Index: include/clang/Basic/Specifiers.h
===================================================================
--- include/clang/Basic/Specifiers.h
+++ include/clang/Basic/Specifiers.h
@@ -251,6 +251,7 @@
     CC_PreserveMost, // __attribute__((preserve_most))
     CC_PreserveAll,  // __attribute__((preserve_all))
     CC_AArch64VectorCall, // __attribute__((aarch64_vector_pcs))
+    CC_AMDGPUVectorCall, // __attribute__((vectorcall))
   };
 
   /// Checks whether the given calling convention supports variadic
Index: lib/AST/ItaniumMangle.cpp
===================================================================
--- lib/AST/ItaniumMangle.cpp
+++ lib/AST/ItaniumMangle.cpp
@@ -2648,6 +2648,7 @@
     return "";
 
   case CC_X86VectorCall:
+  case CC_AMDGPUVectorCall:
   case CC_X86Pascal:
   case CC_X86RegCall:
   case CC_AAPCS:
Index: lib/AST/Type.cpp
===================================================================
--- lib/AST/Type.cpp
+++ lib/AST/Type.cpp
@@ -2826,6 +2826,7 @@
   case CC_X86ThisCall: return "thiscall";
   case CC_X86Pascal: return "pascal";
   case CC_X86VectorCall: return "vectorcall";
+  case CC_AMDGPUVectorCall: return "amdgpu_vectorcall";
   case CC_Win64: return "ms_abi";
   case CC_X86_64SysV: return "sysv_abi";
   case CC_X86RegCall : return "regcall";
Index: lib/AST/TypePrinter.cpp
===================================================================
--- lib/AST/TypePrinter.cpp
+++ lib/AST/TypePrinter.cpp
@@ -862,6 +862,7 @@
       OS << " __attribute__((thiscall))";
       break;
     case CC_X86VectorCall:
+    case CC_AMDGPUVectorCall:
       OS << " __attribute__((vectorcall))";
       break;
     case CC_X86Pascal:
Index: lib/Basic/Targets/AMDGPU.h
===================================================================
--- lib/Basic/Targets/AMDGPU.h
+++ lib/Basic/Targets/AMDGPU.h
@@ -341,6 +341,7 @@
       return CCCR_Warning;
     case CC_C:
     case CC_OpenCLKernel:
+    case CC_AMDGPUVectorCall:
       return CCCR_OK;
     }
   }
Index: lib/CodeGen/CGCall.cpp
===================================================================
--- lib/CodeGen/CGCall.cpp
+++ lib/CodeGen/CGCall.cpp
@@ -58,6 +58,7 @@
   case CC_X86Pascal: return llvm::CallingConv::C;
   // TODO: Add support for __vectorcall to LLVM.
   case CC_X86VectorCall: return llvm::CallingConv::X86_VectorCall;
+  case CC_AMDGPUVectorCall: return llvm::CallingConv::C;
   case CC_AArch64VectorCall: return llvm::CallingConv::AArch64_VectorCall;
   case CC_SpirFunction: return llvm::CallingConv::SPIR_FUNC;
   case CC_OpenCLKernel: return CGM.getTargetCodeGenInfo().getOpenCLKernelCallingConv();
Index: lib/CodeGen/CGDebugInfo.cpp
===================================================================
--- lib/CodeGen/CGDebugInfo.cpp
+++ lib/CodeGen/CGDebugInfo.cpp
@@ -1113,6 +1113,7 @@
 static unsigned getDwarfCC(CallingConv CC) {
   switch (CC) {
   case CC_C:
+  case CC_AMDGPUVectorCall:
     // Avoid emitting DW_AT_calling_convention if the C convention was used.
     return 0;
 
Index: lib/Sema/SemaDeclAttr.cpp
===================================================================
--- lib/Sema/SemaDeclAttr.cpp
+++ lib/Sema/SemaDeclAttr.cpp
@@ -4568,7 +4568,8 @@
     CC = CC_Swift;
     break;
   case ParsedAttr::AT_VectorCall:
-    CC = CC_X86VectorCall;
+    CC = Context.getTargetInfo().getTriple().isAMDGPU()
+      ? CC_AMDGPUVectorCall : CC_X86VectorCall;
     break;
   case ParsedAttr::AT_AArch64VectorPcs:
     CC = CC_AArch64VectorCall;
Index: test/SemaCUDA/amdgpu-windows-vectorcall.cu
===================================================================
--- /dev/null
+++ test/SemaCUDA/amdgpu-windows-vectorcall.cu
@@ -0,0 +1,11 @@
+// RUN: %clang_cc1 -triple amdgcn-amd-amdhsa -aux-triple x86_64-pc-windows-msvc -fms-compatibility -fcuda-is-device -fsyntax-only -verify %s
+
+// expected-no-diagnostics
+template<class _Ty>
+ struct A
+ {
+ };
+
+template<class _Ret, class _Arg0, class _Types> struct A<_Ret (__cdecl _Arg0::*)(_Types) > { }; 
+template<class _Ret, class _Arg0, class _Types> struct A<_Ret (__vectorcall _Arg0::*)(_Types) > {};
+