[Perf] Windows/x64: 2 Regressions on 8/17/2024 7:43:54 PM

[performanceautofiler[bot]]

Run Information

Name	Value
Architecture	x64
OS	Windows 10.0.22621
Queue	TigerWindows
Baseline	0d8e5652d86d5d5aec8125dbf336b3c383c1d124
Compare	217525ae6f6a117a0780620ed4fb1b94e03fd4d6
Diff	Diff
Configs	CompilationMode:tiered, RunKind:micro

Regressions in System.Text.Json.Document.Tests.Perf_EnumerateObject

Benchmark	Baseline	Test	Test/Base	Test Quality	Edge Detector	Baseline IR	Compare IR	IR Ratio
[PropertyIndexer - Duration of single invocation](<https://pvscmdupload.z22.web.core.windows.net/reports/allTestHistory/refs/heads/main_x64_Windows 10.0.22621/System.Text.Json.Document.Tests.Perf_EnumerateObject.PropertyIndexer(TestCase%3a%20NumericProperties).html>) 📝 - Benchmark Source ADX - Test Multi Config Graph	2.36 μs	2.52 μs	1.07	0.01	False
[PropertyIndexer - Duration of single invocation](<https://pvscmdupload.z22.web.core.windows.net/reports/allTestHistory/refs/heads/main_x64_Windows 10.0.22621/System.Text.Json.Document.Tests.Perf_EnumerateObject.PropertyIndexer(TestCase%3a%20StringProperties).html>) 📝 - Benchmark Source ADX - Test Multi Config Graph	2.38 μs	2.52 μs	1.06	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.Text.Json.Document.Tests.Perf_EnumerateObject*'

### System.Text.Json.Document.Tests.Perf_EnumerateObject.PropertyIndexer(TestCase: NumericProperties) #### ETL Files #### Histogram #### JIT Disasms ### System.Text.Json.Document.Tests.Perf_EnumerateObject.PropertyIndexer(TestCase: StringProperties) #### ETL Files #### Histogram #### JIT Disasms ### Docs [Profiling workflow for dotnet/runtime repository](https://github.com/dotnet/performance/blob/master/docs/profiling-workflow-dotnet-runtime.md) [Benchmarking workflow for dotnet/runtime repository](https://github.com/dotnet/performance/blob/master/docs/benchmarking-workflow-dotnet-runtime.md)

[DrewScoggins]

Related to https://github.com/dotnet/runtime/pull/106503

[dotnet-policy-service[bot]]

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

[DrewScoggins]

PTAL @etemi

[etemi]

TBH I don't (yet) see how this regression is related to my changes. I will try to reproduce this locally.

[etemi]

• Regression for System.Text.Json.Document.Tests.Perf_EnumerateObject.PropertyIndexer(TestCase: StringProperties) seems to be caused by other changes

• Regression for System.Text.Json.Document.Tests.Perf_EnumerateObject.PropertyIndexer(TestCase: NumericProperties) seems to be caused by my changes. Unfortunately I wasn't able to reproduce it locally. @eiriktsarpalis Could you please take a look too?

[eiriktsarpalis]

I find it unlikely that this is being caused by #106503. It adds a few instructions to the main JsonDocument parsing method that could, theoretically, incur regressions in the micro level however the benchmark itself measures enumeration of a pre-parsed document which exercises a completely disjoint code-path.

The regression concerns main, so I'm updating the milestone to reflect this.