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/arm64: 5 Regressions on 1/27/2023 2:26:56 PM #81551

Closed
performanceautofiler bot opened this issue Feb 2, 2023 · 5 comments
Closed

[Perf] Linux/arm64: 5 Regressions on 1/27/2023 2:26:56 PM #81551

performanceautofiler bot opened this issue Feb 2, 2023 · 5 comments
Assignees
@kunalspathak
Labels
arch-arm64 area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI os-linux Linux OS (any supported distro) tenet-performance Performance related issue tenet-performance-benchmarks Issue from performance benchmark
Milestone
8.0.0

Comments

@performanceautofiler
Copy link

performanceautofiler bot commented Feb 2, 2023 •
edited by DrewScoggins
Loading

Run Information

Architecture arm64
OS ubuntu 20.04
Baseline f60d188f16f3843eddc0116ee47a7c0f85a08cb2
Compare ddfaf2750e8dcf127a8baf0a3e5a2bd4835f79c7
Diff Diff

Regressions in System.Hashing

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
GetStringHashCode - Duration of single invocation 4.31 ns 5.93 ns 1.38 0.17 False

graph
Test Report

Repro

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

Payloads

Baseline
Compare

Histogram

System.Hashing.GetStringHashCode(BytesCount: 10)

Description of detection logic

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 5.9349663563034 > 4.493280504769283.
IsChangePoint: Marked as a change because one of 1/27/2023 3:43:27 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -72.33306740231566 (T) = (0 -5.9194456174117756) / Math.Sqrt((0.008107597421865047 / (22)) + (0.0026408891834471923 / (17))) is less than -2.026192463026769 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (22) + (17) - 2, .025) and -0.3882814163973431 = (4.263865774975956 - 5.9194456174117756) / 4.263865774975956 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.

Docs

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

### Run Information
Architecture arm64
OS ubuntu 20.04
Baseline f60d188f16f3843eddc0116ee47a7c0f85a08cb2
Compare ddfaf2750e8dcf127a8baf0a3e5a2bd4835f79c7
Diff Diff

Regressions in System.Globalization.Tests.StringSearch

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
IsPrefix_DifferentFirstChar - Duration of single invocation 7.69 ns 8.81 ns 1.15 0.60 False
IsPrefix_DifferentFirstChar - Duration of single invocation 6.05 ns 8.15 ns 1.35 0.42 False

graph
graph
Test Report

Repro

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

Payloads

Baseline
Compare

Histogram

System.Globalization.Tests.StringSearch.IsPrefix_DifferentFirstChar(Options: (, None, False))

Description of detection logic

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 8.813453643175082 > 7.992300890003956.
IsChangePoint: Marked as a change because one of 11/14/2022 3:36:58 AM, 11/14/2022 7:25:10 PM, 1/27/2023 3:43:27 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -4.055596368892988 (T) = (0 -8.755802452278674) / Math.Sqrt((1.0393319274162365 / (22)) + (0.02257802373945518 / (17))) is less than -2.026192463026769 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (22) + (17) - 2, .025) and -0.11368621572679055 = (7.862001278847333 - 8.755802452278674) / 7.862001278847333 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.

```#### System.Globalization.Tests.StringSearch.IsPrefix_DifferentFirstChar(Options: (en-US, Ordinal, False))

```log

Description of detection logic

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 8.1548878509087 > 6.33305152028561.
IsChangePoint: Marked as a change because one of 12/17/2022 3:28:28 PM, 1/7/2023 2:04:07 AM, 1/31/2023 8:24:06 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -5.485274716996756 (T) = (0 -7.601167359486666) / Math.Sqrt((0.3170710389521432 / (36)) + (0.38969506275245686 / (4))) is less than -2.0243941639098457 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (36) + (4) - 2, .025) and -0.30753786929410126 = (5.813343948187366 - 7.601167359486666) / 5.813343948187366 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.

