Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Many test failures in "nohwintrinsics" jobs #38841

Closed
BruceForstall opened this issue Jul 7, 2020 · 5 comments
Closed

Many test failures in "nohwintrinsics" jobs #38841

BruceForstall opened this issue Jul 7, 2020 · 5 comments
Assignees
@tannergooding
Labels
area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI
Milestone
5.0.0

Comments

@BruceForstall
Copy link
Member

In both the runtime-coreclr jitstress-isas-x86 and runtime-coreclr jitstress-isas-arm pipelines, the "nohwintrinsics" (and for arm, the "nosimd") jobs are failing about 40 tests (about 20 for arm).

https://dev.azure.com/dnceng/public/_build/results?buildId=716582&view=ms.vss-test-web.build-test-results-tab&runId=22206304&resultId=108832&paneView=debug
https://dev.azure.com/dnceng/public/_build/results?buildId=716578&view=ms.vss-test-web.build-test-results-tab&runId=22205912&paneView=debug&resultId=105800

@tannergooding @CarolEidt @echesakovMSFT
@BruceForstall BruceForstall added the area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI label Jul 7, 2020
@BruceForstall BruceForstall added this to the 5.0.0 milestone Jul 7, 2020
@Dotnet-GitSync-Bot Dotnet-GitSync-Bot added the untriaged New issue has not been triaged by the area owner label Jul 7, 2020
@BruceForstall BruceForstall removed the untriaged New issue has not been triaged by the area owner label Jul 7, 2020
@tannergooding
Copy link
Member

Will take a look

@tannergooding tannergooding self-assigned this Jul 7, 2020
@tannergooding
Copy link
Member

This is essentially #35605. Basically, EnableHWIntrinsic=0 no longer disables SSE+. This means that the JIT is setting Vector<T> to be 32-bytes but the software fallback is getting executed, which assumes 16-bytes.

The fix should be relatively small and I am testing it out locally.

@BruceForstall
Copy link
Member Author

@tannergooding The proposed PR is #38879? How is this coming along? (Of course, the weekend test runs failed again.)

@tannergooding
Copy link
Member

@BruceForstall, #38879 fixes (I believe) all the x86 issues.

There is still an ARM64 issue tthat I haven't finished investigating that is causing some remaining failures

@tannergooding
Copy link
Member

Closing. The last -arm job fully passed (https://dev.azure.com/dnceng/public/_build?definitionId=665) and the last x86 job only had eventpipe failures which are unrelated: https://dev.azure.com/dnceng/public/_build?definitionId=664

@ghost ghost locked as resolved and limited conversation to collaborators Dec 8, 2020
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Assignees

@tannergooding tannergooding

Labels
area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI
Projects
None yet
Milestone
5.0.0
Development

No branches or pull requests
3 participants
@tannergooding @BruceForstall @Dotnet-GitSync-Bot