Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Perf] Linux/x64: 12 Regressions on 9/16/2023 1:27:38 AM #21885

Open
performanceautofiler bot opened this issue Sep 19, 2023 · 2 comments
Open

[Perf] Linux/x64: 12 Regressions on 9/16/2023 1:27:38 AM #21885

performanceautofiler bot opened this issue Sep 19, 2023 · 2 comments
Assignees
@stephentoub
Labels
arch-wasm arch-x64 branch-refs/heads/main kind-micro os-linux perf-regression PGO Applied if there were any profile guided optimization updates in the observed interval. untriaged wasm-aot

Comments

@performanceautofiler
Copy link

performanceautofiler bot commented Sep 19, 2023 •
edited
Loading

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.Numerics.Tests.Perf_VectorOf<UInt16>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
2.79 ns 4.67 ns 1.67 0.82 False
1.68 ns 3.68 ns 2.19 0.89 False
2.09 ns 6.57 ns 3.14 0.88 False

graph
graph
graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Numerics.Tests.Perf_VectorOf&lt;UInt16&gt;*'

Payloads

Baseline
Compare

System.Numerics.Tests.Perf_VectorOf<UInt16>.SubtractBenchmark

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 4.669234198102709 > 3.402939480490985.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -5.763342730946194 (T) = (0 -5.322413258160948) / Math.Sqrt((0.8323894978974854 / (28)) + (0.3475826596350546 / (10))) is less than -2.028094000977961 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (28) + (10) - 2, .025) and -0.3792711191847272 = (3.8588593526898256 - 5.322413258160948) / 3.8588593526898256 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorOf<UInt16>.ZeroBenchmark

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 3.6822807725387667 > 2.0172856019578247.
IsChangePoint: Marked as a change because one of 9/16/2023 1:27:38 AM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -2.9553785010805047 (T) = (0 -4.303304971502935) / Math.Sqrt((1.5420225269625896 / (29)) + (3.119718711263362 / (8))) is less than -2.0301079282477414 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (29) + (8) - 2, .025) and -0.842207275176109 = (2.3359504815176417 - 4.303304971502935) / 2.3359504815176417 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

System.Numerics.Tests.Perf_VectorOf<UInt16>.AddBenchmark

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 6.5741852485684005 > 4.62789257697812.
IsChangePoint: Marked as a change because one of 9/16/2023 6:40:29 AM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -4.139205724924937 (T) = (0 -5.772312064468867) / Math.Sqrt((2.328372491489724 / (28)) + (0.9261954342187014 / (8))) is less than -2.0322445093148245 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (28) + (8) - 2, .025) and -0.4702187650853108 = (3.9261586109152424 - 5.772312064468867) / 3.9261586109152424 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.Buffers.Text.Tests.Utf8ParserTests

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
84.13 ns 89.66 ns 1.07 0.26 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Buffers.Text.Tests.Utf8ParserTests*'

Payloads

Baseline
Compare

System.Buffers.Text.Tests.Utf8ParserTests.TryParseDecimal(value: 123456.789)

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 89.65670535014439 > 88.21670835620367.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -4.357122173332673 (T) = (0 -90.73714465033112) / Math.Sqrt((10.885536606518079 / (28)) + (7.906919509859351 / (10))) is less than -2.028094000977961 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (28) + (10) - 2, .025) and -0.05501953880850482 = (86.0051793474895 - 90.73714465033112) / 86.0051793474895 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in Benchstone.BenchI.BubbleSort2

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
22.71 ms 25.07 ms 1.10 0.36 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'Benchstone.BenchI.BubbleSort2*'

Payloads

Baseline
Compare

Benchstone.BenchI.BubbleSort2.Test

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 25.071085699999994 > 23.841809793356646.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -5.250128005257674 (T) = (0 -24760404.75256027) / Math.Sqrt((1018788786231.2522 / (27)) + (518021069673.94366 / (10))) is less than -2.0301079282477414 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (27) + (10) - 2, .025) and -0.06774482200997703 = (23189440.25029316 - 24760404.75256027) / 23189440.25029316 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.Collections.ContainsFalse<Int32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
80.89 μs 94.37 μs 1.17 0.35 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Collections.ContainsFalse&lt;Int32&gt;*'

Payloads

Baseline
Compare

System.Collections.ContainsFalse<Int32>.Stack(Size: 512)

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 94.37436274907321 > 84.86444538738986.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -5.11359167622912 (T) = (0 -91473.485306812) / Math.Sqrt((33113643.323564954 / (28)) + (20665531.378724493 / (10))) is less than -2.028094000977961 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (28) + (10) - 2, .025) and -0.11205871579466303 = (82255.98523495786 - 91473.485306812) / 82255.98523495786 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline 0bd362370bdd649f0f09eb5c1e58cf11d9538804
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.Memory.MemoryMarshal<Byte>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
5.47 ns 7.85 ns 1.43 0.82 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Memory.MemoryMarshal&lt;Byte&gt;*'

Payloads

Baseline
Compare

System.Memory.MemoryMarshal<Byte>.AsBytes

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 7.8468254215258915 > 5.196914530036157.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -4.144166757701091 (T) = (0 -7.155900210274351) / Math.Sqrt((1.3844331117415294 / (25)) + (1.6230829799427438 / (10))) is less than -2.034515297446192 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (25) + (10) - 2, .025) and -0.3702413130990484 = (5.222364952703104 - 7.155900210274351) / 5.222364952703104 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.IO.Tests.Perf_Path

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
2.44 μs 2.89 μs 1.18 0.37 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.IO.Tests.Perf_Path*'

