Implementation notes: amd64, rumba5, crypto_hash/shabal256

Computer: rumba5
Architecture: amd64
CPU ID: AuthenticAMD-00800f11-178bfbff
SUPERCOP version: 20171218
Operation: crypto_hash
Primitive: shabal256
TimeImplementationCompilerBenchmark dateSUPERCOP version
11360amd64gcc -m64 -march=core2 -Os -fomit-frame-pointer2018051720171218
11392amd64cc2018051720171218
11392amd64gcc -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -fno-schedule-insns -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -fno-schedule-insns -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -fno-schedule-insns -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops2018051720171218
11392amd64gcc -funroll-loops -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -m64 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -m64 -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -m64 -march=barcelona -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -m64 -march=k8 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -m64 -march=nocona -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -march=barcelona -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -march=k8 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -march=nocona -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -march=nocona -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -march=nocona -O -fomit-frame-pointer2018051720171218
11392amd64gcc -funroll-loops -march=nocona -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -O -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=barcelona -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core-avx-i -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core-avx-i -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core-avx-i -O -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core-avx-i -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core-avx2 -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core-avx2 -O -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core-avx2 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -msse4.1 -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -msse4.1 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -msse4 -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -msse4 -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -msse4 -O -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=core2 -msse4 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=corei7-avx -O -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=corei7-avx -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=corei7 -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=corei7 -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=corei7 -O -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=corei7 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=k8 -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=k8 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=native -mtune=native -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=nocona -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=nocona -O -fomit-frame-pointer2018051720171218
11392amd64gcc -m64 -march=nocona -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -march=barcelona -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -march=k8 -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -march=k8 -O -fomit-frame-pointer2018051720171218
11392amd64gcc -march=k8 -Os -fomit-frame-pointer2018051720171218
11392amd64gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv2018051720171218
11392amd64gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv2018051720171218
11392amd64gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv2018051720171218
11392amd64gcc -march=nocona -O2 -fomit-frame-pointer2018051720171218
11392amd64gcc -march=nocona -O3 -fomit-frame-pointer2018051720171218
11392amd64gcc -march=nocona -Os -fomit-frame-pointer2018051720171218
11424amd64gcc2018051720171218
11424amd64gcc -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -O -fomit-frame-pointer2018051720171218
11424amd64gcc -Os -fomit-frame-pointer2018051720171218
