Implementation notes: armeabi, cubox, crypto_hash/groestl256

Computer: cubox
Architecture: armeabi
CPU ID: unknown CPU ID
SUPERCOP version: 20161026
Operation: crypto_hash
Primitive: groestl256

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
121000	`arm32`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
123353	`arm32`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
125329	`arm32`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
149447	`arm11`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
149581	`arm11`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
149598	`arm11`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
153132	`arm11`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
154350	`arm11`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
154508	`arm11`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
214877	`opt32`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
214942	`opt32`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
218895	`sphlib-small`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
221508	`opt32`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
224584	`opt32`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
260558	`sphlib-small`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
266337	`sphlib-small`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
268413	`sphlib-small`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
271385	`sphlib-small`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
271504	`32bit-2ktable`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
272533	`32bit-2ktable`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
274440	`sphlib-small`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
276801	`32bit-2ktable`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
279364	`32bit-2ktable`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
311744	`opt64`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
314348	`32bit-2ktable`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
341278	`opt64`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
344488	`32bit-2ktable`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
348931	`opt32`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
351408	`opt32`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
413887	`32bit-bytesliced-c-fast`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
442626	`sphlib`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
445418	`sphlib`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
448421	`sphlib-adapted`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
450511	`sphlib-adapted`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
472949	`sphlib`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
480731	`sphlib`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
481926	`sphlib`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
490018	`sphlib-adapted`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
491986	`opt64`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
495578	`sphlib-adapted`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
504561	`opt64`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
507770	`32bit-bytesliced-c-fast`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
527428	`sphlib`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
528355	`sphlib-adapted`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
530874	`32bit-bytesliced-c-fast`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
541670	`8bit_c`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
550928	`8bit_c`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
553077	`8bit_c`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
561828	`32bit-bytesliced-c-small`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
569744	`8bit_c`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
575324	`sphlib-adapted`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
575415	`32bit-bytesliced-c-small`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
581992	`32bit-bytesliced-c-small`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161209	20161026
644308	`32bit-bytesliced-c-fast`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026
651427	`opt64`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161209	20161026
666330	`opt64`	`gcc -mcpu=marvell-pj4 -O3`	20161209	20161026
713748	`8bit_c`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
756360	`32bit-bytesliced-c-fast`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
765331	`8bit_c`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
815424	`32bit-bytesliced-c-fast`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
1050543	`32bit-bytesliced-c-small`	`gcc -mcpu=marvell-pj4 -Os`	20161209	20161026
1056910	`32bit-bytesliced-c-small`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161209	20161026
1194470	`32bit-bytesliced-c-small`	`gcc -mcpu=marvell-pj4 -O2`	20161209	20161026

Checksum failure

Implementation: crypto_hash/groestl256/arm32
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

f079b87636261cf3c9ea6c0c0fa5429569bc7bd103f8d0f0bb23bd4ba5d49053

Number of similar (compiler,implementation) pairs: 3, namely:

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	arm32
gcc -funroll-loops -mcpu=marvell-pj4 -O3	arm32
gcc -mcpu=marvell-pj4 -O3	arm32

Test failure

Implementation: crypto_hash/groestl256/neon-bitslice
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

error 111

Number of similar (compiler,implementation) pairs: 22, namely:

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	neon-bitslice neon-table thumb-asm-fast
gcc -funroll-loops -mcpu=marvell-pj4 -O3	neon-bitslice neon-table thumb-asm-fast thumb-asm-small
gcc -funroll-loops -mcpu=marvell-pj4 -Os	neon-bitslice neon-table thumb-asm-fast thumb-asm-small
gcc -mcpu=marvell-pj4 -O2	neon-bitslice neon-table thumb-asm-fast thumb-asm-small
gcc -mcpu=marvell-pj4 -O3	neon-bitslice neon-table thumb-asm-fast thumb-asm-small
gcc -mcpu=marvell-pj4 -Os	neon-bitslice neon-table thumb-asm-small

Test failure

Implementation: crypto_hash/groestl256/thumb-asm-small
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

error 142
sh: line 1: 10451 Alarm clock killafter 3600 ./try

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

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	thumb-asm-small

Test failure

Implementation: crypto_hash/groestl256/thumb-asm-fast
Compiler: gcc -mcpu=marvell-pj4 -Os

error 142
sh: line 1: 10053 Alarm clock killafter 3600 ./try

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

Compiler	Implementations
gcc -mcpu=marvell-pj4 -Os	thumb-asm-fast

Compiler output

Implementation: crypto_hash/groestl256/vperm-intr
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

hash.c: In file included from hash.c:34:0:
hash.c: groestl-intr-vperm.h:13:23: fatal error: tmmintrin.h: No such file or directory
hash.c: #include gt;
hash.c: ^
hash.c: compilation terminated.

