[llvm] [WebAssembly] Do not form minnum/maxnum (PR #184796)

Thu Mar 5 06:45:18 PST 2026

github-actions[bot] wrote:

<!--PREMERGE ADVISOR COMMENT: Windows-->
# :window: Windows x64 Test Results

* 128321 tests passed
* 2560 tests skipped

All executed tests passed, but another part of the build **failed**. Click on a failure below to see the details.

<details>
<summary>[code=4294967295] tools/mlir/python_packages/mlir_core/mlir/_mlir_libs/MLIRPythonCAPI.dll tools/mlir/python_packages/mlir_core/mlir/_mlir_libs/MLIRPythonCAPI.lib</summary>

```
FAILED: [code=4294967295] tools/mlir/python_packages/mlir_core/mlir/_mlir_libs/MLIRPythonCAPI.dll tools/mlir/python_packages/mlir_core/mlir/_mlir_libs/MLIRPythonCAPI.lib
C:\Windows\system32\cmd.exe /C "C:\Windows\system32\cmd.exe /C "C:\BuildTools\Common7\IDE\CommonExtensions\Microsoft\CMake\CMake\bin\cmake.exe -E __create_def C:\_work\llvm-project\llvm-project\build\tools\mlir\python\CMakeFiles\MLIRPythonCAPI.dir\.\exports.def C:\_work\llvm-project\llvm-project\build\tools\mlir\python\CMakeFiles\MLIRPythonCAPI.dir\.\exports.def.objs --nm=CMAKE_NM-NOTFOUND && cd C:\_work\llvm-project\llvm-project\build" && C:\BuildTools\Common7\IDE\CommonExtensions\Microsoft\CMake\CMake\bin\cmake.exe -E vs_link_dll --msvc-ver=1944 --intdir=tools\mlir\python\CMakeFiles\MLIRPythonCAPI.dir --rc="C:\Program Files (x86)\Windows Kits\10\bin\10.0.26100.0\x64\rc.exe" --mt="C:\Program Files (x86)\Windows Kits\10\bin\10.0.26100.0\x64\mt.exe" --manifests  -- C:\clang\clang-msvc\bin\lld-link.exe /nologo @CMakeFiles\MLIRPythonCAPI.rsp  /out:tools\mlir\python_packages\mlir_core\mlir\_mlir_libs\MLIRPythonCAPI.dll /implib:tools\mlir\python_packages\mlir_core\mlir\_mlir_libs\MLIRPythonCAPI.lib /pdb:tools\mlir\python_packages\mlir_core\mlir\_mlir_libs\MLIRPythonCAPI.pdb /dll /version:0.0 /MANIFEST:NO /INCREMENTAL:NO  /DEF:tools\mlir\python\CMakeFiles\MLIRPythonCAPI.dir\.\exports.def && cd ."
LINK: command "C:\clang\clang-msvc\bin\lld-link.exe /nologo @CMakeFiles\MLIRPythonCAPI.rsp /out:tools\mlir\python_packages\mlir_core\mlir\_mlir_libs\MLIRPythonCAPI.dll /implib:tools\mlir\python_packages\mlir_core\mlir\_mlir_libs\MLIRPythonCAPI.lib /pdb:tools\mlir\python_packages\mlir_core\mlir\_mlir_libs\MLIRPythonCAPI.pdb /dll /version:0.0 /MANIFEST:NO /INCREMENTAL:NO /DEF:tools\mlir\python\CMakeFiles\MLIRPythonCAPI.dir\.\exports.def" failed (exit code 1) with the following output:
lld-link: error: undefined symbol: private: static class mlir::SelfOwningTypeID mlir::detail::TypeIDResolver<class mlir::affine::AffineDialect, void>::id
>>> referenced by MLIRGPUTransforms.lib(DecomposeMemRefs.cpp.obj):(public: virtual void __cdecl mlir::impl::GpuDecomposeMemrefsPassBase<struct `anonymous namespace'::GpuDecomposeMemrefsPass>::getDependentDialects(class mlir::DialectRegistry &) const)
>>> referenced by MLIRGPUTransforms.lib(DecomposeMemRefs.cpp.obj):(private: virtual class mlir::Dialect * __cdecl std::_Func_impl_no_alloc<class `public: void __cdecl mlir::DialectRegistry::insert<class mlir::affine::AffineDialect>(void)'::`1'::<lambda_1>, class mlir::Dialect *, class mlir::MLIRContext *>::_Do_call(class mlir::MLIRContext *&&))
>>> referenced by MLIRLinalgDialect.lib(LinalgDialect.cpp.obj):(private: __cdecl mlir::linalg::LinalgDialect::LinalgDialect(class mlir::MLIRContext *))
>>> referenced 39 more times

lld-link: error: undefined symbol: private: __cdecl mlir::affine::AffineDialect::AffineDialect(class mlir::MLIRContext *)