Docs

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

Run Information

Architecture arm64
OS ubuntu 20.04
Baseline f60d188f16f3843eddc0116ee47a7c0f85a08cb2
Compare ddfaf2750e8dcf127a8baf0a3e5a2bd4835f79c7
Diff Diff

Regressions in System.Collections.IterateForEach<Int32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Dictionary - Duration of single invocation 2.25 μs 2.40 μs 1.07 0.00 False

graph
Test Report

Repro

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

Payloads

Baseline
Compare

Histogram

System.Collections.IterateForEach<Int32>.Dictionary(Size: 512)

Description of detection logic

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 2.3998109775024368 > 2.3598939562945893.
IsChangePoint: Marked as a change because one of 11/14/2022 9:30:20 AM, 1/27/2023 3:43:27 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -27.07311900124489 (T) = (0 -2375.6129219666673) / Math.Sqrt((75.19406458414267 / (22)) + (288.80368289430686 / (18))) is less than -2.0243941639098457 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (22) + (18) - 2, .025) and -0.05293772756643626 = (2256.1760869346144 - 2375.6129219666673) / 2256.1760869346144 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.

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
@performanceautofiler performanceautofiler bot added the untriaged New issue has not been triaged by the area owner label Feb 2, 2023
@DrewScoggins DrewScoggins transferred this issue from dotnet/perf-autofiling-issues Feb 2, 2023
@dotnet-issue-labeler
Copy link

I couldn't figure out the best area label to add to this issue. If you have write-permissions please help me learn by adding exactly one area label.

@DrewScoggins DrewScoggins added arch-arm64 os-linux Linux OS (any supported distro) tenet-performance Performance related issue tenet-performance-benchmarks Issue from performance benchmark labels Feb 2, 2023
@DrewScoggins
Copy link
Member

Seems related to #77540

@jeffschwMSFT jeffschwMSFT added the area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI label Feb 6, 2023
@ghost
Copy link

ghost commented Feb 6, 2023

Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch, @kunalspathak
See info in area-owners.md if you want to be subscribed.

Issue Details

Run Information

Architecture arm64
OS ubuntu 20.04
Baseline f60d188f16f3843eddc0116ee47a7c0f85a08cb2
Compare ddfaf2750e8dcf127a8baf0a3e5a2bd4835f79c7
Diff Diff

Regressions in System.Hashing

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
GetStringHashCode - Duration of single invocation 4.31 ns 5.93 ns 1.38 0.17 False

graph
Test Report

Repro

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

Payloads

Baseline
Compare

Histogram

System.Hashing.GetStringHashCode(BytesCount: 10)

Description of detection logic

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 5.9349663563034 > 4.493280504769283.
IsChangePoint: Marked as a change because one of 1/27/2023 3:43:27 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -72.33306740231566 (T) = (0 -5.9194456174117756) / Math.Sqrt((0.008107597421865047 / (22)) + (0.0026408891834471923 / (17))) is less than -2.026192463026769 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (22) + (17) - 2, .025) and -0.3882814163973431 = (4.263865774975956 - 5.9194456174117756) / 4.263865774975956 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.

Docs

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

### Run Information
Architecture arm64
OS ubuntu 20.04
Baseline f60d188f16f3843eddc0116ee47a7c0f85a08cb2
Compare ddfaf2750e8dcf127a8baf0a3e5a2bd4835f79c7
Diff Diff

Regressions in System.Globalization.Tests.StringSearch

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
IsPrefix_DifferentFirstChar - Duration of single invocation 7.69 ns 8.81 ns 1.15 0.60 False
IsPrefix_DifferentFirstChar - Duration of single invocation 6.05 ns 8.15 ns 1.35 0.42 False

graph
graph
Test Report

Repro

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

Payloads

Baseline
Compare

Histogram

System.Globalization.Tests.StringSearch.IsPrefix_DifferentFirstChar(Options: (, None, False))

