[llvm] [LV][EVL] Generate negative strided load/store for reversed load/store (PR #123608)

[Mel Chen via llvm-commits]

Mel-Chen wrote:

#128718
I use `instWidening` before VPlan generation to decide between reverse or strided accesses based on cost. This benefits vectorization regardless of whether tail folding with EVL is enabled.
Currently, due to the existing dependency on `VPWidenIntOrFpInductionRecipe`, the VPlan for tail folding with EVL will be bailed out. I will work on eliminating this dependency to resolve the issue this week.

https://github.com/llvm/llvm-project/pull/123608