[Mlir-commits] [mlir] [mlir][ExecutionEngine] Add support for global constructors and destructors (PR #78070)

[Cullen Rhodes]

c-rhodes wrote:

Hi, I've reverted this (#78164) this since it's causing test failures on AArch64 linux, hitting the following assert:

```
# | mlir-cpu-runner: /home/culrho01/llvm-project/llvm/lib/ExecutionEngine/RuntimeDyld/RuntimeDyldELF.cpp:519: void llvm::RuntimeDyldELF::resolveAArch64Relocation(const SectionEntry &, uint64_t, uint64_t, uint32_t, int64_t): Assertion `isInt<33>(Result) && "overflow check failed for relocation"' failed.
# | PLEASE submit a bug report to https://github.com/llvm/llvm-project/issues/ and include the crash backtrace.
# | Stack dump:
# | 0.  Program arguments: /home/culrho01/llvm-project/build/bin/mlir-cpu-runner -march=aarch64 -mattr=+sve,+sme -e entry -entry-point-result=void -shared-libs=/home/culrho01/llvm-project/build/lib/libmlir_runner_utils.so,/home/culrho01/llvm-project/build/lib/libmlir_c_runner_utils.so,/home/culrho01/llvm-project/build/lib/libmlir_arm_sme_abi_stubs.so
```

Seeing the same in buildbot as well, e.g.
https://lab.llvm.org/buildbot/#/builders/179/builds/9094/steps/12/logs/FAIL__MLIR__sparse_codegen_dim_mlir

This seems related to https://github.com/llvm/llvm-project/issues/71963 which there's an open PR https://github.com/llvm/llvm-project/pull/71968 for, I've tested that patch but it doesn't fix the issue.

https://github.com/llvm/llvm-project/pull/78070