/* Generated by Cython 0.29.35 */

/* BEGIN: Cython Metadata
{
    "distutils": {
        "depends": [
            "C:\\Users\\mavaz\\git\\cuda-python-build-utils\\conda-recipe\\include\\nvrtc.h"
        ],
        "include_dirs": [
            "C:\\Users\\mavaz\\AppData\\Local\\Programs\\Python\\Python38",
            "C:\\Users\\mavaz\\git\\cuda-python-build-utils\\conda-recipe/include"
        ],
        "language": "c++",
        "library_dirs": [
            "C:\\Users\\mavaz\\AppData\\Local\\Programs\\Python\\Python38\\Lib\\site-packages",
            "C:\\Users\\mavaz\\AppData\\Local\\Programs\\Python\\Python38\\lib"
        ],
        "name": "cuda._cuda.cnvrtc",
        "sources": [
            "cuda/_cuda\\cnvrtc.pyx",
            "cuda/_cuda/loader.cpp"
        ]
    },
    "module_name": "cuda._cuda.cnvrtc"
}
END: Cython Metadata */

#ifndef PY_SSIZE_T_CLEAN
#define PY_SSIZE_T_CLEAN
#endif /* PY_SSIZE_T_CLEAN */
#include "Python.h"
#ifndef Py_PYTHON_H
    #error Python headers needed to compile C extensions, please install development version of Python.
#elif PY_VERSION_HEX < 0x02060000 || (0x03000000 <= PY_VERSION_HEX && PY_VERSION_HEX < 0x03030000)
    #error Cython requires Python 2.6+ or Python 3.3+.
#else
#define CYTHON_ABI "0_29_35"
#define CYTHON_HEX_VERSION 0x001D23F0
#define CYTHON_FUTURE_DIVISION 1
#include <stddef.h>
#ifndef offsetof
  #define offsetof(type, member) ( (size_t) & ((type*)0) -> member )
#endif
#if !defined(WIN32) && !defined(MS_WINDOWS)
  #ifndef __stdcall
    #define __stdcall
  #endif
  #ifndef __cdecl
    #define __cdecl
  #endif
  #ifndef __fastcall
    #define __fastcall
  #endif
#endif
#ifndef DL_IMPORT
  #define DL_IMPORT(t) t
#endif
#ifndef DL_EXPORT
  #define DL_EXPORT(t) t
#endif
#define __PYX_COMMA ,
#ifndef HAVE_LONG_LONG
  #if PY_VERSION_HEX >= 0x02070000
    #define HAVE_LONG_LONG
  #endif
#endif
#ifndef PY_LONG_LONG
  #define PY_LONG_LONG LONG_LONG
#endif
#ifndef Py_HUGE_VAL
  #define Py_HUGE_VAL HUGE_VAL
#endif
#ifdef PYPY_VERSION
  #define CYTHON_COMPILING_IN_PYPY 1
  #define CYTHON_COMPILING_IN_PYSTON 0
  #define CYTHON_COMPILING_IN_CPYTHON 0
  #define CYTHON_COMPILING_IN_NOGIL 0
  #undef CYTHON_USE_TYPE_SLOTS
  #define CYTHON_USE_TYPE_SLOTS 0
  #undef CYTHON_USE_PYTYPE_LOOKUP
  #define CYTHON_USE_PYTYPE_LOOKUP 0
  #if PY_VERSION_HEX < 0x03050000
    #undef CYTHON_USE_ASYNC_SLOTS
    #define CYTHON_USE_ASYNC_SLOTS 0
  #elif !defined(CYTHON_USE_ASYNC_SLOTS)
    #define CYTHON_USE_ASYNC_SLOTS 1
  #endif
  #undef CYTHON_USE_PYLIST_INTERNALS
  #define CYTHON_USE_PYLIST_INTERNALS 0
  #undef CYTHON_USE_UNICODE_INTERNALS
  #define CYTHON_USE_UNICODE_INTERNALS 0
  #undef CYTHON_USE_UNICODE_WRITER
  #define CYTHON_USE_UNICODE_WRITER 0
  #undef CYTHON_USE_PYLONG_INTERNALS
  #define CYTHON_USE_PYLONG_INTERNALS 0
  #undef CYTHON_AVOID_BORROWED_REFS
  #define CYTHON_AVOID_BORROWED_REFS 1
  #undef CYTHON_ASSUME_SAFE_MACROS
  #define CYTHON_ASSUME_SAFE_MACROS 0
  #undef CYTHON_UNPACK_METHODS
  #define CYTHON_UNPACK_METHODS 0
  #undef CYTHON_FAST_THREAD_STATE
  #define CYTHON_FAST_THREAD_STATE 0
  #undef CYTHON_FAST_PYCALL
  #define CYTHON_FAST_PYCALL 0
  #if PY_VERSION_HEX < 0x03090000
    #undef CYTHON_PEP489_MULTI_PHASE_INIT
    #define CYTHON_PEP489_MULTI_PHASE_INIT 0
  #elif !defined(CYTHON_PEP489_MULTI_PHASE_INIT)
    #define CYTHON_PEP489_MULTI_PHASE_INIT 1
  #endif
  #undef CYTHON_USE_TP_FINALIZE
  #define CYTHON_USE_TP_FINALIZE 0
  #undef CYTHON_USE_DICT_VERSIONS
  #define CYTHON_USE_DICT_VERSIONS 0
  #undef CYTHON_USE_EXC_INFO_STACK
  #define CYTHON_USE_EXC_INFO_STACK 0
  #ifndef CYTHON_UPDATE_DESCRIPTOR_DOC
    #define CYTHON_UPDATE_DESCRIPTOR_DOC 0
  #endif
#elif defined(PYSTON_VERSION)
  #define CYTHON_COMPILING_IN_PYPY 0
  #define CYTHON_COMPILING_IN_PYSTON 1
  #define CYTHON_COMPILING_IN_CPYTHON 0
  #define CYTHON_COMPILING_IN_NOGIL 0
  #ifndef CYTHON_USE_TYPE_SLOTS
    #define CYTHON_USE_TYPE_SLOTS 1
  #endif
  #undef CYTHON_USE_PYTYPE_LOOKUP
  #define CYTHON_USE_PYTYPE_LOOKUP 0
  #undef CYTHON_USE_ASYNC_SLOTS
  #define CYTHON_USE_ASYNC_SLOTS 0
  #undef CYTHON_USE_PYLIST_INTERNALS
  #define CYTHON_USE_PYLIST_INTERNALS 0
  #ifndef CYTHON_USE_UNICODE_INTERNALS
    #define CYTHON_USE_UNICODE_INTERNALS 1
  #endif
  #undef CYTHON_USE_UNICODE_WRITER
  #define CYTHON_USE_UNICODE_WRITER 0
  #undef CYTHON_USE_PYLONG_INTERNALS
  #define CYTHON_USE_PYLONG_INTERNALS 0
  #ifndef CYTHON_AVOID_BORROWED_REFS
    #define CYTHON_AVOID_BORROWED_REFS 0
  #endif
  #ifndef CYTHON_ASSUME_SAFE_MACROS
    #define CYTHON_ASSUME_SAFE_MACROS 1
  #endif
  #ifndef CYTHON_UNPACK_METHODS
    #define CYTHON_UNPACK_METHODS 1
  #endif
  #undef CYTHON_FAST_THREAD_STATE
  #define CYTHON_FAST_THREAD_STATE 0
  #undef CYTHON_FAST_PYCALL
  #define CYTHON_FAST_PYCALL 0
  #undef CYTHON_PEP489_MULTI_PHASE_INIT
  #define CYTHON_PEP489_MULTI_PHASE_INIT 0
  #undef CYTHON_USE_TP_FINALIZE
  #define CYTHON_USE_TP_FINALIZE 0
  #undef CYTHON_USE_DICT_VERSIONS
  #define CYTHON_USE_DICT_VERSIONS 0
  #undef CYTHON_USE_EXC_INFO_STACK
  #define CYTHON_USE_EXC_INFO_STACK 0
  #ifndef CYTHON_UPDATE_DESCRIPTOR_DOC
    #define CYTHON_UPDATE_DESCRIPTOR_DOC 0
  #endif
#elif defined(PY_NOGIL)
  #define CYTHON_COMPILING_IN_PYPY 0
  #define CYTHON_COMPILING_IN_PYSTON 0
  #define CYTHON_COMPILING_IN_CPYTHON 0
  #define CYTHON_COMPILING_IN_NOGIL 1
  #ifndef CYTHON_USE_TYPE_SLOTS
    #define CYTHON_USE_TYPE_SLOTS 1
  #endif
  #undef CYTHON_USE_PYTYPE_LOOKUP
  #define CYTHON_USE_PYTYPE_LOOKUP 0
  #ifndef CYTHON_USE_ASYNC_SLOTS
    #define CYTHON_USE_ASYNC_SLOTS 1
  #endif
  #undef CYTHON_USE_PYLIST_INTERNALS
  #define CYTHON_USE_PYLIST_INTERNALS 0
  #ifndef CYTHON_USE_UNICODE_INTERNALS
    #define CYTHON_USE_UNICODE_INTERNALS 1
  #endif
  #undef CYTHON_USE_UNICODE_WRITER
  #define CYTHON_USE_UNICODE_WRITER 0
  #undef CYTHON_USE_PYLONG_INTERNALS
  #define CYTHON_USE_PYLONG_INTERNALS 0
  #ifndef CYTHON_AVOID_BORROWED_REFS
    #define CYTHON_AVOID_BORROWED_REFS 0
  #endif
  #ifndef CYTHON_ASSUME_SAFE_MACROS
    #define CYTHON_ASSUME_SAFE_MACROS 1
  #endif
  #ifndef CYTHON_UNPACK_METHODS
    #define CYTHON_UNPACK_METHODS 1
  #endif
  #undef CYTHON_FAST_THREAD_STATE
  #define CYTHON_FAST_THREAD_STATE 0
  #undef CYTHON_FAST_PYCALL
  #define CYTHON_FAST_PYCALL 0
  #ifndef CYTHON_PEP489_MULTI_PHASE_INIT
    #define CYTHON_PEP489_MULTI_PHASE_INIT 1
  #endif
  #ifndef CYTHON_USE_TP_FINALIZE
    #define CYTHON_USE_TP_FINALIZE 1
  #endif
  #undef CYTHON_USE_DICT_VERSIONS
  #define CYTHON_USE_DICT_VERSIONS 0
  #undef CYTHON_USE_EXC_INFO_STACK
  #define CYTHON_USE_EXC_INFO_STACK 0
#else
  #define CYTHON_COMPILING_IN_PYPY 0
  #define CYTHON_COMPILING_IN_PYSTON 0
  #define CYTHON_COMPILING_IN_CPYTHON 1
  #define CYTHON_COMPILING_IN_NOGIL 0
  #ifndef CYTHON_USE_TYPE_SLOTS
    #define CYTHON_USE_TYPE_SLOTS 1
  #endif
  #if PY_VERSION_HEX < 0x02070000
    #undef CYTHON_USE_PYTYPE_LOOKUP
    #define CYTHON_USE_PYTYPE_LOOKUP 0
  #elif !defined(CYTHON_USE_PYTYPE_LOOKUP)
    #define CYTHON_USE_PYTYPE_LOOKUP 1
  #endif
  #if PY_MAJOR_VERSION < 3
    #undef CYTHON_USE_ASYNC_SLOTS
    #define CYTHON_USE_ASYNC_SLOTS 0
  #elif !defined(CYTHON_USE_ASYNC_SLOTS)
    #define CYTHON_USE_ASYNC_SLOTS 1
  #endif
  #if PY_VERSION_HEX < 0x02070000
    #undef CYTHON_USE_PYLONG_INTERNALS
    #define CYTHON_USE_PYLONG_INTERNALS 0
  #elif !defined(CYTHON_USE_PYLONG_INTERNALS)
    #define CYTHON_USE_PYLONG_INTERNALS (PY_VERSION_HEX < 0x030C00A5)
  #endif
  #ifndef CYTHON_USE_PYLIST_INTERNALS
    #define CYTHON_USE_PYLIST_INTERNALS 1
  #endif
  #ifndef CYTHON_USE_UNICODE_INTERNALS
    #define CYTHON_USE_UNICODE_INTERNALS 1
  #endif
  #if PY_VERSION_HEX < 0x030300F0 || PY_VERSION_HEX >= 0x030B00A2
    #undef CYTHON_USE_UNICODE_WRITER
    #define CYTHON_USE_UNICODE_WRITER 0
  #elif !defined(CYTHON_USE_UNICODE_WRITER)
    #define CYTHON_USE_UNICODE_WRITER 1
  #endif
  #ifndef CYTHON_AVOID_BORROWED_REFS
    #define CYTHON_AVOID_BORROWED_REFS 0
  #endif
  #ifndef CYTHON_ASSUME_SAFE_MACROS
    #define CYTHON_ASSUME_SAFE_MACROS 1
  #endif
  #ifndef CYTHON_UNPACK_METHODS
    #define CYTHON_UNPACK_METHODS 1
  #endif
  #if PY_VERSION_HEX >= 0x030B00A4
    #undef CYTHON_FAST_THREAD_STATE
    #define CYTHON_FAST_THREAD_STATE 0
  #elif !defined(CYTHON_FAST_THREAD_STATE)
    #define CYTHON_FAST_THREAD_STATE 1
  #endif
  #ifndef CYTHON_FAST_PYCALL
    #define CYTHON_FAST_PYCALL (PY_VERSION_HEX < 0x030A0000)
  #endif
  #ifndef CYTHON_PEP489_MULTI_PHASE_INIT
    #define CYTHON_PEP489_MULTI_PHASE_INIT (PY_VERSION_HEX >= 0x03050000)
  #endif
  #ifndef CYTHON_USE_TP_FINALIZE
    #define CYTHON_USE_TP_FINALIZE (PY_VERSION_HEX >= 0x030400a1)
  #endif
  #ifndef CYTHON_USE_DICT_VERSIONS
    #define CYTHON_USE_DICT_VERSIONS ((PY_VERSION_HEX >= 0x030600B1) && (PY_VERSION_HEX < 0x030C00A5))
  #endif
  #if PY_VERSION_HEX >= 0x030B00A4
    #undef CYTHON_USE_EXC_INFO_STACK
    #define CYTHON_USE_EXC_INFO_STACK 0
  #elif !defined(CYTHON_USE_EXC_INFO_STACK)
    #define CYTHON_USE_EXC_INFO_STACK (PY_VERSION_HEX >= 0x030700A3)
  #endif
  #ifndef CYTHON_UPDATE_DESCRIPTOR_DOC
    #define CYTHON_UPDATE_DESCRIPTOR_DOC 1
  #endif
#endif
#if !defined(CYTHON_FAST_PYCCALL)
#define CYTHON_FAST_PYCCALL  (CYTHON_FAST_PYCALL && PY_VERSION_HEX >= 0x030600B1)
#endif
#if CYTHON_USE_PYLONG_INTERNALS
  #if PY_MAJOR_VERSION < 3
    #include "longintrepr.h"
  #endif
  #undef SHIFT
  #undef BASE
  #undef MASK
  #ifdef SIZEOF_VOID_P
    enum { __pyx_check_sizeof_voidp = 1 / (int)(SIZEOF_VOID_P == sizeof(void*)) };
  #endif
#endif
#ifndef __has_attribute
  #define __has_attribute(x) 0
#endif
#ifndef __has_cpp_attribute
  #define __has_cpp_attribute(x) 0
#endif
#ifndef CYTHON_RESTRICT
  #if defined(__GNUC__)
    #define CYTHON_RESTRICT __restrict__
  #elif defined(_MSC_VER) && _MSC_VER >= 1400
    #define CYTHON_RESTRICT __restrict
  #elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L
    #define CYTHON_RESTRICT restrict
  #else
    #define CYTHON_RESTRICT
  #endif
#endif
#ifndef CYTHON_UNUSED
# if defined(__GNUC__)
#   if !(defined(__cplusplus)) || (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 4))
#     define CYTHON_UNUSED __attribute__ ((__unused__))
#   else
#     define CYTHON_UNUSED
#   endif
# elif defined(__ICC) || (defined(__INTEL_COMPILER) && !defined(_MSC_VER))
#   define CYTHON_UNUSED __attribute__ ((__unused__))
# else
#   define CYTHON_UNUSED
# endif
#endif
#ifndef CYTHON_MAYBE_UNUSED_VAR
#  if defined(__cplusplus)
     template<class T> void CYTHON_MAYBE_UNUSED_VAR( const T& ) { }
#  else
#    define CYTHON_MAYBE_UNUSED_VAR(x) (void)(x)
#  endif
#endif
#ifndef CYTHON_NCP_UNUSED
# if CYTHON_COMPILING_IN_CPYTHON
#  define CYTHON_NCP_UNUSED
# else
#  define CYTHON_NCP_UNUSED CYTHON_UNUSED
# endif
#endif
#define __Pyx_void_to_None(void_result) ((void)(void_result), Py_INCREF(Py_None), Py_None)
#ifdef _MSC_VER
    #ifndef _MSC_STDINT_H_
        #if _MSC_VER < 1300
           typedef unsigned char     uint8_t;
           typedef unsigned int      uint32_t;
        #else
           typedef unsigned __int8   uint8_t;
           typedef unsigned __int32  uint32_t;
        #endif
    #endif
#else
   #include <stdint.h>
#endif
#ifndef CYTHON_FALLTHROUGH
  #if defined(__cplusplus) && __cplusplus >= 201103L
    #if __has_cpp_attribute(fallthrough)
      #define CYTHON_FALLTHROUGH [[fallthrough]]
    #elif __has_cpp_attribute(clang::fallthrough)
      #define CYTHON_FALLTHROUGH [[clang::fallthrough]]
    #elif __has_cpp_attribute(gnu::fallthrough)
      #define CYTHON_FALLTHROUGH [[gnu::fallthrough]]
    #endif
  #endif
  #ifndef CYTHON_FALLTHROUGH
    #if __has_attribute(fallthrough)
      #define CYTHON_FALLTHROUGH __attribute__((fallthrough))
    #else
      #define CYTHON_FALLTHROUGH
    #endif
  #endif
  #if defined(__clang__ ) && defined(__apple_build_version__)
    #if __apple_build_version__ < 7000000
      #undef  CYTHON_FALLTHROUGH
      #define CYTHON_FALLTHROUGH
    #endif
  #endif
#endif

