Implementation notes: aarch64, minimac, crypto_kem/saber2

Computer: minimac
Architecture: aarch64
CPU ID: unknown CPU ID
SUPERCOP version: 20210529
Operation: crypto_kem
Primitive: saber2
TimeObject sizeTest sizeImplementationCompilerBenchmark dateSUPERCOP version
123995? ? ?? ? ?T:neon2gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
146927? ? ?? ? ?T:neon2gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
151061? ? ?? ? ?T:neongcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
151194? ? ?? ? ?T:neon2gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
168127? ? ?? ? ?T:neon2gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
168526? ? ?? ? ?T:neongcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
171594? ? ?? ? ?T:neongcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
189859? ? ?? ? ?T:neongcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE2021060220210529
192259? ? ?? ? ?T:refclang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE2021060220210529

Compiler output

Implementation: T:avx2
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
SABER_indcpa.c: In file included from SABER_indcpa.c:5:
SABER_indcpa.c: In file included from ./SABER_indcpa.h:4:
SABER_indcpa.c: In file included from /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/immintrin.h:15:
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:50:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: return (__m64)__builtin_ia32_vec_init_v2si(__i, 0);
SABER_indcpa.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:129:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: return (__m64)__builtin_ia32_packsswb((__v4hi)__m1, (__v4hi)__m2);
SABER_indcpa.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:159:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: return (__m64)__builtin_ia32_packssdw((__v2si)__m1, (__v2si)__m2);
SABER_indcpa.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:189:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: return (__m64)__builtin_ia32_packuswb((__v4hi)__m1, (__v4hi)__m2);
SABER_indcpa.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:216:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: return (__m64)__builtin_ia32_punpckhbw((__v8qi)__m1, (__v8qi)__m2);
SABER_indcpa.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:239:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: return (__m64)__builtin_ia32_punpckhwd((__v4hi)__m1, (__v4hi)__m2);
SABER_indcpa.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:260:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: return (__m64)__builtin_ia32_punpckhdq((__v2si)__m1, (__v2si)__m2);
SABER_indcpa.c: ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
SABER_indcpa.c: /Library/Developer/CommandLineTools/usr/lib/clang/12.0.5/include/mmintrin.h:287:12: error: invalid conversion between vector type '__m64' (vector of 1 'long long' value) and integer type 'int' of different size
SABER_indcpa.c: ...

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:avx2

Compiler output

Implementation: T:avx2
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE
SABER_indcpa.c: In file included from SABER_indcpa.c:5:
SABER_indcpa.c: SABER_indcpa.h:4:10: fatal error: immintrin.h: No such file or directory
SABER_indcpa.c: 4 | #include <immintrin.h>
SABER_indcpa.c: | ^~~~~~~~~~~~~
SABER_indcpa.c: compilation terminated.

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:avx2
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:avx2
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:avx2
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:avx2

Compiler output

Implementation: T:neon
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
poly.c: poly.c:19:3: error: implicit declaration of function 'cshake128_simple' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
poly.c: cshake128_simple(buf,SABER_N,nonce,seed,SABER_NOISESEEDBYTES);
poly.c: ^
poly.c: poly.c:34:3: error: implicit declaration of function 'cshake128_simple' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
poly.c: cshake128_simple(buf0,SABER_N,nonce0,seed,SABER_NOISESEEDBYTES);
poly.c: ^
poly.c: 2 errors generated.

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:neon

Compiler output

Implementation: T:neon
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE
poly.c: poly.c: In function 'poly_getnoise':
poly.c: poly.c:19:3: warning: implicit declaration of function 'cshake128_simple' [-Wimplicit-function-declaration]
poly.c: 19 | cshake128_simple(buf,SABER_N,nonce,seed,SABER_NOISESEEDBYTES);
poly.c: | ^~~~~~~~~~~~~~~~

