[llvm] [RISCV] Prefer VLS over VLA if costs are equal (PR #100564)

[Philip Reames via llvm-commits]

preames wrote:

> Compile TSVC with `clang -O2 -mcpu=spacemit-x60 -Xclang -target-feature -Xclang +use-fixed-over-scalable-if-equal-cost` and test it on Spacemit K1 board, we can see some improvments (we can reduce 1 scalar instruction in some loops like `s121`).

Can you share actual perf results?  And maybe an example where we actually save the instruction?  I'm wondering if we have an opportunity to improve the scalable lowering.

https://github.com/llvm/llvm-project/pull/100564