#ifndef __cplusplus
  #error "Cython files generated with the C++ option must be compiled with a C++ compiler."
#endif
#ifndef CYTHON_INLINE
  #if defined(__clang__)
    #define CYTHON_INLINE __inline__ __attribute__ ((__unused__))
  #else
    #define CYTHON_INLINE inline
  #endif
#endif
template<typename T>
void __Pyx_call_destructor(T& x) {
    x.~T();
}
template<typename T>
class __Pyx_FakeReference {
  public:
    __Pyx_FakeReference() : ptr(NULL) { }
    __Pyx_FakeReference(const T& ref) : ptr(const_cast<T*>(&ref)) { }
    T *operator->() { return ptr; }
    T *operator&() { return ptr; }
    operator T&() { return *ptr; }
    template<typename U> bool operator ==(U other) { return *ptr == other; }
    template<typename U> bool operator !=(U other) { return *ptr != other; }
  private:
    T *ptr;
};

#if CYTHON_COMPILING_IN_PYPY && PY_VERSION_HEX < 0x02070600 && !defined(Py_OptimizeFlag)
  #define Py_OptimizeFlag 0
#endif
#define __PYX_BUILD_PY_SSIZE_T "n"
#define CYTHON_FORMAT_SSIZE_T "z"
#if PY_MAJOR_VERSION < 3
  #define __Pyx_BUILTIN_MODULE_NAME "__builtin__"
  #define __Pyx_PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\
          PyCode_New(a+k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)
  #define __Pyx_DefaultClassType PyClass_Type
#else
  #define __Pyx_BUILTIN_MODULE_NAME "builtins"
  #define __Pyx_DefaultClassType PyType_Type
#if PY_VERSION_HEX >= 0x030B00A1
    static CYTHON_INLINE PyCodeObject* __Pyx_PyCode_New(int a, int k, int l, int s, int f,
                                                    PyObject *code, PyObject *c, PyObject* n, PyObject *v,
                                                    PyObject *fv, PyObject *cell, PyObject* fn,
                                                    PyObject *name, int fline, PyObject *lnos) {
        PyObject *kwds=NULL, *argcount=NULL, *posonlyargcount=NULL, *kwonlyargcount=NULL;
        PyObject *nlocals=NULL, *stacksize=NULL, *flags=NULL, *replace=NULL, *call_result=NULL, *empty=NULL;
        const char *fn_cstr=NULL;
        const char *name_cstr=NULL;
        PyCodeObject* co=NULL;
        PyObject *type, *value, *traceback;
        PyErr_Fetch(&type, &value, &traceback);
        if (!(kwds=PyDict_New())) goto end;
        if (!(argcount=PyLong_FromLong(a))) goto end;
        if (PyDict_SetItemString(kwds, "co_argcount", argcount) != 0) goto end;
        if (!(posonlyargcount=PyLong_FromLong(0))) goto end;
        if (PyDict_SetItemString(kwds, "co_posonlyargcount", posonlyargcount) != 0) goto end;
        if (!(kwonlyargcount=PyLong_FromLong(k))) goto end;
        if (PyDict_SetItemString(kwds, "co_kwonlyargcount", kwonlyargcount) != 0) goto end;
        if (!(nlocals=PyLong_FromLong(l))) goto end;
        if (PyDict_SetItemString(kwds, "co_nlocals", nlocals) != 0) goto end;
        if (!(stacksize=PyLong_FromLong(s))) goto end;
        if (PyDict_SetItemString(kwds, "co_stacksize", stacksize) != 0) goto end;
        if (!(flags=PyLong_FromLong(f))) goto end;
        if (PyDict_SetItemString(kwds, "co_flags", flags) != 0) goto end;
        if (PyDict_SetItemString(kwds, "co_code", code) != 0) goto end;
        if (PyDict_SetItemString(kwds, "co_consts", c) != 0) goto end;
        if (PyDict_SetItemString(kwds, "co_names", n) != 0) goto end;
        if (PyDict_SetItemString(kwds, "co_varnames", v) != 0) goto end;
        if (PyDict_SetItemString(kwds, "co_freevars", fv) != 0) goto end;
        if (PyDict_SetItemString(kwds, "co_cellvars", cell) != 0) goto end;
        if (PyDict_SetItemString(kwds, "co_linetable", lnos) != 0) goto end;
        if (!(fn_cstr=PyUnicode_AsUTF8AndSize(fn, NULL))) goto end;
        if (!(name_cstr=PyUnicode_AsUTF8AndSize(name, NULL))) goto end;
        if (!(co = PyCode_NewEmpty(fn_cstr, name_cstr, fline))) goto end;
        if (!(replace = PyObject_GetAttrString((PyObject*)co, "replace"))) goto cleanup_code_too;
        if (!(empty = PyTuple_New(0))) goto cleanup_code_too; // unfortunately __pyx_empty_tuple isn't available here
        if (!(call_result = PyObject_Call(replace, empty, kwds))) goto cleanup_code_too;
        Py_XDECREF((PyObject*)co);
        co = (PyCodeObject*)call_result;
        call_result = NULL;
        if (0) {
            cleanup_code_too:
            Py_XDECREF((PyObject*)co);
            co = NULL;
        }
        end:
        Py_XDECREF(kwds);
        Py_XDECREF(argcount);
        Py_XDECREF(posonlyargcount);
        Py_XDECREF(kwonlyargcount);
        Py_XDECREF(nlocals);
        Py_XDECREF(stacksize);
        Py_XDECREF(replace);
        Py_XDECREF(call_result);
        Py_XDECREF(empty);
        if (type) {
            PyErr_Restore(type, value, traceback);
        }
        return co;
    }
#else
  #define __Pyx_PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)\
          PyCode_New(a, k, l, s, f, code, c, n, v, fv, cell, fn, name, fline, lnos)
#endif
  #define __Pyx_DefaultClassType PyType_Type
#endif
#ifndef Py_TPFLAGS_CHECKTYPES
  #define Py_TPFLAGS_CHECKTYPES 0
#endif
#ifndef Py_TPFLAGS_HAVE_INDEX
  #define Py_TPFLAGS_HAVE_INDEX 0
#endif
#ifndef Py_TPFLAGS_HAVE_NEWBUFFER
  #define Py_TPFLAGS_HAVE_NEWBUFFER 0
#endif
#ifndef Py_TPFLAGS_HAVE_FINALIZE
  #define Py_TPFLAGS_HAVE_FINALIZE 0
#endif
#ifndef METH_STACKLESS
  #define METH_STACKLESS 0
#endif
#if PY_VERSION_HEX <= 0x030700A3 || !defined(METH_FASTCALL)
  #ifndef METH_FASTCALL
     #define METH_FASTCALL 0x80
  #endif
  typedef PyObject *(*__Pyx_PyCFunctionFast) (PyObject *self, PyObject *const *args, Py_ssize_t nargs);
  typedef PyObject *(*__Pyx_PyCFunctionFastWithKeywords) (PyObject *self, PyObject *const *args,
                                                          Py_ssize_t nargs, PyObject *kwnames);
#else
  #define __Pyx_PyCFunctionFast _PyCFunctionFast
  #define __Pyx_PyCFunctionFastWithKeywords _PyCFunctionFastWithKeywords
#endif
#if CYTHON_FAST_PYCCALL
#define __Pyx_PyFastCFunction_Check(func)\
    ((PyCFunction_Check(func) && (METH_FASTCALL == (PyCFunction_GET_FLAGS(func) & ~(METH_CLASS | METH_STATIC | METH_COEXIST | METH_KEYWORDS | METH_STACKLESS)))))
#else
#define __Pyx_PyFastCFunction_Check(func) 0
#endif
#if CYTHON_COMPILING_IN_PYPY && !defined(PyObject_Malloc)
  #define PyObject_Malloc(s)   PyMem_Malloc(s)
  #define PyObject_Free(p)     PyMem_Free(p)
  #define PyObject_Realloc(p)  PyMem_Realloc(p)
#endif
#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX < 0x030400A1
  #define PyMem_RawMalloc(n)           PyMem_Malloc(n)
  #define PyMem_RawRealloc(p, n)       PyMem_Realloc(p, n)
  #define PyMem_RawFree(p)             PyMem_Free(p)
#endif
#if CYTHON_COMPILING_IN_PYSTON
  #define __Pyx_PyCode_HasFreeVars(co)  PyCode_HasFreeVars(co)
  #define __Pyx_PyFrame_SetLineNumber(frame, lineno) PyFrame_SetLineNumber(frame, lineno)
#else
  #define __Pyx_PyCode_HasFreeVars(co)  (PyCode_GetNumFree(co) > 0)
  #define __Pyx_PyFrame_SetLineNumber(frame, lineno)  (frame)->f_lineno = (lineno)
#endif
#if !CYTHON_FAST_THREAD_STATE || PY_VERSION_HEX < 0x02070000
  #define __Pyx_PyThreadState_Current PyThreadState_GET()
#elif PY_VERSION_HEX >= 0x03060000
  #define __Pyx_PyThreadState_Current _PyThreadState_UncheckedGet()
#elif PY_VERSION_HEX >= 0x03000000
  #define __Pyx_PyThreadState_Current PyThreadState_GET()
#else
  #define __Pyx_PyThreadState_Current _PyThreadState_Current
#endif
#if PY_VERSION_HEX < 0x030700A2 && !defined(PyThread_tss_create) && !defined(Py_tss_NEEDS_INIT)
#include "pythread.h"
#define Py_tss_NEEDS_INIT 0
typedef int Py_tss_t;
static CYTHON_INLINE int PyThread_tss_create(Py_tss_t *key) {
  *key = PyThread_create_key();
  return 0;
}
static CYTHON_INLINE Py_tss_t * PyThread_tss_alloc(void) {
  Py_tss_t *key = (Py_tss_t *)PyObject_Malloc(sizeof(Py_tss_t));
  *key = Py_tss_NEEDS_INIT;
  return key;
}
static CYTHON_INLINE void PyThread_tss_free(Py_tss_t *key) {
  PyObject_Free(key);
}
static CYTHON_INLINE int PyThread_tss_is_created(Py_tss_t *key) {
  return *key != Py_tss_NEEDS_INIT;
}
static CYTHON_INLINE void PyThread_tss_delete(Py_tss_t *key) {
  PyThread_delete_key(*key);
  *key = Py_tss_NEEDS_INIT;
}
static CYTHON_INLINE int PyThread_tss_set(Py_tss_t *key, void *value) {
  return PyThread_set_key_value(*key, value);
}
static CYTHON_INLINE void * PyThread_tss_get(Py_tss_t *key) {
  return PyThread_get_key_value(*key);
}
#endif
#if CYTHON_COMPILING_IN_CPYTHON || defined(_PyDict_NewPresized)
#define __Pyx_PyDict_NewPresized(n)  ((n <= 8) ? PyDict_New() : _PyDict_NewPresized(n))
#else
#define __Pyx_PyDict_NewPresized(n)  PyDict_New()
#endif
#if PY_MAJOR_VERSION >= 3 || CYTHON_FUTURE_DIVISION
  #define __Pyx_PyNumber_Divide(x,y)         PyNumber_TrueDivide(x,y)
  #define __Pyx_PyNumber_InPlaceDivide(x,y)  PyNumber_InPlaceTrueDivide(x,y)
#else
  #define __Pyx_PyNumber_Divide(x,y)         PyNumber_Divide(x,y)
  #define __Pyx_PyNumber_InPlaceDivide(x,y)  PyNumber_InPlaceDivide(x,y)
#endif
#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX >= 0x030500A1 && CYTHON_USE_UNICODE_INTERNALS
#define __Pyx_PyDict_GetItemStr(dict, name)  _PyDict_GetItem_KnownHash(dict, name, ((PyASCIIObject *) name)->hash)
#else
#define __Pyx_PyDict_GetItemStr(dict, name)  PyDict_GetItem(dict, name)
#endif
#if PY_VERSION_HEX > 0x03030000 && defined(PyUnicode_KIND)
  #define CYTHON_PEP393_ENABLED 1
  #if PY_VERSION_HEX >= 0x030C0000
    #define __Pyx_PyUnicode_READY(op)       (0)
  #else
    #define __Pyx_PyUnicode_READY(op)       (likely(PyUnicode_IS_READY(op)) ?\
                                                0 : _PyUnicode_Ready((PyObject *)(op)))
  #endif
  #define __Pyx_PyUnicode_GET_LENGTH(u)   PyUnicode_GET_LENGTH(u)
  #define __Pyx_PyUnicode_READ_CHAR(u, i) PyUnicode_READ_CHAR(u, i)
  #define __Pyx_PyUnicode_MAX_CHAR_VALUE(u)   PyUnicode_MAX_CHAR_VALUE(u)
  #define __Pyx_PyUnicode_KIND(u)         PyUnicode_KIND(u)
  #define __Pyx_PyUnicode_DATA(u)         PyUnicode_DATA(u)
  #define __Pyx_PyUnicode_READ(k, d, i)   PyUnicode_READ(k, d, i)
  #define __Pyx_PyUnicode_WRITE(k, d, i, ch)  PyUnicode_WRITE(k, d, i, ch)
  #if PY_VERSION_HEX >= 0x030C0000
    #define __Pyx_PyUnicode_IS_TRUE(u)      (0 != PyUnicode_GET_LENGTH(u))
  #else
    #if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX >= 0x03090000
    #define __Pyx_PyUnicode_IS_TRUE(u)      (0 != (likely(PyUnicode_IS_READY(u)) ? PyUnicode_GET_LENGTH(u) : ((PyCompactUnicodeObject *)(u))->wstr_length))
    #else
    #define __Pyx_PyUnicode_IS_TRUE(u)      (0 != (likely(PyUnicode_IS_READY(u)) ? PyUnicode_GET_LENGTH(u) : PyUnicode_GET_SIZE(u)))
    #endif
  #endif
#else
  #define CYTHON_PEP393_ENABLED 0
  #define PyUnicode_1BYTE_KIND  1
  #define PyUnicode_2BYTE_KIND  2
  #define PyUnicode_4BYTE_KIND  4
  #define __Pyx_PyUnicode_READY(op)       (0)
  #define __Pyx_PyUnicode_GET_LENGTH(u)   PyUnicode_GET_SIZE(u)
  #define __Pyx_PyUnicode_READ_CHAR(u, i) ((Py_UCS4)(PyUnicode_AS_UNICODE(u)[i]))
  #define __Pyx_PyUnicode_MAX_CHAR_VALUE(u)   ((sizeof(Py_UNICODE) == 2) ? 65535 : 1114111)
  #define __Pyx_PyUnicode_KIND(u)         (sizeof(Py_UNICODE))
  #define __Pyx_PyUnicode_DATA(u)         ((void*)PyUnicode_AS_UNICODE(u))
  #define __Pyx_PyUnicode_READ(k, d, i)   ((void)(k), (Py_UCS4)(((Py_UNICODE*)d)[i]))
  #define __Pyx_PyUnicode_WRITE(k, d, i, ch)  (((void)(k)), ((Py_UNICODE*)d)[i] = ch)
  #define __Pyx_PyUnicode_IS_TRUE(u)      (0 != PyUnicode_GET_SIZE(u))
#endif
#if CYTHON_COMPILING_IN_PYPY
  #define __Pyx_PyUnicode_Concat(a, b)      PyNumber_Add(a, b)
  #define __Pyx_PyUnicode_ConcatSafe(a, b)  PyNumber_Add(a, b)
#else
  #define __Pyx_PyUnicode_Concat(a, b)      PyUnicode_Concat(a, b)
  #define __Pyx_PyUnicode_ConcatSafe(a, b)  ((unlikely((a) == Py_None) || unlikely((b) == Py_None)) ?\
      PyNumber_Add(a, b) : __Pyx_PyUnicode_Concat(a, b))
#endif
#if CYTHON_COMPILING_IN_PYPY && !defined(PyUnicode_Contains)
  #define PyUnicode_Contains(u, s)  PySequence_Contains(u, s)
#endif
#if CYTHON_COMPILING_IN_PYPY && !defined(PyByteArray_Check)
  #define PyByteArray_Check(obj)  PyObject_TypeCheck(obj, &PyByteArray_Type)
#endif
#if CYTHON_COMPILING_IN_PYPY && !defined(PyObject_Format)
  #define PyObject_Format(obj, fmt)  PyObject_CallMethod(obj, "__format__", "O", fmt)
#endif
#define __Pyx_PyString_FormatSafe(a, b)   ((unlikely((a) == Py_None || (PyString_Check(b) && !PyString_CheckExact(b)))) ? PyNumber_Remainder(a, b) : __Pyx_PyString_Format(a, b))
#define __Pyx_PyUnicode_FormatSafe(a, b)  ((unlikely((a) == Py_None || (PyUnicode_Check(b) && !PyUnicode_CheckExact(b)))) ? PyNumber_Remainder(a, b) : PyUnicode_Format(a, b))
#if PY_MAJOR_VERSION >= 3
  #define __Pyx_PyString_Format(a, b)  PyUnicode_Format(a, b)
#else
  #define __Pyx_PyString_Format(a, b)  PyString_Format(a, b)
#endif
#if PY_MAJOR_VERSION < 3 && !defined(PyObject_ASCII)
  #define PyObject_ASCII(o)            PyObject_Repr(o)
#endif
#if PY_MAJOR_VERSION >= 3
  #define PyBaseString_Type            PyUnicode_Type
  #define PyStringObject               PyUnicodeObject
  #define PyString_Type                PyUnicode_Type
  #define PyString_Check               PyUnicode_Check
  #define PyString_CheckExact          PyUnicode_CheckExact
#ifndef PyObject_Unicode
  #define PyObject_Unicode             PyObject_Str
#endif
#endif
#if PY_MAJOR_VERSION >= 3
  #define __Pyx_PyBaseString_Check(obj) PyUnicode_Check(obj)
  #define __Pyx_PyBaseString_CheckExact(obj) PyUnicode_CheckExact(obj)
#else
  #define __Pyx_PyBaseString_Check(obj) (PyString_Check(obj) || PyUnicode_Check(obj))
  #define __Pyx_PyBaseString_CheckExact(obj) (PyString_CheckExact(obj) || PyUnicode_CheckExact(obj))
#endif
#ifndef PySet_CheckExact
  #define PySet_CheckExact(obj)        (Py_TYPE(obj) == &PySet_Type)
#endif
#if PY_VERSION_HEX >= 0x030900A4
  #define __Pyx_SET_REFCNT(obj, refcnt) Py_SET_REFCNT(obj, refcnt)
  #define __Pyx_SET_SIZE(obj, size) Py_SET_SIZE(obj, size)
#else
  #define __Pyx_SET_REFCNT(obj, refcnt) Py_REFCNT(obj) = (refcnt)
  #define __Pyx_SET_SIZE(obj, size) Py_SIZE(obj) = (size)
#endif
#if CYTHON_ASSUME_SAFE_MACROS
  #define __Pyx_PySequence_SIZE(seq)  Py_SIZE(seq)
#else
  #define __Pyx_PySequence_SIZE(seq)  PySequence_Size(seq)
#endif
#if PY_MAJOR_VERSION >= 3
  #define PyIntObject                  PyLongObject
  #define PyInt_Type                   PyLong_Type
  #define PyInt_Check(op)              PyLong_Check(op)
  #define PyInt_CheckExact(op)         PyLong_CheckExact(op)
  #define PyInt_FromString             PyLong_FromString
  #define PyInt_FromUnicode            PyLong_FromUnicode
  #define PyInt_FromLong               PyLong_FromLong
  #define PyInt_FromSize_t             PyLong_FromSize_t
  #define PyInt_FromSsize_t            PyLong_FromSsize_t
  #define PyInt_AsLong                 PyLong_AsLong
  #define PyInt_AS_LONG                PyLong_AS_LONG
  #define PyInt_AsSsize_t              PyLong_AsSsize_t
  #define PyInt_AsUnsignedLongMask     PyLong_AsUnsignedLongMask
  #define PyInt_AsUnsignedLongLongMask PyLong_AsUnsignedLongLongMask
  #define PyNumber_Int                 PyNumber_Long
#endif
#if PY_MAJOR_VERSION >= 3
  #define PyBoolObject                 PyLongObject
#endif
#if PY_MAJOR_VERSION >= 3 && CYTHON_COMPILING_IN_PYPY
  #ifndef PyUnicode_InternFromString
    #define PyUnicode_InternFromString(s) PyUnicode_FromString(s)
  #endif
#endif
#if PY_VERSION_HEX < 0x030200A4
  typedef long Py_hash_t;
  #define __Pyx_PyInt_FromHash_t PyInt_FromLong
  #define __Pyx_PyInt_AsHash_t   __Pyx_PyIndex_AsHash_t
#else
  #define __Pyx_PyInt_FromHash_t PyInt_FromSsize_t
  #define __Pyx_PyInt_AsHash_t   __Pyx_PyIndex_AsSsize_t
#endif
#if PY_MAJOR_VERSION >= 3
  #define __Pyx_PyMethod_New(func, self, klass) ((self) ? ((void)(klass), PyMethod_New(func, self)) : __Pyx_NewRef(func))
#else
  #define __Pyx_PyMethod_New(func, self, klass) PyMethod_New(func, self, klass)
#endif
#if CYTHON_USE_ASYNC_SLOTS
  #if PY_VERSION_HEX >= 0x030500B1
    #define __Pyx_PyAsyncMethodsStruct PyAsyncMethods
    #define __Pyx_PyType_AsAsync(obj) (Py_TYPE(obj)->tp_as_async)
  #else
    #define __Pyx_PyType_AsAsync(obj) ((__Pyx_PyAsyncMethodsStruct*) (Py_TYPE(obj)->tp_reserved))
  #endif
#else
  #define __Pyx_PyType_AsAsync(obj) NULL
#endif
#ifndef __Pyx_PyAsyncMethodsStruct
    typedef struct {
        unaryfunc am_await;
        unaryfunc am_aiter;
        unaryfunc am_anext;
    } __Pyx_PyAsyncMethodsStruct;
#endif

