The LLVM Project Blog

LLVM Project News and Details from the Trenches

Cling -- Beyond Just Interpreting C++

Interactive C++ with Cling

In our previous blog post “Interactive C++ for Data Science” we described eval-style programming, interactive C++ in Notebooks and CUDA. This post will discuss some developed applications of Cling supporting interoperability and extensibility. We aim to demonstrate template instantiation on demand; embedding Cling as a service; and showcase an extension enabling on-the-fly automatic differentiation.

Template Instantiation on Demand

Cling implements a facility called LookupHelper, which takes C++ code and checks if a declaration with that qualified name already exists. For instance:

[cling] #include "cling/Interpreter/Interpreter.h"
[cling] #include "cling/Interpreter/LookupHelper.h"
[cling] #include "clang/AST/Decl.h"
[cling] struct S{};
[cling] cling::LookupHelper& LH = gCling->getLookupHelper()
(cling::LookupHelper &) @0x7fcba3c0bfc0
[cling] auto D = LH.findScope("std::vector<S>",
(const clang::Decl *) 0x1216bdcd8
[cling] D->getDeclKindName()
(const char *) "ClassTemplateSpecialization"

In this particular case, findScope instantiates the template and returns its clang AST representation. Template instantiation on demand addresses the common library problem of template combinatorial explosion. Template instantiation on demand and conversion of textual qualified C++ names into entity meta information has proven to be a very powerful mechanism aiding data serialization and language interoperability.

Language Interop on Demand

An example is cppyy, which provides automatic Python bindings, at runtime, to C++ code through Cling. Python is itself a dynamic language executed by an interpreter, thus making the interaction with C++ code more natural when intermediated by Cling. Examples include runtime template instantiations, function (pointer) callbacks, cross-language inheritance, automatic downcasting, and exception mapping. Many advanced C++ features such as placement new, multiple virtual inheritance, variadic templates, etc., are naturally resolved by the LookupHelper.

cppyy achieves high performance through an all-lazy approach to runtime bindings construction and specializations of common cases through runtime reflection. As such, it has a much lower call overhead than e.g. pybind11, and looping over a std::vector through cppyy is faster than looping over a numpy array of the same type. Taking it a step further, its implementation for PyPy, a fully compatible Python interpreter sporting at tracing JIT, can in many cases provide native access to C++ code in PyPy’s JIT, including overload resolution and JIT hints that allow for aggressive optimizations.

Thanks to Cling’s runtime reflection, cppyy makes maintaining a large software stack simpler: except for cppyy’s own python-interpreter binding, it does not have any compiled code that is Python-dependent. I.e., cppyy-based extension modules require no recompilation when switching Python versions (or even when switching between the CPython and PyPy interpreters, say).

The example below shows the tight integration of C++ and python; shows the tight back and forth communication of the template instantiation and the cross-inheritance overrides; and the runtime behaviors (everything happens at runtime, there is no compiled code here).

import cppyy

template<typename T> class Producer {
  T m_value;

  virtual T produce_imp() = 0;

  Producer(const T& value) : m_value(value) {}
  virtual ~Producer() {}

  T produce_total() { return m_value + produce_imp(); }

class Consumer {
  template<typename T>
  void consume(Producer<T>& p) {
    std::cout << "received: \"" << p.produce_total() << "\"\n";

def factory(base_v, *derived_v):
  class _F(cppyy.gbl.Producer[type(base_v)]):
    def __init__(self, base_v, *derived_v):
      self._values = derived_v

    def produce_imp(self):
      return type(base_v)(sum(self._values))

    return _F(base_v, *derived_v)

consumer = cppyy.gbl.Consumer()
for producer in [factory(*x) for x in \
                  (("hello ", 42), (3., 0.14, 0.0015))]:


received: "hello 42"
received: "3.1415"

In the snippet we create python classes based on python arguments, which derive from a templated C++ class instantiated with a type. The python class provides the implementation for a protected function that is called from a public function, resulting in the expected return value, which is printed. We aim to highlight:

