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: 32 Regressions on 4/16/2024 12:58:48 AM #33133

Closed
performanceautofiler bot opened this issue Apr 23, 2024 · 13 comments
Closed

[Perf] Linux/arm64: 32 Regressions on 4/16/2024 12:58:48 AM #33133

performanceautofiler bot opened this issue Apr 23, 2024 · 13 comments
Labels
ampere arch-arm64 branch-refs/heads/main kind-micro_mono LinkedFromMaster Is this issue linked from the master perf issue mono-aot mono-llvm os-linux perf-regression runtime-mono

Comments

@performanceautofiler
Copy link

performanceautofiler bot commented Apr 23, 2024 •
edited
Loading

Run Information

Name Value
Architecture arm64
OS ubuntu 22.04
Queue AmpereUbuntu
Baseline 5aee4f43c9389239c72d3931a4b2bd5bf123243b
Compare f8afaea32baa10207e6d4299b1b13b46b16af556
Diff Diff
Configs CompilationMode:tiered, LLVM:true, MonoAOT:true, MonoInterpreter:false, RunKind:micro_mono

Regressions in System.IO.Tests.Perf_FileStream

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
739.77 ns 831.35 ns 1.12 0.73 False

graph
Test Report

Repro

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

Repro Steps

Prerequisites (Files either built locally (with build.(sh/cmd) or downloaded from payload above (if same system setup) (in this order))

  • Libraries build extracted to runtime/artifacts or build instructions: Libraries README args: -subset libs+libs.tests -rc release -configuration Release -arch $RunArch -framework net8.0
  • CoreCLR product build extracted to runtime/artifacts/bin/coreclr/$RunOS.$RunArch.Release, build instructions: CoreCLR README args: -subset clr+libs -rc release -configuration Release -arch $RunArch -framework net8.0
  • AOT MONO build extracted to runtime/artifacts/bin/mono/$RunOS.$RunArch.Release, build instructions: MONO README args: -arch $RunArch -os $RunOS -s mono+libs+host+packs -c Release /p:CrossBuild=false /p:MonoLLVMUseCxx11Abi=false
  • Dotnet SDK installed for dotnet commands
  • Running commands from the runtime folder

Linux

# Set $RunDir to the runtime directory
RunDir=`pwd`

# Set the OS, arch, and OSId
RunOS='linux'
RunOSId='linux'
RunArch='x64'

# Create aot directory 
mkdir -p $RunDir/artifacts/bin/aot/sgen
mkdir -p $RunDir/artifacts/bin/aot/pack
cp -r $RunDir/artifacts/obj/mono/$RunOS.$RunArch.Release/mono/* $RunDir/artifacts/bin/aot/sgen
cp -r $RunDir/artifacts/bin/microsoft.netcore.app.runtime.$RunOS-$RunArch/Release/* $RunDir/artifacts/bin/aot/pack

# Create Core Root
$RunDir/src/tests/build.sh release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir/performance

# One line run:
python3 $RunDir/performance/scripts/benchmarks_ci.py --csproj $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.IO.Tests.Perf_FileStream*' --bdn-artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --packages $RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.IO.Tests.Perf_FileStream* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --packages $RunDir/performance/artifacts/packages --buildTimeout 1200

Windows

# Set $RunDir to the runtime directory
$RunDir="FullPathHere"

# Set the OS, arch, and OSId
RunOS='windows'
RunOSId='win'
RunArch='x64'

# Create aot directory
mkdir $RunDir\artifacts\bin\aot\sgen
mkdir $RunDir\artifacts\bin\aot\pack
xcopy $RunDir\artifacts\obj\mono\$RunOS.$RunArch.Release\mono $RunDir\artifacts\bin\aot\sgen\ /e /y
xcopy $RunDir\artifacts\bin\microsoft.netcore.app.runtime.$RunOSId-$RunArch\Release $RunDir\artifacts\bin\aot\pack\ /e /y

# Create Core Root
$RunDir\src\tests\build.cmd release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir\performance

