From fcbe3d950148f01ed8ed47afda78a13f81a8010b Mon Sep 17 00:00:00 2001
From: Andrew Trick <atrick@apple.com>
Date: Thu, 2 Jan 2014 19:29:38 +0000
Subject: [PATCH] indvars: insert truncate at loop boundary to avoid redundant
 IVs.

When widening an IV to remove s/zext, we generally try to eliminate
the original narrow IV. However, LCSSA phi nodes outside the loop were
still using the original IV. Clean this up more aggressively to avoid
redundancy in generated code.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@198338 91177308-0d34-0410-b5e6-96231b3b80d8
---
 lib/Transforms/Scalar/IndVarSimplify.cpp   | 17 +++++++---
 test/Transforms/IndVarSimplify/iv-widen.ll | 39 ++++++++++++++++++++++
 2 files changed, 51 insertions(+), 5 deletions(-)
 create mode 100644 test/Transforms/IndVarSimplify/iv-widen.ll

diff --git a/lib/Transforms/Scalar/IndVarSimplify.cpp b/lib/Transforms/Scalar/IndVarSimplify.cpp
index c291f68bd63..c992f8c12a6 100644
--- a/lib/Transforms/Scalar/IndVarSimplify.cpp
+++ b/lib/Transforms/Scalar/IndVarSimplify.cpp
@@ -897,15 +897,24 @@ const SCEVAddRecExpr *WidenIV::GetWideRecurrence(Instruction *NarrowUse) {
   return AddRec;
 }
 
+/// This IV user cannot be widen. Replace this use of the original narrow IV
+/// with a truncation of the new wide IV to isolate and eliminate the narrow IV.
+static void truncateIVUse(NarrowIVDefUse DU, DominatorTree *DT) {
+  IRBuilder<> Builder(getInsertPointForUses(DU.NarrowUse, DU.NarrowDef, DT));
+  Value *Trunc = Builder.CreateTrunc(DU.WideDef, DU.NarrowDef->getType());
+  DU.NarrowUse->replaceUsesOfWith(DU.NarrowDef, Trunc);
+}
+
 /// WidenIVUse - Determine whether an individual user of the narrow IV can be
 /// widened. If so, return the wide clone of the user.
 Instruction *WidenIV::WidenIVUse(NarrowIVDefUse DU, SCEVExpander &Rewriter) {
 
   // Stop traversing the def-use chain at inner-loop phis or post-loop phis.
   if (isa<PHINode>(DU.NarrowUse) &&
-      LI->getLoopFor(DU.NarrowUse->getParent()) != L)
+      LI->getLoopFor(DU.NarrowUse->getParent()) != L) {
+    truncateIVUse(DU, DT);
     return 0;
-
+  }
   // Our raison d'etre! Eliminate sign and zero extension.
   if (IsSigned ? isa<SExtInst>(DU.NarrowUse) : isa<ZExtInst>(DU.NarrowUse)) {
     Value *NewDef = DU.WideDef;
@@ -953,9 +962,7 @@ Instruction *WidenIV::WidenIVUse(NarrowIVDefUse DU, SCEVExpander &Rewriter) {
     // This user does not evaluate to a recurence after widening, so don't
     // follow it. Instead insert a Trunc to kill off the original use,
     // eventually isolating the original narrow IV so it can be removed.
-    IRBuilder<> Builder(getInsertPointForUses(DU.NarrowUse, DU.NarrowDef, DT));
-    Value *Trunc = Builder.CreateTrunc(DU.WideDef, DU.NarrowDef->getType());
-    DU.NarrowUse->replaceUsesOfWith(DU.NarrowDef, Trunc);
+    truncateIVUse(DU, DT);
     return 0;
   }
   // Assume block terminators cannot evaluate to a recurrence. We can't to
diff --git a/test/Transforms/IndVarSimplify/iv-widen.ll b/test/Transforms/IndVarSimplify/iv-widen.ll
new file mode 100644
index 00000000000..12067ab6fe8
--- /dev/null
+++ b/test/Transforms/IndVarSimplify/iv-widen.ll
@@ -0,0 +1,39 @@
+; RUN: opt < %s -indvars -S | FileCheck %s
+
+target triple = "x86_64-apple-darwin"
+
+; CHECK-LABEL: @sloop
+; CHECK-LABEL: B18:
+; Only one phi now.
+; CHECK: phi
+; CHECK-NOT: phi
+; We now get 2 trunc, one for the gep and one for the lcssa phi.
+; CHECK: trunc i64 %indvars.iv to i32
+; CHECK: trunc i64 %indvars.iv to i32
+; CHECK-LABEL: B24:
+define void @sloop(i32* %a) {
+Prologue:
+  br i1 undef, label %B18, label %B6
+
+B18:                                        ; preds = %B24, %Prologue
+  %.02 = phi i32 [ 0, %Prologue ], [ %tmp33, %B24 ]
+  %tmp23 = zext i32 %.02 to i64
+  %tmp33 = add i32 %.02, 1
+  %o = getelementptr i32* %a, i32 %.02
+  %v = load i32* %o
+  %t = icmp eq i32 %v, 0
+  br i1 %t, label %exit24, label %B24
+
+B24:                                        ; preds = %B18
+  %t2 = icmp eq i32 %tmp33, 20
+  br i1 %t2, label %B6, label %B18
+
+B6:                                       ; preds = %Prologue
+  ret void
+
+exit24:                      ; preds = %B18
+  call void @dummy(i32 %.02)
+  unreachable
+}
+
+declare void @dummy(i32)