>>> referenced by MLIRGPUTransforms.lib(DecomposeMemRefs.cpp.obj):(private: static class std::unique_ptr<class mlir::Dialect, struct std::default_delete<class mlir::Dialect>> __cdecl llvm::function_ref<class std::unique_ptr<class mlir::Dialect, struct std::default_delete<class mlir::Dialect>> __cdecl(void)>::callback_fn<class `public: class mlir::affine::AffineDialect * __cdecl mlir::MLIRContext::getOrLoadDialect<class mlir::affine::AffineDialect>(void)'::`1'::<lambda_1>>(__int64))
>>> referenced by MLIRLinalgDialect.lib(LinalgDialect.cpp.obj)
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj)

lld-link: error: undefined symbol: class mlir::OpFoldResult __cdecl mlir::affine::makeComposedFoldedAffineApply(class mlir::OpBuilder &, class mlir::Location, class mlir::AffineExpr, class llvm::ArrayRef<class mlir::OpFoldResult>, bool)
>>> referenced by MLIRGPUTransforms.lib(DecomposeMemRefs.cpp.obj):(class std::tuple<class mlir::Value, class mlir::OpFoldResult, class llvm::SmallVector<class mlir::OpFoldResult, 6>> __cdecl getFlatOffsetAndStrides(class mlir::OpBuilder &, class mlir::Location, class mlir::Value, class llvm::ArrayRef<class mlir::OpFoldResult>, class llvm::ArrayRef<class mlir::OpFoldResult>))
>>> referenced by MLIRGPUTransforms.lib(DecomposeMemRefs.cpp.obj):(class std::tuple<class mlir::Value, class mlir::OpFoldResult, class llvm::SmallVector<class mlir::OpFoldResult, 6>> __cdecl getFlatOffsetAndStrides(class mlir::OpBuilder &, class mlir::Location, class mlir::Value, class llvm::ArrayRef<class mlir::OpFoldResult>, class llvm::ArrayRef<class mlir::OpFoldResult>))
>>> referenced by MLIRLinalgTransforms.lib(ElementwiseOpFusion.cpp.obj):(class std::optional<class llvm::SmallVector<class mlir::Value, 6>> __cdecl fuseWithReshapeByExpansion(class mlir::linalg::LinalgOp, class mlir::Operation *, class mlir::OpOperand *, class mlir::PatternRewriter &))
>>> referenced 278 more times

lld-link: error: undefined symbol: class llvm::SmallVector<class mlir::OpFoldResult, 6> __cdecl mlir::affine::makeComposedFoldedMultiResultAffineApply(class mlir::OpBuilder &, class mlir::Location, class mlir::AffineMap, class llvm::ArrayRef<class mlir::OpFoldResult>, bool)
>>> referenced by MLIRLinalgDialect.lib(LinalgInterfaces.cpp.obj):(public: struct llvm::LogicalResult __cdecl mlir::linalg::LinalgOp::reifyResultShapes(class mlir::OpBuilder &, class llvm::SmallVector<class llvm::SmallVector<class mlir::OpFoldResult, 6>, 1> &))
>>> referenced by MLIRSparseTensorTransforms.lib(SparseTensorRewriting.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::FuseExtractSliceWithConcat::matchAndRewrite(class mlir::tensor::ExtractSliceOp, class tensor::PatternRewriter &) const)
>>> referenced by MLIRLinalgTransforms.lib(RuntimeOpVerification.cpp.obj):(public: static void __cdecl mlir::detail::RuntimeVerifiableOpInterfaceInterfaceTraits::FallbackModel<struct mlir::linalg::`anonymous namespace'::StructuredOpInterface<class mlir::linalg::AbsOp>>::generateRuntimeVerification(struct mlir::detail::RuntimeVerifiableOpInterfaceInterfaceTraits::Concept const *, class mlir::Operation *, class mlir::OpBuilder &, class mlir::Location, class llvm::function_ref<class std::basic_string<char, struct std::char_traits<char>, class std::allocator<char>> __cdecl(class mlir::Operation *, class llvm::StringRef)>))
>>> referenced 186 more times

lld-link: error: undefined symbol: public: static void __cdecl mlir::affine::AffineApplyOp::getCanonicalizationPatterns(class mlir::RewritePatternSet &, class mlir::MLIRContext *)
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(public: virtual void __cdecl `anonymous namespace'::LowerToAffineLoops::runOnOperation(void))
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(public: virtual void __cdecl `anonymous namespace'::LowerToLoops::runOnOperation(void))