# One line run:
python3 $RunDir\performance\scripts\benchmarks_ci.py --csproj $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.IO.Tests.Perf_FileStream*' --bdn-artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --packages $RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.IO.Tests.Perf_FileStream* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack -aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --packages $RunDir\performance\artifacts\packages --buildTimeout 1200

System.IO.Tests.Perf_FileStream.GetLength(options: Read)

ETL Files

Histogram

JIT Disasms

Docs

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

Run Information

Name Value
Architecture arm64
OS ubuntu 22.04
Queue AmpereUbuntu
Baseline 5aee4f43c9389239c72d3931a4b2bd5bf123243b
Compare f8afaea32baa10207e6d4299b1b13b46b16af556
Diff Diff
Configs CompilationMode:tiered, LLVM:true, MonoAOT:true, MonoInterpreter:false, RunKind:micro_mono

Regressions in System.Memory.Span<Byte>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
12.16 ns 14.18 ns 1.17 0.05 True
16.73 ns 18.19 ns 1.09 0.01 True
14.83 ns 16.35 ns 1.10 0.05 False
14.90 ns 16.34 ns 1.10 0.06 False
11.50 ns 13.07 ns 1.14 0.03 True
23.05 ns 24.50 ns 1.06 0.00 True

graph
graph
graph
graph
graph
graph
Test Report

Repro

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

Repro Steps

Prerequisites (Files either built locally (with build.(sh/cmd) or downloaded from payload above (if same system setup) (in this order))

  • Libraries build extracted to runtime/artifacts or build instructions: Libraries README args: -subset libs+libs.tests -rc release -configuration Release -arch $RunArch -framework net8.0
  • CoreCLR product build extracted to runtime/artifacts/bin/coreclr/$RunOS.$RunArch.Release, build instructions: CoreCLR README args: -subset clr+libs -rc release -configuration Release -arch $RunArch -framework net8.0
  • AOT MONO build extracted to runtime/artifacts/bin/mono/$RunOS.$RunArch.Release, build instructions: MONO README args: -arch $RunArch -os $RunOS -s mono+libs+host+packs -c Release /p:CrossBuild=false /p:MonoLLVMUseCxx11Abi=false
  • Dotnet SDK installed for dotnet commands
  • Running commands from the runtime folder

Linux

# Set $RunDir to the runtime directory
RunDir=`pwd`

# Set the OS, arch, and OSId
RunOS='linux'
RunOSId='linux'
RunArch='x64'

# Create aot directory 
mkdir -p $RunDir/artifacts/bin/aot/sgen
mkdir -p $RunDir/artifacts/bin/aot/pack
cp -r $RunDir/artifacts/obj/mono/$RunOS.$RunArch.Release/mono/* $RunDir/artifacts/bin/aot/sgen
cp -r $RunDir/artifacts/bin/microsoft.netcore.app.runtime.$RunOS-$RunArch/Release/* $RunDir/artifacts/bin/aot/pack

# Create Core Root
$RunDir/src/tests/build.sh release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir/performance

# One line run:
python3 $RunDir/performance/scripts/benchmarks_ci.py --csproj $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Memory.Span&lt;Byte&gt;*' --bdn-artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --packages $RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Memory.Span&lt;Byte&gt;* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --packages $RunDir/performance/artifacts/packages --buildTimeout 1200

Windows

# Set $RunDir to the runtime directory
$RunDir="FullPathHere"

# Set the OS, arch, and OSId
RunOS='windows'
RunOSId='win'
RunArch='x64'

# Create aot directory
mkdir $RunDir\artifacts\bin\aot\sgen
mkdir $RunDir\artifacts\bin\aot\pack
xcopy $RunDir\artifacts\obj\mono\$RunOS.$RunArch.Release\mono $RunDir\artifacts\bin\aot\sgen\ /e /y
xcopy $RunDir\artifacts\bin\microsoft.netcore.app.runtime.$RunOSId-$RunArch\Release $RunDir\artifacts\bin\aot\pack\ /e /y

# Create Core Root
$RunDir\src\tests\build.cmd release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir\performance

# One line run:
python3 $RunDir\performance\scripts\benchmarks_ci.py --csproj $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Memory.Span&lt;Byte&gt;*' --bdn-artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --packages $RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Memory.Span&lt;Byte&gt;* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack -aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --packages $RunDir\performance\artifacts\packages --buildTimeout 1200

