Implementation notes: armeabi, cubox, crypto_core/aes256decrypt

Computer: cubox
Architecture: armeabi
CPU ID: unknown CPU ID
SUPERCOP version: 20161026
Operation: crypto_core
Primitive: aes256decrypt

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
2307	`dolbeau/std-4rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161208	20161026
2315	`dolbeau/std-4rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161208	20161026
2424	`dolbeau/std-4rt-nodk`	`gcc -mcpu=marvell-pj4 -O2`	20161208	20161026
2428	`dolbeau/std-4rt-nodk`	`gcc -mcpu=marvell-pj4 -O3`	20161208	20161026
2448	`dolbeau/std-2rt-nodk`	`gcc -mcpu=marvell-pj4 -O2`	20161208	20161026
2448	`dolbeau/std-2rt-nodk`	`gcc -mcpu=marvell-pj4 -O3`	20161208	20161026
2467	`dolbeau/std-2rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161208	20161026
2468	`dolbeau/std-2rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161208	20161026
2636	`dolbeau/std-1rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161208	20161026
2655	`dolbeau/std-1rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161208	20161026
2662	`dolbeau/std-1rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161208	20161026
2676	`dolbeau/std-1rt-nodk`	`gcc -mcpu=marvell-pj4 -O2`	20161208	20161026
2704	`dolbeau/std-1rt-nodk`	`gcc -mcpu=marvell-pj4 -O3`	20161208	20161026
2752	`dolbeau/std-4rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161208	20161026
2764	`dolbeau/std-2rt-nodk`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161208	20161026
2764	`dolbeau/std-1rt-nodk`	`gcc -mcpu=marvell-pj4 -Os`	20161208	20161026
2828	`dolbeau/std-2rt-nodk`	`gcc -mcpu=marvell-pj4 -Os`	20161208	20161026
2848	`dolbeau/std-4rt-nodk`	`gcc -mcpu=marvell-pj4 -Os`	20161208	20161026
3012	`openssl`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161208	20161026
3016	`openssl`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161208	20161026
3016	`openssl`	`gcc -mcpu=marvell-pj4 -O2`	20161208	20161026
3053	`openssl`	`gcc -mcpu=marvell-pj4 -O3`	20161208	20161026
3110	`openssl`	`gcc -mcpu=marvell-pj4 -Os`	20161208	20161026
3112	`openssl`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161208	20161026
935801	`ref`	`gcc -funroll-loops -mcpu=marvell-pj4 -O3`	20161208	20161026
955871	`ref`	`gcc -mcpu=marvell-pj4 -O3`	20161208	20161026
1345666	`ref`	`gcc -funroll-loops -mcpu=marvell-pj4 -O2`	20161208	20161026
3477660	`ref`	`gcc -mcpu=marvell-pj4 -O2`	20161208	20161026
3573185	`ref`	`gcc -funroll-loops -mcpu=marvell-pj4 -Os`	20161208	20161026
3677488	`ref`	`gcc -mcpu=marvell-pj4 -Os`	20161208	20161026

Compiler output

Implementation: crypto_core/aes256decrypt/dolbeau/armv8crypto
Compiler: gcc -funroll-loops -mcpu=marvell-pj4 -O2

core.c: In file included from core.c:9:0:
core.c: core.c: In function 'aes256_setkey_armv8_decrypt':
core.c: /usr/lib/gcc/armv7l-unknown-linux-gnueabihf/6.2.1/include/arm_neon.h:14742:1: error: inlining failed in call to always_inline 'vaesimcq_u8': target specific option mismatch
core.c: vaesimcq_u8 (uint8x16_t __data)
core.c: ^~~~~~~~~~~
core.c: core.c:21:16: note: called from here
core.c: rkeys[i] = vaesimcq_u8(tkeys[14-i]);
core.c: ^~~~~~~~~~~~~~~~~~~~~~~~

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

Compiler	Implementations
gcc -funroll-loops -mcpu=marvell-pj4 -O2	dolbeau/armv8crypto
gcc -funroll-loops -mcpu=marvell-pj4 -O3	dolbeau/armv8crypto
gcc -funroll-loops -mcpu=marvell-pj4 -Os	dolbeau/armv8crypto
gcc -mcpu=marvell-pj4 -O2	dolbeau/armv8crypto
gcc -mcpu=marvell-pj4 -O3	dolbeau/armv8crypto
gcc -mcpu=marvell-pj4 -Os	dolbeau/armv8crypto