#if defined(_WIN32) || defined(WIN32) || defined(MS_WINDOWS)
  #if !defined(_USE_MATH_DEFINES)
    #define _USE_MATH_DEFINES
  #endif
#endif
#include <math.h>
#ifdef NAN
#define __PYX_NAN() ((float) NAN)
#else
static CYTHON_INLINE float __PYX_NAN() {
  float value;
  memset(&value, 0xFF, sizeof(value));
  return value;
}
#endif
#if defined(__CYGWIN__) && defined(_LDBL_EQ_DBL)
#define __Pyx_truncl trunc
#else
#define __Pyx_truncl truncl
#endif

#define __PYX_MARK_ERR_POS(f_index, lineno) \
    { __pyx_filename = __pyx_f[f_index]; (void)__pyx_filename; __pyx_lineno = lineno; (void)__pyx_lineno; __pyx_clineno = __LINE__; (void)__pyx_clineno; }
#define __PYX_ERR(f_index, lineno, Ln_error) \
    { __PYX_MARK_ERR_POS(f_index, lineno) goto Ln_error; }

#ifndef __PYX_EXTERN_C
  #ifdef __cplusplus
    #define __PYX_EXTERN_C extern "C"
  #else
    #define __PYX_EXTERN_C extern
  #endif
#endif

#define __PYX_HAVE__cuda___cuda__cnvrtc
#define __PYX_HAVE_API__cuda___cuda__cnvrtc
/* Early includes */
#include <stdint.h>
#include "nvrtc.h"
#ifdef _OPENMP
#include <omp.h>
#endif /* _OPENMP */

#if defined(PYREX_WITHOUT_ASSERTIONS) && !defined(CYTHON_WITHOUT_ASSERTIONS)
#define CYTHON_WITHOUT_ASSERTIONS
#endif

typedef struct {PyObject **p; const char *s; const Py_ssize_t n; const char* encoding;
                const char is_unicode; const char is_str; const char intern; } __Pyx_StringTabEntry;

