Implementation notes: amd64, skylake, crypto_stream/cryptmtv3

Computer: skylake
Architecture: amd64
CPU ID: GenuineIntel-000506e3-bfebfbff
SUPERCOP version: 20161026
Operation: crypto_stream
Primitive: cryptmtv3
TimeImplementationCompilerBenchmark dateSUPERCOP version
6254e/v3gcc -funroll-loops -m64 -march=nocona -O3 -fomit-frame-pointer2016121520161026
6254e/v3gcc -funroll-loops -march=nocona -O3 -fomit-frame-pointer2016121520161026
6288e/v3gcc -funroll-loops -m64 -march=nocona -O2 -fomit-frame-pointer2016121520161026
6292e/v3clang -O3 -fomit-frame-pointer -Qunused-arguments2016121520161026
6292e/v3gcc -funroll-loops -march=nocona -O2 -fomit-frame-pointer2016121520161026
6410e/v3gcc -funroll-loops -march=nocona -O -fomit-frame-pointer2016121520161026
6414e/v3gcc -funroll-loops -m64 -march=nocona -O -fomit-frame-pointer2016121520161026
6446e/v3gcc -funroll-loops -march=barcelona -O2 -fomit-frame-pointer2016121520161026
6448e/v3gcc -funroll-loops -m64 -march=barcelona -O2 -fomit-frame-pointer2016121520161026
6448e/v3gcc -funroll-loops -m64 -march=barcelona -O3 -fomit-frame-pointer2016121520161026
6456e/v3gcc -funroll-loops -march=barcelona -O3 -fomit-frame-pointer2016121520161026
6456e/v3gcc -m64 -march=corei7-avx -O3 -fomit-frame-pointer2016121520161026
6464e/v3gcc -m64 -march=core-avx-i -O3 -fomit-frame-pointer2016121520161026
6470e/v3gcc -m64 -march=core-avx2 -O3 -fomit-frame-pointer2016121520161026
6472e/v3gcc -m64 -march=native -mtune=native -O3 -fomit-frame-pointer2016121520161026
6484e/v3gcc -funroll-loops -m64 -O3 -fomit-frame-pointer2016121520161026
6488e/v3gcc -funroll-loops -O3 -fomit-frame-pointer2016121520161026
6488e/v3gcc -funroll-loops -fno-schedule-insns -O3 -fomit-frame-pointer2016121520161026
6490e/v3gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv2016121520161026
6504e/v3gcc -m64 -march=core-avx-i -O2 -fomit-frame-pointer2016121520161026
6504e/v3gcc -m64 -march=corei7 -O2 -fomit-frame-pointer2016121520161026
6510e/v3gcc -m64 -march=corei7-avx -O2 -fomit-frame-pointer2016121520161026
6512e/v3gcc -funroll-loops -fno-schedule-insns -O -fomit-frame-pointer2016121520161026
6512e/v3gcc -funroll-loops -m64 -march=k8 -O -fomit-frame-pointer2016121520161026
6516e/v3gcc -funroll-loops -march=k8 -O -fomit-frame-pointer2016121520161026
6518e/v3gcc -funroll-loops -m64 -O -fomit-frame-pointer2016121520161026
6520e/v3gcc -m64 -march=native -mtune=native -O2 -fomit-frame-pointer2016121520161026
6522e/v3gcc -m64 -march=core-avx2 -O2 -fomit-frame-pointer2016121520161026
6524e/v3gcc -m64 -march=core-avx-i -Os -fomit-frame-pointer2016121520161026
6524e/v3gcc -m64 -march=corei7-avx -Os -fomit-frame-pointer2016121520161026
6532e/v3gcc -funroll-loops -O -fomit-frame-pointer2016121520161026
6544e/v3gcc -funroll-loops -m64 -march=barcelona -O -fomit-frame-pointer2016121520161026
6544e/v3gcc -funroll-loops -march=barcelona -O -fomit-frame-pointer2016121520161026
6550e/v3gcc -m64 -march=corei7 -O3 -fomit-frame-pointer2016121520161026
6560e/v3gcc -funroll-loops -m64 -march=k8 -O3 -fomit-frame-pointer2016121520161026
6566e/v3gcc -funroll-loops -m64 -march=k8 -O2 -fomit-frame-pointer2016121520161026
6568e/v3gcc -m64 -march=core-avx2 -Os -fomit-frame-pointer2016121520161026
6572e/v3gcc -funroll-loops -m64 -O2 -fomit-frame-pointer2016121520161026
6572e/v3gcc -funroll-loops -march=k8 -O2 -fomit-frame-pointer2016121520161026
6572e/v3gcc -m64 -march=native -mtune=native -Os -fomit-frame-pointer2016121520161026
6574e/v3gcc -funroll-loops -O2 -fomit-frame-pointer2016121520161026
6584e/v3gcc -funroll-loops -march=k8 -O3 -fomit-frame-pointer2016121520161026
6588e/v3gcc -funroll-loops -fno-schedule-insns -O2 -fomit-frame-pointer2016121520161026
6588e/v3gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv2016121520161026
