[clang] [llvm] [clang][OpenMP] Improve loop structure for distributed loops (PR #201670)

[via cfe-commits]

carlobertolli wrote:

I don't see a condition over the loop having reductions: does it apply to any ttdpf, even without reductions?
If so, do you have performance numbers for those cases? The ones you show have the label "red" in front of them, but I see at least one test in the suite that does not: "evict_device_cache".
FYI: I am expecting this change to apply to non reduction cases.

https://github.com/llvm/llvm-project/pull/201670