#define __PYX_DEFAULT_STRING_ENCODING_IS_ASCII 0
#define __PYX_DEFAULT_STRING_ENCODING_IS_UTF8 0
#define __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT (PY_MAJOR_VERSION >= 3 && __PYX_DEFAULT_STRING_ENCODING_IS_UTF8)
#define __PYX_DEFAULT_STRING_ENCODING ""
#define __Pyx_PyObject_FromString __Pyx_PyBytes_FromString
#define __Pyx_PyObject_FromStringAndSize __Pyx_PyBytes_FromStringAndSize
#define __Pyx_uchar_cast(c) ((unsigned char)c)
#define __Pyx_long_cast(x) ((long)x)
#define __Pyx_fits_Py_ssize_t(v, type, is_signed)  (\
    (sizeof(type) < sizeof(Py_ssize_t))  ||\
    (sizeof(type) > sizeof(Py_ssize_t) &&\
          likely(v < (type)PY_SSIZE_T_MAX ||\
                 v == (type)PY_SSIZE_T_MAX)  &&\
          (!is_signed || likely(v > (type)PY_SSIZE_T_MIN ||\
                                v == (type)PY_SSIZE_T_MIN)))  ||\
    (sizeof(type) == sizeof(Py_ssize_t) &&\
          (is_signed || likely(v < (type)PY_SSIZE_T_MAX ||\
                               v == (type)PY_SSIZE_T_MAX)))  )
static CYTHON_INLINE int __Pyx_is_valid_index(Py_ssize_t i, Py_ssize_t limit) {
    return (size_t) i < (size_t) limit;
}
#if defined (__cplusplus) && __cplusplus >= 201103L
    #include <cstdlib>
    #define __Pyx_sst_abs(value) std::abs(value)
#elif SIZEOF_INT >= SIZEOF_SIZE_T
    #define __Pyx_sst_abs(value) abs(value)
#elif SIZEOF_LONG >= SIZEOF_SIZE_T
    #define __Pyx_sst_abs(value) labs(value)
#elif defined (_MSC_VER)
    #define __Pyx_sst_abs(value) ((Py_ssize_t)_abs64(value))
#elif defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L
    #define __Pyx_sst_abs(value) llabs(value)
#elif defined (__GNUC__)
    #define __Pyx_sst_abs(value) __builtin_llabs(value)
#else
    #define __Pyx_sst_abs(value) ((value<0) ? -value : value)
#endif
static CYTHON_INLINE const char* __Pyx_PyObject_AsString(PyObject*);
static CYTHON_INLINE const char* __Pyx_PyObject_AsStringAndSize(PyObject*, Py_ssize_t* length);
#define __Pyx_PyByteArray_FromString(s) PyByteArray_FromStringAndSize((const char*)s, strlen((const char*)s))
#define __Pyx_PyByteArray_FromStringAndSize(s, l) PyByteArray_FromStringAndSize((const char*)s, l)
#define __Pyx_PyBytes_FromString        PyBytes_FromString
#define __Pyx_PyBytes_FromStringAndSize PyBytes_FromStringAndSize
static CYTHON_INLINE PyObject* __Pyx_PyUnicode_FromString(const char*);
#if PY_MAJOR_VERSION < 3
    #define __Pyx_PyStr_FromString        __Pyx_PyBytes_FromString
    #define __Pyx_PyStr_FromStringAndSize __Pyx_PyBytes_FromStringAndSize
#else
    #define __Pyx_PyStr_FromString        __Pyx_PyUnicode_FromString
    #define __Pyx_PyStr_FromStringAndSize __Pyx_PyUnicode_FromStringAndSize
#endif
#define __Pyx_PyBytes_AsWritableString(s)     ((char*) PyBytes_AS_STRING(s))
#define __Pyx_PyBytes_AsWritableSString(s)    ((signed char*) PyBytes_AS_STRING(s))
#define __Pyx_PyBytes_AsWritableUString(s)    ((unsigned char*) PyBytes_AS_STRING(s))
#define __Pyx_PyBytes_AsString(s)     ((const char*) PyBytes_AS_STRING(s))
#define __Pyx_PyBytes_AsSString(s)    ((const signed char*) PyBytes_AS_STRING(s))
#define __Pyx_PyBytes_AsUString(s)    ((const unsigned char*) PyBytes_AS_STRING(s))
#define __Pyx_PyObject_AsWritableString(s)    ((char*) __Pyx_PyObject_AsString(s))
#define __Pyx_PyObject_AsWritableSString(s)    ((signed char*) __Pyx_PyObject_AsString(s))
#define __Pyx_PyObject_AsWritableUString(s)    ((unsigned char*) __Pyx_PyObject_AsString(s))
#define __Pyx_PyObject_AsSString(s)    ((const signed char*) __Pyx_PyObject_AsString(s))
#define __Pyx_PyObject_AsUString(s)    ((const unsigned char*) __Pyx_PyObject_AsString(s))
#define __Pyx_PyObject_FromCString(s)  __Pyx_PyObject_FromString((const char*)s)
#define __Pyx_PyBytes_FromCString(s)   __Pyx_PyBytes_FromString((const char*)s)
#define __Pyx_PyByteArray_FromCString(s)   __Pyx_PyByteArray_FromString((const char*)s)
#define __Pyx_PyStr_FromCString(s)     __Pyx_PyStr_FromString((const char*)s)
#define __Pyx_PyUnicode_FromCString(s) __Pyx_PyUnicode_FromString((const char*)s)
static CYTHON_INLINE size_t __Pyx_Py_UNICODE_strlen(const Py_UNICODE *u) {
    const Py_UNICODE *u_end = u;
    while (*u_end++) ;
    return (size_t)(u_end - u - 1);
}
#define __Pyx_PyUnicode_FromUnicode(u)       PyUnicode_FromUnicode(u, __Pyx_Py_UNICODE_strlen(u))
#define __Pyx_PyUnicode_FromUnicodeAndLength PyUnicode_FromUnicode
#define __Pyx_PyUnicode_AsUnicode            PyUnicode_AsUnicode
#define __Pyx_NewRef(obj) (Py_INCREF(obj), obj)
#define __Pyx_Owned_Py_None(b) __Pyx_NewRef(Py_None)
static CYTHON_INLINE PyObject * __Pyx_PyBool_FromLong(long b);
static CYTHON_INLINE int __Pyx_PyObject_IsTrue(PyObject*);
static CYTHON_INLINE int __Pyx_PyObject_IsTrueAndDecref(PyObject*);
static CYTHON_INLINE PyObject* __Pyx_PyNumber_IntOrLong(PyObject* x);
#define __Pyx_PySequence_Tuple(obj)\
    (likely(PyTuple_CheckExact(obj)) ? __Pyx_NewRef(obj) : PySequence_Tuple(obj))
static CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject*);
static CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t);
static CYTHON_INLINE Py_hash_t __Pyx_PyIndex_AsHash_t(PyObject*);
#if CYTHON_ASSUME_SAFE_MACROS
#define __pyx_PyFloat_AsDouble(x) (PyFloat_CheckExact(x) ? PyFloat_AS_DOUBLE(x) : PyFloat_AsDouble(x))
#else
#define __pyx_PyFloat_AsDouble(x) PyFloat_AsDouble(x)
#endif
#define __pyx_PyFloat_AsFloat(x) ((float) __pyx_PyFloat_AsDouble(x))
#if PY_MAJOR_VERSION >= 3
#define __Pyx_PyNumber_Int(x) (PyLong_CheckExact(x) ? __Pyx_NewRef(x) : PyNumber_Long(x))
#else
#define __Pyx_PyNumber_Int(x) (PyInt_CheckExact(x) ? __Pyx_NewRef(x) : PyNumber_Int(x))
#endif
#define __Pyx_PyNumber_Float(x) (PyFloat_CheckExact(x) ? __Pyx_NewRef(x) : PyNumber_Float(x))
#if PY_MAJOR_VERSION < 3 && __PYX_DEFAULT_STRING_ENCODING_IS_ASCII
static int __Pyx_sys_getdefaultencoding_not_ascii;
static int __Pyx_init_sys_getdefaultencoding_params(void) {
    PyObject* sys;
    PyObject* default_encoding = NULL;
    PyObject* ascii_chars_u = NULL;
    PyObject* ascii_chars_b = NULL;
    const char* default_encoding_c;
    sys = PyImport_ImportModule("sys");
    if (!sys) goto bad;
    default_encoding = PyObject_CallMethod(sys, (char*) "getdefaultencoding", NULL);
    Py_DECREF(sys);
    if (!default_encoding) goto bad;
    default_encoding_c = PyBytes_AsString(default_encoding);
    if (!default_encoding_c) goto bad;
    if (strcmp(default_encoding_c, "ascii") == 0) {
        __Pyx_sys_getdefaultencoding_not_ascii = 0;
    } else {
        char ascii_chars[128];
        int c;
        for (c = 0; c < 128; c++) {
            ascii_chars[c] = c;
        }
        __Pyx_sys_getdefaultencoding_not_ascii = 1;
        ascii_chars_u = PyUnicode_DecodeASCII(ascii_chars, 128, NULL);
        if (!ascii_chars_u) goto bad;
        ascii_chars_b = PyUnicode_AsEncodedString(ascii_chars_u, default_encoding_c, NULL);
        if (!ascii_chars_b || !PyBytes_Check(ascii_chars_b) || memcmp(ascii_chars, PyBytes_AS_STRING(ascii_chars_b), 128) != 0) {
            PyErr_Format(
                PyExc_ValueError,
                "This module compiled with c_string_encoding=ascii, but default encoding '%.200s' is not a superset of ascii.",
                default_encoding_c);
            goto bad;
        }
        Py_DECREF(ascii_chars_u);
        Py_DECREF(ascii_chars_b);
    }
    Py_DECREF(default_encoding);
    return 0;
bad:
    Py_XDECREF(default_encoding);
    Py_XDECREF(ascii_chars_u);
    Py_XDECREF(ascii_chars_b);
    return -1;
}
#endif
#if __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT && PY_MAJOR_VERSION >= 3
#define __Pyx_PyUnicode_FromStringAndSize(c_str, size) PyUnicode_DecodeUTF8(c_str, size, NULL)
#else
#define __Pyx_PyUnicode_FromStringAndSize(c_str, size) PyUnicode_Decode(c_str, size, __PYX_DEFAULT_STRING_ENCODING, NULL)
#if __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT
static char* __PYX_DEFAULT_STRING_ENCODING;
static int __Pyx_init_sys_getdefaultencoding_params(void) {
    PyObject* sys;
    PyObject* default_encoding = NULL;
    char* default_encoding_c;
    sys = PyImport_ImportModule("sys");
    if (!sys) goto bad;
    default_encoding = PyObject_CallMethod(sys, (char*) (const char*) "getdefaultencoding", NULL);
    Py_DECREF(sys);
    if (!default_encoding) goto bad;
    default_encoding_c = PyBytes_AsString(default_encoding);
    if (!default_encoding_c) goto bad;
    __PYX_DEFAULT_STRING_ENCODING = (char*) malloc(strlen(default_encoding_c) + 1);
    if (!__PYX_DEFAULT_STRING_ENCODING) goto bad;
    strcpy(__PYX_DEFAULT_STRING_ENCODING, default_encoding_c);
    Py_DECREF(default_encoding);
    return 0;
bad:
    Py_XDECREF(default_encoding);
    return -1;
}
#endif
#endif


/* Test for GCC > 2.95 */
#if defined(__GNUC__)     && (__GNUC__ > 2 || (__GNUC__ == 2 && (__GNUC_MINOR__ > 95)))
  #define likely(x)   __builtin_expect(!!(x), 1)
  #define unlikely(x) __builtin_expect(!!(x), 0)
#else /* !__GNUC__ or GCC < 2.95 */
  #define likely(x)   (x)
  #define unlikely(x) (x)
#endif /* __GNUC__ */
static CYTHON_INLINE void __Pyx_pretend_to_initialize(void* ptr) { (void)ptr; }

static PyObject *__pyx_m = NULL;
static PyObject *__pyx_d;
static PyObject *__pyx_b;
static PyObject *__pyx_cython_runtime = NULL;
static PyObject *__pyx_empty_tuple;
static PyObject *__pyx_empty_bytes;
static PyObject *__pyx_empty_unicode;
static int __pyx_lineno;
static int __pyx_clineno = 0;
static const char * __pyx_cfilenm= __FILE__;
static const char *__pyx_filename;


static const char *__pyx_f[] = {
  "cuda\\_cuda\\cnvrtc.pyx",
};
/* ForceInitThreads.proto */
#ifndef __PYX_FORCE_INIT_THREADS
  #define __PYX_FORCE_INIT_THREADS 0
#endif

/* NoFastGil.proto */
#define __Pyx_PyGILState_Ensure PyGILState_Ensure
#define __Pyx_PyGILState_Release PyGILState_Release
#define __Pyx_FastGIL_Remember()
#define __Pyx_FastGIL_Forget()
#define __Pyx_FastGilFuncInit()


/*--- Type declarations ---*/

/* --- Runtime support code (head) --- */
/* Refnanny.proto */
#ifndef CYTHON_REFNANNY
  #define CYTHON_REFNANNY 0
#endif
#if CYTHON_REFNANNY
  typedef struct {
    void (*INCREF)(void*, PyObject*, int);
    void (*DECREF)(void*, PyObject*, int);
    void (*GOTREF)(void*, PyObject*, int);
    void (*GIVEREF)(void*, PyObject*, int);
    void* (*SetupContext)(const char*, int, const char*);
    void (*FinishContext)(void**);
  } __Pyx_RefNannyAPIStruct;
  static __Pyx_RefNannyAPIStruct *__Pyx_RefNanny = NULL;
  static __Pyx_RefNannyAPIStruct *__Pyx_RefNannyImportAPI(const char *modname);
  #define __Pyx_RefNannyDeclarations void *__pyx_refnanny = NULL;
#ifdef WITH_THREAD
  #define __Pyx_RefNannySetupContext(name, acquire_gil)\
          if (acquire_gil) {\
              PyGILState_STATE __pyx_gilstate_save = PyGILState_Ensure();\
              __pyx_refnanny = __Pyx_RefNanny->SetupContext((name), __LINE__, __FILE__);\
              PyGILState_Release(__pyx_gilstate_save);\
          } else {\
              __pyx_refnanny = __Pyx_RefNanny->SetupContext((name), __LINE__, __FILE__);\
          }
#else
  #define __Pyx_RefNannySetupContext(name, acquire_gil)\
          __pyx_refnanny = __Pyx_RefNanny->SetupContext((name), __LINE__, __FILE__)
#endif
  #define __Pyx_RefNannyFinishContext()\
          __Pyx_RefNanny->FinishContext(&__pyx_refnanny)
  #define __Pyx_INCREF(r)  __Pyx_RefNanny->INCREF(__pyx_refnanny, (PyObject *)(r), __LINE__)
  #define __Pyx_DECREF(r)  __Pyx_RefNanny->DECREF(__pyx_refnanny, (PyObject *)(r), __LINE__)
  #define __Pyx_GOTREF(r)  __Pyx_RefNanny->GOTREF(__pyx_refnanny, (PyObject *)(r), __LINE__)
  #define __Pyx_GIVEREF(r) __Pyx_RefNanny->GIVEREF(__pyx_refnanny, (PyObject *)(r), __LINE__)
  #define __Pyx_XINCREF(r)  do { if((r) != NULL) {__Pyx_INCREF(r); }} while(0)
  #define __Pyx_XDECREF(r)  do { if((r) != NULL) {__Pyx_DECREF(r); }} while(0)
  #define __Pyx_XGOTREF(r)  do { if((r) != NULL) {__Pyx_GOTREF(r); }} while(0)
  #define __Pyx_XGIVEREF(r) do { if((r) != NULL) {__Pyx_GIVEREF(r);}} while(0)
#else
  #define __Pyx_RefNannyDeclarations
  #define __Pyx_RefNannySetupContext(name, acquire_gil)
  #define __Pyx_RefNannyFinishContext()
  #define __Pyx_INCREF(r) Py_INCREF(r)
  #define __Pyx_DECREF(r) Py_DECREF(r)
  #define __Pyx_GOTREF(r)
  #define __Pyx_GIVEREF(r)
  #define __Pyx_XINCREF(r) Py_XINCREF(r)
  #define __Pyx_XDECREF(r) Py_XDECREF(r)
  #define __Pyx_XGOTREF(r)
  #define __Pyx_XGIVEREF(r)
#endif
#define __Pyx_XDECREF_SET(r, v) do {\
        PyObject *tmp = (PyObject *) r;\
        r = v; __Pyx_XDECREF(tmp);\
    } while (0)
#define __Pyx_DECREF_SET(r, v) do {\
        PyObject *tmp = (PyObject *) r;\
        r = v; __Pyx_DECREF(tmp);\
    } while (0)
#define __Pyx_CLEAR(r)    do { PyObject* tmp = ((PyObject*)(r)); r = NULL; __Pyx_DECREF(tmp);} while(0)
#define __Pyx_XCLEAR(r)   do { if((r) != NULL) {PyObject* tmp = ((PyObject*)(r)); r = NULL; __Pyx_DECREF(tmp);}} while(0)

/* PyObjectGetAttrStr.proto */
#if CYTHON_USE_TYPE_SLOTS
static CYTHON_INLINE PyObject* __Pyx_PyObject_GetAttrStr(PyObject* obj, PyObject* attr_name);
#else
#define __Pyx_PyObject_GetAttrStr(o,n) PyObject_GetAttr(o,n)
#endif

/* GetBuiltinName.proto */
static PyObject *__Pyx_GetBuiltinName(PyObject *name);

/* PyThreadStateGet.proto */
#if CYTHON_FAST_THREAD_STATE
#define __Pyx_PyThreadState_declare  PyThreadState *__pyx_tstate;
#define __Pyx_PyThreadState_assign  __pyx_tstate = __Pyx_PyThreadState_Current;
#define __Pyx_PyErr_Occurred()  __pyx_tstate->curexc_type
#else
#define __Pyx_PyThreadState_declare
#define __Pyx_PyThreadState_assign
#define __Pyx_PyErr_Occurred()  PyErr_Occurred()
#endif

/* PyErrFetchRestore.proto */
#if CYTHON_FAST_THREAD_STATE
#define __Pyx_PyErr_Clear() __Pyx_ErrRestore(NULL, NULL, NULL)
#define __Pyx_ErrRestoreWithState(type, value, tb)  __Pyx_ErrRestoreInState(PyThreadState_GET(), type, value, tb)
#define __Pyx_ErrFetchWithState(type, value, tb)    __Pyx_ErrFetchInState(PyThreadState_GET(), type, value, tb)
#define __Pyx_ErrRestore(type, value, tb)  __Pyx_ErrRestoreInState(__pyx_tstate, type, value, tb)
#define __Pyx_ErrFetch(type, value, tb)    __Pyx_ErrFetchInState(__pyx_tstate, type, value, tb)
static CYTHON_INLINE void __Pyx_ErrRestoreInState(PyThreadState *tstate, PyObject *type, PyObject *value, PyObject *tb);
static CYTHON_INLINE void __Pyx_ErrFetchInState(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb);
#if CYTHON_COMPILING_IN_CPYTHON
#define __Pyx_PyErr_SetNone(exc) (Py_INCREF(exc), __Pyx_ErrRestore((exc), NULL, NULL))
#else
#define __Pyx_PyErr_SetNone(exc) PyErr_SetNone(exc)
#endif
#else
#define __Pyx_PyErr_Clear() PyErr_Clear()
#define __Pyx_PyErr_SetNone(exc) PyErr_SetNone(exc)
#define __Pyx_ErrRestoreWithState(type, value, tb)  PyErr_Restore(type, value, tb)
#define __Pyx_ErrFetchWithState(type, value, tb)  PyErr_Fetch(type, value, tb)
#define __Pyx_ErrRestoreInState(tstate, type, value, tb)  PyErr_Restore(type, value, tb)
#define __Pyx_ErrFetchInState(tstate, type, value, tb)  PyErr_Fetch(type, value, tb)
#define __Pyx_ErrRestore(type, value, tb)  PyErr_Restore(type, value, tb)
#define __Pyx_ErrFetch(type, value, tb)  PyErr_Fetch(type, value, tb)
#endif

