Remove obsolete SIMD code #57
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Probably due to compiler optimizations the Go code is now faster than the SSSE3/AVX/AVX2 code: ``` BenchmarkHash/GEN_/8Bytes-32 6486468 184 ns/op 43.46 MB/s BenchmarkHash/GEN_/1K-32 545470 2172 ns/op 471.36 MB/s BenchmarkHash/GEN_/8K-32 74073 16106 ns/op 508.64 MB/s BenchmarkHash/GEN_/1M-32 584 2034247 ns/op 515.46 MB/s BenchmarkHash/GEN_/5M-32 100 10190003 ns/op 514.51 MB/s BenchmarkHash/GEN_/10M-32 56 20357139 ns/op 515.09 MB/s BenchmarkHash/AVX2/8Bytes-32 5263258 226 ns/op 35.44 MB/s BenchmarkHash/AVX2/1K-32 444441 2633 ns/op 388.98 MB/s BenchmarkHash/AVX2/8K-32 61855 19513 ns/op 419.81 MB/s BenchmarkHash/AVX2/1M-32 487 2462013 ns/op 425.90 MB/s BenchmarkHash/AVX2/5M-32 91 12384626 ns/op 423.34 MB/s BenchmarkHash/AVX2/10M-32 44 26636364 ns/op 393.66 MB/s BenchmarkHash/AVX_/8Bytes-32 6349206 188 ns/op 42.54 MB/s BenchmarkHash/AVX_/1K-32 461538 2620 ns/op 390.91 MB/s BenchmarkHash/AVX_/8K-32 61224 19567 ns/op 418.65 MB/s BenchmarkHash/AVX_/1M-32 484 2473140 ns/op 423.99 MB/s BenchmarkHash/AVX_/5M-32 99 12505052 ns/op 419.26 MB/s BenchmarkHash/AVX_/10M-32 46 24869557 ns/op 421.63 MB/s BenchmarkHash/SSSE/8Bytes-32 6282679 192 ns/op 41.71 MB/s BenchmarkHash/SSSE/1K-32 461614 2628 ns/op 389.69 MB/s BenchmarkHash/SSSE/8K-32 60913 19651 ns/op 416.88 MB/s BenchmarkHash/SSSE/1M-32 481 2488563 ns/op 421.36 MB/s BenchmarkHash/SSSE/5M-32 91 12516477 ns/op 418.88 MB/s BenchmarkHash/SSSE/10M-32 46 24869561 ns/op 421.63 MB/s ``` Remove obsolete and slower code. Simplify CPUID code.
harshavardhana
approved these changes
Feb 22, 2021
|
For those of you running Go master/tip, this PR and v1.0.0 fixed the following panic: Hopefully the error mention makes it easier to find the solution by googling :) |
MichaelEischer
added a commit
to MichaelEischer/restic
that referenced
this pull request
Jul 10, 2021
Apparently the standard Go sha256 implementation is now faster than the assembly implementation. The library now only adds support for SHA extensions available in some processors. See minio/sha256-simd#57 for more details.
6 tasks
mfrischknecht
pushed a commit
to mfrischknecht/restic
that referenced
this pull request
Jun 14, 2022
Apparently the standard Go sha256 implementation is now faster than the assembly implementation. The library now only adds support for SHA extensions available in some processors. See minio/sha256-simd#57 for more details.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Probably due to compiler optimizations the Go code is now faster than the SSSE3/AVX/AVX2 code:
Remove obsolete and slower code. Simplify CPUID code.