11424amd64gcc -fno-schedule-insns -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=barcelona -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=k8 -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=k8 -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=k8 -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=nocona -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=nocona -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -m64 -march=nocona -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -march=barcelona -O -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -march=k8 -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -march=k8 -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -funroll-loops -march=k8 -O -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -Os -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=barcelona -O -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=core-avx2 -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=core2 -O -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=core2 -msse4.1 -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=core2 -msse4.1 -O -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=corei7-avx -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=corei7-avx -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=k8 -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=k8 -O -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=native -mtune=native -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=native -mtune=native -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=native -mtune=native -O -fomit-frame-pointer2018051720171218
11424amd64gcc -m64 -march=nocona -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11424amd64gcc -march=barcelona -O -fomit-frame-pointer2018051720171218
11424amd64gcc -march=k8 -O3 -fomit-frame-pointer2018051720171218
11424amd64gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv2018051720171218
11424amd64gcc -march=nocona -O -fomit-frame-pointer2018051720171218
11456sphlibgcc -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer2018051720171218
11520sphlibgcc -funroll-loops -m64 -O3 -fomit-frame-pointer2018051720171218
11584sphlibclang -O3 -fomit-frame-pointer -Qunused-arguments2018051720171218
11584sphlibclang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer -Qunused-arguments2018051720171218
11584sphlibgcc -fno-schedule-insns -O3 -fomit-frame-pointer2018051720171218
11584sphlibgcc -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer2018051720171218
11584sphlibgcc -funroll-loops -m64 -O2 -fomit-frame-pointer2018051720171218
11584sphlibgcc -funroll-loops -march=k8 -O2 -fomit-frame-pointer2018051720171218
11584sphlibgcc -m64 -march=core-avx2 -O2 -fomit-frame-pointer2018051720171218
11616sphlibgcc -funroll-loops -O2 -fomit-frame-pointer2018051720171218
11616sphlibgcc -funroll-loops -m64 -march=k8 -O2 -fomit-frame-pointer2018051720171218
11616sphlibgcc -funroll-loops -m64 -march=k8 -O3 -fomit-frame-pointer2018051720171218
11648sphlibclang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2018051720171218
11648sphlibclang -mcpu=cortex-a9 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2018051720171218
11648sphlibclang -mcpu=native -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2018051720171218
11648sphlibgcc -m64 -march=native -mtune=native -O2 -fomit-frame-pointer2018051720171218
11680sphlibclang -O3 -fwrapv -march=native -fomit-frame-pointer -Qunused-arguments2018051720171218
11680sphlibclang -O3 -fwrapv -mavx -fomit-frame-pointer -Qunused-arguments2018051720171218
11680sphlibgcc -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -fno-schedule-insns -O2 -fomit-frame-pointer2018051720171218
11680sphlibgcc -funroll-loops -m64 -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11680sphlibgcc -funroll-loops -m64 -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -funroll-loops -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11680sphlibgcc -funroll-loops -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -m64 -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -m64 -march=core-avx-i -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -m64 -march=core2 -O2 -fomit-frame-pointer2018051720171218
