Implementation notes: armeabi, cubie2, crypto_aead/joltikneq8048v1

Computer: cubie2
Architecture: armeabi
CPU ID: unknown CPU ID
SUPERCOP version: 20161026
Operation: crypto_aead
Primitive: joltikneq8048v1

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
11794581	`ref`	`gcc -mcpu=native -mfpu=neon-vfpv4 -O3`	20161220	20161026
11815732	`ref`	`gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3`	20161220	20161026
21877356	`ref`	`gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2`	20161220	20161026
35234132	`ref`	`gcc -mcpu=native -mfpu=neon-vfpv4 -O2`	20161220	20161026
36719465	`ref`	`gcc -mcpu=native -mfpu=neon-vfpv4 -Os`	20161220	20161026
37340211	`ref`	`gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os`	20161220	20161026

Compiler output

Implementation: crypto_aead/joltikneq8048v1/vperm
Compiler: gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2

joltik.c: joltik.c:29:3: error: #error "This implementation requires either SSSE3 or AVX2 instruction sets."
joltik.c: #error "This implementation requires either SSSE3 or AVX2 instruction sets."
joltik.c: ^~~~~

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

Compiler	Implementations
gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O2	vperm
gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -O3	vperm
gcc -funroll-loops -mcpu=native -mfpu=neon-vfpv4 -Os	vperm
gcc -mcpu=native -mfpu=neon-vfpv4 -O2	vperm
gcc -mcpu=native -mfpu=neon-vfpv4 -O3	vperm
gcc -mcpu=native -mfpu=neon-vfpv4 -Os	vperm