[flang-commits] [flang] [mlir] [Flang][OpenMP] Support lowering of simd reductions (PR #112194)

[Sergio Afonso via flang-commits]

skatrak wrote:

This PR triggered a Fujistu testsuite regression: [LLVM-1403](https://linaro.atlassian.net/browse/LLVM-1403). I've looked at it and the problem is related to the fact that we currently ignore the inner loop wrapper in 'do simd' during MLIR to LLVM IR translation. I'll create another PR to address this shortly.

https://github.com/llvm/llvm-project/pull/112194