System.Memory.Span<Byte>.IndexOfValue(Size: 4)

ETL Files

Histogram

JIT Disasms

System.Memory.Span<Byte>.IndexOfAnyThreeValues(Size: 4)

ETL Files

Histogram

JIT Disasms

System.Memory.Span<Byte>.IndexOfAnyTwoValues(Size: 4)

ETL Files

Histogram

JIT Disasms

System.Memory.Span<Byte>.LastIndexOfAnyValues(Size: 4)

ETL Files

Histogram

JIT Disasms

System.Memory.Span<Byte>.LastIndexOfValue(Size: 4)

ETL Files

Histogram

JIT Disasms

System.Memory.Span<Byte>.IndexOfAnyTwoValues(Size: 33)

ETL Files

Histogram

JIT Disasms

Docs

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

Run Information

Name Value
Architecture arm64
OS ubuntu 22.04
Queue AmpereUbuntu
Baseline 5aee4f43c9389239c72d3931a4b2bd5bf123243b
Compare f8afaea32baa10207e6d4299b1b13b46b16af556
Diff Diff
Configs CompilationMode:tiered, LLVM:true, MonoAOT:true, MonoInterpreter:false, RunKind:micro_mono

Regressions in System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
8.08 μs 8.55 μs 1.06 0.01 True
3.90 μs 4.14 μs 1.06 0.01 True
83.73 μs 90.12 μs 1.08 0.01 True
2.91 μs 3.08 μs 1.06 0.01 True
90.52 μs 96.60 μs 1.07 0.01 True
3.62 μs 3.88 μs 1.07 0.01 True
187.27 μs 198.02 μs 1.06 0.01 True
67.41 μs 72.13 μs 1.07 0.01 True
189.13 μs 199.37 μs 1.05 0.00 True

graph
graph
graph
graph
graph
graph
graph
graph
graph
Test Report

Repro

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

Repro Steps

Prerequisites (Files either built locally (with build.(sh/cmd) or downloaded from payload above (if same system setup) (in this order))

  • Libraries build extracted to runtime/artifacts or build instructions: Libraries README args: -subset libs+libs.tests -rc release -configuration Release -arch $RunArch -framework net8.0
  • CoreCLR product build extracted to runtime/artifacts/bin/coreclr/$RunOS.$RunArch.Release, build instructions: CoreCLR README args: -subset clr+libs -rc release -configuration Release -arch $RunArch -framework net8.0
  • AOT MONO build extracted to runtime/artifacts/bin/mono/$RunOS.$RunArch.Release, build instructions: MONO README args: -arch $RunArch -os $RunOS -s mono+libs+host+packs -c Release /p:CrossBuild=false /p:MonoLLVMUseCxx11Abi=false
  • Dotnet SDK installed for dotnet commands
  • Running commands from the runtime folder

Linux

# Set $RunDir to the runtime directory
RunDir=`pwd`

# Set the OS, arch, and OSId
RunOS='linux'
RunOSId='linux'
RunArch='x64'

# Create aot directory 
mkdir -p $RunDir/artifacts/bin/aot/sgen
mkdir -p $RunDir/artifacts/bin/aot/pack
cp -r $RunDir/artifacts/obj/mono/$RunOS.$RunArch.Release/mono/* $RunDir/artifacts/bin/aot/sgen
cp -r $RunDir/artifacts/bin/microsoft.netcore.app.runtime.$RunOS-$RunArch/Release/* $RunDir/artifacts/bin/aot/pack

# Create Core Root
$RunDir/src/tests/build.sh release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir/performance

# One line run:
python3 $RunDir/performance/scripts/benchmarks_ci.py --csproj $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Double&gt;*' --bdn-artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --packages $RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Double&gt;* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --packages $RunDir/performance/artifacts/packages --buildTimeout 1200