/* Profile.proto */
#ifndef CYTHON_PROFILE
#if CYTHON_COMPILING_IN_PYPY || CYTHON_COMPILING_IN_PYSTON
  #define CYTHON_PROFILE 0
#else
  #define CYTHON_PROFILE 1
#endif
#endif
#ifndef CYTHON_TRACE_NOGIL
  #define CYTHON_TRACE_NOGIL 0
#else
  #if CYTHON_TRACE_NOGIL && !defined(CYTHON_TRACE)
    #define CYTHON_TRACE 1
  #endif
#endif
#ifndef CYTHON_TRACE
  #define CYTHON_TRACE 0
#endif
#if CYTHON_TRACE
  #undef CYTHON_PROFILE_REUSE_FRAME
#endif
#ifndef CYTHON_PROFILE_REUSE_FRAME
  #define CYTHON_PROFILE_REUSE_FRAME 0
#endif
#if CYTHON_PROFILE || CYTHON_TRACE
  #include "compile.h"
  #include "frameobject.h"
  #include "traceback.h"
#if PY_VERSION_HEX >= 0x030b00a6
  #ifndef Py_BUILD_CORE
    #define Py_BUILD_CORE 1
  #endif
  #include "internal/pycore_frame.h"
#endif
  #if CYTHON_PROFILE_REUSE_FRAME
    #define CYTHON_FRAME_MODIFIER static
    #define CYTHON_FRAME_DEL(frame)
  #else
    #define CYTHON_FRAME_MODIFIER
    #define CYTHON_FRAME_DEL(frame) Py_CLEAR(frame)
  #endif
  #define __Pyx_TraceDeclarations\
      static PyCodeObject *__pyx_frame_code = NULL;\
      CYTHON_FRAME_MODIFIER PyFrameObject *__pyx_frame = NULL;\
      int __Pyx_use_tracing = 0;
  #define __Pyx_TraceFrameInit(codeobj)\
      if (codeobj) __pyx_frame_code = (PyCodeObject*) codeobj;
#if PY_VERSION_HEX >= 0x030b00a2
  #define __Pyx_IsTracing(tstate, check_tracing, check_funcs)\
     (unlikely((tstate)->cframe->use_tracing) &&\
         (!(check_tracing) || !(tstate)->tracing) &&\
         (!(check_funcs) || (tstate)->c_profilefunc || (CYTHON_TRACE && (tstate)->c_tracefunc)))
  #define __Pyx_EnterTracing(tstate) PyThreadState_EnterTracing(tstate)
  #define __Pyx_LeaveTracing(tstate) PyThreadState_LeaveTracing(tstate)
#elif PY_VERSION_HEX >= 0x030a00b1
  #define __Pyx_IsTracing(tstate, check_tracing, check_funcs)\
     (unlikely((tstate)->cframe->use_tracing) &&\
         (!(check_tracing) || !(tstate)->tracing) &&\
         (!(check_funcs) || (tstate)->c_profilefunc || (CYTHON_TRACE && (tstate)->c_tracefunc)))
  #define __Pyx_EnterTracing(tstate)\
      do { tstate->tracing++; tstate->cframe->use_tracing = 0; } while (0)
  #define __Pyx_LeaveTracing(tstate)\
      do {\
          tstate->tracing--;\
          tstate->cframe->use_tracing = ((CYTHON_TRACE && tstate->c_tracefunc != NULL)\
                                 || tstate->c_profilefunc != NULL);\
      } while (0)
#else
  #define __Pyx_IsTracing(tstate, check_tracing, check_funcs)\
     (unlikely((tstate)->use_tracing) &&\
         (!(check_tracing) || !(tstate)->tracing) &&\
         (!(check_funcs) || (tstate)->c_profilefunc || (CYTHON_TRACE && (tstate)->c_tracefunc)))
  #define __Pyx_EnterTracing(tstate)\
      do { tstate->tracing++; tstate->use_tracing = 0; } while (0)
  #define __Pyx_LeaveTracing(tstate)\
      do {\
          tstate->tracing--;\
          tstate->use_tracing = ((CYTHON_TRACE && tstate->c_tracefunc != NULL)\
                                         || tstate->c_profilefunc != NULL);\
      } while (0)
#endif
  #ifdef WITH_THREAD
  #define __Pyx_TraceCall(funcname, srcfile, firstlineno, nogil, goto_error)\
  if (nogil) {\
      if (CYTHON_TRACE_NOGIL) {\
          PyThreadState *tstate;\
          PyGILState_STATE state = PyGILState_Ensure();\
          tstate = __Pyx_PyThreadState_Current;\
          if (__Pyx_IsTracing(tstate, 1, 1)) {\
              __Pyx_use_tracing = __Pyx_TraceSetupAndCall(&__pyx_frame_code, &__pyx_frame, tstate, funcname, srcfile, firstlineno);\
          }\
          PyGILState_Release(state);\
          if (unlikely(__Pyx_use_tracing < 0)) goto_error;\
      }\
  } else {\
      PyThreadState* tstate = PyThreadState_GET();\
      if (__Pyx_IsTracing(tstate, 1, 1)) {\
          __Pyx_use_tracing = __Pyx_TraceSetupAndCall(&__pyx_frame_code, &__pyx_frame, tstate, funcname, srcfile, firstlineno);\
          if (unlikely(__Pyx_use_tracing < 0)) goto_error;\
      }\
  }
  #else
  #define __Pyx_TraceCall(funcname, srcfile, firstlineno, nogil, goto_error)\
  {   PyThreadState* tstate = PyThreadState_GET();\
      if (__Pyx_IsTracing(tstate, 1, 1)) {\
          __Pyx_use_tracing = __Pyx_TraceSetupAndCall(&__pyx_frame_code, &__pyx_frame, tstate, funcname, srcfile, firstlineno);\
          if (unlikely(__Pyx_use_tracing < 0)) goto_error;\
      }\
  }
  #endif
  #define __Pyx_TraceException()\
  if (likely(!__Pyx_use_tracing)); else {\
      PyThreadState* tstate = __Pyx_PyThreadState_Current;\
      if (__Pyx_IsTracing(tstate, 0, 1)) {\
          __Pyx_EnterTracing(tstate);\
          PyObject *exc_info = __Pyx_GetExceptionTuple(tstate);\
          if (exc_info) {\
              if (CYTHON_TRACE && tstate->c_tracefunc)\
                  tstate->c_tracefunc(\
                      tstate->c_traceobj, __pyx_frame, PyTrace_EXCEPTION, exc_info);\
              tstate->c_profilefunc(\
                  tstate->c_profileobj, __pyx_frame, PyTrace_EXCEPTION, exc_info);\
              Py_DECREF(exc_info);\
          }\
          __Pyx_LeaveTracing(tstate);\
      }\
  }
  static void __Pyx_call_return_trace_func(PyThreadState *tstate, PyFrameObject *frame, PyObject *result) {
      PyObject *type, *value, *traceback;
      __Pyx_ErrFetchInState(tstate, &type, &value, &traceback);
      __Pyx_EnterTracing(tstate);
      if (CYTHON_TRACE && tstate->c_tracefunc)
          tstate->c_tracefunc(tstate->c_traceobj, frame, PyTrace_RETURN, result);
      if (tstate->c_profilefunc)
          tstate->c_profilefunc(tstate->c_profileobj, frame, PyTrace_RETURN, result);
      CYTHON_FRAME_DEL(frame);
      __Pyx_LeaveTracing(tstate);
      __Pyx_ErrRestoreInState(tstate, type, value, traceback);
  }
  #ifdef WITH_THREAD
  #define __Pyx_TraceReturn(result, nogil)\
  if (likely(!__Pyx_use_tracing)); else {\
      if (nogil) {\
          if (CYTHON_TRACE_NOGIL) {\
              PyThreadState *tstate;\
              PyGILState_STATE state = PyGILState_Ensure();\
              tstate = __Pyx_PyThreadState_Current;\
              if (__Pyx_IsTracing(tstate, 0, 0)) {\
                  __Pyx_call_return_trace_func(tstate, __pyx_frame, (PyObject*)result);\
              }\
              PyGILState_Release(state);\
          }\
      } else {\
          PyThreadState* tstate = __Pyx_PyThreadState_Current;\
          if (__Pyx_IsTracing(tstate, 0, 0)) {\
              __Pyx_call_return_trace_func(tstate, __pyx_frame, (PyObject*)result);\
          }\
      }\
  }
  #else
  #define __Pyx_TraceReturn(result, nogil)\
  if (likely(!__Pyx_use_tracing)); else {\
      PyThreadState* tstate = __Pyx_PyThreadState_Current;\
      if (__Pyx_IsTracing(tstate, 0, 0)) {\
          __Pyx_call_return_trace_func(tstate, __pyx_frame, (PyObject*)result);\
      }\
  }
  #endif
  static PyCodeObject *__Pyx_createFrameCodeObject(const char *funcname, const char *srcfile, int firstlineno);
  static int __Pyx_TraceSetupAndCall(PyCodeObject** code, PyFrameObject** frame, PyThreadState* tstate, const char *funcname, const char *srcfile, int firstlineno);
#else
  #define __Pyx_TraceDeclarations
  #define __Pyx_TraceFrameInit(codeobj)
  #define __Pyx_TraceCall(funcname, srcfile, firstlineno, nogil, goto_error)   if ((1)); else goto_error;
  #define __Pyx_TraceException()
  #define __Pyx_TraceReturn(result, nogil)
#endif
#if CYTHON_TRACE
  static int __Pyx_call_line_trace_func(PyThreadState *tstate, PyFrameObject *frame, int lineno) {
      int ret;
      PyObject *type, *value, *traceback;
      __Pyx_ErrFetchInState(tstate, &type, &value, &traceback);
      __Pyx_PyFrame_SetLineNumber(frame, lineno);
      __Pyx_EnterTracing(tstate);
      ret = tstate->c_tracefunc(tstate->c_traceobj, frame, PyTrace_LINE, NULL);
      __Pyx_LeaveTracing(tstate);
      if (likely(!ret)) {
          __Pyx_ErrRestoreInState(tstate, type, value, traceback);
      } else {
          Py_XDECREF(type);
          Py_XDECREF(value);
          Py_XDECREF(traceback);
      }
      return ret;
  }
  #ifdef WITH_THREAD
  #define __Pyx_TraceLine(lineno, nogil, goto_error)\
  if (likely(!__Pyx_use_tracing)); else {\
      if (nogil) {\
          if (CYTHON_TRACE_NOGIL) {\
              int ret = 0;\
              PyThreadState *tstate;\
              PyGILState_STATE state = PyGILState_Ensure();\
              tstate = __Pyx_PyThreadState_Current;\
              if (__Pyx_IsTracing(tstate, 0, 0) && tstate->c_tracefunc && __pyx_frame->f_trace) {\
                  ret = __Pyx_call_line_trace_func(tstate, __pyx_frame, lineno);\
              }\
              PyGILState_Release(state);\
              if (unlikely(ret)) goto_error;\
          }\
      } else {\
          PyThreadState* tstate = __Pyx_PyThreadState_Current;\
          if (__Pyx_IsTracing(tstate, 0, 0) && tstate->c_tracefunc && __pyx_frame->f_trace) {\
              int ret = __Pyx_call_line_trace_func(tstate, __pyx_frame, lineno);\
              if (unlikely(ret)) goto_error;\
          }\
      }\
  }
  #else
  #define __Pyx_TraceLine(lineno, nogil, goto_error)\
  if (likely(!__Pyx_use_tracing)); else {\
      PyThreadState* tstate = __Pyx_PyThreadState_Current;\
      if (__Pyx_IsTracing(tstate, 0, 0) && tstate->c_tracefunc && __pyx_frame->f_trace) {\
          int ret = __Pyx_call_line_trace_func(tstate, __pyx_frame, lineno);\
          if (unlikely(ret)) goto_error;\
      }\
  }
  #endif
#else
  #define __Pyx_TraceLine(lineno, nogil, goto_error)   if ((1)); else goto_error;
#endif

/* PyDictVersioning.proto */
#if CYTHON_USE_DICT_VERSIONS && CYTHON_USE_TYPE_SLOTS
#define __PYX_DICT_VERSION_INIT  ((PY_UINT64_T) -1)
#define __PYX_GET_DICT_VERSION(dict)  (((PyDictObject*)(dict))->ma_version_tag)
#define __PYX_UPDATE_DICT_CACHE(dict, value, cache_var, version_var)\
    (version_var) = __PYX_GET_DICT_VERSION(dict);\
    (cache_var) = (value);
#define __PYX_PY_DICT_LOOKUP_IF_MODIFIED(VAR, DICT, LOOKUP) {\
    static PY_UINT64_T __pyx_dict_version = 0;\
    static PyObject *__pyx_dict_cached_value = NULL;\
    if (likely(__PYX_GET_DICT_VERSION(DICT) == __pyx_dict_version)) {\
        (VAR) = __pyx_dict_cached_value;\
    } else {\
        (VAR) = __pyx_dict_cached_value = (LOOKUP);\
        __pyx_dict_version = __PYX_GET_DICT_VERSION(DICT);\
    }\
}
static CYTHON_INLINE PY_UINT64_T __Pyx_get_tp_dict_version(PyObject *obj);
static CYTHON_INLINE PY_UINT64_T __Pyx_get_object_dict_version(PyObject *obj);
static CYTHON_INLINE int __Pyx_object_dict_version_matches(PyObject* obj, PY_UINT64_T tp_dict_version, PY_UINT64_T obj_dict_version);
#else
#define __PYX_GET_DICT_VERSION(dict)  (0)
#define __PYX_UPDATE_DICT_CACHE(dict, value, cache_var, version_var)
#define __PYX_PY_DICT_LOOKUP_IF_MODIFIED(VAR, DICT, LOOKUP)  (VAR) = (LOOKUP);
#endif

/* GetModuleGlobalName.proto */
#if CYTHON_USE_DICT_VERSIONS
#define __Pyx_GetModuleGlobalName(var, name)  do {\
    static PY_UINT64_T __pyx_dict_version = 0;\
    static PyObject *__pyx_dict_cached_value = NULL;\
    (var) = (likely(__pyx_dict_version == __PYX_GET_DICT_VERSION(__pyx_d))) ?\
        (likely(__pyx_dict_cached_value) ? __Pyx_NewRef(__pyx_dict_cached_value) : __Pyx_GetBuiltinName(name)) :\
        __Pyx__GetModuleGlobalName(name, &__pyx_dict_version, &__pyx_dict_cached_value);\
} while(0)
#define __Pyx_GetModuleGlobalNameUncached(var, name)  do {\
    PY_UINT64_T __pyx_dict_version;\
    PyObject *__pyx_dict_cached_value;\
    (var) = __Pyx__GetModuleGlobalName(name, &__pyx_dict_version, &__pyx_dict_cached_value);\
} while(0)
static PyObject *__Pyx__GetModuleGlobalName(PyObject *name, PY_UINT64_T *dict_version, PyObject **dict_cached_value);
#else
#define __Pyx_GetModuleGlobalName(var, name)  (var) = __Pyx__GetModuleGlobalName(name)
#define __Pyx_GetModuleGlobalNameUncached(var, name)  (var) = __Pyx__GetModuleGlobalName(name)
static CYTHON_INLINE PyObject *__Pyx__GetModuleGlobalName(PyObject *name);
#endif

/* PyFunctionFastCall.proto */
#if CYTHON_FAST_PYCALL
#define __Pyx_PyFunction_FastCall(func, args, nargs)\
    __Pyx_PyFunction_FastCallDict((func), (args), (nargs), NULL)
#if 1 || PY_VERSION_HEX < 0x030600B1
static PyObject *__Pyx_PyFunction_FastCallDict(PyObject *func, PyObject **args, Py_ssize_t nargs, PyObject *kwargs);
#else
#define __Pyx_PyFunction_FastCallDict(func, args, nargs, kwargs) _PyFunction_FastCallDict(func, args, nargs, kwargs)
#endif
#define __Pyx_BUILD_ASSERT_EXPR(cond)\
    (sizeof(char [1 - 2*!(cond)]) - 1)
#ifndef Py_MEMBER_SIZE
#define Py_MEMBER_SIZE(type, member) sizeof(((type *)0)->member)
#endif
#if CYTHON_FAST_PYCALL
  static size_t __pyx_pyframe_localsplus_offset = 0;
  #include "frameobject.h"
#if PY_VERSION_HEX >= 0x030b00a6
  #ifndef Py_BUILD_CORE
    #define Py_BUILD_CORE 1
  #endif
  #include "internal/pycore_frame.h"
#endif
  #define __Pxy_PyFrame_Initialize_Offsets()\
    ((void)__Pyx_BUILD_ASSERT_EXPR(sizeof(PyFrameObject) == offsetof(PyFrameObject, f_localsplus) + Py_MEMBER_SIZE(PyFrameObject, f_localsplus)),\
     (void)(__pyx_pyframe_localsplus_offset = ((size_t)PyFrame_Type.tp_basicsize) - Py_MEMBER_SIZE(PyFrameObject, f_localsplus)))
  #define __Pyx_PyFrame_GetLocalsplus(frame)\
    (assert(__pyx_pyframe_localsplus_offset), (PyObject **)(((char *)(frame)) + __pyx_pyframe_localsplus_offset))
#endif // CYTHON_FAST_PYCALL
#endif

