[Mlir-commits] [mlir] [MLIR][OpenMP] Add omp.simd operation (PR #79843)

[Michael Klemm]

mjklemm wrote:

> Thinking about it a bit more, it would be possible to make some sort of 2-level representation system for composite constructs. The first level would be to just represent each allowed composite construct as its own MLIR operation, which is what the frontend would produce.

While it's not strictly related to this PR, since this is a short-term implementation that likely will replace with the final implementation, I'd be careful with this in the long run.  The OpenMP API has several composite constructs and the list might be growing with 6.0.  If each of them needs its own MLIR operation, we will end up with many MLIR operations.  So, we need to find a way to come up with a scalable representation, that can accommodate for all combinations possible w/o polluting the MLIR operations space too much.

https://github.com/llvm/llvm-project/pull/79843