Implementation notes: armeabi, cubox, crypto_hash/keccakc448

Computer: cubox
Architecture: armeabi
CPU ID: unknown CPU ID
SUPERCOP version: 20161026
Operation: crypto_hash
Primitive: keccakc448

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
92315	`opt32bi-rvku2`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
103369	`opt32bi-rvku2`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
108993	`opt32bi-s2lcu4`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
120150	`opt32biT-s2lcu4`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
128709	`opt32bi-s2lcu4`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
139542	`opt32biT-s2lcu4`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
142730	`opt32bi-rvku2`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
145596	`opt32bi-s2lcu4`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
146184	`opt32bi-s2lcu4`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
153370	`opt32biT-s2lcu4`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
153828	`opt32bi-rvku2`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
157292	`opt32bi-rvku2`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
158493	`opt32bi-s2lcu4`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
160313	`opt32bi-s2lcu4`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
160946	`opt32biT-s2lcu4`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
161725	`opt32bi-rvku2`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
166054	`opt32biT-s2lcu4`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
177200	`opt32biT-s2lcu4`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
195702	`opt64lcu6`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
197253	`simple`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
198623	`inplace`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
201457	`opt64u6`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
202417	`simple`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
206084	`inplace`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
213624	`simple`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
213856	`simple`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
218442	`opt64lcu6`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
219255	`simple`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
219531	`opt64u6`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
220563	`opt64lcu6`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
220761	`opt64lcu6`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
220958	`simple`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
221696	`inplace`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
223032	`opt64u6`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
225564	`inplace`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
230174	`opt64u6`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
231589	`inplace`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
236558	`inplace`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
238788	`opt64lcu24`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
244912	`opt64lcu24`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
248554	`opt64lcu6`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
251297	`compact`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
254343	`opt64u6`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
256113	`compact`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
257432	`opt64lcu6`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
263363	`opt64u6`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
278599	`compact8`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
281284	`compact8`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
319849	`compact`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
404709	`compact`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
424280	`opt64lcu24`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161210	20161026
426270	`opt64lcu24`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
440398	`opt64lcu24`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
441840	`opt64lcu24`	`gcc -mcpu=marvell-pj4 -O3`	20161210	20161026
495291	`compact8`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161210	20161026
521385	`compact8`	`gcc -mcpu=marvell-pj4 -O2`	20161210	20161026
548616	`compact`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026
585843	`compact`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
776496	`compact8`	`gcc -mcpu=marvell-pj4 -Os`	20161210	20161026
788934	`compact8`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161210	20161026

Test failure

Implementation: crypto_hash/keccakc448/inplace32bi-armv7a
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

error 111
crypto_hash does not handle m=h overlap

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

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	inplace32bi-armv7a
gcc -funroll-loops -mcpu=marvell-pj4 -O3	inplace32bi-armv7a
gcc -funroll-loops -mcpu=marvell-pj4 -Os	inplace32bi-armv7a
gcc -mcpu=marvell-pj4 -O2	inplace32bi-armv7a
gcc -mcpu=marvell-pj4 -O3	inplace32bi-armv7a
gcc -mcpu=marvell-pj4 -Os	inplace32bi-armv7a

Compiler output

Implementation: crypto_hash/keccakc448/inplace32bi
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

Keccak-inplace32BI.c: Keccak-inplace32BI.c: In function 'crypto_hash_keccakc448_inplace32bi':
Keccak-inplace32BI.c: Keccak-inplace32BI.c:73:6: error: #error "The output size must be a multiple of the lane size in this simple implementation."
Keccak-inplace32BI.c: #error "The output size must be a multiple of the lane size in this simple implementation."
Keccak-inplace32BI.c: ^~~~~

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

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	inplace32bi
gcc -funroll-loops -mcpu=marvell-pj4 -O3	inplace32bi
gcc -funroll-loops -mcpu=marvell-pj4 -Os	inplace32bi
gcc -mcpu=marvell-pj4 -O2	inplace32bi
gcc -mcpu=marvell-pj4 -O3	inplace32bi
gcc -mcpu=marvell-pj4 -Os	inplace32bi

Compiler output

Implementation: crypto_hash/keccakc448/simple32bi
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

Keccak-simple32BI.c: Keccak-simple32BI.c: In function 'crypto_hash_keccakc448_simple32bi':
Keccak-simple32BI.c: Keccak-simple32BI.c:73:6: error: #error "The output size must be a multiple of the lane size in this simple implementation."
Keccak-simple32BI.c: #error "The output size must be a multiple of the lane size in this simple implementation."
Keccak-simple32BI.c: ^~~~~

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

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	simple32bi
gcc -funroll-loops -mcpu=marvell-pj4 -O3	simple32bi
gcc -funroll-loops -mcpu=marvell-pj4 -Os	simple32bi
gcc -mcpu=marvell-pj4 -O2	simple32bi
gcc -mcpu=marvell-pj4 -O3	simple32bi
gcc -mcpu=marvell-pj4 -Os	simple32bi

Compiler output

Implementation: crypto_hash/keccakc448/inplace32bi-armv6m
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

try.c: /usr/bin/ld: error: crypto_hash_keccakc448.a(Keccak-inplace32BI-armgcc-ARMv6M.o): Conflicting architecture profiles M/A
try.c: /usr/bin/ld: failed to merge target specific data of file crypto_hash_keccakc448.a(Keccak-inplace32BI-armgcc-ARMv6M.o)
try.c: collect2: error: ld returned 1 exit status

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

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	inplace32bi-armv6m
gcc -funroll-loops -mcpu=marvell-pj4 -O3	inplace32bi-armv6m
gcc -funroll-loops -mcpu=marvell-pj4 -Os	inplace32bi-armv6m
gcc -mcpu=marvell-pj4 -O2	inplace32bi-armv6m
gcc -mcpu=marvell-pj4 -O3	inplace32bi-armv6m
gcc -mcpu=marvell-pj4 -Os	inplace32bi-armv6m