Index: include/llvm/IR/IntrinsicsAMDGPU.td
===================================================================
--- include/llvm/IR/IntrinsicsAMDGPU.td
+++ include/llvm/IR/IntrinsicsAMDGPU.td
@@ -472,6 +472,14 @@
   GCCBuiltin<"__builtin_amdgcn_s_getreg">,
   Intrinsic<[llvm_i32_ty], [llvm_i32_ty], [IntrReadMem]>;
 
+// __builtin_amdgcn_interp_mov <param>, <attr_chan>, <attr>, <m0>
+// param values: 0 = P10, 1 = P20, 2 = P0
+def int_amdgcn_interp_mov :
+  GCCBuiltin<"__builtin_amdgcn_interp_mov">,
+  Intrinsic<[llvm_float_ty],
+            [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty, llvm_i32_ty],
+            [IntrNoMem]>;
+
 // __builtin_amdgcn_interp_p1 <i>, <attr_chan>, <attr>, <m0>
 def int_amdgcn_interp_p1 :
   GCCBuiltin<"__builtin_amdgcn_interp_p1">,
Index: lib/Target/AMDGPU/SIISelLowering.cpp
===================================================================
--- lib/Target/AMDGPU/SIISelLowering.cpp
+++ lib/Target/AMDGPU/SIISelLowering.cpp
@@ -2525,6 +2525,12 @@
     return DAG.getNode(AMDGPUISD::INTERP_P2, DL, MVT::f32, P1, J,
                              Op.getOperand(1), Op.getOperand(2), Glue);
   }
+  case Intrinsic::amdgcn_interp_mov: {
+    SDValue M0 = copyToM0(DAG, DAG.getEntryNode(), DL, Op.getOperand(4));
+    SDValue Glue = M0.getValue(1);
+    return DAG.getNode(AMDGPUISD::INTERP_MOV, DL, MVT::f32, Op.getOperand(1),
+                       Op.getOperand(2), Op.getOperand(3), Glue);
+  }
   case Intrinsic::amdgcn_interp_p1: {
     SDValue M0 = copyToM0(DAG, DAG.getEntryNode(), DL, Op.getOperand(4));
     SDValue Glue = M0.getValue(1);
Index: test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll
===================================================================
--- test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll
+++ test/CodeGen/AMDGPU/llvm.amdgcn.interp.ll
@@ -6,6 +6,7 @@
 ;GCN: s_mov_b32 m0, s{{[0-9]+}}
 ;GCN: v_interp_p1_f32
 ;GCN: v_interp_p2_f32
+;GCN: v_interp_mov_f32
 define amdgpu_ps void @v_interp(<16 x i8> addrspace(2)* inreg, <16 x i8> addrspace(2)* inreg, <32 x i8> addrspace(2)* inreg, i32 inreg, <2 x float>) {
 main_body:
   %i = extractelement <2 x float> %4, i32 0
@@ -14,7 +15,9 @@
   %p1_0 = call float @llvm.amdgcn.interp.p2(float %p0_0, float %j, i32 0, i32 0, i32 %3)
   %p0_1 = call float @llvm.amdgcn.interp.p1(float %i, i32 1, i32 0, i32 %3)
   %p1_1 = call float @llvm.amdgcn.interp.p2(float %p0_1, float %j, i32 1, i32 0, i32 %3)
-  call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 1, float %p0_0, float %p0_0, float %p1_1, float %p1_1)
+  %const = call float @llvm.amdgcn.interp.mov(i32 2, i32 0, i32 0, i32 %3)
+  %w = fadd float %p1_1, %const
+  call void @llvm.SI.export(i32 15, i32 1, i32 1, i32 0, i32 1, float %p0_0, float %p0_0, float %p1_1, float %w)
   ret void
 }
 
@@ -24,6 +27,8 @@
 ; Function Attrs: nounwind readnone
 declare float @llvm.amdgcn.interp.p2(float, float, i32, i32, i32) #0
 
+declare float @llvm.amdgcn.interp.mov(i32, i32, i32, i32) #0
+
 declare void @llvm.SI.export(i32, i32, i32, i32, i32, float, float, float, float)
 
 attributes #0 = { nounwind readnone }