  • Python creates classes at runtime, as can Cling, even when they are declared in a module (the relevant classes are here created in a factory method);
  • Templated C++ classes can be instantiated on the fly from Python, by taking the type of the argument (i.e. using introspection at runtime in Python) to create the C++ base class for the Python class.
  • Cross-language derivation is at runtime, with no support needed from the C++ class, other than a virtual destructor and virtual methods;
  • C++ “protected” methods can be overridden in Python, even though Python has no such concept and you can not actually call protected methods from bound C++ objects in Python;
  • It all works straight out of the box.

cppyy is used in several large code bases in physics, chemistry, mathematics, and biology. It is readily installable through [pip from PyPI] ( and through conda.

Another example is Symmetry Integration Language (SIL), a D-based domain-specific language of functional flavor developed and used internally by Symmetry Investments. One of the main goals of SIL is to be easily interoperable with all sorts of languages and systems, and this is achieved through various plugins. To call C++ code, SIL uses a plugin called sil-cling, which acts as a middle ground between SIL and Cling. However, sil-cling does not interact directly with Cling, but through cppyy-backend, that is cppyy’s C/C++ wrapper around Cling that provides a stable C/C++ reflection API.

There are two core types that are exposed from sil-cling to SIL. One is CPPNamespace, which exposes a C++ namespace and allows free function calling, access to namespace’s variables, and object instantiation for the classes defined in that namespace. The other is ClingObj, which is a proxy for a C++ object, allowing construction, method calling and the manipulation of the object’s data members. Given that cppyy represents C++ classes, structs and namespaces as ‘scopes’ and reflection information about any of these C++ entities is obtained through its associated ‘scope’ object, both wrapper types exposed to SIL hold a reference to their associated scope object which is queried whenever the wrapper types are used to call C++ code.

All the calls that are done from SIL through the two wrapper types have 3 arguments: the wrapper object used, the name of the C++ function that needs to be called, and (if needed) a sequence of arguments for that function. Once the overload resolution and the argument conversion are done, sil-cling calls the appropriate cppyy function that will wrap the call and dispatch it to Cling for JIT compilation. At the moment, sil-cling can be used to call C++ libraries like Boost.Asio, dlib or Xapian.

The example below creates a Boost Asio-based client-server application written in SIL using the sil-cling plugin. Server.sil contains the SIL code for the server. It starts by including the relevant header files, using cppCompile. The next step is to create wrapper objects for the namespaces that are needed, and this is done by calling cppNamespace with the names of the namespaces that one needs to access. These CPPNamespace wrappers are used to instantiate classes that are defined inside the C++ namespaces that they wrap. Using these wrappers, an endpoint, an acceptor socket (which listens for incoming connections) and an active socket (which handles communication with the client) are created. Then the server waits for a connection and, once a client connects, it reads its message and sends a reply.

// Server.sil
import * from silcling
import format from format

cppCompile ("#include <boost/asio.hpp>")
cppCompile ("#include \"helper.hpp\"")

// CPPNamespace wrappers
asio = cppNamespace("boost::asio")
tcp = cppNamespace("boost::asio::ip::tcp")
helpers = cppNamespace("helpme")

// Using namespace wrappers to instantiate classes - creates ClingObj(s)
ioService = asio.obj("io_service")
endpoint = tcp.obj("endpoint", tcp.v4(), 9999)

// Acceptor socket - incoming connections
acceptorSocket = tcp.obj("acceptor", ioService, endpoint)
// Active socket - communication with client
activeSocket = tcp.obj("socket", ioService)

// Waiting for connection and use the activeSocket to connect with the client
helpers.accept(acceptorSocket, activeSocket)

// Waiting for message
message = helpers.getData(activeSocket);
print(format("[Server]: Received \"%s\" from client.", message.getString()))

// Send reply
reply = "Hello \'" ~ message.getString() ~ "\'!"
helpers.sendData(activeSocket, reply)
print(format("[Server]: Sent \"%s\" to client.", reply))

Client.sil contains the SIL code for the client. As the server, it includes the relevant headers, creates wrappers for the required namespaces and uses them to create an endpoint and a socket. Then the client connects to the server, sends a message, and waits for the server’s reply. Once the reply arrives, the client prints it to the screen.