Windows

# Set $RunDir to the runtime directory
$RunDir="FullPathHere"

# Set the OS, arch, and OSId
RunOS='windows'
RunOSId='win'
RunArch='x64'

# Create aot directory
mkdir $RunDir\artifacts\bin\aot\sgen
mkdir $RunDir\artifacts\bin\aot\pack
xcopy $RunDir\artifacts\obj\mono\$RunOS.$RunArch.Release\mono $RunDir\artifacts\bin\aot\sgen\ /e /y
xcopy $RunDir\artifacts\bin\microsoft.netcore.app.runtime.$RunOSId-$RunArch\Release $RunDir\artifacts\bin\aot\pack\ /e /y

# Create Core Root
$RunDir\src\tests\build.cmd release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir\performance

# One line run:
python3 $RunDir\performance\scripts\benchmarks_ci.py --csproj $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Double&gt;*' --bdn-artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --packages $RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Double&gt;* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack -aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --packages $RunDir\performance\artifacts\packages --buildTimeout 1200

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Pow_ScalarExponent(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Log(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Sigmoid(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Exp(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Log(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Sigmoid(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Pow_ScalarExponent(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Exp(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Double>.Pow_Vectors(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

Docs

Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
@performanceautofiler PerformanceAutoFiler
Copy link
Author

performanceautofiler bot commented Apr 23, 2024 •
edited
Loading

Run Information

Name Value
Architecture arm64
OS ubuntu 22.04
Queue AmpereUbuntu
Baseline 5aee4f43c9389239c72d3931a4b2bd5bf123243b
Compare f8afaea32baa10207e6d4299b1b13b46b16af556
Diff Diff
Configs CompilationMode:tiered, LLVM:true, MonoAOT:true, MonoInterpreter:false, RunKind:micro_mono

Regressions in System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
141.53 μs 155.94 μs 1.10 0.01 True
134.05 μs 153.01 μs 1.14 0.01 True
74.35 μs 83.20 μs 1.12 0.01 True
1.87 μs 2.06 μs 1.10 0.01 True
5.74 μs 6.51 μs 1.14 0.01 True
59.56 μs 67.17 μs 1.13 0.01 True
2.72 μs 3.01 μs 1.11 0.01 True
41.33 μs 45.29 μs 1.10 0.01 True
69.92 μs 78.60 μs 1.12 0.01 True
143.94 μs 159.15 μs 1.11 0.01 True
3.13 μs 3.52 μs 1.12 0.01 True
6.19 μs 6.77 μs 1.09 0.02 True
5.88 μs 6.74 μs 1.15 0.01 True
3.26 μs 3.60 μs 1.10 0.01 True

graph
graph
graph
graph
graph
graph
graph
graph
graph
graph
graph
graph
graph
graph
Test Report

Repro

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

Repro Steps

Prerequisites (Files either built locally (with build.(sh/cmd) or downloaded from payload above (if same system setup) (in this order))

  • Libraries build extracted to runtime/artifacts or build instructions: Libraries README args: -subset libs+libs.tests -rc release -configuration Release -arch $RunArch -framework net8.0
  • CoreCLR product build extracted to runtime/artifacts/bin/coreclr/$RunOS.$RunArch.Release, build instructions: CoreCLR README args: -subset clr+libs -rc release -configuration Release -arch $RunArch -framework net8.0
  • AOT MONO build extracted to runtime/artifacts/bin/mono/$RunOS.$RunArch.Release, build instructions: MONO README args: -arch $RunArch -os $RunOS -s mono+libs+host+packs -c Release /p:CrossBuild=false /p:MonoLLVMUseCxx11Abi=false
  • Dotnet SDK installed for dotnet commands
  • Running commands from the runtime folder

Linux

# Set $RunDir to the runtime directory
RunDir=`pwd`

# Set the OS, arch, and OSId
RunOS='linux'
RunOSId='linux'
RunArch='x64'

# Create aot directory 
mkdir -p $RunDir/artifacts/bin/aot/sgen
mkdir -p $RunDir/artifacts/bin/aot/pack
cp -r $RunDir/artifacts/obj/mono/$RunOS.$RunArch.Release/mono/* $RunDir/artifacts/bin/aot/sgen
cp -r $RunDir/artifacts/bin/microsoft.netcore.app.runtime.$RunOS-$RunArch/Release/* $RunDir/artifacts/bin/aot/pack