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon

Compiler output

Implementation: T:neon2
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
SABER_indcpa.c: In file included from SABER_indcpa.c:22:
SABER_indcpa.c: ./rq_mul/neon_poly_rq_mul.c:272:15: warning: initializing 'uint16_t *' (aka 'unsigned short *') with an expression of type 'const uint16_t *' (aka 'const unsigned short *') discards qualifiers [-Wincompatible-pointer-types-discards-qualifiers]
SABER_indcpa.c: uint16_t *c0 = poly,
SABER_indcpa.c: ^ ~~~~
SABER_indcpa.c: ./rq_mul/neon_poly_rq_mul.c:273:15: warning: initializing 'uint16_t *' (aka 'unsigned short *') with an expression of type 'const uint16_t *' (aka 'const unsigned short *') discards qualifiers [-Wincompatible-pointer-types-discards-qualifiers]
SABER_indcpa.c: *c1 = &poly[1 * SB1],
SABER_indcpa.c: ^ ~~~~~~~~~~~~~~
SABER_indcpa.c: ./rq_mul/neon_poly_rq_mul.c:274:15: warning: initializing 'uint16_t *' (aka 'unsigned short *') with an expression of type 'const uint16_t *' (aka 'const unsigned short *') discards qualifiers [-Wincompatible-pointer-types-discards-qualifiers]
SABER_indcpa.c: *c2 = &poly[2 * SB1],
SABER_indcpa.c: ^ ~~~~~~~~~~~~~~
SABER_indcpa.c: ./rq_mul/neon_poly_rq_mul.c:275:15: warning: initializing 'uint16_t *' (aka 'unsigned short *') with an expression of type 'const uint16_t *' (aka 'const unsigned short *') discards qualifiers [-Wincompatible-pointer-types-discards-qualifiers]
SABER_indcpa.c: *c3 = &poly[3 * SB1],
SABER_indcpa.c: ^ ~~~~~~~~~~~~~~
SABER_indcpa.c: 4 warnings generated.
poly.c: poly.c:19:3: error: implicit declaration of function 'cshake128_simple' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
poly.c: cshake128_simple(buf,SABER_N,nonce,seed,SABER_NOISESEEDBYTES);
poly.c: ^
poly.c: poly.c:34:3: error: implicit declaration of function 'cshake128_simple' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
poly.c: cshake128_simple(buf0,SABER_N,nonce0,seed,SABER_NOISESEEDBYTES);
poly.c: ^
poly.c: 2 errors generated.

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:neon2

Compiler output

Implementation: T:neon2
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE
SABER_indcpa.c: In file included from SABER_indcpa.c:22:
SABER_indcpa.c: rq_mul/neon_poly_rq_mul.c: In function 'tc4_evaluate_neon_SB1':
SABER_indcpa.c: rq_mul/neon_poly_rq_mul.c:272:20: warning: initialization discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
SABER_indcpa.c: 272 | uint16_t *c0 = poly,
SABER_indcpa.c: | ^~~~
SABER_indcpa.c: rq_mul/neon_poly_rq_mul.c:273:20: warning: initialization discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
SABER_indcpa.c: 273 | *c1 = &poly[1 * SB1],
SABER_indcpa.c: | ^
SABER_indcpa.c: rq_mul/neon_poly_rq_mul.c:274:20: warning: initialization discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
SABER_indcpa.c: 274 | *c2 = &poly[2 * SB1],
SABER_indcpa.c: | ^
SABER_indcpa.c: rq_mul/neon_poly_rq_mul.c:275:20: warning: initialization discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
SABER_indcpa.c: 275 | *c3 = &poly[3 * SB1],
SABER_indcpa.c: | ^
poly.c: poly.c: In function 'poly_getnoise':
poly.c: poly.c:19:3: warning: implicit declaration of function 'cshake128_simple' [-Wimplicit-function-declaration]
poly.c: 19 | cshake128_simple(buf,SABER_N,nonce,seed,SABER_NOISESEEDBYTES);
poly.c: | ^~~~~~~~~~~~~~~~

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon2
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon2
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon2
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:neon2