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(public: virtual void __cdecl `anonymous namespace'::LowerToParallelLoops::runOnOperation(void))
>>> referenced 3 more times

lld-link: error: undefined symbol: public: class mlir::AffineMap __cdecl mlir::affine::AffineApplyOp::getMap(void)
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::FoldAffineOp::matchAndRewrite(class mlir::Operation *, class Operation::PatternRewriter &) const)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineApplyLowering::matchAndRewrite(class mlir::affine::AffineApplyOp, class affine::PatternRewriter &) const)
>>> referenced by MLIRTensorTransforms.lib(ReshapePatterns.cpp.obj):(bool __cdecl isMultipleOf(class mlir::OpFoldResult, __int64))
>>> referenced 7 more times

lld-link: error: undefined symbol: private: static class mlir::SelfOwningTypeID mlir::detail::TypeIDResolver<class mlir::affine::AffineApplyOp, void>::id
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(public: static bool __cdecl mlir::Op<class mlir::affine::AffineApplyOp, class mlir::OpTrait::ZeroRegions, class mlir::OpTrait::OneResult, class mlir::OpTrait::OneTypedResult<class mlir::IndexType>::Impl, class mlir::OpTrait::ZeroSuccessors, class mlir::OpTrait::VariadicOperands, class mlir::OpTrait::OpInvariants, struct mlir::BytecodeOpInterface::Trait, struct mlir::ConditionallySpeculatable::Trait, struct mlir::OpTrait::AlwaysSpeculatableImplTrait, struct mlir::MemoryEffectOpInterface::Trait, struct mlir::InferIntRangeInterface::Trait, struct mlir::InferTypeOpInterface::Trait>::classof(class mlir::Operation *))
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj)
>>> referenced by MLIRLinalgUtils.lib(Utils.cpp.obj)
>>> referenced 3 more times

lld-link: error: undefined symbol: public: static class mlir::affine::AffineLoadOp __cdecl mlir::affine::AffineLoadOp::create(class mlir::OpBuilder &, class mlir::Location, class mlir::Value, class mlir::ValueRange)
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(private: static class llvm::SmallVector<class mlir::Value, 6> __cdecl llvm::function_ref<class llvm::SmallVector<class mlir::Value, 6> __cdecl(class mlir::OpBuilder &, class mlir::Location, class mlir::ValueRange, class mlir::ValueRange)>::callback_fn<class `class llvm::FailureOr<class llvm::SmallVector<class mlir::Operation *, 4>> __cdecl linalgOpToLoopsImpl<class mlir::affine::AffineForOp>(class mlir::RewriterBase &, class mlir::linalg::LinalgOp)'::`1'::<lambda_1>>(__int64, class mlir::OpBuilder &, class mlir::Location, class mlir::ValueRange, class mlir::ValueRange))
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(private: static class llvm::SmallVector<class mlir::Value, 6> __cdecl llvm::function_ref<class llvm::SmallVector<class mlir::Value, 6> __cdecl(class mlir::OpBuilder &, class mlir::Location, class mlir::ValueRange, class mlir::ValueRange)>::callback_fn<class `class llvm::FailureOr<class llvm::SmallVector<class mlir::Operation *, 4>> __cdecl linalgOpToLoopsImpl<class mlir::affine::AffineForOp>(class mlir::RewriterBase &, class mlir::linalg::LinalgOp)'::`1'::<lambda_1>>(__int64, class mlir::OpBuilder &, class mlir::Location, class mlir::ValueRange, class mlir::ValueRange))
>>> referenced by MLIRAffineTransforms.lib(FoldMemRefAliasOps.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineLoadOpOfSubViewOpFolder::matchAndRewrite(class mlir::affine::AffineLoadOp, class affine::PatternRewriter &) const)
>>> referenced 3 more times

lld-link: error: undefined symbol: public: static class mlir::affine::AffineStoreOp __cdecl mlir::affine::AffineStoreOp::create(class mlir::OpBuilder &, class mlir::Location, class mlir::Value, class mlir::Value, class mlir::ValueRange)
