Implementation notes: amd64, mangetsu, crypto_sign/mqqsig160

Computer: mangetsu
Architecture: amd64
CPU ID: GenuineIntel-000206a7-bfebfbff
SUPERCOP version: 20141014
Operation: crypto_sign
Primitive: mqqsig160

Time	Implementation	Compiler	Benchmark date	SUPERCOP version
86872	`ref`	`gcc-4.8 -m64 -march=core2 -Os -fomit-frame-pointer`	20140927	20140907
87232	`ref`	`gcc-4.8 -m64 -march=core2 -msse4 -Os -fomit-frame-pointer`	20140927	20140907
87344	`ref`	`gcc-4.8 -m64 -march=core2 -msse4.1 -Os -fomit-frame-pointer`	20140927	20140907
87524	`ref`	`gcc-4.8 -m64 -march=corei7 -Os -fomit-frame-pointer`	20140927	20140907
88052	`ref`	`gcc-4.8 -m64 -march=native -mtune=native -Os -fomit-frame-pointer`	20140927	20140907
88060	`ref`	`gcc-4.8 -m64 -march=core-avx-i -Os -fomit-frame-pointer`	20140927	20140907
88128	`ref`	`gcc-4.8 -m64 -march=corei7-avx -Os -fomit-frame-pointer`	20140927	20140907
88864	`ref`	`gcc-4.7 -m64 -march=corei7-avx -Os -fomit-frame-pointer`	20140926	20140907
88932	`ref`	`gcc-4.7 -m64 -march=native -mtune=native -Os -fomit-frame-pointer`	20140926	20140907
89392	`ref`	`gcc-4.7 -m64 -march=core-avx-i -Os -fomit-frame-pointer`	20140926	20140907
90188	`ref`	`gcc-4.7 -m64 -march=core2 -msse4.1 -Os -fomit-frame-pointer`	20140926	20140907
90376	`ref`	`gcc-4.7 -m64 -march=core2 -msse4 -Os -fomit-frame-pointer`	20140926	20140907
90748	`ref`	`gcc-4.7 -m64 -march=core2 -Os -fomit-frame-pointer`	20140926	20140907
90756	`ref`	`gcc-4.7 -m64 -march=corei7 -Os -fomit-frame-pointer`	20140926	20140907
204036	`ref`	`gcc-4.8 -march=nocona -Os -fomit-frame-pointer`	20140927	20140907
206392	`ref`	`gcc-4.8 -m64 -march=nocona -Os -fomit-frame-pointer`	20140927	20140907
207200	`ref`	`gcc-4.8 -m64 -march=k8 -Os -fomit-frame-pointer`	20140927	20140907
207296	`ref`	`gcc-4.8 -fno-schedule-insns -Os -fomit-frame-pointer`	20140927	20140907
207420	`ref`	`gcc-4.8 -m64 -Os -fomit-frame-pointer`	20140927	20140907
207584	`ref`	`gcc-4.8 -Os -fomit-frame-pointer`	20140927	20140907
207824	`ref`	`gcc-4.8 -funroll-loops -march=nocona -Os -fomit-frame-pointer`	20140927	20140907
208052	`ref`	`gcc-4.8 -m64 -march=barcelona -Os -fomit-frame-pointer`	20140927	20140907
208160	`ref`	`gcc-4.8 -funroll-loops -m64 -march=nocona -Os -fomit-frame-pointer`	20140927	20140907
208432	`ref`	`gcc-4.8 -march=k8 -Os -fomit-frame-pointer`	20140927	20140907
