Implementation notes: amd64, saber214, crypto_hash/sha3224

Computer: saber214
Microarchitecture: amd64; Bulldozer (600f20)
Architecture: amd64
CPU ID: AuthenticAMD-00600f20-1789c3f5
SUPERCOP version: 20231107
Operation: crypto_hash
Primitive: sha3224

Time	Object size	Test size	Implementation	Compiler	Benchmark date	SUPERCOP version
21060	753 0 0	68645 784 800	`oncore64bits`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
21114	617 0 0	67389 776 768	`oncore64bits`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
21130	736 0 0	68413 784 800	`oncore64bits`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
21137	679 0 0	67551 824 728	`oncore64bits`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
21195	2030 0 0	71349 784 800	`oncore64bits`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
21207	1213 0 0	70071 824 728	`oncore64bits`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
21209	637 0 0	67145 816 728	`oncore64bits`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
21253	945 0 0	68911 824 728	`oncore64bits`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
21302	945 0 0	71303 824 744	`oncore64bits`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
25410	243 0 0	10895 872 728	`openssl`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230607	20230530
25439	307 0 0	10787 824 768	`openssl`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
25572	245 0 0	14359 872 744	`openssl`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230607	20230530
25590	328 0 0	12128 848 768	`openssl`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
25707	245 0 0	10505 864 728	`openssl`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230607	20230530
25762	245 0 0	12855 872 728	`openssl`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230607	20230530
25776	245 0 0	11967 872 728	`openssl`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20230607	20230530
25778	325 0 0	11823 840 768	`openssl`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
28027	328 0 0	13608 848 768	`openssl`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
47732	2382 0 0	26943 824 728	`oncore32bits`	`clang_-mcpu=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
47795	1892 0 0	25639 824 728	`oncore32bits`	`clang_-march=native_-O2_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
48014	5597 0 0	30852 792 800	`oncore32bits`	`gcc_-march=native_-mtune=native_-O3_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
48059	2196 0 0	28287 824 744	`oncore32bits`	`clang_-march=native_-O3_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
51384	1368 0 0	24015 824 728	`oncore32bits`	`clang_-march=native_-O_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
52230	1401 0 0	23689 816 728	`oncore32bits`	`clang_-march=native_-Os_-fomit-frame-pointer_-fwrapv_-Qunused-arguments_-fPIC_-fPIE`	20231108	20231107
52288	1817 0 0	25596 792 800	`oncore32bits`	`gcc_-march=native_-mtune=native_-O2_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
52554	2091 0 0	24431 768 768	`oncore32bits`	`gcc_-march=native_-mtune=native_-Os_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107
52859	1847 0 0	25228 776 800	`oncore32bits`	`gcc_-march=native_-mtune=native_-O_-fomit-frame-pointer_-fwrapv_-fPIC_-fPIE`	20231108	20231107

Compiler output

Implementation: kcp/optimized1600ARMv7A
Security model: constbranchindex
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

keccak.s: keccak.s:1:1: error: unexpected token at start of statement
keccak.s: @
keccak.s: ^
keccak.s: keccak.s:2:1: error: unexpected token at start of statement
keccak.s: @ Implementation by the Keccak, Keyak and Ketje Teams, namely, Guido Bertoni,
keccak.s: ^
keccak.s: keccak.s:3:1: error: unexpected token at start of statement
keccak.s: @ Joan Daemen, Michaël Peeters, Gilles Van Assche and Ronny Van Keer, hereby
keccak.s: keccak.s:4:1: error: unexpected token at start of statement
keccak.s: @ denoted as "the implementer".
keccak.s: ^
keccak.s: keccak.s:5:1: error: unexpected token at start of statement
keccak.s: @
keccak.s: ^
keccak.s: keccak.s:6:1: error: unexpected token at start of statement
keccak.s: @ For more information, feedback or questions, please refer to our websites:
keccak.s: ^
keccak.s: keccak.s:7:1: error: unexpected token at start of statement
keccak.s: @ http://keccak.noekeon.org/
keccak.s: ^
keccak.s: keccak.s:8:1: error: unexpected token at start of statement
keccak.s: @ http://keyak.noekeon.org/
keccak.s: ^
keccak.s: keccak.s:9:1: error: unexpected token at start of statement
keccak.s: @ http://ketje.noekeon.org/
keccak.s: ...

