| Issue |
177779
|
| Summary |
mlir-opt --sparse-tensor-codegen crashes on uninitialized SparseTensorEncodingAttr
|
| Labels |
new issue
|
| Assignees |
|
| Reporter |
Subway2023
|
# Description
Running mlir-opt --sparse-tensor-codegen on a MLIR module with sparse_tensor.convert operations crashes with an assertion in SparseTensorDialect.cpp
# Reproduce
mlir-opt version: 21.1.8
location: If the **sparse_tensor.convert** operation is commented out, the crash no longer occurs.
```
module {
spirv.EntryPoint "GLCompute" @main
func.func @main() {
%0 = gpu.num_subgroups : index
%1 = tensor.generate %0 {
^bb0(%arg0: index):
%cst = arith.constant 0.000000e+00 : f32
tensor.yield %cst : f32
} : tensor<?xf32>
%2 = sparse_tensor.convert %1 : tensor<?xf32> to tensor<?xf32, #spirv.storage_class<StorageBuffer>>
return
}
}
```
```
mlir-opt --sparse-tensor-codegen test.mlir
```
```
mlir-opt: /mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/mlir/lib/Dialect/SparseTensor/IR/SparseTensorDialect.cpp:460: bool mlir::sparse_tensor::SparseTensorEncodingAttr::isSlice() const: Assertion `getImpl() && "Uninitialized SparseTensorEncodingAttr"' failed.
PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace.
Stack dump:
0. Program arguments: /mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt --sparse-tensor-codegen /mnt/sdd1/sbw/mlir/data/genProgram_0123_filerted/genProgram_0123_valid/232137_47e4_5.mlir
#0 0x000056f26ca3a612 llvm::sys::PrintStackTrace(llvm::raw_ostream&, int) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x20a1612)
#1 0x000056f26ca36d7f llvm::sys::RunSignalHandlers() (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x209dd7f)
#2 0x000056f26ca36ecc SignalHandler(int, siginfo_t*, void*) Signals.cpp:0:0
#3 0x00007e19cf445330 (/lib/x86_64-linux-gnu/libc.so.6+0x45330)
#4 0x00007e19cf49eb2c __pthread_kill_implementation ./nptl/pthread_kill.c:44:76
#5 0x00007e19cf49eb2c __pthread_kill_internal ./nptl/pthread_kill.c:78:10
#6 0x00007e19cf49eb2c pthread_kill ./nptl/pthread_kill.c:89:10
#7 0x00007e19cf44527e raise ./signal/../sysdeps/posix/raise.c:27:6
#8 0x00007e19cf4288ff abort ./stdlib/abort.c:81:7
#9 0x00007e19cf42881b _nl_load_domain ./intl/loadmsgcat.c:1177:9
#10 0x00007e19cf43b517 (/lib/x86_64-linux-gnu/libc.so.6+0x3b517)
#11 0x000056f26ee68f95 (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x44cff95)
#12 0x000056f26ee8055b (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x44e755b)
#13 0x000056f26f014fe9 (anonymous namespace)::SparseConvertConverter::matchAndRewrite(mlir::sparse_tensor::ConvertOp, mlir::sparse_tensor::ConvertOpGenericAdaptor<llvm::ArrayRef<mlir::ValueRange>>, mlir::ConversionPatternRewriter&) const SparseTensorCodegen.cpp:0:0
#14 0x000056f26f00645c mlir::OpConversionPattern<mlir::sparse_tensor::ConvertOp>::matchAndRewrite(mlir::Operation*, llvm::ArrayRef<mlir::ValueRange>, mlir::ConversionPatternRewriter&) const (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x466d45c)
#15 0x000056f2702970ab mlir::ConversionPattern::matchAndRewrite(mlir::Operation*, mlir::PatternRewriter&) const (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x58fe0ab)
#16 0x000056f27335c08a mlir::PatternApplicator::matchAndRewrite(mlir::Operation*, mlir::PatternRewriter&, llvm::function_ref<bool (mlir::Pattern const&)>, llvm::function_ref<void (mlir::Pattern const&)>, llvm::function_ref<llvm::LogicalResult (mlir::Pattern const&)>) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x89c308a)
#17 0x000056f27029a6fe (anonymous namespace)::OperationLegalizer::legalize(mlir::Operation*, mlir::ConversionPatternRewriter&) DialectConversion.cpp:0:0
#18 0x000056f27029acae mlir::OperationConverter::convert(mlir::ConversionPatternRewriter&, mlir::Operation*) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x5901cae)