Description of detection logic

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 8.813453643175082 > 7.992300890003956.
IsChangePoint: Marked as a change because one of 11/14/2022 3:36:58 AM, 11/14/2022 7:25:10 PM, 1/27/2023 3:43:27 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -4.055596368892988 (T) = (0 -8.755802452278674) / Math.Sqrt((1.0393319274162365 / (22)) + (0.02257802373945518 / (17))) is less than -2.026192463026769 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (22) + (17) - 2, .025) and -0.11368621572679055 = (7.862001278847333 - 8.755802452278674) / 7.862001278847333 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.

```#### System.Globalization.Tests.StringSearch.IsPrefix_DifferentFirstChar(Options: (en-US, Ordinal, False))

```log

Description of detection logic

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 8.1548878509087 > 6.33305152028561.
IsChangePoint: Marked as a change because one of 12/17/2022 3:28:28 PM, 1/7/2023 2:04:07 AM, 1/31/2023 8:24:06 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -5.485274716996756 (T) = (0 -7.601167359486666) / Math.Sqrt((0.3170710389521432 / (36)) + (0.38969506275245686 / (4))) is less than -2.0243941639098457 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (36) + (4) - 2, .025) and -0.30753786929410126 = (5.813343948187366 - 7.601167359486666) / 5.813343948187366 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.

Docs

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

Run Information

Architecture arm64
OS ubuntu 20.04
Baseline f60d188f16f3843eddc0116ee47a7c0f85a08cb2
Compare ddfaf2750e8dcf127a8baf0a3e5a2bd4835f79c7
Diff Diff

Regressions in System.Collections.IterateForEach<Int32>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio Baseline ETL Compare ETL
Dictionary - Duration of single invocation 2.25 μs 2.40 μs 1.07 0.00 False

graph
Test Report

Repro

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

Payloads

Baseline
Compare

Histogram

System.Collections.IterateForEach<Int32>.Dictionary(Size: 512)

Description of detection logic

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 2.3998109775024368 > 2.3598939562945893.
IsChangePoint: Marked as a change because one of 11/14/2022 9:30:20 AM, 1/27/2023 3:43:27 AM, 2/2/2023 2:17:49 AM falls between 1/24/2023 11:54:43 AM and 2/2/2023 2:17:49 AM.
IsRegressionStdDev: Marked as regression because -27.07311900124489 (T) = (0 -2375.6129219666673) / Math.Sqrt((75.19406458414267 / (22)) + (288.80368289430686 / (18))) is less than -2.0243941639098457 = MathNet.Numerics.Distributions.StudentT.InvCDF(0, 1, (22) + (18) - 2, .025) and -0.05293772756643626 = (2256.1760869346144 - 2375.6129219666673) / 2256.1760869346144 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.

Docs

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

Author: performanceautofiler[bot]
Assignees: -
Labels:

arch-arm64, os-linux, tenet-performance, tenet-performance-benchmarks, area-CodeGen-coreclr, untriaged
Milestone: -

@JulieLeeMSFT JulieLeeMSFT mentioned this issue Feb 6, 2023
Merged
@JulieLeeMSFT JulieLeeMSFT removed the untriaged New issue has not been triaged by the area owner label Feb 6, 2023
@JulieLeeMSFT JulieLeeMSFT added this to the 8.0.0 milestone Feb 6, 2023
@JulieLeeMSFT
Copy link
Member

Assigned to @kunalspathak to work with AndyJGraham.

@kunalspathak
Copy link
Member

Most of the regressions came back without us fixing any of them, which tells that the benchmark can be prone to memory alignment. #77540 (comment)

@ghost ghost locked as resolved and limited conversation to collaborators Mar 18, 2023
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees

@kunalspathak kunalspathak

Labels
arch-arm64 area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI os-linux Linux OS (any supported distro) tenet-performance Performance related issue tenet-performance-benchmarks Issue from performance benchmark
Projects
None yet
Milestone
8.0.0
Development

No branches or pull requests
5 participants
@DrewScoggins @jeffschwMSFT @AndyAyersMS @kunalspathak @JulieLeeMSFT