[Perf] Linux/x64: Regressions in System.Collections.Perf_LengthBucketsFrozenDictionary

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	ubuntu 22.04
Queue	TigerUbuntu
Baseline	078530371f31d23945990b47314f11d6e9f1f7c1
Compare	a3fd0957239a6cf44edee68e9bc0a158f15425c9
Diff	Diff
Configs	CompilationMode:tiered, RunKind:micro

Regressions in System.Collections.Perf_LengthBucketsFrozenDictionary

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
TryGetValue_True_FrozenDictionary - Duration of single invocation ADX - Test Multi Config Graph	46.48 μs	50.63 μs	1.09	0.04	False
TryGetValue_False_FrozenDictionary - Duration of single invocation ADX - Test Multi Config Graph	1.13 μs	1.99 μs	1.76	0.22	False
TryGetValue_False_FrozenDictionary - Duration of single invocation ADX - Test Multi Config Graph	100.12 ns	181.76 ns	1.82	0.01	True
TryGetValue_False_FrozenDictionary - Duration of single invocation ADX - Test Multi Config Graph	100.14 ns	180.20 ns	1.80	0.01	True
TryGetValue_False_FrozenDictionary - Duration of single invocation ADX - Test Multi Config Graph	10.08 ns	19.76 ns	1.96	0.03	True
TryGetValue_False_FrozenDictionary - Duration of single invocation ADX - Test Multi Config Graph	1.19 μs	1.98 μs	1.66	0.05	True
TryGetValue_False_FrozenDictionary - Duration of single invocation ADX - Test Multi Config Graph	9.94 ns	20.07 ns	2.02	0.03	True

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.Perf_LengthBucketsFrozenDictionary*'

Details

System.Collections.Perf_LengthBucketsFrozenDictionary.TryGetValue_True_FrozenDictionary(Count: 10000, ItemsPerBucket: 1)

ETL Files

Histogram

JIT Disasms

System.Collections.Perf_LengthBucketsFrozenDictionary.TryGetValue_False_FrozenDictionary(Count: 1000, ItemsPerBucket: 5)

ETL Files

Histogram

JIT Disasms

System.Collections.Perf_LengthBucketsFrozenDictionary.TryGetValue_False_FrozenDictionary(Count: 100, ItemsPerBucket: 1)

ETL Files

Histogram

JIT Disasms

System.Collections.Perf_LengthBucketsFrozenDictionary.TryGetValue_False_FrozenDictionary(Count: 100, ItemsPerBucket: 5)

ETL Files

Histogram

JIT Disasms

System.Collections.Perf_LengthBucketsFrozenDictionary.TryGetValue_False_FrozenDictionary(Count: 10, ItemsPerBucket: 1)

ETL Files

Histogram

JIT Disasms

System.Collections.Perf_LengthBucketsFrozenDictionary.TryGetValue_False_FrozenDictionary(Count: 1000, ItemsPerBucket: 1)

ETL Files

Histogram

JIT Disasms

System.Collections.Perf_LengthBucketsFrozenDictionary.TryGetValue_False_FrozenDictionary(Count: 10, ItemsPerBucket: 5)

ETL Files

Histogram

JIT Disasms

Docs

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

[LoopedBard3]

Likely caused by: #105131, linked improvements in PR.

[LoopedBard3]

Related regressions from throughout the strength reduction story:
Linux/x64: dotnet/perf-autofiling-issues#38697, dotnet/perf-autofiling-issues#38692
Windows/x64: dotnet/perf-autofiling-issues#38716, dotnet/perf-autofiling-issues#38689, dotnet/perf-autofiling-issues#38686
Windows/arm64: dotnet/perf-autofiling-issues#39309

[dotnet-policy-service]

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