Number of similar (compiler,implementation) pairs: 5, namely:

Compiler	Implementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv7A
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv7A
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv7A
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv7A
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv7A

Compiler output

Implementation: kcp/optimized1600ARMv7A
Security model: constbranchindex
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

keccak.s: keccak.s: Assembler messages:
keccak.s: keccak.s:1: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:2: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:3: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:4: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:5: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:6: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:7: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:8: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:9: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:10: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:11: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:12: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:13: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:14: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:16: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:17: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:18: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:21: Error: unknown pseudo-op: `.syntax'
keccak.s: keccak.s:22: Error: unknown pseudo-op: `.fpu'
keccak.s: keccak.s:23: Error: unknown pseudo-op: `.arm'
keccak.s: keccak.s:26: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:27: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:29: Error: junk at end of line, first unrecognized character is `@'
keccak.s: keccak.s:56: Error: junk at end of line, first unrecognized character is `@'
keccak.s: ...

Number of similar (compiler,implementation) pairs: 4, namely:

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv7A
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv7A
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv7A
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv7A

Compiler output

Implementation: kcp/optimized1600ARMv8A
Security model: constbranchindex
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

keccak.s: keccak.s:258:20: error: unknown token in expression
keccak.s: movi v0.2d, #0
keccak.s: ^
keccak.s: keccak.s:259:20: error: unknown token in expression
keccak.s: movi v1.2d, #0
keccak.s: ^
keccak.s: keccak.s:260:20: error: unknown token in expression
keccak.s: movi v2.2d, #0
keccak.s: ^
keccak.s: keccak.s:261:20: error: unknown token in expression
keccak.s: movi v3.2d, #0
keccak.s: ^
keccak.s: keccak.s:262:15: error: unknown token in expression
keccak.s: st4 { v0.2d, v1.2d, v2.2d, v3.2d }, [x0], #64 // Clear 8lanes=64 bytes at a time
keccak.s: ^
keccak.s: keccak.s:263:15: error: unknown token in expression
keccak.s: st4 { v0.2d, v1.2d, v2.2d, v3.2d }, [x0], #64
keccak.s: ^
keccak.s: keccak.s:264:15: error: unknown token in expression
keccak.s: st4 { v0.2d, v1.2d, v2.2d, v3.2d }, [x0], #64
keccak.s: ^
keccak.s: keccak.s:265:15: error: unknown token in expression
keccak.s: st1 { v0.d }[0], [x0], #8
keccak.s: ^
keccak.s: keccak.s:276:20: error: expected ']' in brackets expression
keccak.s: ...

Number of similar (compiler,implementation) pairs: 5, namely:

Compiler	Implementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv8A
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv8A
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv8A
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv8A
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600ARMv8A

Compiler output

Implementation: kcp/optimized1600ARMv8A
Security model: constbranchindex
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

keccak.s: keccak.s: Assembler messages:
keccak.s: keccak.s:258: Error: no such instruction: `movi v0.2d,'
keccak.s: keccak.s:259: Error: no such instruction: `movi v1.2d,'
keccak.s: keccak.s:260: Error: no such instruction: `movi v2.2d,'
keccak.s: keccak.s:261: Error: no such instruction: `movi v3.2d,'
keccak.s: keccak.s:262: Error: no such instruction: `st4 { v0.2d,v1.2d,v2.2d,v3.2d },[x0],'
keccak.s: keccak.s:263: Error: no such instruction: `st4 { v0.2d,v1.2d,v2.2d,v3.2d },[x0],'
keccak.s: keccak.s:264: Error: no such instruction: `st4 { v0.2d,v1.2d,v2.2d,v3.2d },[x0],'
keccak.s: keccak.s:265: Error: no such instruction: `st1 { v0.d }[0],[x0],'
keccak.s: keccak.s:276: Error: no such instruction: `ldrb w3,[x0,x2]'
keccak.s: keccak.s:277: Error: no such instruction: `eor w3,w3,w1'
keccak.s: keccak.s:278: Error: too many memory references for `str'
keccak.s: keccak.s:289: Error: too many memory references for `add'
keccak.s: keccak.s:290: Error: too many memory references for `sub'
keccak.s: keccak.s:291: Error: no such instruction: `b.cc KeccakP1600_AddBytes_Exit//length 0,move along'
keccak.s: keccak.s:293: Error: too many memory references for `sub'
keccak.s: keccak.s:294: Error: no such instruction: `b.cc KeccakP1600_AddBytes_Lanes//Jump if length is negative'
keccak.s: keccak.s:295: Error: no such instruction: `ld4 { v0.2d,v1.2d,v2.2d,v3.2d },[x0]'
keccak.s: keccak.s:296: Error: no such instruction: `ld4 { v4.2d,v5.2d,v6.2d,v7.2d },[x1],'
keccak.s: keccak.s:297: Error: no such instruction: `eor v0.16b,v0.16b,v4.16b'
keccak.s: keccak.s:298: Error: no such instruction: `eor v1.16b,v1.16b,v5.16b'
keccak.s: keccak.s:299: Error: no such instruction: `eor v2.16b,v2.16b,v6.16b'
keccak.s: keccak.s:300: Error: no such instruction: `eor v3.16b,v3.16b,v7.16b'
keccak.s: keccak.s:301: Error: no such instruction: `st4 { v0.2d,v1.2d,v2.2d,v3.2d },[x0],'
keccak.s: keccak.s:302: Error: no such instruction: `b KeccakP1600_AddBytes_8LanesLoop'
keccak.s: ...