/* PyCFunctionFastCall.proto */
#if CYTHON_FAST_PYCCALL
static CYTHON_INLINE PyObject *__Pyx_PyCFunction_FastCall(PyObject *func, PyObject **args, Py_ssize_t nargs);
#else
#define __Pyx_PyCFunction_FastCall(func, args, nargs)  (assert(0), NULL)
#endif

/* PyObjectCall.proto */
#if CYTHON_COMPILING_IN_CPYTHON
static CYTHON_INLINE PyObject* __Pyx_PyObject_Call(PyObject *func, PyObject *arg, PyObject *kw);
#else
#define __Pyx_PyObject_Call(func, arg, kw) PyObject_Call(func, arg, kw)
#endif

/* GetTopmostException.proto */
#if CYTHON_USE_EXC_INFO_STACK
static _PyErr_StackItem * __Pyx_PyErr_GetTopmostException(PyThreadState *tstate);
#endif

/* SaveResetException.proto */
#if CYTHON_FAST_THREAD_STATE
#define __Pyx_ExceptionSave(type, value, tb)  __Pyx__ExceptionSave(__pyx_tstate, type, value, tb)
static CYTHON_INLINE void __Pyx__ExceptionSave(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb);
#define __Pyx_ExceptionReset(type, value, tb)  __Pyx__ExceptionReset(__pyx_tstate, type, value, tb)
static CYTHON_INLINE void __Pyx__ExceptionReset(PyThreadState *tstate, PyObject *type, PyObject *value, PyObject *tb);
#else
#define __Pyx_ExceptionSave(type, value, tb)   PyErr_GetExcInfo(type, value, tb)
#define __Pyx_ExceptionReset(type, value, tb)  PyErr_SetExcInfo(type, value, tb)
#endif

/* GetException.proto */
#if CYTHON_FAST_THREAD_STATE
#define __Pyx_GetException(type, value, tb)  __Pyx__GetException(__pyx_tstate, type, value, tb)
static int __Pyx__GetException(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb);
#else
static int __Pyx_GetException(PyObject **type, PyObject **value, PyObject **tb);
#endif

/* RaiseException.proto */
static void __Pyx_Raise(PyObject *type, PyObject *value, PyObject *tb, PyObject *cause);

/* Import.proto */
static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, int level);

/* ImportFrom.proto */
static PyObject* __Pyx_ImportFrom(PyObject* module, PyObject* name);

/* CLineInTraceback.proto */
#ifdef CYTHON_CLINE_IN_TRACEBACK
#define __Pyx_CLineForTraceback(tstate, c_line)  (((CYTHON_CLINE_IN_TRACEBACK)) ? c_line : 0)
#else
static int __Pyx_CLineForTraceback(PyThreadState *tstate, int c_line);
#endif

/* CodeObjectCache.proto */
typedef struct {
    PyCodeObject* code_object;
    int code_line;
} __Pyx_CodeObjectCacheEntry;
struct __Pyx_CodeObjectCache {
    int count;
    int max_count;
    __Pyx_CodeObjectCacheEntry* entries;
};
static struct __Pyx_CodeObjectCache __pyx_code_cache = {0,0,NULL};
static int __pyx_bisect_code_objects(__Pyx_CodeObjectCacheEntry* entries, int count, int code_line);
static PyCodeObject *__pyx_find_code_object(int code_line);
static void __pyx_insert_code_object(int code_line, PyCodeObject* code_object);

/* AddTraceback.proto */
static void __Pyx_AddTraceback(const char *funcname, int c_line,
                               int py_line, const char *filename);

/* GCCDiagnostics.proto */
#if defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 6))
#define __Pyx_HAS_GCC_DIAGNOSTIC
#endif

/* CIntToPy.proto */
static CYTHON_INLINE PyObject* __Pyx_PyInt_From_long(long value);

/* CIntFromPy.proto */
static CYTHON_INLINE unsigned PY_LONG_LONG __Pyx_PyInt_As_unsigned_PY_LONG_LONG(PyObject *);

/* CIntFromPy.proto */
static CYTHON_INLINE long __Pyx_PyInt_As_long(PyObject *);

/* CIntFromPy.proto */
static CYTHON_INLINE int __Pyx_PyInt_As_int(PyObject *);

/* FastTypeChecks.proto */
#if CYTHON_COMPILING_IN_CPYTHON
#define __Pyx_TypeCheck(obj, type) __Pyx_IsSubtype(Py_TYPE(obj), (PyTypeObject *)type)
static CYTHON_INLINE int __Pyx_IsSubtype(PyTypeObject *a, PyTypeObject *b);
static CYTHON_INLINE int __Pyx_PyErr_GivenExceptionMatches(PyObject *err, PyObject *type);
static CYTHON_INLINE int __Pyx_PyErr_GivenExceptionMatches2(PyObject *err, PyObject *type1, PyObject *type2);
#else
#define __Pyx_TypeCheck(obj, type) PyObject_TypeCheck(obj, (PyTypeObject *)type)
#define __Pyx_PyErr_GivenExceptionMatches(err, type) PyErr_GivenExceptionMatches(err, type)
#define __Pyx_PyErr_GivenExceptionMatches2(err, type1, type2) (PyErr_GivenExceptionMatches(err, type1) || PyErr_GivenExceptionMatches(err, type2))
#endif
#define __Pyx_PyException_Check(obj) __Pyx_TypeCheck(obj, PyExc_Exception)

/* CheckBinaryVersion.proto */
static int __Pyx_check_binary_version(void);

/* FunctionExport.proto */
static int __Pyx_ExportFunction(const char *name, void (*f)(void), const char *sig);

/* InitStrings.proto */
static int __Pyx_InitStrings(__Pyx_StringTabEntry *t);


/* Module declarations from 'libc.stdint' */

/* Module declarations from 'cuda.cnvrtc' */

/* Module declarations from 'cuda._cuda.cnvrtc' */
static int __pyx_v_4cuda_5_cuda_6cnvrtc___cuPythonInit;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetErrorString;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcVersion;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNumSupportedArchs;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetSupportedArchs;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCreateProgram;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcDestroyProgram;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCompileProgram;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTXSize;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTX;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBINSize;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBIN;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVMSize;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVM;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIRSize;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIR;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIRSize;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIR;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLogSize;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLog;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcAddNameExpression;
static void *__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLoweredName;
static int __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(void); /*proto*/
#define __Pyx_MODULE_NAME "cuda._cuda.cnvrtc"
extern int __pyx_module_is_main_cuda___cuda__cnvrtc;
int __pyx_module_is_main_cuda___cuda__cnvrtc = 0;

/* Implementation of 'cuda._cuda.cnvrtc' */
static PyObject *__pyx_builtin_RuntimeError;
static const char __pyx_k_main[] = "__main__";
static const char __pyx_k_name[] = "__name__";
static const char __pyx_k_test[] = "__test__";
static const char __pyx_k_error[] = "error";
static const char __pyx_k_import[] = "__import__";
static const char __pyx_k_struct[] = "struct";
static const char __pyx_k_win32api[] = "win32api";
static const char __pyx_k_pywintypes[] = "pywintypes";
static const char __pyx_k_nvrtcGetPTX[] = "nvrtcGetPTX";
static const char __pyx_k_RuntimeError[] = "RuntimeError";
static const char __pyx_k_nvrtcGetNVVM[] = "nvrtcGetNVVM";
static const char __pyx_k_nvrtcVersion[] = "nvrtcVersion";
static const char __pyx_k_LoadLibraryEx[] = "LoadLibraryEx";
static const char __pyx_k_nvrtcGetCUBIN[] = "nvrtcGetCUBIN";
static const char __pyx_k_nvrtcGetLTOIR[] = "nvrtcGetLTOIR";
static const char __pyx_k_GetProcAddress[] = "GetProcAddress";
static const char __pyx_k_nvrtcGetOptiXIR[] = "nvrtcGetOptiXIR";
static const char __pyx_k_nvrtcGetPTXSize[] = "nvrtcGetPTXSize";
static const char __pyx_k_nvrtcGetNVVMSize[] = "nvrtcGetNVVMSize";
static const char __pyx_k_nvrtc64_120_0_dll[] = "nvrtc64_120_0.dll";
static const char __pyx_k_nvrtcGetCUBINSize[] = "nvrtcGetCUBINSize";
static const char __pyx_k_nvrtcGetLTOIRSize[] = "nvrtcGetLTOIRSize";
static const char __pyx_k_cline_in_traceback[] = "cline_in_traceback";
static const char __pyx_k_nvrtcCreateProgram[] = "nvrtcCreateProgram";
static const char __pyx_k_nvrtcGetProgramLog[] = "nvrtcGetProgramLog";
static const char __pyx_k_nvrtcCompileProgram[] = "nvrtcCompileProgram";
static const char __pyx_k_nvrtcDestroyProgram[] = "nvrtcDestroyProgram";
static const char __pyx_k_nvrtcGetErrorString[] = "nvrtcGetErrorString";
static const char __pyx_k_nvrtcGetLoweredName[] = "nvrtcGetLoweredName";
static const char __pyx_k_nvrtcGetOptiXIRSize[] = "nvrtcGetOptiXIRSize";
static const char __pyx_k_nvrtcAddNameExpression[] = "nvrtcAddNameExpression";
static const char __pyx_k_nvrtcGetProgramLogSize[] = "nvrtcGetProgramLogSize";
static const char __pyx_k_nvrtcGetSupportedArchs[] = "nvrtcGetSupportedArchs";
static const char __pyx_k_nvrtcGetNumSupportedArchs[] = "nvrtcGetNumSupportedArchs";
static const char __pyx_k_Function_nvrtcGetPTX_not_found[] = "Function \"nvrtcGetPTX\" not found";
static const char __pyx_k_Failed_to_LoadLibraryEx_nvrtc64[] = "Failed to LoadLibraryEx nvrtc64_120_0.dll";
static const char __pyx_k_Function_nvrtcAddNameExpression[] = "Function \"nvrtcAddNameExpression\" not found";
static const char __pyx_k_Function_nvrtcCreateProgram_not[] = "Function \"nvrtcCreateProgram\" not found";
static const char __pyx_k_Function_nvrtcGetNVVM_not_found[] = "Function \"nvrtcGetNVVM\" not found";
static const char __pyx_k_Function_nvrtcGetProgramLogSize[] = "Function \"nvrtcGetProgramLogSize\" not found";
static const char __pyx_k_Function_nvrtcGetProgramLog_not[] = "Function \"nvrtcGetProgramLog\" not found";
static const char __pyx_k_Function_nvrtcGetSupportedArchs[] = "Function \"nvrtcGetSupportedArchs\" not found";
static const char __pyx_k_Function_nvrtcVersion_not_found[] = "Function \"nvrtcVersion\" not found";
static const char __pyx_k_Function_nvrtcCompileProgram_not[] = "Function \"nvrtcCompileProgram\" not found";
static const char __pyx_k_Function_nvrtcDestroyProgram_not[] = "Function \"nvrtcDestroyProgram\" not found";
static const char __pyx_k_Function_nvrtcGetCUBINSize_not_f[] = "Function \"nvrtcGetCUBINSize\" not found";
static const char __pyx_k_Function_nvrtcGetCUBIN_not_found[] = "Function \"nvrtcGetCUBIN\" not found";
static const char __pyx_k_Function_nvrtcGetErrorString_not[] = "Function \"nvrtcGetErrorString\" not found";
static const char __pyx_k_Function_nvrtcGetLTOIRSize_not_f[] = "Function \"nvrtcGetLTOIRSize\" not found";
static const char __pyx_k_Function_nvrtcGetLTOIR_not_found[] = "Function \"nvrtcGetLTOIR\" not found";
static const char __pyx_k_Function_nvrtcGetLoweredName_not[] = "Function \"nvrtcGetLoweredName\" not found";
static const char __pyx_k_Function_nvrtcGetNVVMSize_not_fo[] = "Function \"nvrtcGetNVVMSize\" not found";
static const char __pyx_k_Function_nvrtcGetNumSupportedArc[] = "Function \"nvrtcGetNumSupportedArchs\" not found";
static const char __pyx_k_Function_nvrtcGetOptiXIRSize_not[] = "Function \"nvrtcGetOptiXIRSize\" not found";
static const char __pyx_k_Function_nvrtcGetOptiXIR_not_fou[] = "Function \"nvrtcGetOptiXIR\" not found";
static const char __pyx_k_Function_nvrtcGetPTXSize_not_fou[] = "Function \"nvrtcGetPTXSize\" not found";
static PyObject *__pyx_kp_u_Failed_to_LoadLibraryEx_nvrtc64;
static PyObject *__pyx_kp_u_Function_nvrtcAddNameExpression;
static PyObject *__pyx_kp_u_Function_nvrtcCompileProgram_not;
static PyObject *__pyx_kp_u_Function_nvrtcCreateProgram_not;
static PyObject *__pyx_kp_u_Function_nvrtcDestroyProgram_not;
static PyObject *__pyx_kp_u_Function_nvrtcGetCUBINSize_not_f;
static PyObject *__pyx_kp_u_Function_nvrtcGetCUBIN_not_found;
static PyObject *__pyx_kp_u_Function_nvrtcGetErrorString_not;
static PyObject *__pyx_kp_u_Function_nvrtcGetLTOIRSize_not_f;
static PyObject *__pyx_kp_u_Function_nvrtcGetLTOIR_not_found;
static PyObject *__pyx_kp_u_Function_nvrtcGetLoweredName_not;
static PyObject *__pyx_kp_u_Function_nvrtcGetNVVMSize_not_fo;
static PyObject *__pyx_kp_u_Function_nvrtcGetNVVM_not_found;
static PyObject *__pyx_kp_u_Function_nvrtcGetNumSupportedArc;
static PyObject *__pyx_kp_u_Function_nvrtcGetOptiXIRSize_not;
static PyObject *__pyx_kp_u_Function_nvrtcGetOptiXIR_not_fou;
static PyObject *__pyx_kp_u_Function_nvrtcGetPTXSize_not_fou;
static PyObject *__pyx_kp_u_Function_nvrtcGetPTX_not_found;
static PyObject *__pyx_kp_u_Function_nvrtcGetProgramLogSize;
static PyObject *__pyx_kp_u_Function_nvrtcGetProgramLog_not;
static PyObject *__pyx_kp_u_Function_nvrtcGetSupportedArchs;
static PyObject *__pyx_kp_u_Function_nvrtcVersion_not_found;
static PyObject *__pyx_n_s_GetProcAddress;
static PyObject *__pyx_n_s_LoadLibraryEx;
static PyObject *__pyx_n_s_RuntimeError;
static PyObject *__pyx_n_s_cline_in_traceback;
static PyObject *__pyx_n_s_error;
static PyObject *__pyx_n_s_import;
static PyObject *__pyx_n_s_main;
static PyObject *__pyx_n_s_name;
static PyObject *__pyx_kp_u_nvrtc64_120_0_dll;
static PyObject *__pyx_n_u_nvrtcAddNameExpression;
static PyObject *__pyx_n_u_nvrtcCompileProgram;
static PyObject *__pyx_n_u_nvrtcCreateProgram;
static PyObject *__pyx_n_u_nvrtcDestroyProgram;
static PyObject *__pyx_n_u_nvrtcGetCUBIN;
static PyObject *__pyx_n_u_nvrtcGetCUBINSize;
static PyObject *__pyx_n_u_nvrtcGetErrorString;
static PyObject *__pyx_n_u_nvrtcGetLTOIR;
static PyObject *__pyx_n_u_nvrtcGetLTOIRSize;
static PyObject *__pyx_n_u_nvrtcGetLoweredName;
static PyObject *__pyx_n_u_nvrtcGetNVVM;
static PyObject *__pyx_n_u_nvrtcGetNVVMSize;
static PyObject *__pyx_n_u_nvrtcGetNumSupportedArchs;
static PyObject *__pyx_n_u_nvrtcGetOptiXIR;
static PyObject *__pyx_n_u_nvrtcGetOptiXIRSize;
static PyObject *__pyx_n_u_nvrtcGetPTX;
static PyObject *__pyx_n_u_nvrtcGetPTXSize;
static PyObject *__pyx_n_u_nvrtcGetProgramLog;
static PyObject *__pyx_n_u_nvrtcGetProgramLogSize;
static PyObject *__pyx_n_u_nvrtcGetSupportedArchs;
static PyObject *__pyx_n_u_nvrtcVersion;
static PyObject *__pyx_n_s_pywintypes;
static PyObject *__pyx_n_s_struct;
static PyObject *__pyx_n_s_test;
static PyObject *__pyx_n_s_win32api;
static PyObject *__pyx_int_0;
static PyObject *__pyx_tuple_;
static PyObject *__pyx_tuple__2;
static PyObject *__pyx_tuple__3;
static PyObject *__pyx_tuple__4;
static PyObject *__pyx_tuple__5;
static PyObject *__pyx_tuple__6;
static PyObject *__pyx_tuple__7;
static PyObject *__pyx_tuple__8;
static PyObject *__pyx_tuple__9;
static PyObject *__pyx_tuple__10;
static PyObject *__pyx_tuple__11;
static PyObject *__pyx_tuple__12;
static PyObject *__pyx_tuple__13;
static PyObject *__pyx_tuple__14;
static PyObject *__pyx_tuple__15;
static PyObject *__pyx_tuple__16;
static PyObject *__pyx_tuple__17;
static PyObject *__pyx_tuple__18;
static PyObject *__pyx_tuple__19;
static PyObject *__pyx_tuple__20;
static PyObject *__pyx_tuple__21;
static PyObject *__pyx_tuple__22;
/* Late includes */

