Implementation notes: amd64, jasper3, crypto_stream/speck64128ctr

Computer: jasper3
Microarchitecture: amd64; Tremont (906c0)
Architecture: amd64
CPU ID: GenuineIntel-000906c0-20-bfebfbff
SUPERCOP version: 20240625
Operation: crypto_stream
Primitive: speck64128ctr
TimeObject sizeTest sizeImplementationCompilerBenchmark dateSUPERCOP version
496729697 0 043752 812 952T:sse4clang_-march=native_-O2_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625
497229697 0 046040 812 952T:sse4clang_-march=native_-O3_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625
497528941 0 041118 804 952T:sse4clang_-march=native_-Os_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625
502329012 0 040763 756 984T:sse4gcc_-march=native_-mtune=native_-Os_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625
506729044 0 041710 804 952T:sse4clang_-march=native_-O_-fwrapv_-Qunused-arguments_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625
515532928 0 047600 780 1016T:sse4gcc_-march=native_-mtune=native_-O3_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625
515731536 0 044976 780 1016T:sse4gcc_-march=native_-mtune=native_-O2_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625
522431504 0 044463 772 1016T:sse4gcc_-march=native_-mtune=native_-O_-fwrapv_-fPIC_-fPIE_-gdwarf-4_-Wall2024062620240625

Compiler output


stream.c: stream.c:332:3: error: always_inline function '_mm256_set_epi32' requires target feature 'avx', but would be inlined into function 'ExpandKey' that is compiled without support for 'avx'
stream.c:   EK(A,B,C,D,rk,key);
stream.c:   ^
stream.c: ./Speck64128AVX2.h:53:28: note: expanded from macro 'EK'
stream.c: #define EK(A,B,C,D,k,key) (RK(B,A,k,key,0),  RK(C,A,k,key,1),  RK(D,A,k,key,2),  RK(B,A,k,key,3),  RK(C,A,k,key,4),  RK(D,A,k,key,5),  RK(B,A,k,key,6), \
stream.c:                            ^
stream.c: ./Speck64128AVX2.h:51:28: note: expanded from macro 'RK'
stream.c: #define RK(X,Y,k,key,i)   (SET1(k[i],Y), key[i]=Y, X=RCS(X,8), X+=Y, X^=i, Y=LCS(Y,3), Y^=X)
stream.c:                            ^
stream.c: ./Intrinsics_AVX2_64block.h:25:22: note: expanded from macro 'SET1'
stream.c: #define SET1(X,c) (X=SET(c,c,c,c,c,c,c,c))
stream.c:                      ^
stream.c: ./Intrinsics_AVX2_64block.h:24:13: note: expanded from macro 'SET'
stream.c: #define SET _mm256_set_epi32
stream.c:             ^
stream.c: stream.c:332:3: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
stream.c: ./Speck64128AVX2.h:53:28: note: expanded from macro 'EK'
stream.c: #define EK(A,B,C,D,k,key) (RK(B,A,k,key,0),  RK(C,A,k,key,1),  RK(D,A,k,key,2),  RK(B,A,k,key,3),  RK(C,A,k,key,4),  RK(D,A,k,key,5),  RK(B,A,k,key,6), \
stream.c:                            ^
stream.c: ./Speck64128AVX2.h:51:28: note: expanded from macro 'RK'
stream.c: #define RK(X,Y,k,key,i)   (SET1(k[i],Y), key[i]=Y, X=RCS(X,8), X+=Y, X^=i, Y=LCS(Y,3), Y^=X)
stream.c:                            ^
stream.c: ./Intrinsics_AVX2_64block.h:25:22: note: expanded from macro 'SET1'
stream.c: #define SET1(X,c) (X=SET(c,c,c,c,c,c,c,c))
stream.c:                      ^
stream.c: ...

Number of similar (implementation,compiler) pairs: 5, namely:
ImplementationCompiler
T:avx2clang -march=native -O2 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx2clang -march=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx2clang -march=native -O -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx2clang -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx2clang -mcpu=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)

