[llvm] [NFCI][VPlan] Split initial mem-widening into a separate transformation (PR #182592)

[Florian Hahn via llvm-commits]

================
@@ -90,6 +78,26 @@ class VPRecipeBuilder {
   VPRecipeBase *tryToCreateWidenNonPhiRecipe(VPSingleDefRecipe *R,
                                              VFRange &Range);
 
+  /// Check if the load or store instruction \p VPI should widened for \p
+  /// Range.Start and potentially masked. Such instructions are handled by a
+  /// recipe that takes an additional VPInstruction for the mask.
+  VPRecipeBase *tryToWidenMemory(VPInstruction *VPI, VFRange &Range);
+
+  /// If \p VPI represents a histogram operation (as determined by
+  /// LoopVectorizationLegality) make that safe for vectorization, by emitting a
+  /// llvm.experimental.vector.histogram.add intrinsic in place of the Load +
+  /// Add|Sub + Store operations that perform the histogram in the original
+  /// scalar loop.
+  VPHistogramRecipe *widenIfHistogram(VPInstruction *VPI);
+
+  /// The stores with invariant address inside the loop will be deleted, and in
----------------
fhahn wrote:


nit: would be good to tie the comment to the passed in VPI, perhaps something along the lines below.

```suggestion
  /// If \p VPI is a store with an invariant address, delete it. If it is the final store of a reduction result,  a uniform store recipe will be created for the final invariant store of the reduction result in the middle block.
```

https://github.com/llvm/llvm-project/pull/182592