6600e/v3gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv2016121520161026
6642e/v3gcc -m64 -march=core2 -O3 -fomit-frame-pointer2016121520161026
6642e/v3gcc -m64 -march=core2 -msse4 -O3 -fomit-frame-pointer2016121520161026
6644e/v3gcc -m64 -march=core2 -msse4.1 -O3 -fomit-frame-pointer2016121520161026
6676e/v3gcc -m64 -march=corei7-avx -O -fomit-frame-pointer2016121520161026
6678e/v3gcc -m64 -march=core2 -O2 -fomit-frame-pointer2016121520161026
6682e/v3gcc -m64 -march=core2 -msse4 -O2 -fomit-frame-pointer2016121520161026
6684e/v3gcc -m64 -march=core-avx-i -O -fomit-frame-pointer2016121520161026
6690e/v3gcc -m64 -march=core2 -msse4.1 -O2 -fomit-frame-pointer2016121520161026
6700e/v3gcc -m64 -march=native -mtune=native -O -fomit-frame-pointer2016121520161026
6720e/v3gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv2016121520161026
6724e/v3gcc -march=nocona -O3 -fomit-frame-pointer2016121520161026
6732e/v3gcc -m64 -march=core-avx2 -O -fomit-frame-pointer2016121520161026
6734e/v3gcc -m64 -march=corei7 -O -fomit-frame-pointer2016121520161026
6740e/v3gcc -m64 -march=nocona -O3 -fomit-frame-pointer2016121520161026
6748e/v3gcc -fno-schedule-insns -O3 -fomit-frame-pointer2016121520161026
6762e/v3gcc -O3 -fomit-frame-pointer2016121520161026
6786e/v3gcc -m64 -O3 -fomit-frame-pointer2016121520161026
6786e/v3gcc -march=k8 -O3 -fomit-frame-pointer2016121520161026
6788e/v3gcc -m64 -march=nocona -O2 -fomit-frame-pointer2016121520161026
6794e/v3gcc -march=nocona -O2 -fomit-frame-pointer2016121520161026
6810e/v3gcc -march=barcelona -O3 -fomit-frame-pointer2016121520161026
6814e/v3gcc -march=barcelona -O2 -fomit-frame-pointer2016121520161026
6818e/v3gcc -m64 -march=barcelona -O2 -fomit-frame-pointer2016121520161026
6824e/v3gcc -m64 -march=barcelona -O3 -fomit-frame-pointer2016121520161026
6834e/v3gcc -m64 -march=k8 -O3 -fomit-frame-pointer2016121520161026
6850e/v3gcc -m64 -march=nocona -O -fomit-frame-pointer2016121520161026
6858e/v3gcc -march=nocona -O -fomit-frame-pointer2016121520161026
6870e/v3gcc -O2 -fomit-frame-pointer2016121520161026
6870e/v3gcc -m64 -march=k8 -O2 -fomit-frame-pointer2016121520161026
6870e/v3gcc -march=k8 -O2 -fomit-frame-pointer2016121520161026
6882e/v3gcc -fno-schedule-insns -O2 -fomit-frame-pointer2016121520161026
6886e/v3gcc -m64 -O2 -fomit-frame-pointer2016121520161026
6986e/v3gcc -fno-schedule-insns -O -fomit-frame-pointer2016121520161026
6990e/v3gcc -O -fomit-frame-pointer2016121520161026
6992e/v3gcc -m64 -O -fomit-frame-pointer2016121520161026
6996e/v3gcc -m64 -march=core2 -msse4.1 -O -fomit-frame-pointer2016121520161026
6998e/v3gcc -m64 -march=barcelona -O -fomit-frame-pointer2016121520161026
7000e/v3gcc -march=k8 -O -fomit-frame-pointer2016121520161026
7006e/v3gcc -m64 -march=core2 -msse4 -O -fomit-frame-pointer2016121520161026
7010e/v3gcc -m64 -march=k8 -O -fomit-frame-pointer2016121520161026
7010e/v3gcc -march=barcelona -O -fomit-frame-pointer2016121520161026
7016e/v3gcc -m64 -march=core2 -O -fomit-frame-pointer2016121520161026
7042e/v3clang -O3 -fwrapv -mavx -maes -mpclmul -fomit-frame-pointer -Qunused-arguments2016121520161026
7064e/v3clang -O3 -fwrapv -mavx -fomit-frame-pointer -Qunused-arguments2016121520161026
7070e/v3clang -mcpu=cortex-a9 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2016121520161026
7070e/v3clang -mcpu=native -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2016121520161026
7080e/v3clang -mcpu=cortex-a8 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2016121520161026
7138e/v3clang -O3 -fwrapv -mavx2 -fomit-frame-pointer -Qunused-arguments2016121520161026
7140e/v3clang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer -Qunused-arguments2016121520161026
7178e/v3clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments2016121520161026
