Implementation notes: aarch64, jetsontx1, crypto_hash/blake2b

Computer: jetsontx1
Architecture: aarch64
CPU ID: unknown CPU ID
SUPERCOP version: 20180818
Operation: crypto_hash
Primitive: blake2b
TimeImplementationCompilerBenchmark dateSUPERCOP version
7597regsclang-6.0 -mcpu=cortex-a57 -Os -fomit-frame-pointer -fwrapv2018101320180818
8420regsgcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv2018101320180818
8420regsgcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv2018101320180818
8878regsclang-6.0 -mcpu=cortex-a57 -O3 -fomit-frame-pointer -fwrapv2018101320180818
8889regsclang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv2018101320180818
9703refclang-6.0 -mcpu=cortex-a57 -Os -fomit-frame-pointer -fwrapv2018101320180818
9760regsgcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv2018101320180818
9766refgcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv2018101320180818
10692refgcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv2018101320180818
11052refclang-6.0 -mcpu=cortex-a57 -O3 -fomit-frame-pointer -fwrapv2018101320180818
11120refclang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv2018101320180818
11620refgcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv2018101320180818

Compiler output

Implementation: crypto_hash/blake2b/avx2-1
Compiler: clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv
blake2b.c: In file included from blake2b.c:11:
blake2b.c: In file included from /usr/lib/llvm-6.0/lib/clang/6.0.0/include/immintrin.h:28:
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:64:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_vec_init_v2si(__i, 0);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:143:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packsswb((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:173:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packssdw((__v2si)__m1, (__v2si)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:203:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packuswb((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:230:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhbw((__v8qi)__m1, (__v8qi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:253:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhwd((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:274:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhdq((__v2si)__m1, (__v2si)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:301:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpcklbw((__v8qi)__m1, (__v8qi)__m2);
blake2b.c: ...

Number of similar (compiler,implementation) pairs: 3, namely:
CompilerImplementations
clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv avx2-1
clang-6.0 -mcpu=cortex-a57 -O3 -fomit-frame-pointer -fwrapv avx2-1
clang-6.0 -mcpu=cortex-a57 -Os -fomit-frame-pointer -fwrapv avx2-1

Compiler output

Implementation: crypto_hash/blake2b/avx2-2
Compiler: clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv
blake2b.c: In file included from blake2b.c:12:
blake2b.c: In file included from /usr/lib/llvm-6.0/lib/clang/6.0.0/include/immintrin.h:28:
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:64:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_vec_init_v2si(__i, 0);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:143:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packsswb((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:173:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packssdw((__v2si)__m1, (__v2si)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:203:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packuswb((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:230:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhbw((__v8qi)__m1, (__v8qi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:253:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhwd((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:274:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhdq((__v2si)__m1, (__v2si)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:301:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpcklbw((__v8qi)__m1, (__v8qi)__m2);
blake2b.c: ...

Number of similar (compiler,implementation) pairs: 6, namely:
CompilerImplementations
clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv avx2-2 avx2-3
clang-6.0 -mcpu=cortex-a57 -O3 -fomit-frame-pointer -fwrapv avx2-2 avx2-3
clang-6.0 -mcpu=cortex-a57 -Os -fomit-frame-pointer -fwrapv avx2-2 avx2-3

Compiler output

Implementation: crypto_hash/blake2b/xmm
Compiler: clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv
blake2b.c: In file included from blake2b.c:6:
blake2b.c: ./blake2-config.h:68:2: error: "This code requires at least SSE2."
blake2b.c: #error "This code requires at least SSE2."
blake2b.c: ^
blake2b.c: In file included from blake2b.c:11:
blake2b.c: In file included from /usr/lib/llvm-6.0/lib/clang/6.0.0/include/immintrin.h:28:
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:64:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_vec_init_v2si(__i, 0);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:143:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packsswb((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:173:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packssdw((__v2si)__m1, (__v2si)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:203:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_packuswb((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:230:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhbw((__v8qi)__m1, (__v8qi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:253:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: return (__m64)__builtin_ia32_punpckhwd((__v4hi)__m1, (__v4hi)__m2);
blake2b.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blake2b.c: /usr/lib/llvm-6.0/lib/clang/6.0.0/include/mmintrin.h:274:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
blake2b.c: ...