/* "cuda/_cuda/cnvrtc.pyx":35
 * cdef void *__nvrtcGetLoweredName = NULL
 * 
 * cdef int cuPythonInit() except -1 nogil:             # <<<<<<<<<<<<<<
 *     global __cuPythonInit
 *     if __cuPythonInit:
 */

static int __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(void) {
  long __pyx_v_LOAD_LIBRARY_SAFE_CURRENT_DIRS;
  PyObject *__pyx_v_handle = NULL;
  int __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  PyObject *__pyx_t_2 = NULL;
  PyObject *__pyx_t_3 = NULL;
  PyObject *__pyx_t_4 = NULL;
  PyObject *__pyx_t_5 = NULL;
  PyObject *__pyx_t_6 = NULL;
  PyObject *__pyx_t_7 = NULL;
  PyObject *__pyx_t_8 = NULL;
  int __pyx_t_9;
  PyObject *__pyx_t_10 = NULL;
  unsigned PY_LONG_LONG __pyx_t_11;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("cuPythonInit", 1);
  __Pyx_TraceCall("cuPythonInit", __pyx_f[0], 35, 1, __PYX_ERR(0, 35, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":36
 * 
 * cdef int cuPythonInit() except -1 nogil:
 *     global __cuPythonInit             # <<<<<<<<<<<<<<
 *     if __cuPythonInit:
 *         return 0
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":37
 * cdef int cuPythonInit() except -1 nogil:
 *     global __cuPythonInit
 *     if __cuPythonInit:             # <<<<<<<<<<<<<<
 *         return 0
 *     __cuPythonInit = True
 */
    __pyx_t_1 = (__pyx_v_4cuda_5_cuda_6cnvrtc___cuPythonInit != 0);
    if (__pyx_t_1) {

      /* "cuda/_cuda/cnvrtc.pyx":38
 *     global __cuPythonInit
 *     if __cuPythonInit:
 *         return 0             # <<<<<<<<<<<<<<
 *     __cuPythonInit = True
 * 
 */
      __pyx_r = 0;
      goto __pyx_L3_return;

      /* "cuda/_cuda/cnvrtc.pyx":37
 * cdef int cuPythonInit() except -1 nogil:
 *     global __cuPythonInit
 *     if __cuPythonInit:             # <<<<<<<<<<<<<<
 *         return 0
 *     __cuPythonInit = True
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":39
 *     if __cuPythonInit:
 *         return 0
 *     __cuPythonInit = True             # <<<<<<<<<<<<<<
 * 
 *     # Load library
 */
    __pyx_v_4cuda_5_cuda_6cnvrtc___cuPythonInit = 1;

    /* "cuda/_cuda/cnvrtc.pyx":42
 * 
 *     # Load library
 *     LOAD_LIBRARY_SAFE_CURRENT_DIRS = 0x00002000             # <<<<<<<<<<<<<<
 *     with gil:
 *         try:
 */
    __pyx_v_LOAD_LIBRARY_SAFE_CURRENT_DIRS = 0x00002000;

    /* "cuda/_cuda/cnvrtc.pyx":43
 *     # Load library
 *     LOAD_LIBRARY_SAFE_CURRENT_DIRS = 0x00002000
 *     with gil:             # <<<<<<<<<<<<<<
 *         try:
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 */
    {
        #ifdef WITH_THREAD
        PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
        #endif
        /*try:*/ {

          /* "cuda/_cuda/cnvrtc.pyx":44
 *     LOAD_LIBRARY_SAFE_CURRENT_DIRS = 0x00002000
 *     with gil:
 *         try:             # <<<<<<<<<<<<<<
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 *         except:
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":45
 *     with gil:
 *         try:
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)             # <<<<<<<<<<<<<<
 *         except:
 *             raise RuntimeError('Failed to LoadLibraryEx nvrtc64_120_0.dll')
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 45, __pyx_L10_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_LoadLibraryEx); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 45, __pyx_L10_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = __Pyx_PyInt_From_long(__pyx_v_LOAD_LIBRARY_SAFE_CURRENT_DIRS); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 45, __pyx_L10_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_8 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_8 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_8)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_8);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[4] = {__pyx_t_8, __pyx_kp_u_nvrtc64_120_0_dll, __pyx_int_0, __pyx_t_6};
                __pyx_t_5 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 3+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 45, __pyx_L10_error)
                __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;
                __Pyx_GOTREF(__pyx_t_5);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[4] = {__pyx_t_8, __pyx_kp_u_nvrtc64_120_0_dll, __pyx_int_0, __pyx_t_6};
                __pyx_t_5 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 3+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 45, __pyx_L10_error)
                __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;
                __Pyx_GOTREF(__pyx_t_5);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              } else
              #endif
              {
                __pyx_t_10 = PyTuple_New(3+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 45, __pyx_L10_error)
                __Pyx_GOTREF(__pyx_t_10);
                if (__pyx_t_8) {
                  __Pyx_GIVEREF(__pyx_t_8); PyTuple_SET_ITEM(__pyx_t_10, 0, __pyx_t_8); __pyx_t_8 = NULL;
                }
                __Pyx_INCREF(__pyx_kp_u_nvrtc64_120_0_dll);
                __Pyx_GIVEREF(__pyx_kp_u_nvrtc64_120_0_dll);
                PyTuple_SET_ITEM(__pyx_t_10, 0+__pyx_t_9, __pyx_kp_u_nvrtc64_120_0_dll);
                __Pyx_INCREF(__pyx_int_0);
                __Pyx_GIVEREF(__pyx_int_0);
                PyTuple_SET_ITEM(__pyx_t_10, 1+__pyx_t_9, __pyx_int_0);
                __Pyx_GIVEREF(__pyx_t_6);
                PyTuple_SET_ITEM(__pyx_t_10, 2+__pyx_t_9, __pyx_t_6);
                __pyx_t_6 = 0;
                __pyx_t_5 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_10, NULL); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 45, __pyx_L10_error)
                __Pyx_GOTREF(__pyx_t_5);
                __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_v_handle = __pyx_t_5;
              __pyx_t_5 = 0;

              /* "cuda/_cuda/cnvrtc.pyx":44
 *     LOAD_LIBRARY_SAFE_CURRENT_DIRS = 0x00002000
 *     with gil:
 *         try:             # <<<<<<<<<<<<<<
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 *         except:
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L15_try_end;
            __pyx_L10_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":46
 *         try:
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 *         except:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Failed to LoadLibraryEx nvrtc64_120_0.dll')
 * 
 */
            /*except:*/ {
              __Pyx_AddTraceback("cuda._cuda.cnvrtc.cuPythonInit", __pyx_clineno, __pyx_lineno, __pyx_filename);
              if (__Pyx_GetException(&__pyx_t_5, &__pyx_t_7, &__pyx_t_10) < 0) __PYX_ERR(0, 46, __pyx_L12_except_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_GOTREF(__pyx_t_10);

              /* "cuda/_cuda/cnvrtc.pyx":47
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 *         except:
 *             raise RuntimeError('Failed to LoadLibraryEx nvrtc64_120_0.dll')             # <<<<<<<<<<<<<<
 * 
 * 
 */
              __pyx_t_6 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple_, NULL); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 47, __pyx_L12_except_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_Raise(__pyx_t_6, 0, 0, 0);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __PYX_ERR(0, 47, __pyx_L12_except_error)
            }
            __pyx_L12_except_error:;

            /* "cuda/_cuda/cnvrtc.pyx":44
 *     LOAD_LIBRARY_SAFE_CURRENT_DIRS = 0x00002000
 *     with gil:
 *         try:             # <<<<<<<<<<<<<<
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 *         except:
 */
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            goto __pyx_L8_error;
            __pyx_L15_try_end:;
          }
        }

        /* "cuda/_cuda/cnvrtc.pyx":43
 *     # Load library
 *     LOAD_LIBRARY_SAFE_CURRENT_DIRS = 0x00002000
 *     with gil:             # <<<<<<<<<<<<<<
 *         try:
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 */
        /*finally:*/ {
          /*normal exit:*/{
            #ifdef WITH_THREAD
            __Pyx_PyGILState_Release(__pyx_gilstate_save);
            #endif
            goto __pyx_L9;
          }
          __pyx_L8_error: {
            #ifdef WITH_THREAD
            __Pyx_PyGILState_Release(__pyx_gilstate_save);
            #endif
            goto __pyx_L4_error;
          }
          __pyx_L9:;
        }
    }

    /* "cuda/_cuda/cnvrtc.pyx":51
 * 
 *     # Load function
 *     with gil:             # <<<<<<<<<<<<<<
 *         try:
 *             global __nvrtcGetErrorString
 */
    {
        #ifdef WITH_THREAD
        PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
        #endif
        /*try:*/ {

