[Perf] Windows/x64: Regressions in System.Collections on 8/2/2024 1:16:56 PM

[performanceautofiler]

Run Information

Name	Value
Architecture	x64
OS	Windows 10.0.22621
Queue	TigerWindows
Baseline	30b34e6d7b475c1de91623d3589723181a260523
Compare	3e572e9d09026f386a57302bfd624abd7b660fb5
Diff	Diff
Configs	CompilationMode:tiered, RunKind:micro

Regressions in System.Collections.CtorFromCollection<Int32>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
ConcurrentDictionary - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	25.07 μs	26.45 μs	1.05	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
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Collections.CtorFromCollection<Int32>*'

System.Collections.CtorFromCollection<Int32>.ConcurrentDictionary(Size: 512)

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	Windows 10.0.22621
Queue	TigerWindows
Baseline	30b34e6d7b475c1de91623d3589723181a260523
Compare	3e572e9d09026f386a57302bfd624abd7b660fb5
Diff	Diff
Configs	CompilationMode:tiered, RunKind:micro

Regressions in System.Collections.TryAddGiventSize<Int32>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Dictionary - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	4.32 μs	6.30 μs	1.46	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
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Collections.TryAddGiventSize<Int32>*'

System.Collections.TryAddGiventSize<Int32>.Dictionary(Count: 512)

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	Windows 10.0.22621
Queue	TigerWindows
Baseline	30b34e6d7b475c1de91623d3589723181a260523
Compare	3e572e9d09026f386a57302bfd624abd7b660fb5
Diff	Diff
Configs	CompilationMode:tiered, RunKind:micro

Regressions in System.Collections.TryGetValueFalse<Int32, Int32>

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
Dictionary - Duration of single invocation 📝 - Benchmark Source ADX - Test Multi Config Graph	2.80 μs	6.96 μs	2.48	0.08	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
py .\performance\scripts\benchmarks_ci.py -f net8.0 --filter 'System.Collections.TryGetValueFalse<Int32, Int32>*'

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

ETL Files

Histogram

JIT Disasms

Docs

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

[LoopedBard3]

Related Concurrent dictionary regressions: dotnet/perf-autofiling-issues#39502, dotnet/perf-autofiling-issues#39463

Mono related regressions: dotnet/perf-autofiling-issues#39525

[LoopedBard3]

Likely caused by: #105380
Comparison range: e60db07...fd99ad5

[dotnet-policy-service]

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

[stephentoub]

Likely caused by: #105380

It's possible my change impacted ConcurrentDictionary, but it wouldn't have had any effect on Dictionary, which according to the above graph is showing a huge regression.

[stephentoub]

Re-opening to help track that the regressions are indeed fixed. The Dictionary one is very likely something else.

[eiriktsarpalis]

I see the numbers have recovered in all three benchmarks, so we can close this.