208864	`ref`	`gcc-4.8 -march=barcelona -Os -fomit-frame-pointer`	20140927	20140907
235172	`ref`	`gcc-4.8 -funroll-loops -Os -fomit-frame-pointer`	20140927	20140907
235180	`ref`	`gcc-4.8 -funroll-loops -march=k8 -Os -fomit-frame-pointer`	20140927	20140907
235352	`ref`	`gcc-4.8 -funroll-loops -march=barcelona -Os -fomit-frame-pointer`	20140927	20140907
235356	`ref`	`gcc-4.8 -funroll-loops -m64 -Os -fomit-frame-pointer`	20140927	20140907
235592	`ref`	`gcc-4.8 -funroll-loops -m64 -march=k8 -Os -fomit-frame-pointer`	20140927	20140907
235608	`ref`	`gcc-4.8 -funroll-loops -m64 -march=barcelona -Os -fomit-frame-pointer`	20140927	20140907
236760	`ref`	`gcc-4.8 -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer`	20140927	20140907
259472	`ref`	`gcc-4.7 -funroll-loops -m64 -march=k8 -Os -fomit-frame-pointer`	20140926	20140907
259804	`ref`	`gcc-4.7 -funroll-loops -march=k8 -Os -fomit-frame-pointer`	20140926	20140907
259848	`ref`	`gcc-4.7 -m64 -march=k8 -Os -fomit-frame-pointer`	20140926	20140907
260172	`ref`	`gcc-4.7 -m64 -march=barcelona -Os -fomit-frame-pointer`	20140926	20140907
260268	`ref`	`gcc-4.7 -funroll-loops -m64 -Os -fomit-frame-pointer`	20140926	20140907
260348	`ref`	`gcc-4.7 -m64 -Os -fomit-frame-pointer`	20140926	20140907
260496	`ref`	`gcc-4.7 -march=barcelona -Os -fomit-frame-pointer`	20140926	20140907
260524	`ref`	`gcc-4.7 -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer`	20140927	20140907
260544	`ref`	`gcc-4.7 -funroll-loops -Os -fomit-frame-pointer`	20140926	20140907
260596	`ref`	`gcc-4.7 -m64 -march=nocona -Os -fomit-frame-pointer`	20140926	20140907
260608	`ref`	`gcc-4.7 -funroll-loops -march=barcelona -Os -fomit-frame-pointer`	20140927	20140907
260628	`ref`	`gcc-4.7 -fno-schedule-insns -Os -fomit-frame-pointer`	20140927	20140907
260992	`ref`	`gcc-4.7 -funroll-loops -march=nocona -Os -fomit-frame-pointer`	20140926	20140907
261000	`ref`	`gcc-4.7 -Os -fomit-frame-pointer`	20140926	20140907
261024	`ref`	`gcc-4.7 -march=k8 -Os -fomit-frame-pointer`	20140926	20140907
261052	`ref`	`gcc-4.7 -funroll-loops -m64 -march=barcelona -Os -fomit-frame-pointer`	20140926	20140907
261156	`ref`	`gcc-4.7 -march=nocona -Os -fomit-frame-pointer`	20140926	20140907
261160	`ref`	`gcc-4.7 -funroll-loops -m64 -march=nocona -Os -fomit-frame-pointer`	20140926	20140907