7182e/v3clang -O3 -fwrapv -march=native -fomit-frame-pointer -Qunused-arguments2016121520161026
7674e/v3gcc -m64 -march=core2 -msse4 -Os -fomit-frame-pointer2016121520161026
7678e/v3gcc -m64 -march=core2 -msse4.1 -Os -fomit-frame-pointer2016121520161026
7682e/v3gcc -m64 -march=core2 -Os -fomit-frame-pointer2016121520161026
7780e/v3gcc -funroll-loops -march=barcelona -Os -fomit-frame-pointer2016121520161026
7786e/v3gcc -funroll-loops -m64 -march=barcelona -Os -fomit-frame-pointer2016121520161026
7786e/v3gcc -m64 -march=nocona -Os -fomit-frame-pointer2016121520161026
7788e/v3gcc -funroll-loops -Os -fomit-frame-pointer2016121520161026
7788e/v3gcc -march=nocona -Os -fomit-frame-pointer2016121520161026
7792e/v3gcc -funroll-loops -m64 -march=k8 -Os -fomit-frame-pointer2016121520161026
7794e/v3gcc -march=barcelona -Os -fomit-frame-pointer2016121520161026
7796e/v3gcc -funroll-loops -march=k8 -Os -fomit-frame-pointer2016121520161026
7798e/v3gcc -funroll-loops -fno-schedule-insns -Os -fomit-frame-pointer2016121520161026
7800e/v3gcc -m64 -march=barcelona -Os -fomit-frame-pointer2016121520161026
7804e/v3gcc -funroll-loops -m64 -Os -fomit-frame-pointer2016121520161026
7814e/v3gcc -m64 -march=k8 -Os -fomit-frame-pointer2016121520161026
7816e/v3gcc -march=k8 -Os -fomit-frame-pointer2016121520161026
7820e/v3gcc -fno-schedule-insns -Os -fomit-frame-pointer2016121520161026
7822e/v3gcc -Os -fomit-frame-pointer2016121520161026
7834e/v3gcc -funroll-loops -m64 -march=nocona -Os -fomit-frame-pointer2016121520161026
7834e/v3gcc -m64 -Os -fomit-frame-pointer2016121520161026
7836e/v3gcc -funroll-loops -march=nocona -Os -fomit-frame-pointer2016121520161026
8114e/v3gcc -m64 -march=corei7 -Os -fomit-frame-pointer2016121520161026
40760e/v3cc2016121520161026
40860e/v3gcc -funroll-loops2016121520161026
40866e/v3gcc2016121520161026

Compiler output

Implementation: crypto_stream/cryptmtv3/e/v3
Compiler: clang -O3 -fomit-frame-pointer -Qunused-arguments
cryptmt-v3.c: In file included from cryptmt-v3.c:1:
cryptmt-v3.c: ./e/cryptmt-v3.c:189:18: warning: array index 156 is past the end of the array (which contains 156 elements) [-Warray-bounds]
cryptmt-v3.c: do_recursion(ps->gt;sfmt[N], ps->gt;sfmt[POS1], ps->gt;sfmt[N - 1]);
cryptmt-v3.c: ^ ~
cryptmt-v3.c: ./e/cryptmt-v3.c:25:5: note: array 'sfmt' declared here
cryptmt-v3.c: u32 sfmt[N][4];
cryptmt-v3.c: ^
cryptmt-v3.c: ./e/cryptmt-v3.c:403:27: warning: 'memset' call operates on objects of type 'ECRYPT_ctx' while the size is based on a different type 'ECRYPT_ctx *' [-Wsizeof-pointer-memaccess]
cryptmt-v3.c: memset(ctx, 0, sizeof(ctx));
cryptmt-v3.c: ~~~ ^~~
cryptmt-v3.c: ./e/cryptmt-v3.c:403:27: note: did you mean to dereference the argument to 'sizeof' (and multiply it by the number of elements)?
cryptmt-v3.c: memset(ctx, 0, sizeof(ctx));
cryptmt-v3.c: ^~~
cryptmt-v3.c: 2 warnings generated.

Number of similar (compiler,implementation) pairs: 10, namely:
CompilerImplementations
clang -O3 -fomit-frame-pointer -Qunused-arguments e/v3
clang -O3 -fwrapv -march=native -fomit-frame-pointer -Qunused-arguments e/v3
clang -O3 -fwrapv -march=x86-64 -mcpu=core-avx2 -mavx2 -maes -mpclmul -fomit-frame-pointer -Qunused-arguments e/v3
clang -O3 -fwrapv -mavx2 -fomit-frame-pointer -Qunused-arguments e/v3
clang -O3 -fwrapv -mavx -fomit-frame-pointer -Qunused-arguments e/v3
clang -O3 -fwrapv -mavx -maes -mpclmul -fomit-frame-pointer -Qunused-arguments e/v3
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments e/v3
clang -mcpu=cortex-a8 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments e/v3
clang -mcpu=cortex-a9 -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments e/v3
clang -mcpu=native -mfpu=neon -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments e/v3