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(private: static class llvm::SmallVector<class mlir::Value, 6> __cdecl llvm::function_ref<class llvm::SmallVector<class mlir::Value, 6> __cdecl(class mlir::OpBuilder &, class mlir::Location, class mlir::ValueRange, class mlir::ValueRange)>::callback_fn<class `class llvm::FailureOr<class llvm::SmallVector<class mlir::Operation *, 4>> __cdecl linalgOpToLoopsImpl<class mlir::affine::AffineForOp>(class mlir::RewriterBase &, class mlir::linalg::LinalgOp)'::`1'::<lambda_1>>(__int64, class mlir::OpBuilder &, class mlir::Location, class mlir::ValueRange, class mlir::ValueRange))
>>> referenced by MLIRAffineTransforms.lib(FoldMemRefAliasOps.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineStoreOpOfSubViewOpFolder::matchAndRewrite(class mlir::affine::AffineStoreOp, class affine::PatternRewriter &) const)
>>> referenced by MLIRAffineTransforms.lib(FoldMemRefAliasOps.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineStoreOpOfExpandShapeOpFolder::matchAndRewrite(class mlir::affine::AffineStoreOp, class affine::PatternRewriter &) const)
>>> referenced 2 more times

lld-link: error: undefined symbol: void __cdecl mlir::affine::canonicalizeMapAndOperands(class mlir::AffineMap *, class llvm::SmallVectorImpl<class mlir::Value> *)
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(class llvm::SmallVector<class mlir::Value, 6> __cdecl makeCanonicalAffineApplies(class mlir::OpBuilder &, class mlir::Location, class mlir::AffineMap, class llvm::ArrayRef<class mlir::Value>))
>>> referenced by MLIRAffineUtils.lib(Utils.cpp.obj):(struct llvm::LogicalResult __cdecl mlir::affine::replaceAllMemRefUsesWith(class mlir::Value, class mlir::Value, class mlir::Operation *, class llvm::ArrayRef<class mlir::Value>, class mlir::AffineMap, class llvm::ArrayRef<class mlir::Value>, class llvm::ArrayRef<class mlir::Value>, bool))
>>> referenced by MLIRAffineUtils.lib(LoopUtils.cpp.obj):(struct llvm::LogicalResult __cdecl generateCleanupLoopForUnroll(class mlir::affine::AffineForOp, unsigned __int64))
>>> referenced 10 more times

lld-link: error: undefined symbol: public: static class mlir::affine::AffineApplyOp __cdecl mlir::affine::AffineApplyOp::create(class mlir::OpBuilder &, class mlir::Location, class mlir::AffineMap, class mlir::ValueRange)
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(class llvm::SmallVector<class mlir::Value, 6> __cdecl makeCanonicalAffineApplies(class mlir::OpBuilder &, class mlir::Location, class mlir::AffineMap, class llvm::ArrayRef<class mlir::Value>))
>>> referenced by MLIRLinalgTransforms.lib(ElementwiseOpFusion.cpp.obj):(class llvm::FailureOr<struct mlir::linalg::ElementwiseOpFusionResult> __cdecl mlir::linalg::fuseElementwiseOps(class mlir::RewriterBase &, class mlir::OpOperand *))
>>> referenced by MLIRSparseTensorTransforms.lib(SparseTensorRewriting.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::SparseTensorDimOpRewriter::matchAndRewrite(class mlir::tensor::DimOp, class tensor::PatternRewriter &) const)
>>> referenced 46 more times

lld-link: error: undefined symbol: private: static class mlir::SelfOwningTypeID mlir::detail::TypeIDResolver<class mlir::affine::AffineForOp, void>::id