11680sphlibgcc -m64 -march=core2 -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -m64 -march=corei7 -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -m64 -march=native -mtune=native -O3 -fomit-frame-pointer2018051720171218
11680sphlibgcc -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11680sphlibgcc -march=k8 -O2 -fomit-frame-pointer2018051720171218
11712sphlibclang -O3 -fwrapv -mavx -maes -mpclmul -fomit-frame-pointer -Qunused-arguments2018051720171218
11712sphlibclang -mcpu=cortex-a8 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2018051720171218
11712sphlibgcc -funroll-loops -O3 -fomit-frame-pointer2018051720171218
11712sphlibgcc -m64 -O2 -fomit-frame-pointer2018051720171218
11712sphlibgcc -m64 -march=barcelona -O2 -fomit-frame-pointer2018051720171218
11712sphlibgcc -m64 -march=core-avx-i -O2 -fomit-frame-pointer2018051720171218
11712sphlibgcc -m64 -march=core2 -msse4.1 -O2 -fomit-frame-pointer2018051720171218
11712sphlibgcc -m64 -march=corei7-avx -O2 -fomit-frame-pointer2018051720171218
11712sphlibgcc -m64 -march=corei7 -O2 -fomit-frame-pointer2018051720171218
11712sphlibgcc -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11744sphlibgcc -m64 -march=barcelona -O3 -fomit-frame-pointer2018051720171218
11744sphlibgcc -m64 -march=core-avx2 -O3 -fomit-frame-pointer2018051720171218
11744sphlibgcc -m64 -march=core2 -msse4 -O2 -fomit-frame-pointer2018051720171218
11744sphlibgcc -m64 -march=corei7-avx -O3 -fomit-frame-pointer2018051720171218
11744sphlibgcc -march=k8 -O3 -fomit-frame-pointer2018051720171218
11776sphlibgcc -funroll-loops -march=k8 -O3 -fomit-frame-pointer2018051720171218
11776sphlibgcc -m64 -march=core-avx2 -O -fomit-frame-pointer2018051720171218
11776sphlibgcc -m64 -march=core2 -msse4.1 -O3 -fomit-frame-pointer2018051720171218
11776sphlibgcc -m64 -march=k8 -O3 -fomit-frame-pointer2018051720171218
11776sphlibgcc -m64 -march=native -mtune=native -O -fomit-frame-pointer2018051720171218
11776sphlibgcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv2018051720171218
11776sphlibgcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv2018051720171218
11808sphlibgcc -m64 -march=core2 -msse4 -O3 -fomit-frame-pointer2018051720171218
11808sphlibgcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv2018051720171218
11840sphlibgcc -O2 -fomit-frame-pointer2018051720171218
11840sphlibgcc -m64 -march=k8 -O2 -fomit-frame-pointer2018051720171218
11872sphlibclang -O3 -fwrapv -mavx2 -fomit-frame-pointer -Qunused-arguments2018051720171218
11872sphlibgcc -funroll-loops -m64 -march=k8 -O -fomit-frame-pointer2018051720171218
11872sphlibgcc -funroll-loops -march=barcelona -O -fomit-frame-pointer2018051720171218
11872sphlibgcc -march=nocona -O3 -fomit-frame-pointer2018051720171218
11904sphlibgcc -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer2018051720171218
11904sphlibgcc -funroll-loops -march=k8 -O -fomit-frame-pointer2018051720171218
11904sphlibgcc -funroll-loops -march=nocona -O2 -fomit-frame-pointer2018051720171218
11936sphlibgcc -funroll-loops -m64 -O -fomit-frame-pointer2018051720171218
11936sphlibgcc -funroll-loops -m64 -march=barcelona -O -fomit-frame-pointer2018051720171218
11936sphlibgcc -funroll-loops -m64 -march=nocona -O2 -fomit-frame-pointer2018051720171218
11936sphlibgcc -m64 -O -fomit-frame-pointer2018051720171218
11936sphlibgcc -m64 -march=core2 -msse4.1 -O -fomit-frame-pointer2018051720171218
11936sphlibgcc -m64 -march=corei7-avx -O -fomit-frame-pointer2018051720171218
11936sphlibgcc -m64 -march=nocona -O2 -fomit-frame-pointer2018051720171218
11936sphlibgcc -m64 -march=nocona -O -fomit-frame-pointer2018051720171218
11936sphlibgcc -march=nocona -O2 -fomit-frame-pointer2018051720171218
11936sphlibgcc -march=nocona -O -fomit-frame-pointer2018051720171218
11968sphlibgcc -O -fomit-frame-pointer2018051720171218
11968sphlibgcc -funroll-loops -O -fomit-frame-pointer2018051720171218
11968sphlibgcc -funroll-loops -m64 -march=nocona -O -fomit-frame-pointer2018051720171218
11968sphlibgcc -funroll-loops -march=nocona -O -fomit-frame-pointer2018051720171218
11968sphlibgcc -m64 -march=core2 -msse4 -O -fomit-frame-pointer2018051720171218