Compiler output


stream.c: In file included from Speck64128AVX2.h:17,
stream.c:                  from stream.c:20:
stream.c: stream.c: In function 'Encrypt':
stream.c: Intrinsics_AVX2_64block.h:25:21: warning: AVX vector return without AVX enabled changes the ABI [-Wpsabi]
stream.c:    25 | #define SET1(X,c) (X=SET(c,c,c,c,c,c,c,c))
stream.c:       |                   ~~^~~~~~~~~~~~~~~~~~~~~~
stream.c: stream.c:150:3: note: in expansion of macro 'SET1'
stream.c:   150 |   SET1(X[0],nonce[1]); SET8(Y[0],nonce[0]);
stream.c:       |   ^~~~
stream.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:53,
stream.c:                  from Intrinsics_AVX2_64block.h:5,
stream.c:                  from Speck64128AVX2.h:17,
stream.c:                  from stream.c:20:
stream.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx2intrin.h:119:1: error: inlining failed in call to 'always_inline' '_mm256_add_epi32': target specific option mismatch
stream.c:   119 | _mm256_add_epi32 (__m256i __A, __m256i __B)
stream.c:       | ^~~~~~~~~~~~~~~~
stream.c: In file included from Speck64128AVX2.h:17,
stream.c:                  from stream.c:20:
stream.c: Intrinsics_AVX2_64block.h:17:13: note: called from here
stream.c:    17 | #define ADD _mm256_add_epi32
stream.c:       |             ^
stream.c: Intrinsics_AVX2_64block.h:26:46: note: in expansion of macro 'ADD'
stream.c:    26 | #define SET8(X,c) (X=SET(c,c,c,c,c,c,c,c), X=ADD(X,_q))
stream.c:       |                                              ^~~
stream.c: stream.c:150:24: note: in expansion of macro 'SET8'
stream.c: ...

Number of similar (implementation,compiler) pairs: 4, namely:
ImplementationCompiler
T:avx2gcc -march=native -mtune=native -O2 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)
T:avx2gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)
T:avx2gcc -march=native -mtune=native -O -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)
T:avx2gcc -march=native -mtune=native -Os -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)

Compiler output


stream.c: stream.c:348:3: error: always_inline function '_mm512_set_epi32' requires target feature 'avx512f', but would be inlined into function 'ExpandKey' that is compiled without support for 'avx512f'
stream.c:   EK(A,B,C,D,rk,key);
stream.c:   ^
stream.c: ./Speck64128AVX512.h:53:28: note: expanded from macro 'EK'
stream.c: #define EK(A,B,C,D,k,key) (RK(B,A,k,key,0),  RK(C,A,k,key,1),  RK(D,A,k,key,2),  RK(B,A,k,key,3),  RK(C,A,k,key,4),  RK(D,A,k,key,5),  RK(B,A,k,key,6), \
stream.c:                            ^
stream.c: ./Speck64128AVX512.h:51:26: note: expanded from macro 'RK'
stream.c: #define RK(X,Y,k,key,i) (SET1(k[i],Y), key[i]=Y, X=RCS(X,8), X+=Y, X^=i, Y=LCS(Y,3), Y^=X)
stream.c:                          ^
stream.c: ./Intrinsics_AVX512_64block.h:25:22: note: expanded from macro 'SET1'
stream.c: #define SET1(X,c) (X=SET(c,c,c,c,c,c,c,c,c,c,c,c,c,c,c,c))
stream.c:                      ^
stream.c: ./Intrinsics_AVX512_64block.h:14:13: note: expanded from macro 'SET'
stream.c: #define SET _mm512_set_epi32
stream.c:             ^
stream.c: stream.c:348:3: error: AVX vector return of type '__m512i' (vector of 8 'long long' values) without 'avx512f' enabled changes the ABI
stream.c: ./Speck64128AVX512.h:53:28: note: expanded from macro 'EK'
stream.c: #define EK(A,B,C,D,k,key) (RK(B,A,k,key,0),  RK(C,A,k,key,1),  RK(D,A,k,key,2),  RK(B,A,k,key,3),  RK(C,A,k,key,4),  RK(D,A,k,key,5),  RK(B,A,k,key,6), \
stream.c:                            ^
stream.c: ./Speck64128AVX512.h:51:26: note: expanded from macro 'RK'
stream.c: #define RK(X,Y,k,key,i) (SET1(k[i],Y), key[i]=Y, X=RCS(X,8), X+=Y, X^=i, Y=LCS(Y,3), Y^=X)
stream.c:                          ^
stream.c: ./Intrinsics_AVX512_64block.h:25:22: note: expanded from macro 'SET1'
stream.c: #define SET1(X,c) (X=SET(c,c,c,c,c,c,c,c,c,c,c,c,c,c,c,c))
stream.c:                      ^
stream.c: ...