Number of similar (compiler,implementation) pairs: 4, namely:

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv8A
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv8A
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv8A
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600ARMv8A

Compiler output

Implementation: kcp/optimized1600AVX2
Security model: constbranchindex
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:643:5: error: always_inline function '_mm256_xor_si256' requires target feature 'avx2', but would be inlined into function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_Permute_Nrounds' that is compiled without support for 'avx2'
KeccakP-1600-AVX2.c: KECCAK_PERMUTE_LOOP(, nrounds)
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:162:14: note: expanded from macro 'KECCAK_PERMUTE_LOOP'
KeccakP-1600-AVX2.c: r0 = XOR(a0, a1); \
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:54:41: note: expanded from macro 'XOR'
KeccakP-1600-AVX2.c: #define XOR(a, b) _mm256_xor_si256(a, b)
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:643:5: error: always_inline function '_mm256_xor_si256' requires target feature 'avx2', but would be inlined into function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_Permute_Nrounds' that is compiled without support for 'avx2'
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:163:14: note: expanded from macro 'KECCAK_PERMUTE_LOOP'
KeccakP-1600-AVX2.c: r0 = XOR(r0, a2); \
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:54:41: note: expanded from macro 'XOR'
KeccakP-1600-AVX2.c: #define XOR(a, b) _mm256_xor_si256(a, b)
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:643:5: error: always_inline function '_mm256_xor_si256' requires target feature 'avx2', but would be inlined into function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_Permute_Nrounds' that is compiled without support for 'avx2'
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:164:14: note: expanded from macro 'KECCAK_PERMUTE_LOOP'
KeccakP-1600-AVX2.c: r0 = XOR(r0, a3); \
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:54:41: note: expanded from macro 'XOR'
KeccakP-1600-AVX2.c: #define XOR(a, b) _mm256_xor_si256(a, b)
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:643:5: error: always_inline function '_mm256_xor_si256' requires target feature 'avx2', but would be inlined into function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_Permute_Nrounds' that is compiled without support for 'avx2'
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:165:14: note: expanded from macro 'KECCAK_PERMUTE_LOOP'
KeccakP-1600-AVX2.c: ...

Number of similar (compiler,implementation) pairs: 4, namely:

Compiler	Implementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX2
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX2
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX2
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX2

Compiler output

Implementation: kcp/optimized1600AVX2
Security model: constbranchindex
Compiler: clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:506:13: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_AddBytes' that is compiled without support for 'avx'
KeccakP-1600-AVX2.c: s->a0 = LOAD(t + 0*5);
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:62:41: note: expanded from macro 'LOAD'
KeccakP-1600-AVX2.c: #define LOAD(p) _mm256_loadu_si256((const __m256i *)(p))
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:506:13: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:62:41: note: expanded from macro 'LOAD'
KeccakP-1600-AVX2.c: #define LOAD(p) _mm256_loadu_si256((const __m256i *)(p))
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:507:13: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_AddBytes' that is compiled without support for 'avx'
KeccakP-1600-AVX2.c: s->a1 = LOAD(t + 1*5);
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:62:41: note: expanded from macro 'LOAD'
KeccakP-1600-AVX2.c: #define LOAD(p) _mm256_loadu_si256((const __m256i *)(p))
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:507:13: error: AVX vector return of type '__m256i' (vector of 4 'long long' values) without 'avx' enabled changes the ABI
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:62:41: note: expanded from macro 'LOAD'
KeccakP-1600-AVX2.c: #define LOAD(p) _mm256_loadu_si256((const __m256i *)(p))
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:508:13: error: always_inline function '_mm256_loadu_si256' requires target feature 'avx', but would be inlined into function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_AddBytes' that is compiled without support for 'avx'
KeccakP-1600-AVX2.c: s->a2 = LOAD(t + 2*5);
KeccakP-1600-AVX2.c: ^
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:62:41: note: expanded from macro 'LOAD'
KeccakP-1600-AVX2.c: #define LOAD(p) _mm256_loadu_si256((const __m256i *)(p))
KeccakP-1600-AVX2.c: ...

