From f79c5ce90bae0f60ffa44491a6eccf5af18deb60 Mon Sep 17 00:00:00 2001
From: Elena Demikhovsky <elena.demikhovsky@intel.com>
Date: Sun, 15 Nov 2015 08:19:35 +0000
Subject: [PATCH] Fixed GEP visitor in the InstCombine pass. The current
 implementation of GEP visitor in InstCombine fails with assertion on Vector
 GEP with mix of scalar and vector types, like this:

getelementptr double, double* %a, <8 x i32> %i
(It fails to create a "sext" from <8 x i32> to <8 x i64>)

I fixed it and added some tests.

Differential Revision: http://reviews.llvm.org/D14485



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@253162 91177308-0d34-0410-b5e6-96231b3b80d8
---
 .../InstCombine/InstructionCombining.cpp           | 15 +++++++++-----
 test/Transforms/InstCombine/vector_gep2.ll         | 23 ++++++++++++++++++++++
 2 files changed, 33 insertions(+), 5 deletions(-)

diff --git a/lib/Transforms/InstCombine/InstructionCombining.cpp b/lib/Transforms/InstCombine/InstructionCombining.cpp
index f024ab1795e..3ae7f08238b 100644
--- a/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -1355,7 +1355,8 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
   // Eliminate unneeded casts for indices, and replace indices which displace
   // by multiples of a zero size type with zero.
   bool MadeChange = false;
-  Type *IntPtrTy = DL.getIntPtrType(GEP.getPointerOperandType());
+  Type *IntPtrTy =
+    DL.getIntPtrType(GEP.getPointerOperandType()->getScalarType());
 
   gep_type_iterator GTI = gep_type_begin(GEP);
   for (User::op_iterator I = GEP.op_begin() + 1, E = GEP.op_end(); I != E;
@@ -1365,21 +1366,25 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
     if (!SeqTy)
       continue;
 
+    // Index type should have the same width as IntPtr
+    Type *IndexTy = (*I)->getType();
+    Type *NewIndexType = IndexTy->isVectorTy() ?
+      VectorType::get(IntPtrTy, IndexTy->getVectorNumElements()) : IntPtrTy;
+ 
     // If the element type has zero size then any index over it is equivalent
     // to an index of zero, so replace it with zero if it is not zero already.
     if (SeqTy->getElementType()->isSized() &&
         DL.getTypeAllocSize(SeqTy->getElementType()) == 0)
       if (!isa<Constant>(*I) || !cast<Constant>(*I)->isNullValue()) {
-        *I = Constant::getNullValue(IntPtrTy);
+        *I = Constant::getNullValue(NewIndexType);
         MadeChange = true;
       }
 
-    Type *IndexTy = (*I)->getType();
-    if (IndexTy != IntPtrTy) {
+    if (IndexTy != NewIndexType) {
       // If we are using a wider index than needed for this platform, shrink
       // it to what we need.  If narrower, sign-extend it to what we need.
       // This explicit cast can make subsequent optimizations more obvious.
-      *I = Builder->CreateIntCast(*I, IntPtrTy, true);
+      *I = Builder->CreateIntCast(*I, NewIndexType, true);
       MadeChange = true;
     }
   }
diff --git a/test/Transforms/InstCombine/vector_gep2.ll b/test/Transforms/InstCombine/vector_gep2.ll
index d76a7d56cc7..1b80ffd101c 100644
--- a/test/Transforms/InstCombine/vector_gep2.ll
+++ b/test/Transforms/InstCombine/vector_gep2.ll
@@ -9,3 +9,26 @@ define <2 x i8*> @testa(<2 x i8*> %a) {
 ; CHECK: getelementptr i8, <2 x i8*> %a, <2 x i64> <i64 0, i64 1>
   ret <2 x i8*> %g
 }
+
+define <8 x double*> @vgep_s_v8i64(double* %a, <8 x i64>%i) {
+; CHECK-LABEL: @vgep_s_v8i64
+; CHECK: getelementptr double, double* %a, <8 x i64> %i
+  %VectorGep = getelementptr double, double* %a, <8 x i64> %i
+  ret <8 x double*> %VectorGep
+}
+
+define <8 x double*> @vgep_s_v8i32(double* %a, <8 x i32>%i) {
+; CHECK-LABEL: @vgep_s_v8i32
+; CHECK: %1 = sext <8 x i32> %i to <8 x i64>
+; CHECK: getelementptr double, double* %a, <8 x i64> %1
+  %VectorGep = getelementptr double, double* %a, <8 x i32> %i
+  ret <8 x double*> %VectorGep
+}
+
+define <8 x i8*> @vgep_v8iPtr_i32(<8 x i8*> %a, i32 %i) {
+; CHECK-LABEL: @vgep_v8iPtr_i32
+; CHECK:  %1 = sext i32 %i to i64
+; CHECK:  %VectorGep = getelementptr i8, <8 x i8*> %a, i64 %1
+  %VectorGep = getelementptr i8, <8 x i8*> %a, i32 %i
+  ret <8 x i8*> %VectorGep
+}
-- 
2.11.0