# Create Core Root
$RunDir/src/tests/build.sh release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir/performance

# One line run:
python3 $RunDir/performance/scripts/benchmarks_ci.py --csproj $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Single&gt;*' --bdn-artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --packages $RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Single&gt;* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --packages $RunDir/performance/artifacts/packages --buildTimeout 1200

Windows

# Set $RunDir to the runtime directory
$RunDir="FullPathHere"

# Set the OS, arch, and OSId
RunOS='windows'
RunOSId='win'
RunArch='x64'

# Create aot directory
mkdir $RunDir\artifacts\bin\aot\sgen
mkdir $RunDir\artifacts\bin\aot\pack
xcopy $RunDir\artifacts\obj\mono\$RunOS.$RunArch.Release\mono $RunDir\artifacts\bin\aot\sgen\ /e /y
xcopy $RunDir\artifacts\bin\microsoft.netcore.app.runtime.$RunOSId-$RunArch\Release $RunDir\artifacts\bin\aot\pack\ /e /y

# Create Core Root
$RunDir\src\tests\build.cmd release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir\performance

# One line run:
python3 $RunDir\performance\scripts\benchmarks_ci.py --csproj $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Single&gt;*' --bdn-artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --packages $RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives&lt;Single&gt;* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack -aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --packages $RunDir\performance\artifacts\packages --buildTimeout 1200

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Pow_Vectors(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Pow_ScalarExponent(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Sigmoid(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Log(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Pow_ScalarExponent(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Exp(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Exp(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Log(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Sinh(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Pow_ScalarBase(BufferLength: 3079)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Sinh(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Pow_ScalarBase(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Pow_Vectors(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

System.Numerics.Tensors.Tests.Perf_FloatingPointTensorPrimitives<Single>.Sigmoid(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

Docs

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

Run Information

Name Value
Architecture arm64
OS ubuntu 22.04
Queue AmpereUbuntu
Baseline 5aee4f43c9389239c72d3931a4b2bd5bf123243b
Compare f8afaea32baa10207e6d4299b1b13b46b16af556
Diff Diff
Configs CompilationMode:tiered, LLVM:true, MonoAOT:true, MonoInterpreter:false, RunKind:micro_mono

Regressions in System.Numerics.Tests.Perf_Vector3

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
29.11 ns 35.95 ns 1.23 0.18 False

graph
Test Report

Repro

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

Repro Steps

Prerequisites (Files either built locally (with build.(sh/cmd) or downloaded from payload above (if same system setup) (in this order))

  • Libraries build extracted to runtime/artifacts or build instructions: Libraries README args: -subset libs+libs.tests -rc release -configuration Release -arch $RunArch -framework net8.0
  • CoreCLR product build extracted to runtime/artifacts/bin/coreclr/$RunOS.$RunArch.Release, build instructions: CoreCLR README args: -subset clr+libs -rc release -configuration Release -arch $RunArch -framework net8.0
  • AOT MONO build extracted to runtime/artifacts/bin/mono/$RunOS.$RunArch.Release, build instructions: MONO README args: -arch $RunArch -os $RunOS -s mono+libs+host+packs -c Release /p:CrossBuild=false /p:MonoLLVMUseCxx11Abi=false
  • Dotnet SDK installed for dotnet commands
  • Running commands from the runtime folder

Linux

# Set $RunDir to the runtime directory
RunDir=`pwd`

# Set the OS, arch, and OSId
RunOS='linux'
RunOSId='linux'
RunArch='x64'

# Create aot directory 
mkdir -p $RunDir/artifacts/bin/aot/sgen
mkdir -p $RunDir/artifacts/bin/aot/pack
cp -r $RunDir/artifacts/obj/mono/$RunOS.$RunArch.Release/mono/* $RunDir/artifacts/bin/aot/sgen
cp -r $RunDir/artifacts/bin/microsoft.netcore.app.runtime.$RunOS-$RunArch/Release/* $RunDir/artifacts/bin/aot/pack