Test failure

Implementation: crypto_sign/mqqsig160/ref
Compiler: cc

error 111
crypto_sign_open does not handle c=t overlap

Number of similar (compiler,implementation) pairs: 156, namely:

Compiler	Implementations
cc	ref
clang -O3 -fomit-frame-pointer	ref
clang -O3 -fwrapv -march=native -fomit-frame-pointer	ref
clang -O3 -fwrapv -march=native -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer	ref
clang -O3 -fwrapv -mavx -fomit-frame-pointer	ref
clang -O3 -fwrapv -mavx -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer	ref
clang -O3 -fwrapv -mavx -maes -mpclmul -fomit-frame-pointer	ref
clang -O3 -fwrapv -mavx -maes -mpclmul -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer	ref
gcc-4.7	ref
gcc-4.7 -O2 -fomit-frame-pointer	ref
gcc-4.7 -O3 -fomit-frame-pointer	ref
gcc-4.7 -O -fomit-frame-pointer	ref
gcc-4.7 -fno-schedule-insns -O2 -fomit-frame-pointer	ref
gcc-4.7 -fno-schedule-insns -O3 -fomit-frame-pointer	ref
gcc-4.7 -fno-schedule-insns -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops	ref
gcc-4.7 -funroll-loops -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=barcelona -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=k8 -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -m64 -march=nocona -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=barcelona -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=k8 -O -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.7 -funroll-loops -march=nocona -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core-avx-i -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core-avx-i -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core-avx-i -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -msse4.1 -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -msse4.1 -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -msse4.1 -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -msse4 -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -msse4 -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core2 -msse4 -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=corei7-avx -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=corei7-avx -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=corei7-avx -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=corei7 -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=corei7 -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=corei7 -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=k8 -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=native -mtune=native -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=native -mtune=native -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=native -mtune=native -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=nocona -O -fomit-frame-pointer	ref
gcc-4.7 -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.7 -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.7 -march=barcelona -O -fomit-frame-pointer	ref
gcc-4.7 -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.7 -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.7 -march=k8 -O -fomit-frame-pointer	ref
gcc-4.7 -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.7 -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.7 -march=nocona -O -fomit-frame-pointer	ref
gcc-4.8	ref
gcc-4.8 -O2 -fomit-frame-pointer	ref
gcc-4.8 -O3 -fomit-frame-pointer	ref
gcc-4.8 -O -fomit-frame-pointer	ref
gcc-4.8 -fno-schedule-insns -O2 -fomit-frame-pointer	ref
gcc-4.8 -fno-schedule-insns -O3 -fomit-frame-pointer	ref
gcc-4.8 -fno-schedule-insns -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops	ref
gcc-4.8 -funroll-loops -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=barcelona -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=k8 -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -m64 -march=nocona -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=barcelona -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=k8 -O -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.8 -funroll-loops -march=nocona -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core-avx-i -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core-avx-i -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core-avx-i -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -msse4.1 -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -msse4.1 -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -msse4.1 -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -msse4 -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -msse4 -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core2 -msse4 -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=corei7-avx -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=corei7-avx -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=corei7-avx -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=corei7 -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=corei7 -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=corei7 -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=k8 -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=native -mtune=native -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=native -mtune=native -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=native -mtune=native -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=nocona -O -fomit-frame-pointer	ref
gcc-4.8 -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.8 -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.8 -march=barcelona -O -fomit-frame-pointer	ref
gcc-4.8 -march=k8 -O2 -fomit-frame-pointer	ref
gcc-4.8 -march=k8 -O3 -fomit-frame-pointer	ref
gcc-4.8 -march=k8 -O -fomit-frame-pointer	ref
gcc-4.8 -march=nocona -O2 -fomit-frame-pointer	ref
gcc-4.8 -march=nocona -O3 -fomit-frame-pointer	ref
gcc-4.8 -march=nocona -O -fomit-frame-pointer	ref

Test failure

Implementation: crypto_sign/mqqsig160/ref
Compiler: clang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer

error 111

Number of similar (compiler,implementation) pairs: 8, namely:

Compiler	Implementations
clang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer	ref
clang -O3 -fwrapv -mavx2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core-avx2 -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core-avx2 -O -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=core-avx2 -Os -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core-avx2 -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core-avx2 -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=core-avx2 -Os -fomit-frame-pointer	ref

Test failure

Implementation: crypto_sign/mqqsig160/ref
Compiler: gcc-4.7 -m64 -march=barcelona -O2 -fomit-frame-pointer

error 111
crypto_sign_open does not handle c=t overlap
error 111
crypto_sign_open does not handle c=t overlap

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

Compiler	Implementations
gcc-4.7 -m64 -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.7 -m64 -march=barcelona -O -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=barcelona -O2 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=barcelona -O3 -fomit-frame-pointer	ref
gcc-4.8 -m64 -march=barcelona -O -fomit-frame-pointer	ref

Compiler output

Implementation: crypto_sign/mqqsig160/ref
Compiler: clang -O3 -fwrapv -march=native -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer

keygen.c: clang-3.3: warning: argument unused during compilation: '-fpolly'
mqq-sign.c: clang-3.3: warning: argument unused during compilation: '-fpolly'
mqq-verify.c: clang-3.3: warning: argument unused during compilation: '-fpolly'

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

Compiler	Implementations
clang -O3 -fwrapv -march=native -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer	ref
clang -O3 -fwrapv -mavx -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer	ref
clang -O3 -fwrapv -mavx -maes -mpclmul -fpolly -funroll-loops -fvectorize -fslp-vectorize -fslp-vectorize-aggressive -fomit-frame-pointer	ref

Compiler output

Implementation: crypto_sign/mqqsig160/ref
Compiler: clang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer

keygen.c: clang-3.3: warning: argument unused during compilation: '-mcpu=core-avx2'
mqq-sign.c: clang-3.3: warning: argument unused during compilation: '-mcpu=core-avx2'
mqq-verify.c: clang-3.3: warning: argument unused during compilation: '-mcpu=core-avx2'

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

Compiler	Implementations
clang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer	ref