu=core-avx2'
ThreeIC_p-test.cpp: In file included from ThreeIC_p-test.cpp:16:
ThreeIC_p-test.cpp: In file included from ./ThreeIC_p.h:6:
ThreeIC_p-test.cpp: ./tower.h:364:9: error: argument to '__builtin_ia32_palignr128' must be a constant integer
ThreeIC_p-test.cpp: return _mm_alignr_epi8(a1, a0, i*2);
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: /usr/bin/../lib64/clang/3.3/include/tmmintrin.h:72:12: note: expanded from macro '_mm_alignr_epi8'
ThreeIC_p-test.cpp: (__m128i)__builtin_ia32_palignr128((__v16qi)__a, (__v16qi)__b, (n)); })
ThreeIC_p-test.cpp: ^ ~~~
ThreeIC_p-test.cpp: In file included from ThreeIC_p-test.cpp:16:
ThreeIC_p-test.cpp: In file included from ./ThreeIC_p.h:6:
ThreeIC_p-test.cpp: ./tower.h:364:9: error: cannot initialize return object of type '__m128i' with an rvalue of type 'void'
ThreeIC_p-test.cpp: return _mm_alignr_epi8(a1, a0, i*2);
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: /usr/bin/../lib64/clang/3.3/include/tmmintrin.h:69:34: note: expanded from macro '_mm_alignr_epi8'
ThreeIC_p-test.cpp: #define _mm_alignr_epi8(a, b, n) __extension__ ({ \
ThreeIC_p-test.cpp: ^~~~~~~~~~~~~~~~~~
ThreeIC_p-test.cpp: 2 errors generated.

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang++ -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer ref