-
Notifications
You must be signed in to change notification settings - Fork 4.8k
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.
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
[Perf][Mono] 50 % Regression System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue, other Queue Benchmarks #46684
Comments
Tagging subscribers to this area: @BrzVlad Issue Details50 % regression in System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue and 4 other concurrent data structure benchmarks. Based on the time stamps and changes, it looks like the only plausible cause for this is: 8e6415d @kouvel Could that change have caused this regression? Although it could theoretically be anything else in this date range: 8e913a2...acd4855 @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
|
It's likely 8e6415d that caused the Not sure about the other tests, will have to see. |
I filed an issue with all the tests that we are seeing with this issue, it is here #46714. Mostly they regressed tests are in Concurrent* classes and we had a few in RentReturnArrayPool tests.
Given this can we get a work item cut to get rid of these tests and put in place more realistic ones? Doesn't seem super useful to have around if when they regress we can't take any meaningful action. |
Yea it would probably be reasonable to remove those tests. The ASP.NET tests seem to be better and seem to sufficiently cover this type of change for |
Actually it might be ok to keep the tests. They may not be good at assessing policy changes like in spin-waiting, but they could still identify regressions in pure overhead not involving policy changes. For a policy-type change we can just consider it a re-baseline since it's actually measuring something different from before. |
Closing, as these are one-off differences and we can re-baseline on the new numbers |
50 % regression in System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue and 4 other concurrent data structure benchmarks.
Based on the time stamps and changes, it looks like the only plausible cause for this is: 8e6415d
@kouvel Could that change have caused this regression?
Although it could theoretically be anything else in this date range: 8e913a2...acd4855
Run Information
Regressions in System.Collections.Concurrent.AddRemoveFromSameThreads
Related Issue on x64 Windows
[Perf 54%] System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue
Related Issue on x86 Windows
[Perf 55%] System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue
Historical Data in Reporting System
Repro
.
Payloads
Baseline
Compare
Histogram
System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue(Size: 2000000)
Docs
Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
Run Information
Regressions in System.Buffers.Tests.RentReturnArrayPoolTests
Related Issue on x64 Windows
[Perf -49%] System.Buffers.Tests.RentReturnArrayPoolTests.SingleSerial
Related Issue on x64 Windows
[Perf 30%] System.Buffers.Tests.RentReturnArrayPoolTests.MultipleSerial
Related Issue on x86 Windows
[Perf 66%] System.Buffers.Tests.RentReturnArrayPoolTests.SingleParallel
Related Issue on x86 Windows
[Perf 41%] System.Buffers.Tests.RentReturnArrayPoolTests (6)
Historical Data in Reporting System
Repro
.
Payloads
Baseline
Compare
Histogram
System.Buffers.Tests.RentReturnArrayPoolTests.SingleParallel(RentalSize: 4096, ManipulateArray: False, Async: True, UseSharedPool: True)
System.Buffers.Tests.RentReturnArrayPoolTests.ProducerConsumer(RentalSize: 4096, ManipulateArray: False, Async: True, UseSharedPool: True)
Docs
Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
Run Information
Regressions in System.Collections.Concurrent.AddRemoveFromSameThreads
Related Issue on x64 Windows
[Perf 55%] System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue
Related Issue on x86 Windows
[Perf 55%] System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue
Historical Data in Reporting System
Repro
.
Payloads
Baseline
Compare
Histogram
System.Collections.Concurrent.AddRemoveFromSameThreads.ConcurrentQueue(Size: 2000000)
Docs
Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
Run Information
Regressions in System.Threading.Tests.Perf_ThreadPool
Related Issue on x64 Windows
[Perf 25%] System.Threading.Tests.Perf_ThreadPool.QueueUserWorkItem_WaitCallback_Throughput
Historical Data in Reporting System
Repro
.
Payloads
Baseline
Compare
Histogram
System.Threading.Tests.Perf_ThreadPool.QueueUserWorkItem_WaitCallback_Throughput(WorkItemsPerCore: 20000000)
Docs
Profiling workflow for dotnet/runtime repository
Benchmarking workflow for dotnet/runtime repository
The text was updated successfully, but these errors were encountered: