Don't try to fold gep when the idx is a vector

[davide]

Summary

The code in ConstantFoldGetElementPtr() assumes integer, and therefore it crashes trying to get the integer bidwith of a vector type (in this case <4 x i32>. I just changed the code to prevent the folding in case of vectors and I didn't bother to generalize as this doesn't seem to me something that really happens in practice, but I'm willing to change the patch if you think it's worth it.
This is hard to trigger from -instsimplify or -instcombine only as the second instruction is dead, so the test uses loop-unroll.

[davide]

Update comment(s).

[RKSimon]

LGTM

llvm/trunk/lib/IR/ConstantFold.cpp

if (CE->getOpcode() == Instruction::GetElementPtr) {
       // an i8, we should *not* get:
       // i8* getelementptr ({ [2 x i8], i32, i8, [3 x i8] }* @main.c,
       //                    i32 0, i32 0, i64 8)
       //
       // This GEP tries to index array element '8  which runs out-of-bounds.
       // Subsequent evaluation would get confused and produce erroneous results.
       //
       // The following prohibits such a GEP from being formed by checking to see
-      // if the index is in-range with respect to an array or vector.
+      // if the index is in-range with respect to an array.
+      // TODO: This code may be extended to handle vectors as well.
       bool PerformFold = false;
       if (Idx0->isNullValue())
         PerformFold = true;
       else if (LastI.isSequential())
         if (ConstantInt *CI = dyn_cast<ConstantInt>(Idx0))
-          PerformFold =
-              !LastI.isBoundedSequential() ||
-              isIndexInRangeOfArrayType(LastI.getSequentialNumElements(), CI);
+          PerformFold = (!LastI.isBoundedSequential() ||
+                         isIndexInRangeOfArrayType(
+                             LastI.getSequentialNumElements(), CI)) &&
+                        !CE->getOperand(CE->getNumOperands() - 1)
+                             ->getType()
+                             ->isVectorTy();
 
       if (PerformFold) {
         SmallVector<Value*, 16> NewIndices;
         NewIndices.reserve(Idxs.size() + CE->getNumOperands());
         NewIndices.append(CE->op_begin() + 1, CE->op_end() - 1);
 
         // Add the last index of the source with the first index of the new GEP.
         // Make sure to handle the case when they are actually different types.

llvm/trunk/test/Transforms/InstSimplify/pr33957.ll

+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt -loop-unroll -S %s | FileCheck %s
+
+%struct.bar = type { i32 }
+
+@global = external constant [78 x %struct.bar], align 4
+
+define void @patatino(i32 %x) {
+; CHECK-LABEL: @patatino(
+; CHECK-NEXT:  bb:
+; CHECK-NEXT:    br i1 true, label [[BB1_PREHEADER:%.*]], label [[BB3:%.*]]
+; CHECK:       bb1.preheader:
+; CHECK-NEXT:    br label [[BB1:%.*]]
+; CHECK:       bb1:
+; CHECK-NEXT:    br label [[BB3]]
+; CHECK:       bb3:
+; CHECK-NEXT:    ret void
+;
+bb:
+  br i1 true, label %bb1, label %bb3
+
+bb1:
+  %tmp = getelementptr inbounds [78 x %struct.bar], [78 x %struct.bar]* @global, i32 0, <4 x i32> undef
+  %tmp2 = getelementptr inbounds %struct.bar, <4 x %struct.bar*> %tmp, i32 1
+  br i1 true, label %bb3, label %bb1
+
+bb3:
+  ret void
+}