[VPlan] Place VPExpandSCEVRecipe in pre-header.

[fhahn]

Summary

After D121624 models the pre-header in VPlan, VPExpandSCEVRecipes can be
placed there. This ensures SCEV expansion happens before modifying the
CFG during VPlan execution, when CFG is incomplete.

Depends on D121624.

[Ayal]

Looks good to me!

[fhahn]

Rebase to trigger pre-commit tests before landing.

llvm/lib/Transforms/Vectorize/VPlan.h

bool onlyFirstLaneUsed(const VPValue *Op) const override {
 
     // Widened, consecutive memory operations only demand the first lane of
     // their address.
     return Op == getAddr() && isConsecutive();
   }
 };
 
 /// Recipe to expand a SCEV expression.
-/// TODO: Currently the recipe always expands the expression in the loop
-/// pre-header, but the recipe is currently placed in the header; place it in
-/// the pre-header once the latter is modeled in VPlan as a VPBasicBlock.
 class VPExpandSCEVRecipe : public VPRecipeBase, public VPValue {
   const SCEV *Expr;
   ScalarEvolution &SE;
 
 public:
   VPExpandSCEVRecipe(const SCEV *Expr, ScalarEvolution &SE)
       : VPRecipeBase(VPExpandSCEVSC, {}), VPValue(nullptr, this), Expr(Expr),
         SE(SE) {}

llvm/lib/Transforms/Vectorize/VPlan.cpp

 }
 #endif
 
 void VPExpandSCEVRecipe::execute(VPTransformState &State) {
   assert(!State.Instance && "cannot be used in per-lane");
   const DataLayout &DL = State.CFG.PrevBB->getModule()->getDataLayout();
   SCEVExpander Exp(SE, DL, "induction");
 
-  BasicBlock *VectorPH = State.CFG.getPreheaderBBFor(this);
-  Value *Res =
-      Exp.expandCodeFor(Expr, Expr->getType(), VectorPH->getTerminator());
+  Value *Res = Exp.expandCodeFor(Expr, Expr->getType(),
+                                 &*State.Builder.GetInsertPoint());
 
   for (unsigned Part = 0, UF = State.UF; Part < UF; ++Part)
     State.set(this, Res, Part);
 }
 
 #if !defined(NDEBUG) || defined(LLVM_ENABLE_DUMP)
 void VPExpandSCEVRecipe::print(raw_ostream &O, const Twine &Indent,
                                VPSlotTracker &SlotTracker) const {

llvm/lib/Transforms/Vectorize/VPlanTransforms.cpp

for (VPRecipeBase &Phi : HeaderVPBB->phis()) {
     }
 
     Instruction *TruncI = IV->getTruncInst();
     VPScalarIVStepsRecipe *Steps = new VPScalarIVStepsRecipe(
         IV->getPHINode()->getType(), ID, Plan.getCanonicalIV(),
         IV->getStartValue(), Step, TruncI ? TruncI->getType() : nullptr);
 
     HeaderVPBB->insert(Steps, HeaderVPBB->getFirstNonPhi());
-    if (Step->getDef()) {
-      // TODO: Place the step in the preheader, once it is explicitly modeled in
-      // VPlan.
-      HeaderVPBB->insert(cast<VPRecipeBase>(Step->getDef()),
-                         HeaderVPBB->getFirstNonPhi());
-    }
+    if (Step->getDef())
+      Plan.getEntry()->getEntryBasicBlock()->appendRecipe(
+          cast<VPRecipeBase>(Step->getDef()));
 
     // If there are no vector users of IV, simply update all users to use Step
     // instead.
     if (!IV->needsVectorIV()) {
       IV->replaceAllUsesWith(Steps);
       continue;
     }
 

llvm/test/Transforms/LoopVectorize/vplan-printing.ll

 declare float @llvm.fmuladd.f32(float, float, float)
 
 define void @print_expand_scev(i64 %y, i8* %ptr) {
 ; CHECK-LABEL: Checking a loop in 'print_expand_scev'
 ; CHECK: VPlan 'Initial VPlan for VF={4},UF>=1' {
 ; CHECK-NEXT: Live-in vp<%0> = vector-trip-count
 ; CHECK-EMPTY:
 ; CHECK-NEXT: vector.ph:
+; CHECK-NEXT:   EMIT vp<[[EXP_SCEV:%.+]]> = EXPAND SCEV (1 + (%y /u 492802768830814060))<nuw><nsw>
 ; CHECK-NEXT: Successor(s): vector loop
 ; CHECK-EMPTY:
 ; CHECK-NEXT: <x1> vector loop: {
 ; CHECK-NEXT:   vector.body:
 ; CHECK-NEXT:     EMIT vp<[[CAN_IV:%.+]]> = CANONICAL-INDUCTION
 ; CHECK-NEXT:     WIDEN-INDUCTION\l" +
 ; CHECK-NEXT:     "  %iv = phi %iv.next, 0\l" +
 ; CHECK-NEXT:     "  ir<%v2>
-; CHECK-NEXT:     EMIT vp<[[EXP_SCEV:%.+]]> = EXPAND SCEV (1 + (%y /u 492802768830814060))<nuw><nsw>
 ; CHECK-NEXT:     vp<[[STEPS:%.+]]> = SCALAR-STEPS vp<[[CAN_IV]]>, ir<0>, vp<[[EXP_SCEV]]>
 ; CHECK-NEXT:     WIDEN ir<%v3> = add ir<%v2>, ir<1>
 ; CHECK-NEXT:     REPLICATE ir<%gep> = getelementptr ir<%ptr>, vp<[[STEPS]]>
 ; CHECK-NEXT:     REPLICATE store ir<%v3>, ir<%gep>
 ; CHECK-NEXT:     EMIT vp<[[CAN_INC:%.+]]> = VF * UF +(nuw)  vp<[[CAN_IV]]>
 ; CHECK-NEXT:     EMIT branch-on-count  vp<[[CAN_INC]]> vp<%0>
 ; CHECK-NEXT:   No successors
 ; CHECK-NEXT: }