Payloads

Baseline
Compare

System.IO.Tests.Perf_Path.GetFullPathForReallyLongPath

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 2.8878997010372616 > 2.56214480796792.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -4.874759717623636 (T) = (0 -2847.0569468067642) / Math.Sqrt((40068.64979890235 / (27)) + (22116.47122328643 / (10))) is less than -2.0301079282477414 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (27) + (10) - 2, .025) and -0.1161819666022665 = (2550.71039668684 - 2847.0569468067642) / 2550.71039668684 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.Collections.ContainsTrue<Int32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
47.95 μs 54.27 μs 1.13 0.41 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Collections.ContainsTrue&lt;Int32&gt;*'

Payloads

Baseline
Compare

System.Collections.ContainsTrue<Int32>.Stack(Size: 512)

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 54.271282986111125 > 49.604257135460465.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -5.129591646354223 (T) = (0 -53568.161113424765) / Math.Sqrt((9402736.78854179 / (28)) + (5449414.567837811 / (10))) is less than -2.028094000977961 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (28) + (10) - 2, .025) and -0.09874126226307436 = (48754.117965034435 - 53568.161113424765) / 48754.117965034435 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.Security.Cryptography.Primitives.Tests.Performance.Perf_FixedTimeEquals

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
42.16 ns 56.97 ns 1.35 0.41 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Security.Cryptography.Primitives.Tests.Performance.Perf_FixedTimeEquals*'

Payloads

Baseline
Compare

System.Security.Cryptography.Primitives.Tests.Performance.Perf_FixedTimeEquals.FixedTimeEquals_256Bit_AllBitsDifferent

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 56.97260881541608 > 46.25640315140038.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -6.057561019744553 (T) = (0 -57.150327924601825) / Math.Sqrt((31.222922332085982 / (26)) + (16.797272135427963 / (10))) is less than -2.0322445093148245 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (26) + (10) - 2, .025) and -0.21935736550782886 = (46.86921942756321 - 57.150327924601825) / 46.86921942756321 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in System.Text.Json.Serialization.Tests.WriteJson<IndexViewModel>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
28.42 μs 30.99 μs 1.09 0.30 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Text.Json.Serialization.Tests.WriteJson&lt;IndexViewModel&gt;*'

Payloads

Baseline
Compare

System.Text.Json.Serialization.Tests.WriteJson<IndexViewModel>.SerializeToWriter(Mode: SourceGen)

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 30.988830869709 > 29.705660825246532.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -5.190772840971554 (T) = (0 -31571.929023273482) / Math.Sqrt((1847375.5678164237 / (27)) + (1611905.3848469523 / (10))) is less than -2.0301079282477414 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (27) + (10) - 2, .025) and -0.08551961470015344 = (29084.623249294676 - 31571.929023273482) / 29084.623249294676 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository

Run Information

Name Value
Architecture x64
OS ubuntu 22.04
Queue TigerUbuntu
Baseline a47b51b4c2421f8e8557d1712708144fdf2ce4cb
Compare fd20f42f1f4fa77a0facde02e0c1527fa26e95cb
Diff Diff
Configs AOT:true, CompilationMode:wasm, RunKind:micro

Regressions in Benchstone.MDBenchI.MDPuzzle

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
883.83 ms 977.03 ms 1.11 0.36 False

graph
Test Report

Repro

General Docs link: https://github.com/dotnet/performance/blob/main/docs/benchmarking-workflow-dotnet-runtime.md

git clone https://github.com/dotnet/performance.git
python3 .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'Benchstone.MDBenchI.MDPuzzle*'

Payloads

Baseline
Compare

Benchstone.MDBenchI.MDPuzzle.Test

ETL Files

Histogram

Description of detection logic

IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsRegressionBase: Marked as regression because the compare was 5% greater than the baseline, and the value was not too small.
IsRegressionChecked: Marked as regression because the three check build points were 0.05 greater than the baseline.
IsRegressionWindowed: Marked as regression because 977.0313846153846 > 931.9303277575001.
IsChangePoint: Marked as a change because one of 9/15/2023 9:01:53 PM, 9/18/2023 8:11:22 PM falls between 9/9/2023 9:29:58 PM and 9/18/2023 8:11:22 PM.
IsRegressionStdDev: Marked as regression because -5.328563144942166 (T) = (0 -969516193.8227838) / Math.Sqrt((1452233374666260.5 / (27)) + (904757114137304.1 / (10))) is less than -2.0301079282477414 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (27) + (10) - 2, .025) and -0.07067901945616527 = (905515262.9358839 - 969516193.8227838) / 905515262.9358839 is less than -0.05.
IsImprovementBase: Marked as not an improvement because the compare was not 5% less than the baseline, or the value was too small.
IsChangeEdgeDetector: Marked not as a regression because Edge Detector said so.

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
@performanceautofiler performanceautofiler bot added arch-wasm arch-x64 branch-refs/heads/main kind-micro os-linux perf-regression PGO Applied if there were any profile guided optimization updates in the observed interval. untriaged wasm-aot labels Sep 19, 2023
@performanceautofiler performanceautofiler bot mentioned this issue Sep 19, 2023
Closed
@radekdoulik
Copy link
Member

Many of these are unstable and were alternating. Now some stabilized around the longer times. MDPuzzle test might be dotnet/runtime#92154

@stephentoub
Copy link
Member

MDPuzzle test might be dotnet/runtime#92154

That PR is for brand new APIs that aren't used by any perf tests yet, directly or indirectly.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@stephentoub stephentoub

Labels
arch-wasm arch-x64 branch-refs/heads/main kind-micro os-linux perf-regression PGO Applied if there were any profile guided optimization updates in the observed interval. untriaged wasm-aot
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@stephentoub @radekdoulik