11968sphlibgcc -m64 -march=nocona -O3 -fomit-frame-pointer2018051720171218
11968sphlibgcc -march=k8 -O -fomit-frame-pointer2018051720171218
12000sphlibgcc -funroll-loops -m64 -march=nocona -O3 -fomit-frame-pointer2018051720171218
12000sphlibgcc -funroll-loops -march=nocona -O3 -fomit-frame-pointer2018051720171218
12000sphlibgcc -m64 -march=barcelona -O -fomit-frame-pointer2018051720171218
12000sphlibgcc -m64 -march=core-avx-i -O -fomit-frame-pointer2018051720171218
12000sphlibgcc -m64 -march=corei7 -O -fomit-frame-pointer2018051720171218
12000sphlibgcc -m64 -march=k8 -O -fomit-frame-pointer2018051720171218
12032sphlibgcc -m64 -march=core2 -O -fomit-frame-pointer2018051720171218
12064sphlibgcc -fno-schedule-insns -O -fomit-frame-pointer2018051720171218
12064sphlibgcc -march=barcelona -O -fomit-frame-pointer2018051720171218
13920sphlibgcc -m64 -march=core-avx2 -Os -fomit-frame-pointer2018051720171218
13984sphlibgcc -funroll-loops -march=barcelona -Os -fomit-frame-pointer2018051720171218
14016sphlibgcc -funroll-loops -m64 -march=barcelona -Os -fomit-frame-pointer2018051720171218
14016sphlibgcc -funroll-loops -m64 -march=k8 -Os -fomit-frame-pointer2018051720171218
14016sphlibgcc -funroll-loops -march=k8 -Os -fomit-frame-pointer2018051720171218
14016sphlibgcc -m64 -march=native -mtune=native -Os -fomit-frame-pointer2018051720171218
14048sphlibgcc -funroll-loops -m64 -Os -fomit-frame-pointer2018051720171218
14080sphlibgcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv2018051720171218
14144sphlibgcc -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer2018051720171218
14208sphlibgcc -fno-schedule-insns -Os -fomit-frame-pointer2018051720171218
14208sphlibgcc -m64 -Os -fomit-frame-pointer2018051720171218
14240sphlibgcc -funroll-loops -Os -fomit-frame-pointer2018051720171218
14240sphlibgcc -m64 -march=core2 -Os -fomit-frame-pointer2018051720171218
14240sphlibgcc -m64 -march=core2 -msse4 -Os -fomit-frame-pointer2018051720171218
14240sphlibgcc -m64 -march=corei7-avx -Os -fomit-frame-pointer2018051720171218
14240sphlibgcc -m64 -march=k8 -Os -fomit-frame-pointer2018051720171218
14272sphlibgcc -m64 -march=core-avx-i -Os -fomit-frame-pointer2018051720171218
14304sphlibgcc -Os -fomit-frame-pointer2018051720171218
14304sphlibgcc -march=k8 -Os -fomit-frame-pointer2018051720171218
14336sphlibgcc -m64 -march=barcelona -Os -fomit-frame-pointer2018051720171218
14336sphlibgcc -m64 -march=core2 -msse4.1 -Os -fomit-frame-pointer2018051720171218
14336sphlibgcc -m64 -march=corei7 -Os -fomit-frame-pointer2018051720171218
14368sphlibgcc -march=nocona -Os -fomit-frame-pointer2018051720171218
14400sphlibgcc -march=barcelona -Os -fomit-frame-pointer2018051720171218
14432sphlibgcc -m64 -march=nocona -Os -fomit-frame-pointer2018051720171218
14592sphlibgcc -funroll-loops -march=nocona -Os -fomit-frame-pointer2018051720171218
14624sphlibgcc -funroll-loops -m64 -march=nocona -Os -fomit-frame-pointer2018051720171218
30656sphlibcc2018051720171218
30688sphlibgcc -funroll-loops2018051720171218
30816sphlibgcc2018051720171218

Compiler output

Implementation: crypto_hash/shabal256/amd64
Compiler: clang -O3 -fomit-frame-pointer -Qunused-arguments
shabal_amd64.s: shabal_amd64.s:30:14: error: single quote way too long
shabal_amd64.s: # 64 'ptr' field (8 bytes)
shabal_amd64.s: ^

Number of similar (compiler,implementation) pairs: 10, namely:
CompilerImplementations
clang -O3 -fomit-frame-pointer -Qunused-arguments amd64
clang -O3 -fwrapv -march=native -fomit-frame-pointer -Qunused-arguments amd64
clang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer -Qunused-arguments amd64
clang -O3 -fwrapv -mavx2 -fomit-frame-pointer -Qunused-arguments amd64
clang -O3 -fwrapv -mavx -fomit-frame-pointer -Qunused-arguments amd64
clang -O3 -fwrapv -mavx -maes -mpclmul -fomit-frame-pointer -Qunused-arguments amd64
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments amd64
clang -mcpu=cortex-a8 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments amd64
clang -mcpu=cortex-a9 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments amd64
clang -mcpu=native -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments amd64