Number of similar (compiler,implementation) pairs: 1, namely:

Compiler	Implementations
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX2

Compiler output

Implementation: kcp/optimized1600AVX2
Security model: constbranchindex
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

KeccakP-1600-AVX2.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
KeccakP-1600-AVX2.c: from /usr/lib/gcc/x86_64-linux-gnu/11/include/x86intrin.h:32,
KeccakP-1600-AVX2.c: from KeccakP-1600-AVX2.c:20:
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c: In function 'crypto_hash_sha3224_kcp_optimized1600AVX2_constbranchindex_KeccakP1600_Permute_Nrounds':
KeccakP-1600-AVX2.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:1084:1: error: inlining failed in call to 'always_inline' '_mm256_permute2x128_si256': target specific option mismatch
KeccakP-1600-AVX2.c: 1084 | _mm256_permute2x128_si256 (__m256i __X, __m256i __Y, const int __M)
KeccakP-1600-AVX2.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:269:14: note: called from here
KeccakP-1600-AVX2.c: 269 | a4 = _mm256_permute2x128_si256(r0, r1, 0x20); \
KeccakP-1600-AVX2.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:643:5: note: in expansion of macro 'KECCAK_PERMUTE_LOOP'
KeccakP-1600-AVX2.c: 643 | KECCAK_PERMUTE_LOOP(, nrounds)
KeccakP-1600-AVX2.c: | ^~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX2.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:47,
KeccakP-1600-AVX2.c: from /usr/lib/gcc/x86_64-linux-gnu/11/include/x86intrin.h:32,
KeccakP-1600-AVX2.c: from KeccakP-1600-AVX2.c:20:
KeccakP-1600-AVX2.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx2intrin.h:906:1: error: inlining failed in call to 'always_inline' '_mm256_unpacklo_epi64': target specific option mismatch
KeccakP-1600-AVX2.c: 906 | _mm256_unpacklo_epi64 (__m256i __A, __m256i __B)
KeccakP-1600-AVX2.c: | ^~~~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:268:14: note: called from here
KeccakP-1600-AVX2.c: 268 | r1 = _mm256_unpacklo_epi64(a24, a34); \
KeccakP-1600-AVX2.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX2.c: KeccakP-1600-AVX2.c:643:5: note: in expansion of macro 'KECCAK_PERMUTE_LOOP'
KeccakP-1600-AVX2.c: 643 | KECCAK_PERMUTE_LOOP(, nrounds)
KeccakP-1600-AVX2.c: | ^~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX2.c: ...

Number of similar (compiler,implementation) pairs: 4, namely:

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX2
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX2
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX2
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX2

Compiler output

Implementation: kcp/optimized1600AVX512
Security model: constbranchindex
Compiler: clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE

KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:238:41: error: always_inline function '_mm512_maskz_loadu_epi64' requires target feature 'avx512f', but would be inlined into function 'KeccakP1600_AddBytes' that is compiled without support for 'avx512f'
KeccakP-1600-AVX512.c: STORE_8Lanes( stateAsLanes, XOR(LOAD_8Lanes(stateAsLanes), LOAD_8Lanes((const UINT64*)data)));
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:215:37: note: expanded from macro 'LOAD_8Lanes'
KeccakP-1600-AVX512.c: #define LOAD_8Lanes(a) LOAD_Lanes(0xFF,a)
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:212:37: note: expanded from macro 'LOAD_Lanes'
KeccakP-1600-AVX512.c: #define LOAD_Lanes(m,a) _mm512_maskz_loadu_epi64(m,a)
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:238:41: error: AVX vector return of type '__m512i' (vector of 8 'long long' values) without 'avx512f' enabled changes the ABI
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:215:37: note: expanded from macro 'LOAD_8Lanes'
KeccakP-1600-AVX512.c: #define LOAD_8Lanes(a) LOAD_Lanes(0xFF,a)
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:212:37: note: expanded from macro 'LOAD_Lanes'
KeccakP-1600-AVX512.c: #define LOAD_Lanes(m,a) _mm512_maskz_loadu_epi64(m,a)
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:238:68: error: always_inline function '_mm512_maskz_loadu_epi64' requires target feature 'avx512f', but would be inlined into function 'KeccakP1600_AddBytes' that is compiled without support for 'avx512f'
KeccakP-1600-AVX512.c: STORE_8Lanes( stateAsLanes, XOR(LOAD_8Lanes(stateAsLanes), LOAD_8Lanes((const UINT64*)data)));
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:215:37: note: expanded from macro 'LOAD_8Lanes'
KeccakP-1600-AVX512.c: #define LOAD_8Lanes(a) LOAD_Lanes(0xFF,a)
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:212:37: note: expanded from macro 'LOAD_Lanes'
KeccakP-1600-AVX512.c: #define LOAD_Lanes(m,a) _mm512_maskz_loadu_epi64(m,a)
KeccakP-1600-AVX512.c: ^
KeccakP-1600-AVX512.c: ...