Number of similar (implementation,compiler) pairs: 5, namely:
ImplementationCompiler
T:avx512clang -march=native -O2 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx512clang -march=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx512clang -march=native -O -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx512clang -march=native -Os -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)
T:avx512clang -mcpu=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)

Compiler output


stream.c: In file included from Speck64128AVX512.h:17,
stream.c:                  from stream.c:20:
stream.c: stream.c: In function 'Encrypt':
stream.c: Intrinsics_AVX512_64block.h:25:21: warning: AVX512F vector return without AVX512F enabled changes the ABI [-Wpsabi]
stream.c:    25 | #define SET1(X,c) (X=SET(c,c,c,c,c,c,c,c,c,c,c,c,c,c,c,c))
stream.c:       |                   ~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
stream.c: stream.c:135:5: note: in expansion of macro 'SET1'
stream.c:   135 |     SET1(X[0],nonce[1]); SET16(Y[0],nonce[0]);
stream.c:       |     ^~~~
stream.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/10/include/immintrin.h:55,
stream.c:                  from Intrinsics_AVX512_64block.h:5,
stream.c:                  from Speck64128AVX512.h:17,
stream.c:                  from stream.c:20:
stream.c: /usr/lib/gcc/x86_64-linux-gnu/10/include/avx512fintrin.h:6429:1: error: inlining failed in call to 'always_inline' '_mm512_storeu_si512': target specific option mismatch
stream.c:  6429 | _mm512_storeu_si512 (void *__P, __m512i __A)
stream.c:       | ^~~~~~~~~~~~~~~~~~~
stream.c: In file included from Speck64128AVX512.h:17,
stream.c:                  from stream.c:20:
stream.c: Intrinsics_AVX512_64block.h:31:18: note: called from here
stream.c:    31 | #define ST(ip,X) _mm512_storeu_si512((__m512i *)(ip),X)
stream.c:       |                  ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
stream.c: Intrinsics_AVX512_64block.h:32:42: note: in expansion of macro 'ST'
stream.c:    32 | #define STORE(out,X,Y) (ST(out,LOW(Y,X)),ST(out+64,HIGH(Y,X)))
stream.c:       |                                          ^~
stream.c: stream.c:138:5: note: in expansion of macro 'STORE'
stream.c: ...

Number of similar (implementation,compiler) pairs: 4, namely:
ImplementationCompiler
T:avx512gcc -march=native -mtune=native -O2 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)
T:avx512gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)
T:avx512gcc -march=native -mtune=native -O -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)
T:avx512gcc -march=native -mtune=native -Os -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (10.2.1_20210110)

Compiler output


stream.c: stream.c:134:21: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'Encrypt' that is compiled without support for 'ssse3'
stream.c:   if (numbytes==32) Enc(X,Y,rk,4);
stream.c:                     ^
stream.c: ./Speck64128SSE4.h:45:23: note: expanded from macro 'Enc'
stream.c: #define Enc(X,Y,k,n) (Rx##n(X,Y,k[0]),  Rx##n(X,Y,k[1]),  Rx##n(X,Y,k[2]),  Rx##n(X,Y,k[3]),  Rx##n(X,Y,k[4]),  Rx##n(X,Y,k[5]),  Rx##n(X,Y,k[6]),  Rx##n(X,Y,k[7]), \
stream.c:                       ^
stream.c: <scratch space>:108:1: note: expanded from here
stream.c: Rx4
stream.c: ^
stream.c: ./Speck64128SSE4.h:25:22: note: expanded from macro 'Rx4'
stream.c: #define Rx4(X,Y,k)  (R(X[0],Y[0],k))
stream.c:                      ^
stream.c: ./Speck64128SSE4.h:23:29: note: expanded from macro 'R'
stream.c: #define R(X,Y,k) (X=XOR(ADD(ROR8(X),Y),k), Y=XOR(ROL(Y,3),X))
stream.c:                             ^
stream.c: ./Intrinsics_SSE4_64block.h:40:19: note: expanded from macro 'ROR8'
stream.c: #define ROR8(X)  (SHFL(X,R8))
stream.c:                   ^
stream.c: ./Intrinsics_SSE4_64block.h:35:14: note: expanded from macro 'SHFL'
stream.c: #define SHFL _mm_shuffle_epi8
stream.c:              ^
stream.c: stream.c:134:21: error: always_inline function '_mm_shuffle_epi8' requires target feature 'ssse3', but would be inlined into function 'Encrypt' that is compiled without support for 'ssse3'
stream.c: ./Speck64128SSE4.h:45:41: note: expanded from macro 'Enc'
stream.c: #define Enc(X,Y,k,n) (Rx##n(X,Y,k[0]),  Rx##n(X,Y,k[1]),  Rx##n(X,Y,k[2]),  Rx##n(X,Y,k[3]),  Rx##n(X,Y,k[4]),  Rx##n(X,Y,k[5]),  Rx##n(X,Y,k[6]),  Rx##n(X,Y,k[7]), \
stream.c:                                         ^
stream.c: ...

Number of similar (implementation,compiler) pairs: 1, namely:
ImplementationCompiler
T:sse4clang -mcpu=native -O3 -fwrapv -Qunused-arguments -fPIC -fPIE -gdwarf-4 -Wall (Debian_Clang_11.0.1)