[llvm] [VPlan] Run narrowInterleaveGroups during general VPlan optimizations. (PR #149706)

[David Sherwood via llvm-commits]

================
@@ -3173,30 +3174,47 @@ void VPlanTransforms::narrowInterleaveGroups(VPlan &Plan, ElementCount VF,
     //  * recipes writing to memory except interleave groups
     // Only support plans with a canonical induction phi.
     if (R.isPhi())
-      return;
+      return nullptr;
 
     auto *InterleaveR = dyn_cast<VPInterleaveRecipe>(&R);
     if (R.mayWriteToMemory() && !InterleaveR)
-      return;
+      return nullptr;
 
     // Do not narrow interleave groups if there are VectorPointer recipes and
     // the plan was unrolled. The recipe implicitly uses VF from
     // VPTransformState.
     // TODO: Remove restriction once the VF for the VectorPointer offset is
     // modeled explicitly as operand.
     if (isa<VPVectorPointerRecipe>(&R) && Plan.getUF() > 1)
-      return;
+      return nullptr;
 
     // All other ops are allowed, but we reject uses that cannot be converted
     // when checking all allowed consumers (store interleave groups) below.
     if (!InterleaveR)
       continue;
 
-    // Bail out on non-consecutive interleave groups.
-    if (!isConsecutiveInterleaveGroup(InterleaveR, FixedVF, TypeInfo,
-                                      VectorRegWidth))
-      return;
-
+    // Try to find a single VF, where all interleave groups are consecutive and
+    // saturate the full vector width. If we already have a candidate VF, check
+    // if it is applicable for the current InterleaveR, otherwise look for a
+    // suitable VF across the Plans VFs.
+    //
+    if (VFToOptimize) {
+      if (!isConsecutiveInterleaveGroup(InterleaveR, *VFToOptimize, TypeInfo,
+                                        VectorRegWidth))
+        return nullptr;
+    } else {
+      for (ElementCount VF : Plan.vectorFactors()) {
+        if (!VF.isFixed())
----------------
david-arm wrote:

Why are we bailing out for scalable vectors here? Is there a good reason why this cannot work for scalable vectors?

https://github.com/llvm/llvm-project/pull/149706