Implementation notes: amd64, haswell, crypto_sign/3icp

Computer: haswell
Architecture: amd64
CPU ID: GenuineIntel-000306c3-bfebfbff
SUPERCOP version: 20130419
Operation: crypto_sign
Primitive: 3icp

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
788656	`ref`	`g++ -m64 -march=core-avx-i -O3 -fomit-frame-pointer`	20130609	20130419
839432	`ref`	`g++ -m64 -march=corei7-avx -O3 -fomit-frame-pointer`	20130609	20130419
841120	`ref`	`g++ -m64 -march=barcelona -O3 -fomit-frame-pointer`	20130609	20130419
851072	`ref`	`g++ -m64 -march=barcelona -O2 -fomit-frame-pointer`	20130609	20130419
853892	`ref`	`g++ -m64 -march=core2 -msse4.1 -O2 -fomit-frame-pointer`	20130609	20130419
877064	`ref`	`g++ -m64 -march=nocona -O2 -fomit-frame-pointer`	20130609	20130419
883916	`ref`	`g++ -march=k8 -O2 -fomit-frame-pointer`	20130609	20130419
885391	`ref`	`clang++ -O3`	20130609	20130419
896676	`ref`	`g++ -m64 -march=corei7 -O3 -fomit-frame-pointer`	20130609	20130419
900984	`ref`	`g++ -m64 -march=nocona -O3 -fomit-frame-pointer`	20130609	20130419
910296	`ref`	`g++ -march=barcelona -O3 -fomit-frame-pointer`	20130609	20130419
936161	`ref`	`g++ -m64 -march=corei7 -O2 -fomit-frame-pointer`	20130609	20130419
939808	`ref`	`g++ -m64 -march=core2 -msse4 -O2 -fomit-frame-pointer`	20130609	20130419
948348	`ref`	`g++ -m64 -march=core-avx-i -O -fomit-frame-pointer`	20130609	20130419
950432	`ref`	`g++ -m64 -march=core-avx-i -O2 -fomit-frame-pointer`	20130609	20130419
952580	`ref`	`g++ -march=barcelona -O2 -fomit-frame-pointer`	20130609	20130419
964076	`ref`	`g++ -m64 -march=k8 -O3 -fomit-frame-pointer`	20130609	20130419
969796	`ref`	`g++ -m64 -march=corei7-avx -O2 -fomit-frame-pointer`	20130609	20130419
973447	`ref`	`g++ -m64 -march=core2 -msse4.1 -O3 -fomit-frame-pointer`	20130609	20130419
984964	`ref`	`g++ -m64 -march=k8 -O2 -fomit-frame-pointer`	20130609	20130419
990704	`ref`	`g++ -march=k8 -O3 -fomit-frame-pointer`	20130609	20130419
999948	`ref`	`g++ -march=nocona -O3 -fomit-frame-pointer`	20130609	20130419
1014135	`ref`	`g++ -m64 -march=nocona -O -fomit-frame-pointer`	20130609	20130419
1021873	`ref`	`g++ -m64 -march=core2 -O3 -fomit-frame-pointer`	20130609	20130419
1024558	`ref`	`g++ -m64 -march=core2 -O2 -fomit-frame-pointer`	20130609	20130419
1027855	`ref`	`g++ -march=nocona -O -fomit-frame-pointer`	20130609	20130419
1051512	`ref`	`g++ -m64 -march=corei7 -O -fomit-frame-pointer`	20130609	20130419
1066504	`ref`	`g++ -m64 -march=core2 -msse4 -O -fomit-frame-pointer`	20130609	20130419
1068088	`ref`	`g++ -m64 -march=core2 -msse4 -O3 -fomit-frame-pointer`	20130609	20130419
1070412	`ref`	`g++ -march=nocona -O2 -fomit-frame-pointer`	20130609	20130419
1074102	`ref`	`g++ -m64 -march=barcelona -O -fomit-frame-pointer`	20130609	20130419
1091568	`ref`	`g++ -march=barcelona -O -fomit-frame-pointer`	20130609	20130419
1110008	`ref`	`g++ -m64 -march=core2 -msse4.1 -O -fomit-frame-pointer`	20130609	20130419
1120852	`ref`	`g++ -m64 -march=corei7-avx -O -fomit-frame-pointer`	20130609	20130419
1135441	`ref`	`g++ -m64 -march=k8 -O -fomit-frame-pointer`	20130609	20130419
