[clang] [Clang][OpenCL][AMDGPU] Allow a kernel to call another kernel (PR #115821)

[Aniket Lal via cfe-commits]

================
@@ -1162,9 +1162,15 @@ void MicrosoftCXXNameMangler::mangleUnqualifiedName(GlobalDecl GD,
                   ->getTemplatedDecl()
                   ->hasAttr<CUDAGlobalAttr>())) &&
             GD.getKernelReferenceKind() == KernelReferenceKind::Stub;
+        bool IsOCLDeviceStub =
+            ND && (isa<FunctionDecl>(ND) && ND->hasAttr<OpenCLKernelAttr>()) &&
----------------
lalaniket8 wrote:

Fixed

https://github.com/llvm/llvm-project/pull/115821