Index: llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
===================================================================
--- llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
+++ llvm/lib/Target/AMDGPU/AMDGPUAsmPrinter.cpp
@@ -833,6 +833,8 @@
         case AMDGPU::EXEC_HI:
         case AMDGPU::SCC:
         case AMDGPU::M0:
+        case AMDGPU::M0_LO16:
+        case AMDGPU::M0_HI16:
         case AMDGPU::SRC_SHARED_BASE:
         case AMDGPU::SRC_SHARED_LIMIT:
         case AMDGPU::SRC_PRIVATE_BASE:
Index: llvm/test/CodeGen/AMDGPU/indirect-addressing-si-gfx9.ll
===================================================================
--- llvm/test/CodeGen/AMDGPU/indirect-addressing-si-gfx9.ll
+++ llvm/test/CodeGen/AMDGPU/indirect-addressing-si-gfx9.ll
@@ -66,6 +66,23 @@
   ret void
 }
 
+declare hidden void @foo()
+
+; For functions with calls, we were not accounting for m0_lo16/m0_hi16
+; uses on the BUNDLE created when expanding the insert register pseudo.
+; GCN-LABEL: {{^}}insertelement_with_call:
+; GCN: s_set_gpr_idx_on s{{[0-9]+}}, gpr_idx(DST)
+; GCN-NEXT: v_mov_b32_e32 {{v[0-9]+}}, 8
+; GCN-NEXT: s_set_gpr_idx_off
+; GCN: s_swappc_b64
+define amdgpu_kernel void @insertelement_with_call(<16 x i32> addrspace(1)* %ptr, i32 %idx) #0 {
+  %vec = load <16 x i32>, <16 x i32> addrspace(1)* %ptr
+  %i6 = insertelement <16 x i32> %vec, i32 8, i32 %idx
+  call void @foo()
+  store <16 x i32> %i6, <16 x i32> addrspace(1)* null
+  ret void
+}
+
 declare i32 @llvm.amdgcn.workitem.id.x() #1
 declare void @llvm.amdgcn.s.barrier() #2