Compiler output

Implementation: T:ref
Security model: timingleaks
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE
SABER_indcpa.c: In file included from SABER_indcpa.c:6:
SABER_indcpa.c: ./poly_mul.inc:20:43: warning: format specifies type 'unsigned long' but the argument has type 'uint64_t' (aka 'unsigned long long') [-Wformat]
SABER_indcpa.c: printf(" Mod(%d,%lu)*x^%d + ", a[i], p,i);
SABER_indcpa.c: ~~~ ^
SABER_indcpa.c: %llu
SABER_indcpa.c: ./poly_mul.inc:22:41: warning: format specifies type 'unsigned long' but the argument has type 'uint64_t' (aka 'unsigned long long') [-Wformat]
SABER_indcpa.c: printf(" Mod(%d,%lu)*x^%d ", a[i], p,i);
SABER_indcpa.c: ~~~ ^
SABER_indcpa.c: %llu
SABER_indcpa.c: 2 warnings generated.
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/knownrandombytes.o) was built for newer macOS version (11.3) than being linked (11.0)
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/kernelrandombytes.o) was built for newer macOS version (11.3) than being linked (11.0)
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/cpucycles.o) was built for newer macOS version (11.3) than being linked (11.0)
try.c: ld: warning: could not create compact unwind for _init: registers 21 and 22 not saved contiguously in frame
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/libsupercop.a(crypto_rng_chacha20_ref_constbranchindex-rng.o)) was built for newer macOS version (11.3) than being linked (11.0)
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/knownrandombytes.o) was built for newer macOS version (11.3) than being linked (11.0)
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/kernelrandombytes.o) was built for newer macOS version (11.3) than being linked (11.0)
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/cpucycles.o) was built for newer macOS version (11.3) than being linked (11.0)
try.c: ld: warning: could not create compact unwind for _init: registers 21 and 22 not saved contiguously in frame
try.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/libsupercop.a(crypto_rng_chacha20_ref_constbranchindex-rng.o)) was built for newer macOS version (11.3) than being linked (11.0)
measure.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/fastrandombytes.o) was built for newer macOS version (11.3) than being linked (11.0)
measure.c: ld: warning: could not create compact unwind for _randombytes: registers 21 and 22 not saved contiguously in frame
measure.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/kernelrandombytes.o) was built for newer macOS version (11.3) than being linked (11.0)
measure.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/cpucycles.o) was built for newer macOS version (11.3) than being linked (11.0)
measure.c: ld: warning: could not create compact unwind for _init: registers 21 and 22 not saved contiguously in frame
measure.c: ld: warning: object file (/Users/djb/supercop-20210529/supercop-data/minimac/aarch64/lib/libsupercop.a(crypto_rng_chacha20_ref_constbranchindex-rng.o)) was built for newer macOS version (11.3) than being linked (11.0)

Number of similar (compiler,implementation) pairs: 1, namely:
CompilerImplementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE T:ref

Compiler output

Implementation: T:ref
Security model: timingleaks
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE
try.c: duplicate symbol '_clock_kp_mv' in:
try.c: libcrypto_kem_saber2.a(kem.o)
try.c: libcrypto_kem_saber2.a(SABER_indcpa.o)
try.c: duplicate symbol '_clock_cl_mv' in:
try.c: libcrypto_kem_saber2.a(kem.o)
try.c: libcrypto_kem_saber2.a(SABER_indcpa.o)
try.c: duplicate symbol '_clock_kp_sm' in:
try.c: libcrypto_kem_saber2.a(kem.o)
try.c: libcrypto_kem_saber2.a(SABER_indcpa.o)
try.c: duplicate symbol '_clock_cl_sm' in:
try.c: libcrypto_kem_saber2.a(kem.o)
try.c: libcrypto_kem_saber2.a(SABER_indcpa.o)
try.c: duplicate symbol '_clock2' in:
try.c: libcrypto_kem_saber2.a(kem.o)
try.c: libcrypto_kem_saber2.a(SABER_indcpa.o)
try.c: duplicate symbol '_clock1' in:
try.c: libcrypto_kem_saber2.a(kem.o)
try.c: libcrypto_kem_saber2.a(SABER_indcpa.o)
try.c: ld: 6 duplicate symbols for architecture arm64
try.c: collect2: error: ld returned 1 exit status

Number of similar (compiler,implementation) pairs: 4, namely:
CompilerImplementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE T:ref