// Client.sil
import * from silcling
import format from format

cppCompile ("#include <boost/asio.hpp>")
cppCompile ("#include \"helper.hpp\"")

asio = cppNamespace("boost::asio")
tcp = cppNamespace("boost::asio::ip::tcp")
helpers = cppNamespace("helpme")

// Scope resolution operator <-> address::static_method() or address::static_member
address = classScope("boost::asio::ip::address")

ioService = asio.obj("io_service")
endpoint = tcp.obj("endpoint", address.from_string(""), 9999)

// Creating socket
client_socket = tcp.obj("socket", ioService)
// Connect

message = "demo"
helpers.sendData(client_socket, message)
print(format("[Client]: Sent \"%s\" to server.", message))

message = helpers.getData(client_socket);
print(format("[Client]: Received \"%s\" from server.", message.getString()))


[Client]: Sent "demo" to server.
[Server]: Received "demo" from client.
[Server]: Sent "Hello demo" to client.
[Client]: Received "Hello demo" from server.

Interpreter/Compiler as a Service

The design of Cling, just like Clang, allows it to be used as a library. In the next example we show how to incorporate libCling in a C++ program. Cling can be used on-demand, as a service, to compile, modify or describe C++ code. The example program shows several ways in which compiled and interpreted C++ can interact:

  • callCompiledFn – The cling-demo.cpp defines an in global variable, aGlobal; a static float variable, anotherGlobal; and its accessors. The interp argument is an earlier created instance of the Cling interpreter. Just like in standard C++, it is sufficient to forward declare the compiled entities to the interpreter to be able to use them. Then, the execution information from the different calls to process is stored in a generic Cling Value object which is used to exchange information between compiled and interpreted code.
  • callInterpretedFn – Complementing callCompiledFn, compiled code can call an interpreted function by asking Cling to form a function pointer from a given mangled name. Then the call uses the standard C++ syntax.
  • modifyCompiledValue – Cling has full understanding of C++ and thus we can support complex low-level operations on stack-allocated memory. In the example we ask the compiler for the memory address of the local variable loc and ask the interpreter, at runtime, to square its value.
// cling-demo.cpp
// g++ ... cling-demo.cpp; ./cling-demo
#include <cling/Interpreter/Interpreter.h>
#include <cling/Interpreter/Value.h>
#include <cling/Utils/Casting.h>
#include <iostream>
#include <string>
#include <sstream>

/// Definitions of declarations injected also into cling.
/// NOTE: this could also stay in a header #included here and into cling, but
/// for the sake of simplicity we just redeclare them here.
int aGlobal = 42;
static float anotherGlobal = 3.141;
float getAnotherGlobal() { return anotherGlobal; }
void setAnotherGlobal(float val) { anotherGlobal = val; }

///\brief Call compiled functions from the interpreter.
void callCompiledFn(cling::Interpreter& interp) {
  // We could use a header, too...
  interp.declare("int aGlobal;\n"
                 "float getAnotherGlobal();\n"
                 "void setAnotherGlobal(float val);\n");

  cling::Value res; // Will hold the result of the expression evaluation.
  interp.process("aGlobal;", &res);
  std::cout << "aGlobal is " << res.getAs<long long>() << '\n';
  interp.process("getAnotherGlobal();", &res);
  std::cout << "getAnotherGlobal() returned " << res.getAs<float>() << '\n';

  setAnotherGlobal(1.); // We modify the compiled value,
  interp.process("getAnotherGlobal();", &res); // does the interpreter see it?
  std::cout << "getAnotherGlobal() returned " << res.getAs<float>() << '\n';

  // We modify using the interpreter, now the binary sees the new value.
  interp.process("setAnotherGlobal(7.777); getAnotherGlobal();");
  std::cout << "getAnotherGlobal() returned " << getAnotherGlobal() << '\n';

