[llvm] [LoopVectorize] Add cost of generating tail-folding mask to the loop (PR #90191)

[David Sherwood via llvm-commits]

================

----------------
david-arm wrote:

The tests have changed because the active lane mask cost is the same for VF=8 and VF=16. As a result the per scalar iteration cost is lower for VF=16.

https://github.com/llvm/llvm-project/pull/90191