[Perf] Linux/x64: 3 Regressions in IndexOf

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	ViperUbuntu
Baseline	fce7c7d1c316b56b90c6246bc4c91180ca6b3c99
Compare	0372b5080a4aec8d73d6861ad4b993d9d0fe3a11
Diff	Diff
Configs	CompilationMode:tiered, RunKind:micro

Regressions in System.Memory.ReadOnlySpan

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
IndexOfString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	18.71 ns	23.59 ns	1.26	0.06	True
IndexOfString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	131.73 ns	144.67 ns	1.10	0.01	False
IndexOfString - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	23.87 ns	28.75 ns	1.20	0.02	False

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.ReadOnlySpan*'

System.Memory.ReadOnlySpan.IndexOfString(input: "string1", value: "string2", comparisonType: InvariantCulture)

ETL Files

Histogram

JIT Disasms

System.Memory.ReadOnlySpan.IndexOfString(input: "AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAXAAAAAAAAAAAAAAAAAAAAAAAAAAAAA", value: "x", comparisonType: InvariantCultureIgnoreCase)

ETL Files

Histogram

JIT Disasms

System.Memory.ReadOnlySpan.IndexOfString(input: "foobardzsdzs", value: "rddzs", comparisonType: InvariantCulture)

ETL Files

Histogram

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	ViperUbuntu
Baseline	fce7c7d1c316b56b90c6246bc4c91180ca6b3c99
Compare	0372b5080a4aec8d73d6861ad4b993d9d0fe3a11
Diff	Diff
Configs	CompilationMode:tiered, RunKind:micro

Regressions in System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Int32>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Max_Vector - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	13.08 ns	14.44 ns	1.10	0.11	False

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.Tensors.Tests.Perf_NumberTensorPrimitives<Int32>*'

System.Numerics.Tensors.Tests.Perf_NumberTensorPrimitives<Int32>.Max_Vector(BufferLength: 128)

ETL Files

Histogram

JIT Disasms

Docs

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

[dotnet-policy-service]

Tagging subscribers to this area: @dotnet/area-system-memory
See info in area-owners.md if you want to be subscribed.

[LoopedBard3]

Likely commit #108499 @tarekgh, commit range 0ab7efe...6160518

[LoopedBard3]

Related regressions:
Windows x64: dotnet/perf-autofiling-issues#42733, dotnet/perf-autofiling-issues#42703
Linux arm64: dotnet/perf-autofiling-issues#42886

[tarekgh]

We have a PR #108709 out to address this issue.

[tarekgh]

This should be addressed now by #108709. Note, we expect some perf cost in specific scenario when searching all ascii string inside all ascii string while failing to find the search string.