Repository navigation
Test failure in System.Numerics.Tensors.Tests.SingleGenericTensorPrimitives.SpanDestinationFunctions_SpecialValues #101731
Copy link
Copy link
Labels
Known Build ErrorUse this to report build issues in the .NET Helix tabUse this to report build issues in the .NET Helix tabarea-System.Numerics.Tensorsblocking-clean-ciBlocking PR or rolling runs of 'runtime' or 'runtime-extra-platforms'Blocking PR or rolling runs of 'runtime' or 'runtime-extra-platforms'
Milestone
Description
Failed in runtime-coreclr libraries-jitstress2-jitstressregs
Stack Trace:
/_/src/libraries/System.Numerics.Tensors/tests/TensorPrimitives.Generic.cs(444,0): at System.Numerics.Tensors.Tests.GenericFloatingPointNumberTensorPrimitivesTests`1.SpanDestinationFunctions_SpecialValues(SpanDestinationDelegate tensorPrimitivesMethod, Func`2 expectedMethod, Nullable`1 tolerance)
at InvokeStub_GenericFloatingPointNumberTensorPrimitivesTests`1.SpanDestinationFunctions_SpecialValues(Object, Span`1)
at System.Reflection.MethodBaseInvoker.InvokeWithFewArgs(Object obj, BindingFlags invokeAttr, Binder binder, Object[] parameters, CultureInfo culture)
Finished: System.Numerics.Tensors.Net8.Tests
Known Issue Error Message
Fill the error message using step by step known issues guidance.
{
"ErrorMessage": "GenericFloatingPointNumberTensorPrimitivesTests`1.SpanDestinationFunctions_SpecialValues",
"ErrorPattern": "",
"BuildRetry": false,
"ExcludeConsoleLog": false
}Known issue validation
Build: 🔎 https://dev.azure.com/dnceng-public/public/_build/results?buildId=660140
Error message validated: [GenericFloatingPointNumberTensorPrimitivesTests1.SpanDestinationFunctions_SpecialValues`]
Result validation: ✅ Known issue matched with the provided build.
Validation performed at: 5/1/2024 3:48:58 AM UTC
Report
Summary
| 24-Hour Hit Count | 7-Day Hit Count | 1-Month Count |
|---|---|---|
| 21 | 147 | 212 |
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
Known Build ErrorUse this to report build issues in the .NET Helix tabUse this to report build issues in the .NET Helix tabarea-System.Numerics.Tensorsblocking-clean-ciBlocking PR or rolling runs of 'runtime' or 'runtime-extra-platforms'Blocking PR or rolling runs of 'runtime' or 'runtime-extra-platforms'
Activity
dotnet-policy-service commentedon Apr 30, 2024
Tagging subscribers to this area: @dotnet/area-system-numerics
See info in area-owners.md if you want to be subscribed.
dotnet-policy-service commentedon Apr 30, 2024
Tagging subscribers to this area: @dotnet/area-system-numerics-tensors
See info in area-owners.md if you want to be subscribed.
JulieLeeMSFT commentedon Apr 30, 2024
It looks similar to #97297.
JulieLeeMSFT commentedon Apr 30, 2024
Looks similar in runtime-coreclr libraries-jitstressregs
JulieLeeMSFT commentedon Apr 30, 2024
@tannergooding, PTAL.
tannergooding commentedon Apr 30, 2024
I can take a look and see if I can find an initial root cause, but I expect this is likely a JIT issue and unrelated to the libraries side of things.
CC. @jeffhandley
JulieLeeMSFT commentedon May 1, 2024
It is now failing in Runtime pipeline as well. Please see the Known issue report above.
kunalspathak commentedon May 5, 2024
This seems to be still failing in #101859 which included the fix made in #101800
jkotas commentedon May 6, 2024
It is failing on #101767 with the fix.
stephentoub commentedon May 6, 2024
e.g.
tannergooding commentedon May 10, 2024
I'm not able to reproduce the failure locally, on a machine with or without AVX512.
My best guess is that in some edge case scenario the code is calling the fallback instead of allowing the intrinsic implementation to kick in.
Changing it to be
mustExpandwith no managed implementation should avoid any potential for that, so that's what I'll get a fix up to do.EgorBo commentedon May 12, 2024
Should we disable this test for now? it's been blocking clean CI for two weeks now (e.g. I've just hit it in #102132)
tannergooding commentedon May 13, 2024
The issue is fixed with #102098 and so disabling shouldn't be needed, just final sign-off
If, for whatever reason, the test still shows as flaky somewhere after then I think we can disable it then