[llvm] [AArch64] Disconnect FeatureUseScalarIncVL from FeatureSVE2. (PR #114568)

[David Green via llvm-commits]

================
@@ -797,6 +794,9 @@ def FeatureUseFixedOverScalableIfEqualCost: SubtargetFeature<"use-fixed-over-sca
   "UseFixedOverScalableIfEqualCost", "true",
   "Prefer fixed width loop vectorization over scalable if the cost-model assigns equal costs">;
 
+def FeatureDontUseScalarIncVL : SubtargetFeature<"dont-use-scalar-inc-vl",
----------------
davemgreen wrote:

Yeah that should be OK. It should at least make the patch simple and I think just involves inverting all the logic, so turning the option on when we have SVE2/SME (as opposed to turning it off when we have only SVE-1). Having it default to addvl felt like a better default.  Do you have a particular reason to not want to change it?

(I'm not sure if dont-use-scalar-inc-vl is the best name, to be honest. It might be better to have something like preferAddCntToAddvl to avoid the negatives and make it a little more explicit)



https://github.com/llvm/llvm-project/pull/114568