Test results for aarch64, cfarm425, crypto_hashblocks/sha512
[Page version: 20261005 22:53:57]
Measurements for aarch64, cfarm425, crypto_hashblocks
Test results for aarch64, cfarm425, crypto_hashblocks
Test results for crypto_hashblocks/sha512
Computer: cfarm425
Microarchitecture: aarch64; Cortex-A76 (413fd0c1)
Architecture: aarch64
CPU ID: 413fd0c1
SUPERCOP version: 20261004
Operation: crypto_hashblocks
Primitive: sha512
| Time | Object size | Test size | Implementation | Compiler | Benchmark date | SUPERCOP version |
|---|
| 9927 | 3020 0 0 | 22364 1200 1416 | wflip | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall | 20261005 | 20261004 |
| 10612 | 11228 0 0 | 30588 1200 1416 | ref | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall | 20261005 | 20261004 |
| 10614 | 11240 0 0 | 30636 1200 1416 | inplace | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall | 20261005 | 20261004 |
| 10866 | 3928 0 0 | 23340 1200 1416 | compact | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall | 20261005 | 20261004 |
| 14756 | 1852 0 0 | 21228 1200 1416 | compact4 | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall | 20261005 | 20261004 |
| 20131 | 2072 0 0 | 21484 1200 1416 | compact2 | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall | 20261005 | 20261004 |
| 20674 | 1848 0 0 | 21220 1200 1416 | compact3 | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall | 20261005 | 20261004 |
Compiler output
blocks.c: blocks.c:15:2: warning: #warning "FEAT_SHA512 is not enabled in the compiler. To enable FEAT_SHA512, use for instance \"-mcpu=neoverse-v2+crypto+sha3\" on Graviton 4 or NVidia Grace." [-Wcpp]
blocks.c: 15 | #warning "FEAT_SHA512 is not enabled in the compiler. To enable FEAT_SHA512, use for instance \"-mcpu=neoverse-v2+crypto+sha3\" on Graviton 4 or NVidia Grace."
blocks.c: | ^~~~~~~
blocks.c: In file included from blocks.c:11:
blocks.c: /usr/lib/gcc/aarch64-linux-gnu/16/include/arm_neon.h: In function 'crypto_hashblocks_sha512_dolbeau_arm_sha512_constbranchindex':
blocks.c: /usr/lib/gcc/aarch64-linux-gnu/16/include/arm_neon.h:26827:1: error: inlining failed in call to 'always_inline' 'vsha512h2q_u64': target specific option mismatch
blocks.c: 26827 | vsha512h2q_u64 (uint64x2_t __a, uint64x2_t __b, uint64x2_t __c)
blocks.c: | ^~~~~~~~~~~~~~
blocks.c: blocks.c:59:8: note: called from here
blocks.c: 59 | ef = vsha512h2q_u64(half_round_v2, ab, gh); \
blocks.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blocks.c: blocks.c:212:5: note: in expansion of macro 'F4_v2'
blocks.c: 212 | F4_v2(w12, GETK(76), w14, GETK(78), ef, gh, ab, cd);
blocks.c: | ^~~~~
blocks.c: /usr/lib/gcc/aarch64-linux-gnu/16/include/arm_neon.h:26820:1: error: inlining failed in call to 'always_inline' 'vsha512hq_u64': target specific option mismatch
blocks.c: 26820 | vsha512hq_u64 (uint64x2_t __a, uint64x2_t __b, uint64x2_t __c)
blocks.c: | ^~~~~~~~~~~~~
blocks.c: blocks.c:58:19: note: called from here
blocks.c: 58 | half_round_v2 = vsha512hq_u64(k2plusw2_swapped_plusef_v2, vextq_u64(cd, ef, 1), vextq_u64(ab, cd, 1)); \
blocks.c: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
blocks.c: blocks.c:212:5: note: in expansion of macro 'F4_v2'
blocks.c: 212 | F4_v2(w12, GETK(76), w14, GETK(78), ef, gh, ab, cd);
blocks.c: | ^~~~~
blocks.c: /usr/lib/gcc/aarch64-linux-gnu/16/include/arm_neon.h:26827:1: error: inlining failed in call to 'always_inline' 'vsha512h2q_u64': target specific option mismatch
blocks.c: 26827 | vsha512h2q_u64 (uint64x2_t __a, uint64x2_t __b, uint64x2_t __c)
blocks.c: ...
Number of similar (implementation,compiler) pairs: 1, namely:
| Implementation | Compiler |
| dolbeau/arm-sha512 | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |
Passed TIMECOP
TIMECOP iterations: 10
Number of similar (implementation,compiler) pairs: 7, namely:
| Implementation | Compiler |
| compact | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |
| compact2 | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |
| compact3 | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |
| compact4 | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |
| inplace | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |
| ref | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |
| wflip | gcc -march=native -mtune=native -O3 -fwrapv -fPIC -fPIE -gdwarf-4 -Wall (16.2.0) |