Dmitry Babokin

### What happened? Here's the full reproducer: ``` > cat module.torch module { func.func @main(%arg0: !torch.vtensor) -> !torch.vtensor { %int1 = torch.constant.int 1 %true = torch.constant.bool true %none = torch.constant.none...

avx512knl-* targets EOL

2

LLVM is removing support for Xeon Phi in `19.0` (current trunk), so we have to remove/disable `avx512knl-*` targets for builds with LLVM 19.0. No need to remove it with older...

LLVM rebuild on macOS should be switched back to shared workers

1

Current LLVM rebuild pipeline uses self-hosted runner for macOS: https://github.com/ispc/ispc/blob/9e749459630c8fc639d0fca828a6cf681d13b255/.github/workflows/reusable.rebuild.yml#L306 This is because public runners were not able to build LLVM without running out of time. Current Shared workers seem...

Handle OpenMP libraries in portable way (issue #959)

6

Closes #959 See the issue for detailed problem description.

OpenMP libraries a linked using the library name (i.e. `omp`): https://github.com/plaidml/tpp-mlir/blob/9c53a05fb56128cda8869a563b999e90b31f2e17/cmake/modules/xsmm-dnn.cmake#L47-L50 https://github.com/plaidml/tpp-mlir/blob/9c53a05fb56128cda8869a563b999e90b31f2e17/tools/tpp-run/CMakeLists.txt#L12-L14 This is not portable, as it assumes the `libomp.so` is in standard location, which might not be the...

Fixes #3071. The issue has long investigation of the problem, and I think that the proposed fix is the right one, as we need to preserve compatibility with older macOS...

Here's an error message: ``` /Users/dbabokin/ispc/dbabokin_ispc/wt/20_0_fix/src/util.cpp:52:11: error: '__libcpp_verbose_abort' is missing exception specification 'noexcept' 52 | void std::__libcpp_verbose_abort(char const *format, ...) { | ^ | noexcept /Users/dbabokin/llvm/bin-trunk/bin/../include/c++/v1/__verbose_abort:24:49: note: previous declaration is...

When building with `-DTORCH_MLIR_ENABLE_STABLEHLO=OFF` and running lit tests, I see three fails: ``` TORCH_MLIR :: Conversion/TorchToLinalg/sparse.mlir TORCH_MLIR :: Dialect/Torch/invalid.mlir TORCH_MLIR :: Dialect/Torch/ops.mlir ``` All of the fails are due to...

Compiling with clang 16.0 on macOS I have warnings about incorrect printf format (see below). Values to be printed are `int64_t`, but they are printed with `%zu` and `%ld`, which...

There's something strange about how `torch_mlir._mlir_libs._jit_ir_importer` is defined in the build system. When building Python packages, `jit_ir_importer` is declared as an extension: https://github.com/llvm/torch-mlir/blob/9a6fe58a027d701eff6799e86a65535a8c2f3708/setup.py#L238-L242 It's declared without a build recipe or...