[AMDGPU] gfx1010 wave32 icmp/fcmp intrinsic changes for wave32

Differential Revision: https://reviews.llvm.org/D63301

llvm-svn: 363339
diff --git a/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp b/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp
index 4756a77..1c503c2 100644
--- a/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp
+++ b/llvm/lib/Target/AMDGPU/AMDGPUAtomicOptimizer.cpp
@@ -249,7 +249,8 @@
   // We need to know how many lanes are active within the wavefront, and we do
   // this by doing a ballot of active lanes.
   CallInst *const Ballot =
-      B.CreateIntrinsic(Intrinsic::amdgcn_icmp, {B.getInt32Ty()},
+      B.CreateIntrinsic(Intrinsic::amdgcn_icmp,
+                        {B.getInt64Ty(), B.getInt32Ty()},
                         {B.getInt32(1), B.getInt32(0), B.getInt32(33)});
 
   // We need to know how many lanes are active within the wavefront that are