# Create Core Root
$RunDir/src/tests/build.sh release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir/performance

# One line run:
python3 $RunDir/performance/scripts/benchmarks_ci.py --csproj $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tests.Perf_Vector3*' --bdn-artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --packages $RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tests.Perf_Vector3* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --packages $RunDir/performance/artifacts/packages --buildTimeout 1200

Windows

# Set $RunDir to the runtime directory
$RunDir="FullPathHere"

# Set the OS, arch, and OSId
RunOS='windows'
RunOSId='win'
RunArch='x64'

# Create aot directory
mkdir $RunDir\artifacts\bin\aot\sgen
mkdir $RunDir\artifacts\bin\aot\pack
xcopy $RunDir\artifacts\obj\mono\$RunOS.$RunArch.Release\mono $RunDir\artifacts\bin\aot\sgen\ /e /y
xcopy $RunDir\artifacts\bin\microsoft.netcore.app.runtime.$RunOSId-$RunArch\Release $RunDir\artifacts\bin\aot\pack\ /e /y

# Create Core Root
$RunDir\src\tests\build.cmd release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir\performance

# One line run:
python3 $RunDir\performance\scripts\benchmarks_ci.py --csproj $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tests.Perf_Vector3*' --bdn-artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --packages $RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tests.Perf_Vector3* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack -aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --packages $RunDir\performance\artifacts\packages --buildTimeout 1200

System.Numerics.Tests.Perf_Vector3.GetHashCodeBenchmark

ETL Files

Histogram

JIT Disasms

Docs

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

Run Information

Name Value
Architecture arm64
OS ubuntu 22.04
Queue AmpereUbuntu
Baseline 5aee4f43c9389239c72d3931a4b2bd5bf123243b
Compare f8afaea32baa10207e6d4299b1b13b46b16af556
Diff Diff
Configs CompilationMode:tiered, LLVM:true, MonoAOT:true, MonoInterpreter:false, RunKind:micro_mono

Regressions in System.Numerics.Tests.Perf_Vector4

Benchmark Baseline Test Test/Base Test Quality Edge Detector Baseline IR Compare IR IR Ratio
37.27 ns 51.84 ns 1.39 0.17 False

graph
Test Report

Repro

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

Repro Steps

Prerequisites (Files either built locally (with build.(sh/cmd) or downloaded from payload above (if same system setup) (in this order))

  • Libraries build extracted to runtime/artifacts or build instructions: Libraries README args: -subset libs+libs.tests -rc release -configuration Release -arch $RunArch -framework net8.0
  • CoreCLR product build extracted to runtime/artifacts/bin/coreclr/$RunOS.$RunArch.Release, build instructions: CoreCLR README args: -subset clr+libs -rc release -configuration Release -arch $RunArch -framework net8.0
  • AOT MONO build extracted to runtime/artifacts/bin/mono/$RunOS.$RunArch.Release, build instructions: MONO README args: -arch $RunArch -os $RunOS -s mono+libs+host+packs -c Release /p:CrossBuild=false /p:MonoLLVMUseCxx11Abi=false
  • Dotnet SDK installed for dotnet commands
  • Running commands from the runtime folder

Linux

# Set $RunDir to the runtime directory
RunDir=`pwd`

# Set the OS, arch, and OSId
RunOS='linux'
RunOSId='linux'
RunArch='x64'

# Create aot directory 
mkdir -p $RunDir/artifacts/bin/aot/sgen
mkdir -p $RunDir/artifacts/bin/aot/pack
cp -r $RunDir/artifacts/obj/mono/$RunOS.$RunArch.Release/mono/* $RunDir/artifacts/bin/aot/sgen
cp -r $RunDir/artifacts/bin/microsoft.netcore.app.runtime.$RunOS-$RunArch/Release/* $RunDir/artifacts/bin/aot/pack

# Create Core Root
$RunDir/src/tests/build.sh release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir/performance

# One line run:
python3 $RunDir/performance/scripts/benchmarks_ci.py --csproj $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tests.Perf_Vector4*' --bdn-artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --packages $RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir/performance/artifacts/packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir/performance/src/benchmarks/micro/MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tests.Perf_Vector4* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir/artifacts/bin/aot/sgen/mini/mono-sgen --customruntimepack $RunDir/artifacts/bin/aot/pack --aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir/artifacts/BenchmarkDotNet.Artifacts --packages $RunDir/performance/artifacts/packages --buildTimeout 1200

