[LoopVectorize] Pointer indicies may be wider than the pointer

If we are dealing with a pointer induction variable, isInductionPHI
gives back a step value of Stride / size of pointer.  However, we might
be indexing with a legal type wider than the pointer width.
Handle this by inserting casts where appropriate instead of crashing.

This fixes PR23954.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@240877 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Transforms/Vectorize/LoopVectorize.cpp b/lib/Transforms/Vectorize/LoopVectorize.cpp
index 49b07dadd8e2834490bbc7ef6f0d2bf285095a56..eff7c03c7f7e13edc9f9d3398d964515033ad22d 100644 (file)
--- a/lib/Transforms/Vectorize/LoopVectorize.cpp
+++ b/lib/Transforms/Vectorize/LoopVectorize.cpp
@@ -850,6 +850,8 @@ public:
         return B.CreateAdd(StartValue, Index);
 
       case IK_PtrInduction:
+        assert(Index->getType() == StepValue->getType() &&
+               "Index type does not match StepValue type");
         if (StepValue->isMinusOne())
           Index = B.CreateNeg(Index);
         else if (!StepValue->isOne())
@@ -2798,7 +2800,10 @@ void InnerLoopVectorizer::createEmptyLoop() {
       break;
     }
     case LoopVectorizationLegality::IK_PtrInduction: {
-      EndValue = II.transform(BypassBuilder, CountRoundDown);
+      Value *CRD = BypassBuilder.CreateSExtOrTrunc(CountRoundDown,
+                                                   II.StepValue->getType(),
+                                                   "cast.crd");
+      EndValue = II.transform(BypassBuilder, CRD);
       EndValue->setName("ptr.ind.end");
       break;
     }
@@ -3448,12 +3453,14 @@ void InnerLoopVectorizer::widenPHIInstruction(Instruction *PN,
       // This is the normalized GEP that starts counting at zero.
       Value *NormalizedIdx =
           Builder.CreateSub(Induction, ExtendedIdx, "normalized.idx");
+      NormalizedIdx =
+          Builder.CreateSExtOrTrunc(NormalizedIdx, II.StepValue->getType());
       // This is the vector of results. Notice that we don't generate
       // vector geps because scalar geps result in better code.
       for (unsigned part = 0; part < UF; ++part) {
         if (VF == 1) {
           int EltIndex = part;
-          Constant *Idx = ConstantInt::get(Induction->getType(), EltIndex);
+          Constant *Idx = ConstantInt::get(NormalizedIdx->getType(), EltIndex);
           Value *GlobalIdx = Builder.CreateAdd(NormalizedIdx, Idx);
           Value *SclrGep = II.transform(Builder, GlobalIdx);
           SclrGep->setName("next.gep");
@@ -3464,7 +3471,7 @@ void InnerLoopVectorizer::widenPHIInstruction(Instruction *PN,
         Value *VecVal = UndefValue::get(VectorType::get(P->getType(), VF));
         for (unsigned int i = 0; i < VF; ++i) {
           int EltIndex = i + part * VF;
-          Constant *Idx = ConstantInt::get(Induction->getType(), EltIndex);
+          Constant *Idx = ConstantInt::get(NormalizedIdx->getType(), EltIndex);
           Value *GlobalIdx = Builder.CreateAdd(NormalizedIdx, Idx);
           Value *SclrGep = II.transform(Builder, GlobalIdx);
           SclrGep->setName("next.gep");

diff --git a/test/Transforms/LoopVectorize/X86/ptr-indvar-crash.ll b/test/Transforms/LoopVectorize/X86/ptr-indvar-crash.ll
new file mode 100644 (file)
index 0000000..56ea2f4
--- /dev/null
+++ b/test/Transforms/LoopVectorize/X86/ptr-indvar-crash.ll
@@ -0,0 +1,20 @@
+; RUN: opt -slp-vectorizer -S %s
+
+target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
+target triple = "x86_64-unknown-linux-gnu"
+
+define void @f(i128 %p1) {
+entry:
+  br label %while.body
+
+while.body:
+  %p.05 = phi i8* [ %add.ptr, %while.body ], [ null, %entry ]
+  %p1.addr.04 = phi i128 [ %sub, %while.body ], [ %p1, %entry ]
+  %add.ptr = getelementptr inbounds i8, i8* %p.05, i32 2
+  %sub = add nsw i128 %p1.addr.04, -2
+  %tobool = icmp eq i128 %sub, 0
+  br i1 %tobool, label %while.end, label %while.body
+
+while.end:
+  ret void
+}