#19 0x000056f27029f57c mlir::OperationConverter::convertOperations(llvm::ArrayRef<mlir::Operation*>) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x590657c)
#20 0x000056f2702a0bba applyConversion(llvm::ArrayRef<mlir::Operation*>, mlir::ConversionTarget const&, mlir::FrozenRewritePatternSet const&, mlir::ConversionConfig, (anonymous namespace)::OpConversionMode) DialectConversion.cpp:0:0
#21 0x000056f2702a0cf2 mlir::applyPartialConversion(mlir::Operation*, mlir::ConversionTarget const&, mlir::FrozenRewritePatternSet const&, mlir::ConversionConfig) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x5907cf2)
#22 0x000056f26ef40068 (anonymous namespace)::SparseTensorCodegenPass::runOnOperation() SparseTensorPasses.cpp:0:0
#23 0x000056f2701ef07e mlir::detail::OpToOpPassAdaptor::run(mlir::Pass*, mlir::Operation*, mlir::AnalysisManager, bool, unsigned int) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x585607e)
#24 0x000056f2701ef4c1 mlir::detail::OpToOpPassAdaptor::runPipeline(mlir::OpPassManager&, mlir::Operation*, mlir::AnalysisManager, bool, unsigned int, mlir::PassInstrumentor*, mlir::PassInstrumentation::PipelineParentInfo const*) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x58564c1)
#25 0x000056f2701f0523 mlir::PassManager::run(mlir::Operation*) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x5857523)
#26 0x000056f2701e00a7 performActions(llvm::raw_ostream&, std::shared_ptr<llvm::SourceMgr> const&, mlir::MLIRContext*, mlir::MlirOptMainConfig const&) MlirOptMain.cpp:0:0
#27 0x000056f2701e09c6 processBuffer(llvm::raw_ostream&, std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, mlir::MlirOptMainConfig const&, mlir::DialectRegistry&, llvm::ThreadPoolInterface*) MlirOptMain.cpp:0:0
#28 0x000056f2701e0b8c llvm::LogicalResult llvm::function_ref<llvm::LogicalResult (std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, llvm::raw_ostream&)>::callback_fn<mlir::MlirOptMain(llvm::raw_ostream&, std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, mlir::DialectRegistry&, mlir::MlirOptMainConfig const&)::'lambda'(std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, llvm::raw_ostream&)>(long, std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, llvm::raw_ostream&) MlirOptMain.cpp:0:0
#29 0x000056f2702f624e mlir::splitAndProcessBuffer(std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, llvm::function_ref<llvm::LogicalResult (std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, llvm::raw_ostream&)>, llvm::raw_ostream&, llvm::StringRef, llvm::StringRef) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x595d24e)
#30 0x000056f2701d9080 mlir::MlirOptMain(llvm::raw_ostream&, std::unique_ptr<llvm::MemoryBuffer, std::default_delete<llvm::MemoryBuffer>>, mlir::DialectRegistry&, mlir::MlirOptMainConfig const&) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x5840080)
#31 0x000056f2701e0cf0 mlir::MlirOptMain(int, char**, llvm::StringRef, llvm::StringRef, mlir::DialectRegistry&) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x5847cf0)
#32 0x000056f2701e11ef mlir::MlirOptMain(int, char**, llvm::StringRef, mlir::DialectRegistry&) (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x58481ef)
#33 0x000056f26c952577 main (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x1fb9577)
#34 0x00007e19cf42a1ca __libc_start_call_main ./csu/../sysdeps/nptl/libc_start_call_main.h:74:3
#35 0x00007e19cf42a28b call_init ./csu/../csu/libc-start.c:128:20
#36 0x00007e19cf42a28b __libc_start_main ./csu/../csu/libc-start.c:347:5
#37 0x000056f26ca0b665 _start (/mnt/sdd1/sbw/mlir/tool/llvm-project-llvmorg-21.1.8/build/bin/mlir-opt+0x2072665)
Aborted (core dumped)
```
![]()
_______________________________________________
llvm-bugs mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