Number of similar (compiler,implementation) pairs: 6, namely:

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	vperm-intr
gcc -funroll-loops -mcpu=marvell-pj4 -O3	vperm-intr
gcc -funroll-loops -mcpu=marvell-pj4 -Os	vperm-intr
gcc -mcpu=marvell-pj4 -O2	vperm-intr
gcc -mcpu=marvell-pj4 -O3	vperm-intr
gcc -mcpu=marvell-pj4 -Os	vperm-intr

Compiler output

Implementation: crypto_hash/groestl256/neon-bitslice
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

hash.c: hash.c: In function 'crypto_hash_groestl256_neon_bitslice':
hash.c: hash.c:40:12: warning: iteration 64 invokes undefined behavior [-Waggressive-loop-optimizations]
hash.c: ctx[i] = 0;
hash.c: ~~~~~~~^~~
hash.c: hash.c:39:3: note: within this loop
hash.c: for(i=0;i hash.c: ^~~

Number of similar (compiler,implementation) pairs: 6, namely:

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	neon-bitslice
gcc -funroll-loops -mcpu=marvell-pj4 -O3	neon-bitslice
gcc -funroll-loops -mcpu=marvell-pj4 -Os	neon-bitslice
gcc -mcpu=marvell-pj4 -O2	neon-bitslice
gcc -mcpu=marvell-pj4 -O3	neon-bitslice
gcc -mcpu=marvell-pj4 -Os	neon-bitslice

Compiler output

Implementation: crypto_hash/groestl256/neon-vperm
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

hash.c: hash.c: In function 'crypto_hash_groestl256_neon_vperm':
hash.c: hash.c:38:12: warning: iteration 64 invokes undefined behavior [-Waggressive-loop-optimizations]
hash.c: ctx[i] = 0;
hash.c: ~~~~~~~^~~
hash.c: hash.c:37:3: note: within this loop
hash.c: for(i=0;i hash.c: ^~~
vperm-neon.S: vperm-neon.S: Assembler messages:
vperm-neon.S: vperm-neon.S:911: Error: expected symbol name
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q0,q0,q8' in ARM mode
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q1,q1,q9' in ARM mode
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q2,q2,q9' in ARM mode
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q3,q3,q9' in ARM mode
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q4,q4,q9' in ARM mode
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q5,q5,q9' in ARM mode
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q6,q6,q9' in ARM mode
vperm-neon.S: vperm-neon.S:913: Error: selected processor does not support `veor q7,q7,q10' in ARM mode
vperm-neon.S: ...
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.8 d0,d1' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.8 d2,d3' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.8 d4,d5' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.8 d6,d7' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.16 d0,d2' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.16 d1,d3' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.16 d4,d6' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.16 d5,d7' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.32 q0,q2' in ARM mode
vperm-neon.S: vperm-neon.S:1072: Error: selected processor does not support `vtrn.32 q1,q3' in ARM mode

Number of similar (compiler,implementation) pairs: 6, namely:

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	neon-vperm
gcc -funroll-loops -mcpu=marvell-pj4 -O3	neon-vperm
gcc -funroll-loops -mcpu=marvell-pj4 -Os	neon-vperm
gcc -mcpu=marvell-pj4 -O2	neon-vperm
gcc -mcpu=marvell-pj4 -O3	neon-vperm
gcc -mcpu=marvell-pj4 -Os	neon-vperm

Compiler output

Implementation: crypto_hash/groestl256/opt64
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

hash.c: hash.c:194:14: warning: 'inP' is static but declared in inline function 'F1024' which is not static
hash.c: static u64 inP[COLS1024] __attribute__((aligned(16)));
hash.c: ^~~
hash.c: hash.c:193:14: warning: 'outQ' is static but declared in inline function 'F1024' which is not static
hash.c: static u64 outQ[COLS1024] __attribute__((aligned(16)));
hash.c: ^~~~
hash.c: hash.c:192:14: warning: 'z' is static but declared in inline function 'F1024' which is not static
hash.c: static u64 z[COLS1024] __attribute__((aligned(16)));
hash.c: ^
hash.c: hash.c:191:14: warning: 'y' is static but declared in inline function 'F1024' which is not static
hash.c: static u64 y[COLS1024] __attribute__((aligned(16)));
hash.c: ^

Number of similar (compiler,implementation) pairs: 6, namely:

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	opt64
gcc -funroll-loops -mcpu=marvell-pj4 -O3	opt64
gcc -funroll-loops -mcpu=marvell-pj4 -Os	opt64
gcc -mcpu=marvell-pj4 -O2	opt64
gcc -mcpu=marvell-pj4 -O3	opt64
gcc -mcpu=marvell-pj4 -Os	opt64