[polly] [Polly] Reject SCoPs with an escaping scalar defined in an empty-doma… (PR #213656)

[Shikhar Jain via llvm-commits]

ShikharJ-Corp wrote:

> I don't see why we would need to reject the entire SCoP. If the producing ScopStmt is purged because it is known to be never executed and only required because of structural reasons, it can be replaced with a undef (or better: poison) value.
> 
> EDIT: I missed you are already working on it in [#213656 (comment)](https://github.com/llvm/llvm-project/pull/213656#issuecomment-5166632769). I'd prefer if you did not just dump LLM output. By LLVM policy, you must also disclose significant use of AI.

hi .. thanks for the feedback . Regarding long Commit msg and summary, would like to apologies... actually i was trying to understand few important concepts involved with scop-builder itself and given the time-frame in which we usually need to complete the issue , i thought like to have a verification of my understanding itself.. and also let the stakeholders involved to know about the problem as a whole.. thus intentionally i did this and like used LLM to jot down my understanding and present in a simple manner so that like anyone can easily go through it. .. That was the idea.. But i would surely keep this in mind for further commits.. In my next commit for this issue i will be careful and transparent.. 

https://github.com/llvm/llvm-project/pull/213656