Windows

# Set $RunDir to the runtime directory
$RunDir="FullPathHere"

# Set the OS, arch, and OSId
RunOS='windows'
RunOSId='win'
RunArch='x64'

# Create aot directory
mkdir $RunDir\artifacts\bin\aot\sgen
mkdir $RunDir\artifacts\bin\aot\pack
xcopy $RunDir\artifacts\obj\mono\$RunOS.$RunArch.Release\mono $RunDir\artifacts\bin\aot\sgen\ /e /y
xcopy $RunDir\artifacts\bin\microsoft.netcore.app.runtime.$RunOSId-$RunArch\Release $RunDir\artifacts\bin\aot\pack\ /e /y

# Create Core Root
$RunDir\src\tests\build.cmd release $RunArch generatelayoutonly /p:LibrariesConfiguration=Release

# Clone performance 
git clone --branch main --depth 1 --quiet https://github.com/dotnet/performance.git $RunDir\performance

# One line run:
python3 $RunDir\performance\scripts\benchmarks_ci.py --csproj $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --incremental no --architecture $RunArch -f net8.0 --filter 'System.Numerics.Tests.Perf_Vector4*' --bdn-artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --bdn-arguments="--anyCategories Libraries Runtime  --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack --aotcompilermode llvm --logBuildOutput --generateBinLog"

# Individual Commands:
# Restore 
dotnet restore $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --packages $RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Build
dotnet build $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore /p:NuGetPackageRoot=$RunDir\performance\artifacts\packages /p:UseSharedCompilation=false /p:BuildInParallel=false /m:1

# Run
dotnet run --project $RunDir\performance\src\benchmarks\micro\MicroBenchmarks.csproj --configuration Release --framework net8.0 --no-restore --no-build -- --filter System.Numerics.Tests.Perf_Vector4* --anyCategories Libraries Runtime " --category-exclusion-filter NoAOT NoWASM --runtimes monoaotllvm --aotcompilerpath $RunDir\artifacts\bin\aot\sgen\mini\mono-sgen.exe --customruntimepack $RunDir\artifacts\bin\aot\pack -aotcompilermode llvm --logBuildOutput --generateBinLog " --artifacts $RunDir\artifacts\BenchmarkDotNet.Artifacts --packages $RunDir\performance\artifacts\packages --buildTimeout 1200

System.Numerics.Tests.Perf_Vector4.GetHashCodeBenchmark

ETL Files

Histogram

JIT Disasms

Docs

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

@performanceautofiler performanceautofiler bot mentioned this issue Apr 23, 2024
Closed
@matouskozak
Copy link
Member

The range of regression is dotnet/runtime@dd7c82e...08426cd, could it be dotnet/runtime#100842 @MihaZupan ?

@matouskozak matouskozak added LinkedFromMaster Is this issue linked from the master perf issue and removed untriaged labels Apr 24, 2024
@matouskozak matouskozak mentioned this issue Apr 24, 2024
Open
@MihaZupan
Copy link
Member

For the IndexOf benchmarks in Span<Byte> possibly yes, if Mono doesn't handle Unsafe.BitCast as well as Unsafe.As.

For the rest (tensors, hashcode, ...) it's something else.

@matouskozak
Copy link
Member

For the IndexOf benchmarks in Span<Byte> possibly yes, if Mono doesn't handle Unsafe.BitCast as well as Unsafe.As.

For the rest (tensors, hashcode, ...) it's something else.

Thank you, could it be missing intrinsics for Unsafe.BitCast on Mono side? I believe we already hit that in the past @fanyang-mono

@matouskozak
Copy link
Member

matouskozak commented Apr 24, 2024 •
edited
Loading

The tensors and hashcode regressions (+ x64 regressions #33094) could be related to dotnet/runtime#100386. Do you think it could be related @kg ?

