diff --git a/lib/Target/R600/MCTargetDesc/R600MCCodeEmitter.cpp b/lib/Target/R600/MCTargetDesc/R600MCCodeEmitter.cpp
index 01df8087d5f..2171f902630 100644
--- a/lib/Target/R600/MCTargetDesc/R600MCCodeEmitter.cpp
+++ b/lib/Target/R600/MCTargetDesc/R600MCCodeEmitter.cpp
@@ -161,6 +161,7 @@ void R600MCCodeEmitter::EncodeInstruction(const MCInst &MI, raw_ostream &OS,
     case AMDGPU::VTX_READ_PARAM_8_eg:
     case AMDGPU::VTX_READ_PARAM_16_eg:
     case AMDGPU::VTX_READ_PARAM_32_eg:
+    case AMDGPU::VTX_READ_PARAM_128_eg:
     case AMDGPU::VTX_READ_GLOBAL_8_eg:
     case AMDGPU::VTX_READ_GLOBAL_32_eg:
     case AMDGPU::VTX_READ_GLOBAL_128_eg:
diff --git a/lib/Target/R600/R600Instructions.td b/lib/Target/R600/R600Instructions.td
index 373a793c80d..c9885a35115 100644
--- a/lib/Target/R600/R600Instructions.td
+++ b/lib/Target/R600/R600Instructions.td
@@ -1491,6 +1491,10 @@ def VTX_READ_PARAM_32_eg : VTX_READ_32_eg <0,
   [(set (i32 R600_TReg32_X:$dst), (load_param ADDRVTX_READ:$ptr))]
 >;
 
+def VTX_READ_PARAM_128_eg : VTX_READ_128_eg <0,
+  [(set (v4i32 R600_Reg128:$dst), (load_param ADDRVTX_READ:$ptr))]
+>;
+
 //===----------------------------------------------------------------------===//
 // VTX Read from global memory space
 //===----------------------------------------------------------------------===//
diff --git a/test/CodeGen/R600/128bit-kernel-args.ll b/test/CodeGen/R600/128bit-kernel-args.ll
new file mode 100644
index 00000000000..114f9e74474
--- /dev/null
+++ b/test/CodeGen/R600/128bit-kernel-args.ll
@@ -0,0 +1,18 @@
+;RUN: llc < %s -march=r600 -mcpu=redwood | FileCheck %s
+
+; CHECK: @v4i32_kernel_arg
+; CHECK: VTX_READ_128 T{{[0-9]+}}.XYZW, T{{[0-9]+}}.X, 40
+
+define void @v4i32_kernel_arg(<4 x i32> addrspace(1)* %out, <4 x i32>  %in) {
+entry:
+  store <4 x i32> %in, <4 x i32> addrspace(1)* %out
+  ret void
+}
+
+; CHECK: @v4f32_kernel_arg
+; CHECK: VTX_READ_128 T{{[0-9]+}}.XYZW, T{{[0-9]+}}.X, 40
+define void @v4f32_kernel_args(<4 x float> addrspace(1)* %out, <4 x float>  %in) {
+entry:
+  store <4 x float> %in, <4 x float> addrspace(1)* %out
+  ret void
+}