Number of similar (compiler,implementation) pairs: 6, namely:
CompilerImplementations
clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv xmm ymm
clang-6.0 -mcpu=cortex-a57 -O3 -fomit-frame-pointer -fwrapv xmm ymm
clang-6.0 -mcpu=cortex-a57 -Os -fomit-frame-pointer -fwrapv xmm ymm

Compiler output

Implementation: crypto_hash/blake2b/avxicc
Compiler: clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv
blake2b.s: clang: warning: argument unused during compilation: '-fomit-frame-pointer' [-Wunused-command-line-argument]
blake2b.s: clang: warning: argument unused during compilation: '-fwrapv' [-Wunused-command-line-argument]
blake2b.s: clang: warning: argument unused during compilation: '-D SUPERCOP' [-Wunused-command-line-argument]
blake2b.s: blake2b.s:3:2: error: unknown directive
blake2b.s: .intel_syntax noprefix
blake2b.s: ^
blake2b.s: blake2b.s:14:19: error: invalid operand for instruction
blake2b.s: sub rsp, 552
blake2b.s: ^
blake2b.s: blake2b.s:16:9: error: unrecognized instruction mnemonic, did you mean: eor, orn, orr, ror?
blake2b.s: xor r11d, r11d
blake2b.s: ^
blake2b.s: blake2b.s:17:9: error: unrecognized instruction mnemonic, did you mean: eor, orn, orr, ror?
blake2b.s: xor ecx, ecx
blake2b.s: ^
blake2b.s: blake2b.s:18:19: error: invalid operand for instruction
blake2b.s: mov r9, rsi
blake2b.s: ^
blake2b.s: blake2b.s:19:33: error: unexpected token in argument list
blake2b.s: vmovdqu xmm2, XMMWORD PTR .L_2il0floatpacket.13[rip]
blake2b.s: ^
blake2b.s: blake2b.s:20:9: error: unrecognized instruction mnemonic, did you mean: eor, orn, orr, ror?
blake2b.s: xor eax, eax
blake2b.s: ^
blake2b.s: blake2b.s:21:33: error: unexpected token in argument list
blake2b.s: ...

Number of similar (compiler,implementation) pairs: 3, namely:
CompilerImplementations
clang-6.0 -mcpu=cortex-a57 -O2 -fomit-frame-pointer -fwrapv avxicc
clang-6.0 -mcpu=cortex-a57 -O3 -fomit-frame-pointer -fwrapv avxicc
clang-6.0 -mcpu=cortex-a57 -Os -fomit-frame-pointer -fwrapv avxicc

Compiler output

Implementation: crypto_hash/blake2b/xmm
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv
blake2b.c: In file included from blake2b.c:6:
blake2b.c: blake2-config.h:68:2: error: #error "This code requires at least SSE2."
blake2b.c: #error "This code requires at least SSE2."
blake2b.c: ^~~~~
blake2b.c: blake2b.c:11:10: fatal error: immintrin.h: No such file or directory
blake2b.c: #include gt;
blake2b.c: ^~~~~~~~~~~~~
blake2b.c: compilation terminated.

Number of similar (compiler,implementation) pairs: 6, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv xmm ymm
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv xmm ymm
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv xmm ymm

Compiler output

Implementation: crypto_hash/blake2b/avx2-1
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv
blake2b.c: blake2b.c:11:10: fatal error: immintrin.h: No such file or directory
blake2b.c: #include gt;
blake2b.c: ^~~~~~~~~~~~~
blake2b.c: compilation terminated.

Number of similar (compiler,implementation) pairs: 3, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv avx2-1
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv avx2-1
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv avx2-1

Compiler output

Implementation: crypto_hash/blake2b/avx2-2
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv
blake2b.c: blake2b.c:12:10: fatal error: immintrin.h: No such file or directory
blake2b.c: #include gt;
blake2b.c: ^~~~~~~~~~~~~
blake2b.c: compilation terminated.