@matouskozak matouskozak mentioned this issue Apr 24, 2024
Closed
@fanyang-mono
Copy link
Member

Yeah, Mono intrinsifies Unsafe.As but not Unsafe.BitCast. The code lives here: https://github.com/dotnet/runtime/blob/main/src/mono/mono/mini/intrinsics.c#L396

@fanyang-mono
Copy link
Member

I created an issue to tracking the Unsafe API's which weren't intrinsified by Mono yet. dotnet/runtime#101495

@kg
Copy link
Member

kg commented Apr 24, 2024

The tensors and hashcode regressions (+ x64 regressions #33094) could be related to dotnet/runtime#100386. Do you think it could be related @kg ?

No, it only modified classname lookup.

@matouskozak
Copy link
Member

matouskozak commented Apr 30, 2024 •
edited
Loading

For the IndexOf benchmarks in Span<Byte> possibly yes, if Mono doesn't handle Unsafe.BitCast as well as Unsafe.As.

For the rest (tensors, hashcode, ...) it's something else.

@MihaZupan @fanyang-mono I did some local investigation and it seems that the dotnet/runtime#100842 is affecting more scenarios that originally expected.

I was using the local benchmarking script and the following command: python benchmarks_local.py --commits 3c10707ce7e91cd0843e7c7c292b9a9fb41fe44b 617f81f3145ad0829e38101b4abc6c98b201542a --run-types MonoAOTLLVM --filter "*GetHashCodeBenchmark*" and the output logs show clear regression in GetHashCodeBenchmark microbenchmarks after dotnet/runtime@617f81f.

The same can be observe for the Tensor microbenchmarks.

@MihaZupan
Copy link
Member

Looking at System.Numerics.Tests.Perf_Vector3.GetHashCodeBenchmark, the implementation is

public int GetHashCodeBenchmark() => VectorTests.Vector3Value.GetHashCode();

where Vector3.GetHashCode() is

public override readonly int GetHashCode() => HashCode.Combine(X, Y, Z);

where HashCode.Combine<T1, T2, T3> calls GetHashCode on every input.

public static int Combine<T1, T2, T3>(T1 value1, T2 value2, T3 value3)
{
    uint hc1 = (uint)(value1?.GetHashCode() ?? 0);
    uint hc2 = (uint)(value2?.GetHashCode() ?? 0);
    uint hc3 = (uint)(value3?.GetHashCode() ?? 0);
    // ...

where float.GetHashCode is

public override int GetHashCode()
{
    uint bits = BitConverter.SingleToUInt32Bits(m_value);
    // ...

where SingleToUInt32Bits is

public static unsafe uint SingleToUInt32Bits(float value) => Unsafe.BitCast<float, uint>(value);

So... Unsafe.BitCast again :)
Thank you for double checking @matouskozak

dotnet/runtime@617f81f changed the (non-intrinsic) implementation of BitCast like so:

public static TTo BitCast<TFrom, TTo>(TFrom source)
{
-   if (sizeof(TFrom) != sizeof(TTo))
+   if (sizeof(TFrom) != sizeof(TTo) || default(TFrom) is null || default(TTo) is null)
    {
        ThrowHelper.ThrowNotSupportedException();
    }
    return ReadUnaligned<TTo>(ref As<TFrom, byte>(ref source));
}

I'm guessing Mono isn't able to eliminate these type checks even for known types @fanyang-mono?

@matouskozak matouskozak mentioned this issue Apr 30, 2024
Closed
@fanyang-mono
Copy link
Member

Mono doesn't emit hardware intrinsics for Unsafe.BitCast yet. I've opened an issue to track that.

@jkurdek
Copy link
Member

jkurdek commented May 22, 2024

Similar regressions occured on Mono JIT: #33184, #33185, #33187

@matouskozak
Copy link
Member

Should be fixed by dotnet/runtime#105150 @fanyang-mono !

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
ampere arch-arm64 branch-refs/heads/main kind-micro_mono LinkedFromMaster Is this issue linked from the master perf issue mono-aot mono-llvm os-linux perf-regression runtime-mono
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
5 participants
@kg @MihaZupan @fanyang-mono @matouskozak @jkurdek