Implementation notes: aarch64, a72, crypto_hash/bash512

Computer: a72
Architecture: aarch64
CPU ID: unknown CPU ID
SUPERCOP version: 20170904
Operation: crypto_hash
Primitive: bash512

Checksum failure

Implementation: crypto_hash/bash512/ref
Compiler: gcc -funroll-loops -march=native -mtune=native -O2

a397472ba7599635360cf4397e77b42145d47322e3ee531180b21ebf8e99930d

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

Compiler	Implementations
gcc -funroll-loops -march=native -mtune=native -O2	ref
gcc -funroll-loops -march=native -mtune=native -O3	ref
gcc -funroll-loops -march=native -mtune=native -Os	ref
gcc -march=native -mtune=native -O2	ref
gcc -march=native -mtune=native -O3	ref
gcc -march=native -mtune=native -Os	ref

Compiler output

Implementation: crypto_hash/bash512/avx2
Compiler: gcc -funroll-loops -march=native -mtune=native -O2

bash_avx2.c: bash_avx2.c:70:2: error: #error "The compiler does not support AVX2 intrinsics."
bash_avx2.c: #error "The compiler does not support AVX2 intrinsics."
bash_avx2.c: ^~~~~
bash_avx2.c: bash_avx2.c: In function 'bashavx2_F0':
bash_avx2.c: bash_avx2.c:271:5: error: unknown type name 'u256'; did you mean 'm25'?
bash_avx2.c: u256 S1,S2,T0,T1,T2,U0,U1,U2;
bash_avx2.c: ^~~~
bash_avx2.c: m25
bash_avx2.c: bash_avx2.c:272:5: error: unknown type name 'u256'; did you mean 'm25'?
bash_avx2.c: u256 W0 = LOADW( S + 0 );
bash_avx2.c: ...
bash_avx2.c: ^
bash_avx2.c: bash_avx2.c:265:9: note: in expansion of macro 'bashR1'
bash_avx2.c: bashR1(W0,W1,W2,W3,W4,W5,C23 ,S1,S2,T0,T1,T2,U0,U1,U2);
bash_avx2.c: ^~~~~~
bash_avx2.c: bash_avx2.c:279:5: note: in expansion of macro 'bashFP'
bash_avx2.c: bashFP(W0,W1,W2,W3,W4,W5 ,S1,S2,T0,T1,T2,U0,U1,U2);
bash_avx2.c: ^~~~~~
bash_avx2.c: bash_avx2.c:281:5: warning: implicit declaration of function 'STOREW' [-Wimplicit-function-declaration]
bash_avx2.c: STOREW( S + 0, W0 );
bash_avx2.c: ^~~~~~

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

Compiler	Implementations
gcc -funroll-loops -march=native -mtune=native -O2	avx2
gcc -funroll-loops -march=native -mtune=native -O3	avx2
gcc -funroll-loops -march=native -mtune=native -Os	avx2
gcc -march=native -mtune=native -O2	avx2
gcc -march=native -mtune=native -O3	avx2
gcc -march=native -mtune=native -Os	avx2