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

AF: ILC: Assertion failed '((simdSize == 16) && compOpportunisticallyDependsOn(InstructionSet_SSE41)) || ((simdSize == 32) && compOpportunisticallyDependsOn(InstructionSet_AVX2))' #109760

Open
jkotas opened this issue Nov 13, 2024 · 3 comments · May be fixed by #109842
Open

AF: ILC: Assertion failed '((simdSize == 16) && compOpportunisticallyDependsOn(InstructionSet_SSE41)) || ((simdSize == 32) && compOpportunisticallyDependsOn(InstructionSet_AVX2))' #109760

jkotas opened this issue Nov 13, 2024 · 3 comments · May be fixed by #109842
Assignees
@tannergooding
Labels
area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI blocking-clean-ci-optional Blocking optional rolling runs in-pr There is an active PR which will close this issue when it is merged
Milestone
10.0.0

Comments

@jkotas
Copy link
Member

jkotas commented Nov 13, 2024 •
edited
Loading

/azp run runtime-nativeaot-outerloop fails during compilation of System.Numerics.Vectors.Tests with:

ILC: Assertion failed '((simdSize == 16) && compOpportunisticallyDependsOn(InstructionSet_SSE41)) || ((simdSize == 32) && compOpportunisticallyDependsOn(InstructionSet_AVX2))' in 'System.Numerics.Vector`1[long]:System.Runtime.Intrinsics.ISimdVector<System.Numerics.Vector<T>,T>.MultiplyAddEstimate(System.Numerics.Vector`1[long],System.Numerics.Vector`1[long],System.Numerics.Vector`1[long]):System.Numerics.Vector`1[long]' during 'Importation' (IL size 9; hash 0x709b06b8; FullOpts)

Fails on Windows and Linux x64. Example of a log: https://dev.azure.com/dnceng-public/public/_build/results?buildId=867214&view=logs&jobId=7e7cde7f-4669-5f41-976b-370a74291e60&j=7e7cde7f-4669-5f41-976b-370a74291e60&t=9ee39664-b162-579c-6968-bf80271a4c1d
@dotnet-issue-labeler dotnet-issue-labeler bot added the area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI label Nov 13, 2024
@dotnet-policy-service dotnet-policy-service bot added the untriaged New issue has not been triaged by the area owner label Nov 13, 2024
@jkotas jkotas added blocking-clean-ci-optional Blocking optional rolling runs and removed area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI untriaged New issue has not been triaged by the area owner labels Nov 13, 2024
@dotnet-policy-service Dotnet Policy Service
Copy link
Contributor

Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch
See info in area-owners.md if you want to be subscribed.

@jkotas jkotas added area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI untriaged New issue has not been triaged by the area owner labels Nov 13, 2024
@jkotas
Copy link
Member Author

jkotas commented Nov 13, 2024

Regression introduced by #106594 . cc @tannergooding

@tannergooding tannergooding self-assigned this Nov 13, 2024
@tannergooding
Copy link
Member

At first glance it looks like a path getting NAOT is opting to emit Add(Mul(a, b), c) which should also be reproing in the ISA stress jobs for V128/V256, but it looks like it isn't because those the relevant tests live in the libraries side and not the JIT tests side.

We used to exit early for SIMD GT_MUL involving long/ulong on older hardware, but SSE4.1 fallback acceleration support was added back around July, and it looks like updating the Multiply was updated but not MultiplyAddEstimate

Fix should just be mirroring the GT_MUL guard in MultiplyAddEstimate, so I'll get that up early tomorrow.

@tannergooding tannergooding linked a pull request Nov 14, 2024 that will close this issue
Open
@dotnet-policy-service dotnet-policy-service bot added the in-pr There is an active PR which will close this issue when it is merged label Nov 14, 2024
@JulieLeeMSFT JulieLeeMSFT removed the untriaged New issue has not been triaged by the area owner label Nov 14, 2024
@JulieLeeMSFT JulieLeeMSFT added this to the 10.0.0 milestone Nov 14, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@tannergooding tannergooding

Labels
area-CodeGen-coreclr CLR JIT compiler in src/coreclr/src/jit and related components such as SuperPMI blocking-clean-ci-optional Blocking optional rolling runs in-pr There is an active PR which will close this issue when it is merged
Projects
None yet
Milestone
10.0.0
Development

Successfully merging a pull request may close this issue.

Ensure that MultiplyAddEstimate exits for cases where GT_MUL is unsupported tannergooding/runtime
3 participants
@jkotas @tannergooding @JulieLeeMSFT