Number of similar (compiler,implementation) pairs: 5, namely:

Compiler	Implementations
clang -march=native -O2 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX512
clang -march=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX512
clang -march=native -O -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX512
clang -march=native -Os -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX512
clang -mcpu=native -O3 -fomit-frame-pointer -fwrapv -Qunused-arguments -fPIC -fPIE	kcp/optimized1600AVX512

Compiler output

Implementation: kcp/optimized1600AVX512
Security model: constbranchindex
Compiler: gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE

KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c: In function 'KeccakP1600_AddBytes':
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:216:37: warning: AVX512F vector return without AVX512F enabled changes the ABI [-Wpsabi]
KeccakP-1600-AVX512.c: 216 | #define STORE_Lanes(a,m,v) _mm512_mask_storeu_epi64(a,m,v)
KeccakP-1600-AVX512.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:219:37: note: in expansion of macro 'STORE_Lanes'
KeccakP-1600-AVX512.c: 219 | #define STORE_8Lanes(a,v) STORE_Lanes(a,0xFF,v)
KeccakP-1600-AVX512.c: | ^~~~~~~~~~~
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:238:9: note: in expansion of macro 'STORE_8Lanes'
KeccakP-1600-AVX512.c: 238 | STORE_8Lanes( stateAsLanes, XOR(LOAD_8Lanes(stateAsLanes), LOAD_8Lanes((const UINT64*)data)));
KeccakP-1600-AVX512.c: | ^~~~~~~~~~~~
KeccakP-1600-AVX512.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:49,
KeccakP-1600-AVX512.c: from KeccakP-1600-AVX512.c:26:
KeccakP-1600-AVX512.c: /usr/lib/gcc/x86_64-linux-gnu/11/include/avx512fintrin.h:6440:1: error: inlining failed in call to 'always_inline' '_mm512_mask_storeu_epi64': target specific option mismatch
KeccakP-1600-AVX512.c: 6440 | _mm512_mask_storeu_epi64 (void *__P, __mmask8 __U, __m512i __A)
KeccakP-1600-AVX512.c: | ^~~~~~~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:216:37: note: called from here
KeccakP-1600-AVX512.c: 216 | #define STORE_Lanes(a,m,v) _mm512_mask_storeu_epi64(a,m,v)
KeccakP-1600-AVX512.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:219:37: note: in expansion of macro 'STORE_Lanes'
KeccakP-1600-AVX512.c: 219 | #define STORE_8Lanes(a,v) STORE_Lanes(a,0xFF,v)
KeccakP-1600-AVX512.c: | ^~~~~~~~~~~
KeccakP-1600-AVX512.c: KeccakP-1600-AVX512.c:238:9: note: in expansion of macro 'STORE_8Lanes'
KeccakP-1600-AVX512.c: 238 | STORE_8Lanes( stateAsLanes, XOR(LOAD_8Lanes(stateAsLanes), LOAD_8Lanes((const UINT64*)data)));
KeccakP-1600-AVX512.c: | ^~~~~~~~~~~~
KeccakP-1600-AVX512.c: In file included from /usr/lib/gcc/x86_64-linux-gnu/11/include/immintrin.h:49,
KeccakP-1600-AVX512.c: ...

Number of similar (compiler,implementation) pairs: 4, namely:

Compiler	Implementations
gcc -march=native -mtune=native -O2 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX512
gcc -march=native -mtune=native -O3 -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX512
gcc -march=native -mtune=native -O -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX512
gcc -march=native -mtune=native -Os -fomit-frame-pointer -fwrapv -fPIC -fPIE	kcp/optimized1600AVX512