Number of similar (compiler,implementation) pairs: 6, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv avx2-2 avx2-3
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv avx2-2 avx2-3
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv avx2-2 avx2-3

Compiler output

Implementation: crypto_hash/blake2b/avxicc
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv
blake2b.s: blake2b.s: Assembler messages:
blake2b.s: blake2b.s:3: Error: unknown pseudo-op: `.intel_syntax'
blake2b.s: blake2b.s:14: Error: operand 1 must be an integer or stack pointer register -- `sub rsp,552'
blake2b.s: blake2b.s:16: Error: unknown mnemonic `xor' -- `xor r11d,r11d'
blake2b.s: blake2b.s:17: Error: unknown mnemonic `xor' -- `xor ecx,ecx'
blake2b.s: blake2b.s:18: Error: operand 1 must be an integer register -- `mov r9,rsi'
blake2b.s: blake2b.s:19: Error: unknown mnemonic `vmovdqu' -- `vmovdqu xmm2,XMMWORD PTR .L_2il0floatpacket.13[rip]'
blake2b.s: blake2b.s:20: Error: unknown mnemonic `xor' -- `xor eax,eax'
blake2b.s: blake2b.s:21: Error: unknown mnemonic `vmovdqu' -- `vmovdqu xmm3,XMMWORD PTR .L_2il0floatpacket.14[rip]'
blake2b.s: blake2b.s:22: Error: operand 1 must be an integer register -- `mov r8,rdx'
blake2b.s: blake2b.s:23: Error: unknown mnemonic `vmovdqu' -- `vmovdqu xmm4,XMMWORD PTR .L_2il0floatpacket.15[rip]'
blake2b.s: blake2b.s:24: Error: unknown mnemonic `xor' -- `xor r10d,r10d'
blake2b.s: blake2b.s:25: Error: unknown mnemonic `vmovdqu' -- `vmovdqu xmm11,XMMWORD PTR .L_2il0floatpacket.16[rip]'
blake2b.s: blake2b.s:26: Error: unknown mnemonic `vmovdqu' -- `vmovdqu xmm1,XMMWORD PTR .L_2il0floatpacket.11[rip]'
blake2b.s: blake2b.s:27: Error: unknown mnemonic `vmovdqu' -- `vmovdqu xmm0,XMMWORD PTR .L_2il0floatpacket.12[rip]'
blake2b.s: blake2b.s:28: Error: unknown mnemonic `vmovdqu' -- `vmovdqu XMMWORD PTR[448+rsp],xmm2'
blake2b.s: blake2b.s:29: Error: unknown mnemonic `vmovdqu' -- `vmovdqu XMMWORD PTR[464+rsp],xmm3'
blake2b.s: blake2b.s:30: Error: unknown mnemonic `vmovdqu' -- `vmovdqu XMMWORD PTR[496+rsp],xmm4'
blake2b.s: blake2b.s:31: Error: unknown mnemonic `vmovdqu' -- `vmovdqu XMMWORD PTR[480+rsp],xmm11'
blake2b.s: blake2b.s:32: Error: unknown mnemonic `vmovdqu' -- `vmovdqu xmm5,XMMWORD PTR .L_2il0floatpacket.17[rip]'
blake2b.s: blake2b.s:33: Error: operand 1 must be an integer or stack pointer register -- `cmp rdx,128'
blake2b.s: blake2b.s:34: Error: unknown mnemonic `jbe' -- `jbe ..B1.5'
blake2b.s: blake2b.s:37: Error: operand 1 must be an integer register -- `mov QWORD PTR[rsp],rbp'
blake2b.s: blake2b.s:38: Error: unknown mnemonic `lea' -- `lea rsi,QWORD PTR[-1+rdx]'
blake2b.s: blake2b.s:39: Error: unknown mnemonic `sar' -- `sar rsi,6'
blake2b.s: ...

Number of similar (compiler,implementation) pairs: 3, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv avxicc
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv avxicc
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv avxicc