>>> referenced by MLIRLinalgTransforms.lib(Loops.cpp.obj):(public: static bool __cdecl mlir::Op<class mlir::affine::AffineForOp, class mlir::OpTrait::OneRegion, class mlir::OpTrait::VariadicResults, class mlir::OpTrait::ZeroSuccessors, class mlir::OpTrait::VariadicOperands, class mlir::OpTrait::AttrSizedOperandSegments, struct mlir::OpTrait::SingleBlock, class mlir::OpTrait::SingleBlockImplicitTerminator<class mlir::affine::AffineYieldOp>::Impl, class mlir::OpTrait::OpInvariants, struct mlir::BytecodeOpInterface::Trait, class mlir::OpTrait::AutomaticAllocationScope, struct mlir::ConditionallySpeculatable::Trait, class mlir::OpTrait::HasRecursiveMemoryEffects, struct mlir::LoopLikeOpInterface::Trait, struct mlir::RegionBranchOpInterface::Trait>::classof(class mlir::Operation *))
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj)
>>> referenced by MLIRSCFToGPU.lib(SCFToGPUPass.cpp.obj)
>>> referenced 14 more times

lld-link: error: undefined symbol: class mlir::OpFoldResult __cdecl mlir::affine::makeComposedFoldedAffineApply(class mlir::OpBuilder &, class mlir::Location, class mlir::AffineMap, class llvm::ArrayRef<class mlir::OpFoldResult>, bool)
>>> referenced by MLIRLinalgTransforms.lib(ElementwiseOpFusion.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::FoldPadWithProducerReshapeOpByCollapsing::matchAndRewrite(class mlir::tensor::PadOp, class tensor::PatternRewriter &) const)
>>> referenced by MLIRAffineUtils.lib(ViewLikeInterfaceUtils.cpp.obj):(void __cdecl mlir::affine::resolveIndicesIntoOpWithOffsetsAndStrides(class mlir::RewriterBase &, class mlir::Location, class llvm::ArrayRef<class mlir::OpFoldResult>, class llvm::ArrayRef<class mlir::OpFoldResult>, class llvm::SmallBitVector const &, class llvm::ArrayRef<class mlir::OpFoldResult>, class llvm::SmallVectorImpl<class mlir::Value> &))
>>> referenced by MLIRMemRefUtils.lib(MemRefUtils.cpp.obj):(void __cdecl mlir::memref::resolveSourceIndicesCollapseShape(class mlir::Location, class mlir::PatternRewriter &, class mlir::memref::CollapseShapeOp, class mlir::ValueRange, class llvm::SmallVectorImpl<class mlir::Value> &))
>>> referenced 23 more times

lld-link: error: undefined symbol: public: struct std::pair<unsigned int, unsigned int> __cdecl mlir::affine::AffineForOp::getODSOperandIndexAndLength(unsigned int)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(class mlir::Value __cdecl mlir::lowerAffineUpperBound(class mlir::affine::AffineForOp, class mlir::OpBuilder &))
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(class mlir::Value __cdecl mlir::lowerAffineLowerBound(class mlir::affine::AffineForOp, class mlir::OpBuilder &))
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineForLowering::matchAndRewrite(class mlir::affine::AffineForOp, class affine::PatternRewriter &) const)
>>> referenced 33 more times

lld-link: error: undefined symbol: public: class mlir::AffineMap __cdecl mlir::affine::AffineForOp::getUpperBoundMap(void)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(class mlir::Value __cdecl mlir::lowerAffineUpperBound(class mlir::affine::AffineForOp, class mlir::OpBuilder &))
>>> referenced by MLIRAffineUtils.lib(Utils.cpp.obj):(struct llvm::LogicalResult __cdecl mlir::affine::affineParallelize(class mlir::affine::AffineForOp, class llvm::ArrayRef<struct mlir::affine::LoopReduction>, class mlir::affine::AffineParallelOp *))
>>> referenced by MLIRAffineUtils.lib(Utils.cpp.obj):(struct llvm::LogicalResult __cdecl mlir::affine::normalizeAffineFor(class mlir::affine::AffineForOp, bool))
>>> referenced 13 more times

lld-link: error: undefined symbol: public: class mlir::AffineMap __cdecl mlir::affine::AffineForOp::getLowerBoundMap(void)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(class mlir::Value __cdecl mlir::lowerAffineLowerBound(class mlir::affine::AffineForOp, class mlir::OpBuilder &))
>>> referenced by MLIRAffineUtils.lib(Utils.cpp.obj):(struct llvm::LogicalResult __cdecl mlir::affine::affineParallelize(class mlir::affine::AffineForOp, class llvm::ArrayRef<struct mlir::affine::LoopReduction>, class mlir::affine::AffineParallelOp *))
