Is reduction clause in OpenMP supported?

[roastduck]

I am differencing a program with OpenMP's parallel reduction, but I got the following error:

  %115 = call i32 @__kmpc_reduce_nowait(%struct.ident_t* @8, i32 %23, i32 1, i64 8, i8* %114, void (i8*, i8*)* @.omp.reduction.reduction_func, [8 x i32]* @.gomp_critical_user_.reduction.var) #21, !dbg !1316
clang++: /utils/enzyme/Enzyme-0.0.66-llvm-12/enzyme/Enzyme/AdjointGenerator.h:9824: void AdjointGenerator<AugmentedReturnType>::visitCallInst(llvm::CallInst&) [with AugmentedReturnType = AugmentedReturn*]: Assertion `0 && "unhandled openmp function"' failed.                                                                                                                                                                                                                        PLEASE submit a bug report to https://bugs.llvm.org/ and include the crash backtrace, preprocessed source, and associated run script.
Stack dump:
0.      Program arguments: /utils/clang/clang+llvm-12.0.1-x86_64-linux-gnu-ubuntu-/bin/clang++ /root/.freetensor/Y24MOV/run.cpp -I/utils/freetensor/FreeTensor/runtime -I/usr/local/share/runtime_include -std=c++20 -O3 -fPIC -Wall -fopenmp -c -o /root/.freetensor/Y24MOV/run.o -DFT_WITH_MKL -I/usr/include/mkl -march=native -g -Xclang -load -Xclang /utils/enzyme/Enzyme-0.0.66-llvm-12/enzyme/build/Enzyme/ClangEnzyme-12.so -O0
1.      <eof> parser at end of file
2.      Per-module optimization passes
3.      Running pass 'Enzyme Pass' on module '/root/.freetensor/Y24MOV/run.cpp'.

Is reduction clause of OpenMP supported by Enzyme?

Another related issue is: if I call omp_get_max_threads inside the program, I will get a similar error saying @__kmpc_push_num_threads is not handled.

[tgymnich]

Which reduction op are you using?

[roastduck]

Which reduction op are you using?

"+", min and max.

[tgymnich]

From quickly looking over our code it seems like OpenMP reductions are not yet implemented. Correct me if I am wrong @wsmoses @vchuravy @ludgerpaehler

[ludgerpaehler]

@jdoerfert do you happen to remember, if you ever attempted to implement it during the rush back then?

[ludgerpaehler]

@tgymnich shall we open a further project to keep track of OpenMP Support?

[jdoerfert]

@jdoerfert do you happen to remember, if you ever attempted to implement it during the rush back then?

I don’t think we did