On x86_64, currently 3 variants of AVX512, 3 variants of AVX2 and 3 variants of SSE2 are benchmarked on initialization, taking between 144-153 jiffies. Testing across a hardware pool of various generations of intel cpus I could not find a single case where SSE2 won over AVX2 or AVX512. There are cases where AVX2 wins over AVX512 however. Change "prefer" into an integer priority field (similar to how recov selection works) to have more than one ranking level available, which is backwards compatible with existing behavior. Give AVX2/512 variants higher priority over SSE2 in order to skip SSE testing when AVX is available. in a AVX2/x86_64/HZ=250 case this saves in the order of 200ms of initialization time.
Showing the changes to include/linux/raid (1 of 4 files). Show all
Unified · Split
| @@ -81,7 +81,7 @@ struct raid6_calls { · 80 unmodified lines | |||
| 81 | void (*xor_syndrome)(int, int, int, size_t, void **); | 81 | void (*xor_syndrome)(int, int, int, size_t, void **); |
| 82 | int (*valid)(void); /* Returns 1 if this routine set is usable */ | 82 | int (*valid)(void); /* Returns 1 if this routine set is usable */ |
| 83 | const char *name; /* Name of this routine set */ | 83 | const char *name; /* Name of this routine set */ |
| 84 | int prefer; /* Has special performance attribute */ | 84 | int priority; /* Relative priority ranking if non-zero */ |
| 85 | }; | 85 | }; |
| 86 | 86 | ||
| 87 | /* Selected algorithm */ | 87 | /* Selected algorithm */ |