>>> referenced by MLIRAffineUtils.lib(Utils.cpp.obj):(struct llvm::LogicalResult __cdecl mlir::affine::normalizeAffineFor(class mlir::affine::AffineForOp, bool))
>>> referenced 16 more times

lld-link: error: undefined symbol: public: class mlir::AffineMap __cdecl mlir::affine::AffineDmaStartOp::getSrcMap(void)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineDmaStartLowering::matchAndRewrite(class mlir::affine::AffineDmaStartOp, class affine::PatternRewriter &) const)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineDmaStartLowering::matchAndRewrite(class mlir::affine::AffineDmaStartOp, class affine::PatternRewriter &) const)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineDmaStartLowering::matchAndRewrite(class mlir::affine::AffineDmaStartOp, class affine::PatternRewriter &) const)
>>> referenced 16 more times

lld-link: error: undefined symbol: public: class mlir::AffineMap __cdecl mlir::affine::AffineDmaStartOp::getDstMap(void)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineDmaStartLowering::matchAndRewrite(class mlir::affine::AffineDmaStartOp, class affine::PatternRewriter &) const)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineDmaStartLowering::matchAndRewrite(class mlir::affine::AffineDmaStartOp, class affine::PatternRewriter &) const)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: class mlir::Value __cdecl mlir::affine::AffineDmaStartOp::getNumElementsPerStride(void))
>>> referenced 7 more times

lld-link: error: undefined symbol: public: class mlir::AffineMap __cdecl mlir::affine::AffineDmaStartOp::getTagMap(void)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: virtual struct llvm::LogicalResult __cdecl `anonymous namespace'::AffineDmaStartLowering::matchAndRewrite(class mlir::affine::AffineDmaStartOp, class affine::PatternRewriter &) const)
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: class mlir::Value __cdecl mlir::affine::AffineDmaStartOp::getNumElementsPerStride(void))
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: class mlir::Value __cdecl mlir::affine::AffineDmaStartOp::getStride(void))
>>> referenced 2 more times

lld-link: error: undefined symbol: private: static class mlir::SelfOwningTypeID mlir::detail::TypeIDResolver<class mlir::affine::AffineDmaStartOp, void>::id
>>> referenced by MLIRAffineToStandard.lib(AffineToStandard.cpp.obj):(public: static bool __cdecl mlir::Op<class mlir::affine::AffineDmaStartOp, class mlir::OpTrait::ZeroRegions, class mlir::OpTrait::ZeroResults, class mlir::OpTrait::ZeroSuccessors, class mlir::OpTrait::VariadicOperands, class mlir::OpTrait::OpInvariants, struct mlir::BytecodeOpInterface::Trait, struct mlir::OpTrait::MemRefsNormalizable, struct mlir::affine::AffineMapAccessInterface::Trait, struct mlir::MemoryEffectOpInterface::Trait>::classof(class mlir::Operation *))
>>> referenced by MLIRAffineTransforms.lib(PipelineDataTransfer.cpp.obj):(public: virtual void __cdecl `anonymous namespace'::PipelineDataTransfer::runOnOperation(void))
>>> referenced by MLIRAffineTransforms.lib(PipelineDataTransfer.cpp.obj):(public: virtual void __cdecl `anonymous namespace'::PipelineDataTransfer::runOnOperation(void))
>>> referenced 1 more times

lld-link: error: too many errors emitted, stopping now (use /errorlimit:0 to see all errors)
```
</details>

If these failures are unrelated to your changes (for example tests are broken or flaky at HEAD), please open an issue at https://github.com/llvm/llvm-project/issues and add the `infrastructure` label.

https://github.com/llvm/llvm-project/pull/184796