[ConstantFolding] Fix assertion failure on non-power-of-two vector load.

[foad]

Summary

The test case does an (out of bounds) load from a global constant with
type <3 x float>. InstSimplify tried to turn this into an integer load
of the whole alloc size of the vector, which is 128 bits due to
alignment padding, and then bitcast this to <3 x vector> which failed
an assertion due to the type size mismatch.

The fix is to do an integer load of the normal size of the vector, with
no alignment padding.

Change-Id: I911d744b6f048fb18df5814dfef21a9d8259aeb6

[hfinkel]

The test case does an (out of bounds) load from a global constant with type <3 x float>.

Do you need an out-of-bounds access to trigger the problem?

[foad]

In D63375#1544851, @hfinkel wrote:

The test case does an (out of bounds) load from a global constant with type <3 x float>.

Do you need an out-of-bounds access to trigger the problem?

Yes to prevent it from being handled earlier by llvm::ConstantFoldLoadThroughGEPConstantExpr.

[arsenm]

LGTM with nit

test/Transforms/InstSimplify/load.ll

34 ↗

(On Diff #204915)

Can you add a comment that this is an out of bounds load

llvm/trunk/lib/Analysis/ConstantFolding.cpp

if (!IntType) {
     if (LoadTy->isHalfTy())
       MapTy = Type::getInt16Ty(C->getContext());
     else if (LoadTy->isFloatTy())
       MapTy = Type::getInt32Ty(C->getContext());
     else if (LoadTy->isDoubleTy())
       MapTy = Type::getInt64Ty(C->getContext());
     else if (LoadTy->isVectorTy()) {
       MapTy = PointerType::getIntNTy(C->getContext(),
-                                     DL.getTypeAllocSizeInBits(LoadTy));
+                                     DL.getTypeSizeInBits(LoadTy));
     } else
       return nullptr;
 
     C = FoldBitCast(C, MapTy->getPointerTo(AS), DL);
     if (Constant *Res = FoldReinterpretLoadFromConstPtr(C, MapTy, DL))
       return FoldBitCast(Res, LoadTy, DL);
     return nullptr;
   }

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

 @GV = private constant [8 x i32] [i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48, i32 49]
 
 define <8 x i32> @partial_load() {
 ; CHECK-LABEL: @partial_load(
 ; CHECK:         ret <8 x i32> <i32 0, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48>
   %load = load <8 x i32>, <8 x i32>* bitcast (i32* getelementptr ([8 x i32], [8 x i32]* @GV, i64 0, i64 -1) to <8 x i32>*)
   ret <8 x i32> %load
 }
+
+@constvec = internal constant <3 x float> <float 0xBFDA20BC40000000, float 0xBFE6A09EE0000000, float 0x3FE279A760000000>
+
+; This does an out of bounds load from the global constant
+define <3 x float> @load_vec3() {
+; CHECK-LABEL: @load_vec3(
+; CHECK-NEXT:    ret <3 x float> undef
+  %1 = load <3 x float>, <3 x float>* getelementptr inbounds (<3 x float>, <3 x float>* @constvec, i64 1)
+  ret <3 x float> %1
+}