[llvm] [RISCV][CostModel] VPIntrinsics have same cost as their non-vp counterparts (PR #67178)

[Stella Laurenzo via llvm-commits]

stellaraccident wrote:

> > may I push a revert? We can try to generate a better repro asynchronously.
> 
> Sure thing. Can you please show me how I can reproduce this upstream?

Verified that reverting this patch in our branch does fix the crash, so I'm considering this a strong indicator. I'm getting the recipe from my team for how to extract the LLVM-level reproducer.

https://github.com/llvm/llvm-project/pull/67178