Implementation notes: amd64, utrecht, crypto_hash/bmw256

Computer: utrecht
Architecture: amd64
CPU ID: GenuineIntel-000006fb-bfebfbff
SUPERCOP version: 20160806
Operation: crypto_hash
Primitive: bmw256

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
9693	`core_2_65nm/x86-64/ssse3_v1`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
9954	`core_2_65nm/x86-64/ssse3_v1`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
10881	`opt31ssse3_asm64`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
10890	`opt31ssse3_asm64`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
10899	`opt31ssse3_asm64`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
10980	`opt31ssse3_asm64`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
12411	`amd_k10/x86-64/sse2`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
12429	`amd_k10/x86-64/sse2`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
12537	`amd_k10/x86-64/sse2`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
12888	`amd_k8/x86-64/noSSE`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
13032	`amd_k8/x86-64/noSSE`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
13878	`optc01`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
14112	`optc01`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
14121	`optabe`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
14157	`optc01`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
14184	`optbbe`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
14193	`optc03`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
14247	`optc04`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
14283	`optc02`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
14346	`optc01`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
14409	`optbgbe`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
14616	`optbgbe`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
14706	`optbgbe`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
14859	`core_2_65nm/x86-64/ssse3_v1`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15129	`core_2_65nm/x86-64/ssse3_v1`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
15201	`optbbe`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15219	`optbbe`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
15219	`optabe`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15273	`opt29_v5_ssse3_asm64`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
15300	`optabe`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
15300	`opt29_v5_ssse3_asm64`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15327	`opt29_v5_ssse3_asm64`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
15336	`sphlib`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15345	`optabe`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
15426	`optbbe`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
15480	`optc03`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15480	`optbgbe`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
15489	`optc03`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
15489	`sphlib`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
15534	`sphlib`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
15606	`optc04`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
15642	`opt29_v5_ssse3_asm64`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
15687	`optc04`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15849	`optc02`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
15894	`optc04`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
15975	`sphlib-small`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
16029	`optc02`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
16029	`optc02`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
16146	`optc03`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
16326	`sphlib-small`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
16578	`sphlib-small`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
17955	`amd_k10/x86-64/sse2`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
18441	`sphlib`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
18828	`sphlib-small`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731
19386	`amd_k8/x86-64/noSSE`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
19476	`amd_k8/x86-64/noSSE`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
27054	`ref`	`gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv`	20160804	20160731
27225	`ref`	`gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv`	20160804	20160731
27234	`ref`	`gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv`	20160804	20160731
33156	`ref`	`gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv`	20160804	20160731

Test failure

Implementation: crypto_hash/bmw256/core_2_45nm/x86-64/sse4_v1
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv

error 111

Number of similar (compiler,implementation) pairs: 8, namely:

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv	core_2_45nm/x86-64/sse4_v1 core_i7_45nm/x86-64/sse4_v1
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv	core_2_45nm/x86-64/sse4_v1 core_i7_45nm/x86-64/sse4_v1
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv	core_2_45nm/x86-64/sse4_v1 core_i7_45nm/x86-64/sse4_v1
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv	core_2_45nm/x86-64/sse4_v1 core_i7_45nm/x86-64/sse4_v1

Compiler output

Implementation: crypto_hash/bmw256/optc31ssse3
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv

Compress256_sse_31.c: Compress256_sse_31.c: In function 'Compress256':
Compress256_sse_31.c: Compress256_sse_31.c:374:2: warning: passing argument 1 of '_mm_store_si128' from incompatible pointer type [enabled by default]
Compress256_sse_31.c: _mm_store_si128(&p256[00], copyTemp);
Compress256_sse_31.c: ^
Compress256_sse_31.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/4.8/include/pmmintrin.h:35:0,
Compress256_sse_31.c: from /usr/lib/gcc/x86_64-linux-gnu/4.8/include/tmmintrin.h:35,
Compress256_sse_31.c: from Compress256_sse_31.c:12:
Compress256_sse_31.c: /usr/lib/gcc/x86_64-linux-gnu/4.8/include/emmintrin.h:694:1: note: expected '__vector(2) long long int *' but argument is of type 'u_int32_t *'
Compress256_sse_31.c: _mm_store_si128 (__m128i *__P, __m128i __B)
Compress256_sse_31.c: ^
Compress256_sse_31.c: Compress256_sse_31.c:377:2: warning: passing argument 1 of '_mm_load_si128' from incompatible pointer type [enabled by default]
Compress256_sse_31.c: copyTemp = _mm_load_si128(&hashState256(state)->gt;DoublePipe[ 4]);
Compress256_sse_31.c: ^
Compress256_sse_31.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/4.8/include/pmmintrin.h:35:0,
Compress256_sse_31.c: from /usr/lib/gcc/x86_64-linux-gnu/4.8/include/tmmintrin.h:35,
Compress256_sse_31.c: from Compress256_sse_31.c:12:
Compress256_sse_31.c: /usr/lib/gcc/x86_64-linux-gnu/4.8/include/emmintrin.h:676:1: note: expected 'const __vector(2) long long int *' but argument is of type 'u_int32_t *'
Compress256_sse_31.c: _mm_load_si128 (__m128i const *__P)
Compress256_sse_31.c: ^
Compress256_sse_31.c: Compress256_sse_31.c:378:2: warning: passing argument 1 of '_mm_store_si128' from incompatible pointer type [enabled by default]
Compress256_sse_31.c: _mm_store_si128(&p256[04], copyTemp);
Compress256_sse_31.c: ^
Compress256_sse_31.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/4.8/include/pmmintrin.h:35:0,
Compress256_sse_31.c: from /usr/lib/gcc/x86_64-linux-gnu/4.8/include/tmmintrin.h:35,
Compress256_sse_31.c: from Compress256_sse_31.c:12:
Compress256_sse_31.c: ...

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

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv	optc31ssse3
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv	optc31ssse3
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv	optc31ssse3
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv	optc31ssse3

Compiler output

Implementation: crypto_hash/bmw256/opt31ssse3_asm64
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv

wrapper.c: wrapper.c:37:0: warning: "crypto_hash_BYTES" redefined [enabled by default]
wrapper.c: #define crypto_hash_BYTES 32
wrapper.c: ^
wrapper.c: In file included from wrapper.c:20:0:
wrapper.c: crypto_hash.h:7:0: note: this is the location of the previous definition
wrapper.c: #define crypto_hash_BYTES crypto_hash_bmw256_BYTES
wrapper.c: ^
wrapper.c: wrapper.c: In function 'crypto_hash_bmw256_opt31ssse3_asm64':
wrapper.c: wrapper.c:238:2: warning: passing argument 3 of 'Compress256' from incompatible pointer type [enabled by default]
wrapper.c: Compress256(data32, data32 +16, CONST32final-4); // TODO: This is a HACK...
wrapper.c: ^
wrapper.c: wrapper.c:39:8: note: expected 'struct hashState *' but argument is of type 'u_int32_t *'
wrapper.c: extern Compress256(u_int32_t *data32, u_int32_t *data32_end, hashState *state);
wrapper.c: ^

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

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv	opt31ssse3_asm64
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv	opt31ssse3_asm64
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv	opt31ssse3_asm64
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv	opt31ssse3_asm64