/// Call an interpreted function using its symbol address.
void callInterpretedFn(cling::Interpreter& interp) {
  // Declare a function to the interpreter. Make it extern "C" to remove
  // mangling from the game.
  interp.declare("extern \"C\" int plutification(int siss, int sat) "
                 "{ return siss * sat; }");
  void* addr = interp.getAddressOfGlobal("plutification");
  using func_t = int(int, int);
  func_t* pFunc = cling::utils::VoidToFunctionPtr<func_t*>(addr);
  std::cout << "7 * 8 = " << pFunc(7, 8) << '\n';

/// Pass a pointer into cling as a string.
void modifyCompiledValue(cling::Interpreter& interp) {
  int loc = 17; // The value that will be modified

  // Update the value of loc by passing it to the interpreter.
  std::ostringstream sstr;
  // on Windows, to prefix the hexadecimal value of a pointer with '0x',
  // one need to write: std::hex << std::showbase << (size_t)pointer
  sstr << "int& ref = *(int*)" << std::hex << std::showbase << (size_t)&loc << ';';
  sstr << "ref = ref * ref;";
  std::cout << "The square of 17 is " << loc << '\n';

int main(int argc, const char* const* argv) {
  // Create the Interpreter. LLVMDIR is provided as -D during compilation.
  cling::Interpreter interp(argc, argv, LLVMDIR);


  return 0;



aGlobal is 42
getAnotherGlobal() returned 3.141
getAnotherGlobal() returned 1
getAnotherGlobal() returned 7.777
7 * 8 = 56
The square of 17 is 289

Crossing the compiled-interpreted boundary relies on the stability of Clang’s implementation of the host’s application binary interface (ABI). Over the years it has been very reliable for both Unix and Windows however, Cling is heavily used to interact with GCC-compiled codebases and is sensitive to ABI incompatibilities between GCC and Clang with respect to the Itanium ABI specification.


Just like Clang, Cling can be extended by plugins. The next example demonstrates embedded use of Cling’s extension for automatic differentiation, Clad. Clad transforms the clang’s AST to produce derivatives and gradients of mathematical functions. When creating the Cling instance we specify -fplugin and the path to the plugin itself. Then we define a target function, pow2, and ask for its derivative with respect to its first argument.

#include <cling/Interpreter/Interpreter.h>
#include <cling/Interpreter/Value.h>

// Derivatives as a service.

void gimme_pow2dx(cling::Interpreter &interp) {
  // Definitions of declarations injected also into cling.
  interp.declare("double pow2(double x) { return x*x; }");
  interp.declare("#include <clad/Differentiator/Differentiator.h>");
  interp.declare("auto dfdx = clad::differentiate(pow2, 0);");

  cling::Value res; // Will hold the evaluation result.
  interp.process("dfdx.getFunctionPtr();", &res);

  using func_t = double(double);
  func_t* pFunc = res.getAs<func_t*>();
  printf("dfdx at 1 = %f\n", pFunc(1));

int main(int argc, const char* const* argv) {
 std::vector<const char*> argvExt(argv, argv+argc);
  // Create cling. LLVMDIR is provided as -D during compilation.
  cling::Interpreter interp(argvExt.size(), &argvExt[0], LLVMDIR);
  return 0;


dfdx at 1 = 2.000000


We have demonstrated Cling’s capabilities for template instantiation on demand; incorporating an interpreter in third-party code; and facilitating interpreter extension. The lazy template instantiation in an embedded interpreter provides a service which is very suitable for interoperability with C++. Extending such a service with domain-specific capabilities such as automatic differentiation can be an key enabler for various science cases and for other broader communities.


The author would like to thank Sylvain Corlay, Simeon Ehrig, David Lange, Chris Lattner, Javier Lopez Gomez, Wim Lavrijsen, Axel Naumann, Alexander Penev, Xavier Valls Pla, Richard Smith, Martin Vassilev, Ioana Ifrim who contributed to this post.

You can find out more about our activities at and