1138240	`ref`	`g++ -march=k8 -O -fomit-frame-pointer`	20130609	20130419
1167478	`ref`	`g++ -m64 -march=core2 -O -fomit-frame-pointer`	20130609	20130419
1229180	`ref`	`g++ -m64 -march=corei7-avx -Os -fomit-frame-pointer`	20130609	20130419
1240548	`ref`	`g++ -O3 -fomit-frame-pointer`	20130609	20130419
1263087	`ref`	`g++ -m64 -march=native -mtune=native -O -fomit-frame-pointer`	20130609	20130419
1283241	`ref`	`g++ -fno-schedule-insns -O2 -fomit-frame-pointer`	20130609	20130419
1292956	`ref`	`g++ -m64 -march=native -mtune=native -Os -fomit-frame-pointer`	20130609	20130419
1353076	`ref`	`g++ -m64 -march=core2 -Os -fomit-frame-pointer`	20130609	20130419
1365640	`ref`	`g++ -m64 -Os -fomit-frame-pointer`	20130609	20130419
1381612	`ref`	`g++ -m64 -march=native -mtune=native -O2 -fomit-frame-pointer`	20130609	20130419
1389252	`ref`	`g++ -m64 -march=k8 -Os -fomit-frame-pointer`	20130609	20130419
1400576	`ref`	`g++ -O2 -fomit-frame-pointer`	20130609	20130419
1473730	`ref`	`g++ -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer`	20130609	20130419
1479373	`ref`	`g++ -m64 -march=native -mtune=native -O3 -fomit-frame-pointer`	20130609	20130419
1481138	`ref`	`g++ -march=barcelona -Os -fomit-frame-pointer`	20130609	20130419
1483836	`ref`	`g++ -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer`	20130609	20130419
1517024	`ref`	`g++ -m64 -march=core-avx-i -Os -fomit-frame-pointer`	20130609	20130419
1523337	`ref`	`g++ -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer`	20130609	20130419
1535128	`ref`	`g++ -m64 -march=corei7 -Os -fomit-frame-pointer`	20130609	20130419
1552320	`ref`	`g++ -fno-schedule-insns -O3 -fomit-frame-pointer`	20130609	20130419
1572831	`ref`	`g++ -m64 -march=nocona -Os -fomit-frame-pointer`	20130609	20130419
1580832	`ref`	`g++ -m64 -march=core2 -msse4 -Os -fomit-frame-pointer`	20130609	20130419
1595356	`ref`	`g++ -m64 -O3 -fomit-frame-pointer`	20130609	20130419
1608088	`ref`	`g++ -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer`	20130609	20130419
1617493	`ref`	`g++ -m64 -march=core2 -msse4.1 -Os -fomit-frame-pointer`	20130609	20130419
1639600	`ref`	`g++ -m64 -O2 -fomit-frame-pointer`	20130609	20130419
1644932	`ref`	`g++ -fno-schedule-insns -O -fomit-frame-pointer`	20130609	20130419
1650532	`ref`	`g++ -march=k8 -Os -fomit-frame-pointer`	20130609	20130419
1742933	`ref`	`g++ -O -fomit-frame-pointer`	20130609	20130419
1758256	`ref`	`g++ -Os -fomit-frame-pointer`	20130609	20130419
1927192	`ref`	`g++ -march=nocona -Os -fomit-frame-pointer`	20130609	20130419
1951733	`ref`	`g++ -m64 -march=barcelona -Os -fomit-frame-pointer`	20130609	20130419
2090220	`ref`	`g++ -m64 -O -fomit-frame-pointer`	20130609	20130419
2104652	`ref`	`g++ -fno-schedule-insns -Os -fomit-frame-pointer`	20130609	20130419
8402254	`ref`	`g++`	20130609	20130419

Compiler output

Implementation: crypto_sign/3icp/ref
Compiler: clang++ -O3 -mavx

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/../lib/clang/3.2/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/../lib/clang/3.2/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:

Compiler	Implementations
clang++ -O3 -mavx	ref