Add benchmarks for a number of NSStringAPI functions#10914
Merged
phausler merged 1 commit intoswiftlang:masterfrom Jul 13, 2017
Merged
Add benchmarks for a number of NSStringAPI functions#10914phausler merged 1 commit intoswiftlang:masterfrom
phausler merged 1 commit intoswiftlang:masterfrom
Conversation
Contributor
Author
|
@swift-ci please benchmark |
Contributor
Author
|
@swift-ci please smoke test |
Contributor
|
@phausler 'please benchmark' takes forever, if you're interested in just a rough picture 'please smoke benchmark' is way faster. It will run all the benchmarks but with smaller number of samples, IIRC. |
Contributor
Author
|
gotcha, I will let this run-out for now just to get good baselines. I presume that the baselines are recorded when the tests are run? or per pull request doing an A/B execution before and after the patch? My local executions are just hoisting the methods into a benchmark lambda: import Foundation
#if os(macOS) || os(iOS)
import Darwin
let CLOCK_REALTIME = _CLOCK_REALTIME
#else
import Glibc
#endif
func benchmark(_ samples: Int, _ work: (Int) -> Void) -> Double {
var start = timespec()
var end = timespec()
clock_gettime(CLOCK_REALTIME, &start)
work(samples)
clock_gettime(CLOCK_REALTIME, &end)
return (Double(end.tv_sec) * 1.0e9 + Double(end.tv_nsec)) - (Double(start.tv_sec) * 1.0e9 + Double(start.tv_nsec)) / Double(samples)
} |
Contributor
|
I think it's the latter, as in A/B every time. |
Contributor
Author
|
hmm is there a way I can cancel a build on CI? |
Contributor
Author
|
@swift-ci please smoke benchmark |
Contributor
|
@phausler some of the builds get cancelled automatically once the PR changes, not sure about the benchmark ones. |
Contributor
Author
|
I was actually able to abort it |
Contributor
Author
|
interesting followup results: I was able to eek out performance improvements for the following cases: These all were re-writes using either direct So far I am seeing some cases be anywhere from 2x-3x up to 40x improvements. It is also worth noting that some of the tests regress when not using |
Contributor
|
|
Contributor
Build comment file:Build failed before running benchmark. |
seabaylea
reviewed
Jul 13, 2017
ae0cbce to
78cf2aa
Compare
Contributor
Author
|
@swift-ci please smoke benchmark |
Contributor
Author
|
@swift-ci please smoke test |
Contributor
Build comment file:Optimized (O)Regression (18)
Improvement (17)
No Changes (286)
Added (17)
Unoptimized (Onone)Regression (3)
Improvement (3)
No Changes (315)
Added (17)
Hardware Overview |
Contributor
Author
|
We should probably add more of these to get full coverage but these should do for now to get a feeling of how NSStringAPI extensions are performing with the incoming changes |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
This adds a number of simple benchmarks for common NSString APIs that are exposed on String