          /* "cuda/_cuda/cnvrtc.pyx":52
 *     # Load function
 *     with gil:
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetErrorString
 *             __nvrtcGetErrorString = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetErrorString')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":54
 *         try:
 *             global __nvrtcGetErrorString
 *             __nvrtcGetErrorString = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetErrorString')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_win32api); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 54, __pyx_L21_error)
              __Pyx_GOTREF(__pyx_t_7);
              __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 54, __pyx_L21_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_7 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {
                __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_5);
                if (likely(__pyx_t_7)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);
                  __Pyx_INCREF(__pyx_t_7);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_5, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetErrorString};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 54, __pyx_L21_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetErrorString};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 54, __pyx_L21_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_6 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 54, __pyx_L21_error)
                __Pyx_GOTREF(__pyx_t_6);
                if (__pyx_t_7) {
                  __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_7); __pyx_t_7 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_6, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetErrorString);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetErrorString);
                PyTuple_SET_ITEM(__pyx_t_6, 1+__pyx_t_9, __pyx_n_u_nvrtcGetErrorString);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_6, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 54, __pyx_L21_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              }
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 54, __pyx_L21_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetErrorString = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":52
 *     # Load function
 *     with gil:
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetErrorString
 *             __nvrtcGetErrorString = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetErrorString')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L26_try_end;
            __pyx_L21_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":55
 *             global __nvrtcGetErrorString
 *             __nvrtcGetErrorString = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetErrorString')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L22_exception_handled;
            }
            __pyx_L22_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L26_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":57
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcVersion
 *             __nvrtcVersion = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcVersion')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":59
 *         try:
 *             global __nvrtcVersion
 *             __nvrtcVersion = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcVersion')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_win32api); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 59, __pyx_L27_error)
              __Pyx_GOTREF(__pyx_t_5);
              __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 59, __pyx_L27_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_5 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {
                __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_6);
                if (likely(__pyx_t_5)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);
                  __Pyx_INCREF(__pyx_t_5);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_6, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcVersion};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 59, __pyx_L27_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcVersion};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 59, __pyx_L27_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_7 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 59, __pyx_L27_error)
                __Pyx_GOTREF(__pyx_t_7);
                if (__pyx_t_5) {
                  __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcVersion);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcVersion);
                PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_9, __pyx_n_u_nvrtcVersion);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_7, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 59, __pyx_L27_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              }
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 59, __pyx_L27_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcVersion = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":57
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcVersion
 *             __nvrtcVersion = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcVersion')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L32_try_end;
            __pyx_L27_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":60
 *             global __nvrtcVersion
 *             __nvrtcVersion = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcVersion')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L28_exception_handled;
            }
            __pyx_L28_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L32_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":62
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetNumSupportedArchs
 *             __nvrtcGetNumSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNumSupportedArchs')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":64
 *         try:
 *             global __nvrtcGetNumSupportedArchs
 *             __nvrtcGetNumSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNumSupportedArchs')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 64, __pyx_L33_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 64, __pyx_L33_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_6)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_6);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetNumSupportedArchs};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 64, __pyx_L33_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetNumSupportedArchs};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 64, __pyx_L33_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_5 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 64, __pyx_L33_error)
                __Pyx_GOTREF(__pyx_t_5);
                if (__pyx_t_6) {
                  __Pyx_GIVEREF(__pyx_t_6); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_6); __pyx_t_6 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetNumSupportedArchs);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetNumSupportedArchs);
                PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_9, __pyx_n_u_nvrtcGetNumSupportedArchs);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_5, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 64, __pyx_L33_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 64, __pyx_L33_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNumSupportedArchs = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":62
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetNumSupportedArchs
 *             __nvrtcGetNumSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNumSupportedArchs')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L38_try_end;
            __pyx_L33_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":65
 *             global __nvrtcGetNumSupportedArchs
 *             __nvrtcGetNumSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNumSupportedArchs')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L34_exception_handled;
            }
            __pyx_L34_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L38_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":67
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetSupportedArchs
 *             __nvrtcGetSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetSupportedArchs')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":69
 *         try:
 *             global __nvrtcGetSupportedArchs
 *             __nvrtcGetSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetSupportedArchs')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_win32api); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 69, __pyx_L39_error)
              __Pyx_GOTREF(__pyx_t_7);
              __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 69, __pyx_L39_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_7 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {
                __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_5);
                if (likely(__pyx_t_7)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);
                  __Pyx_INCREF(__pyx_t_7);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_5, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetSupportedArchs};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 69, __pyx_L39_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetSupportedArchs};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 69, __pyx_L39_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_6 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 69, __pyx_L39_error)
                __Pyx_GOTREF(__pyx_t_6);
                if (__pyx_t_7) {
                  __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_7); __pyx_t_7 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_6, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetSupportedArchs);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetSupportedArchs);
                PyTuple_SET_ITEM(__pyx_t_6, 1+__pyx_t_9, __pyx_n_u_nvrtcGetSupportedArchs);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_6, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 69, __pyx_L39_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              }
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 69, __pyx_L39_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetSupportedArchs = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":67
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetSupportedArchs
 *             __nvrtcGetSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetSupportedArchs')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L44_try_end;
            __pyx_L39_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":70
 *             global __nvrtcGetSupportedArchs
 *             __nvrtcGetSupportedArchs = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetSupportedArchs')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L40_exception_handled;
            }
            __pyx_L40_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L44_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":72
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcCreateProgram
 *             __nvrtcCreateProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCreateProgram')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":74
 *         try:
 *             global __nvrtcCreateProgram
 *             __nvrtcCreateProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCreateProgram')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_win32api); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 74, __pyx_L45_error)
              __Pyx_GOTREF(__pyx_t_5);
              __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 74, __pyx_L45_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_5 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {
                __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_6);
                if (likely(__pyx_t_5)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);
                  __Pyx_INCREF(__pyx_t_5);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_6, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcCreateProgram};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 74, __pyx_L45_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcCreateProgram};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 74, __pyx_L45_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_7 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 74, __pyx_L45_error)
                __Pyx_GOTREF(__pyx_t_7);
                if (__pyx_t_5) {
                  __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcCreateProgram);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcCreateProgram);
                PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_9, __pyx_n_u_nvrtcCreateProgram);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_7, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 74, __pyx_L45_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              }
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 74, __pyx_L45_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCreateProgram = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":72
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcCreateProgram
 *             __nvrtcCreateProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCreateProgram')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L50_try_end;
            __pyx_L45_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":75
 *             global __nvrtcCreateProgram
 *             __nvrtcCreateProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCreateProgram')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L46_exception_handled;
            }
            __pyx_L46_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L50_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":77
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcDestroyProgram
 *             __nvrtcDestroyProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcDestroyProgram')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":79
 *         try:
 *             global __nvrtcDestroyProgram
 *             __nvrtcDestroyProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcDestroyProgram')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 79, __pyx_L51_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 79, __pyx_L51_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_6)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_6);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcDestroyProgram};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 79, __pyx_L51_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcDestroyProgram};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 79, __pyx_L51_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_5 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 79, __pyx_L51_error)
                __Pyx_GOTREF(__pyx_t_5);
                if (__pyx_t_6) {
                  __Pyx_GIVEREF(__pyx_t_6); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_6); __pyx_t_6 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcDestroyProgram);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcDestroyProgram);
                PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_9, __pyx_n_u_nvrtcDestroyProgram);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_5, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 79, __pyx_L51_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 79, __pyx_L51_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcDestroyProgram = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":77
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcDestroyProgram
 *             __nvrtcDestroyProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcDestroyProgram')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L56_try_end;
            __pyx_L51_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":80
 *             global __nvrtcDestroyProgram
 *             __nvrtcDestroyProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcDestroyProgram')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L52_exception_handled;
            }
            __pyx_L52_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L56_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":82
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcCompileProgram
 *             __nvrtcCompileProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCompileProgram')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":84
 *         try:
 *             global __nvrtcCompileProgram
 *             __nvrtcCompileProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCompileProgram')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_win32api); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 84, __pyx_L57_error)
              __Pyx_GOTREF(__pyx_t_7);
              __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 84, __pyx_L57_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_7 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {
                __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_5);
                if (likely(__pyx_t_7)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);
                  __Pyx_INCREF(__pyx_t_7);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_5, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcCompileProgram};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 84, __pyx_L57_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcCompileProgram};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 84, __pyx_L57_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_6 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 84, __pyx_L57_error)
                __Pyx_GOTREF(__pyx_t_6);
                if (__pyx_t_7) {
                  __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_7); __pyx_t_7 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_6, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcCompileProgram);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcCompileProgram);
                PyTuple_SET_ITEM(__pyx_t_6, 1+__pyx_t_9, __pyx_n_u_nvrtcCompileProgram);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_6, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 84, __pyx_L57_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              }
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 84, __pyx_L57_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCompileProgram = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":82
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcCompileProgram
 *             __nvrtcCompileProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCompileProgram')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L62_try_end;
            __pyx_L57_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":85
 *             global __nvrtcCompileProgram
 *             __nvrtcCompileProgram = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcCompileProgram')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L58_exception_handled;
            }
            __pyx_L58_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L62_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":87
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetPTXSize
 *             __nvrtcGetPTXSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTXSize')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":89
 *         try:
 *             global __nvrtcGetPTXSize
 *             __nvrtcGetPTXSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTXSize')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_win32api); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 89, __pyx_L63_error)
              __Pyx_GOTREF(__pyx_t_5);
              __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 89, __pyx_L63_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_5 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {
                __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_6);
                if (likely(__pyx_t_5)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);
                  __Pyx_INCREF(__pyx_t_5);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_6, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetPTXSize};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 89, __pyx_L63_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetPTXSize};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 89, __pyx_L63_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_7 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 89, __pyx_L63_error)
                __Pyx_GOTREF(__pyx_t_7);
                if (__pyx_t_5) {
                  __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetPTXSize);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetPTXSize);
                PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_9, __pyx_n_u_nvrtcGetPTXSize);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_7, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 89, __pyx_L63_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              }
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 89, __pyx_L63_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTXSize = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":87
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetPTXSize
 *             __nvrtcGetPTXSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTXSize')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L68_try_end;
            __pyx_L63_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":90
 *             global __nvrtcGetPTXSize
 *             __nvrtcGetPTXSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTXSize')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L64_exception_handled;
            }
            __pyx_L64_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L68_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":92
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetPTX
 *             __nvrtcGetPTX = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTX')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":94
 *         try:
 *             global __nvrtcGetPTX
 *             __nvrtcGetPTX = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTX')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 94, __pyx_L69_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 94, __pyx_L69_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_6)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_6);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetPTX};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 94, __pyx_L69_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetPTX};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 94, __pyx_L69_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_5 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 94, __pyx_L69_error)
                __Pyx_GOTREF(__pyx_t_5);
                if (__pyx_t_6) {
                  __Pyx_GIVEREF(__pyx_t_6); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_6); __pyx_t_6 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetPTX);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetPTX);
                PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_9, __pyx_n_u_nvrtcGetPTX);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_5, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 94, __pyx_L69_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 94, __pyx_L69_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTX = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":92
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetPTX
 *             __nvrtcGetPTX = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTX')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L74_try_end;
            __pyx_L69_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":95
 *             global __nvrtcGetPTX
 *             __nvrtcGetPTX = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetPTX')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L70_exception_handled;
            }
            __pyx_L70_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L74_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":97
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetCUBINSize
 *             __nvrtcGetCUBINSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBINSize')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":99
 *         try:
 *             global __nvrtcGetCUBINSize
 *             __nvrtcGetCUBINSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBINSize')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_win32api); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 99, __pyx_L75_error)
              __Pyx_GOTREF(__pyx_t_7);
              __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 99, __pyx_L75_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_7 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {
                __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_5);
                if (likely(__pyx_t_7)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);
                  __Pyx_INCREF(__pyx_t_7);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_5, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetCUBINSize};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 99, __pyx_L75_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetCUBINSize};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 99, __pyx_L75_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_6 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 99, __pyx_L75_error)
                __Pyx_GOTREF(__pyx_t_6);
                if (__pyx_t_7) {
                  __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_7); __pyx_t_7 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_6, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetCUBINSize);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetCUBINSize);
                PyTuple_SET_ITEM(__pyx_t_6, 1+__pyx_t_9, __pyx_n_u_nvrtcGetCUBINSize);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_6, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 99, __pyx_L75_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              }
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 99, __pyx_L75_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBINSize = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":97
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetCUBINSize
 *             __nvrtcGetCUBINSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBINSize')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L80_try_end;
            __pyx_L75_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":100
 *             global __nvrtcGetCUBINSize
 *             __nvrtcGetCUBINSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBINSize')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L76_exception_handled;
            }
            __pyx_L76_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L80_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":102
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetCUBIN
 *             __nvrtcGetCUBIN = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBIN')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":104
 *         try:
 *             global __nvrtcGetCUBIN
 *             __nvrtcGetCUBIN = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBIN')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_win32api); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 104, __pyx_L81_error)
              __Pyx_GOTREF(__pyx_t_5);
              __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 104, __pyx_L81_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_5 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {
                __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_6);
                if (likely(__pyx_t_5)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);
                  __Pyx_INCREF(__pyx_t_5);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_6, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetCUBIN};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 104, __pyx_L81_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetCUBIN};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 104, __pyx_L81_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_7 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 104, __pyx_L81_error)
                __Pyx_GOTREF(__pyx_t_7);
                if (__pyx_t_5) {
                  __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetCUBIN);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetCUBIN);
                PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_9, __pyx_n_u_nvrtcGetCUBIN);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_7, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 104, __pyx_L81_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              }
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 104, __pyx_L81_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBIN = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":102
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetCUBIN
 *             __nvrtcGetCUBIN = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBIN')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L86_try_end;
            __pyx_L81_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":105
 *             global __nvrtcGetCUBIN
 *             __nvrtcGetCUBIN = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetCUBIN')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L82_exception_handled;
            }
            __pyx_L82_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L86_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":107
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetNVVMSize
 *             __nvrtcGetNVVMSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVMSize')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":109
 *         try:
 *             global __nvrtcGetNVVMSize
 *             __nvrtcGetNVVMSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVMSize')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 109, __pyx_L87_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 109, __pyx_L87_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_6)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_6);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetNVVMSize};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 109, __pyx_L87_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetNVVMSize};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 109, __pyx_L87_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_5 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 109, __pyx_L87_error)
                __Pyx_GOTREF(__pyx_t_5);
                if (__pyx_t_6) {
                  __Pyx_GIVEREF(__pyx_t_6); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_6); __pyx_t_6 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetNVVMSize);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetNVVMSize);
                PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_9, __pyx_n_u_nvrtcGetNVVMSize);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_5, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 109, __pyx_L87_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 109, __pyx_L87_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVMSize = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":107
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetNVVMSize
 *             __nvrtcGetNVVMSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVMSize')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L92_try_end;
            __pyx_L87_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":110
 *             global __nvrtcGetNVVMSize
 *             __nvrtcGetNVVMSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVMSize')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L88_exception_handled;
            }
            __pyx_L88_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L92_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":112
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetNVVM
 *             __nvrtcGetNVVM = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVM')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":114
 *         try:
 *             global __nvrtcGetNVVM
 *             __nvrtcGetNVVM = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVM')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_win32api); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 114, __pyx_L93_error)
              __Pyx_GOTREF(__pyx_t_7);
              __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 114, __pyx_L93_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_7 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {
                __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_5);
                if (likely(__pyx_t_7)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);
                  __Pyx_INCREF(__pyx_t_7);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_5, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetNVVM};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 114, __pyx_L93_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetNVVM};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 114, __pyx_L93_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_6 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 114, __pyx_L93_error)
                __Pyx_GOTREF(__pyx_t_6);
                if (__pyx_t_7) {
                  __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_7); __pyx_t_7 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_6, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetNVVM);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetNVVM);
                PyTuple_SET_ITEM(__pyx_t_6, 1+__pyx_t_9, __pyx_n_u_nvrtcGetNVVM);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_6, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 114, __pyx_L93_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              }
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 114, __pyx_L93_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVM = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":112
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetNVVM
 *             __nvrtcGetNVVM = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVM')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L98_try_end;
            __pyx_L93_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":115
 *             global __nvrtcGetNVVM
 *             __nvrtcGetNVVM = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetNVVM')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L94_exception_handled;
            }
            __pyx_L94_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L98_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":117
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetLTOIRSize
 *             __nvrtcGetLTOIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIRSize')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":119
 *         try:
 *             global __nvrtcGetLTOIRSize
 *             __nvrtcGetLTOIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIRSize')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_win32api); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 119, __pyx_L99_error)
              __Pyx_GOTREF(__pyx_t_5);
              __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 119, __pyx_L99_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_5 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {
                __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_6);
                if (likely(__pyx_t_5)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);
                  __Pyx_INCREF(__pyx_t_5);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_6, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetLTOIRSize};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 119, __pyx_L99_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetLTOIRSize};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 119, __pyx_L99_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_7 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 119, __pyx_L99_error)
                __Pyx_GOTREF(__pyx_t_7);
                if (__pyx_t_5) {
                  __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetLTOIRSize);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetLTOIRSize);
                PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_9, __pyx_n_u_nvrtcGetLTOIRSize);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_7, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 119, __pyx_L99_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              }
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 119, __pyx_L99_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIRSize = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":117
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetLTOIRSize
 *             __nvrtcGetLTOIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIRSize')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L104_try_end;
            __pyx_L99_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":120
 *             global __nvrtcGetLTOIRSize
 *             __nvrtcGetLTOIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIRSize')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L100_exception_handled;
            }
            __pyx_L100_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L104_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":122
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetLTOIR
 *             __nvrtcGetLTOIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIR')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":124
 *         try:
 *             global __nvrtcGetLTOIR
 *             __nvrtcGetLTOIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIR')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 124, __pyx_L105_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 124, __pyx_L105_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_6)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_6);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetLTOIR};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 124, __pyx_L105_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetLTOIR};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 124, __pyx_L105_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_5 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 124, __pyx_L105_error)
                __Pyx_GOTREF(__pyx_t_5);
                if (__pyx_t_6) {
                  __Pyx_GIVEREF(__pyx_t_6); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_6); __pyx_t_6 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetLTOIR);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetLTOIR);
                PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_9, __pyx_n_u_nvrtcGetLTOIR);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_5, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 124, __pyx_L105_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 124, __pyx_L105_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIR = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":122
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetLTOIR
 *             __nvrtcGetLTOIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIR')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L110_try_end;
            __pyx_L105_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":125
 *             global __nvrtcGetLTOIR
 *             __nvrtcGetLTOIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLTOIR')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L106_exception_handled;
            }
            __pyx_L106_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L110_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":127
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetOptiXIRSize
 *             __nvrtcGetOptiXIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIRSize')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":129
 *         try:
 *             global __nvrtcGetOptiXIRSize
 *             __nvrtcGetOptiXIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIRSize')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_win32api); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 129, __pyx_L111_error)
              __Pyx_GOTREF(__pyx_t_7);
              __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 129, __pyx_L111_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_7 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {
                __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_5);
                if (likely(__pyx_t_7)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);
                  __Pyx_INCREF(__pyx_t_7);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_5, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetOptiXIRSize};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 129, __pyx_L111_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetOptiXIRSize};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 129, __pyx_L111_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_6 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 129, __pyx_L111_error)
                __Pyx_GOTREF(__pyx_t_6);
                if (__pyx_t_7) {
                  __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_7); __pyx_t_7 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_6, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetOptiXIRSize);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetOptiXIRSize);
                PyTuple_SET_ITEM(__pyx_t_6, 1+__pyx_t_9, __pyx_n_u_nvrtcGetOptiXIRSize);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_6, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 129, __pyx_L111_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              }
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 129, __pyx_L111_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIRSize = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":127
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetOptiXIRSize
 *             __nvrtcGetOptiXIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIRSize')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L116_try_end;
            __pyx_L111_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":130
 *             global __nvrtcGetOptiXIRSize
 *             __nvrtcGetOptiXIRSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIRSize')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L112_exception_handled;
            }
            __pyx_L112_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L116_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":132
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetOptiXIR
 *             __nvrtcGetOptiXIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIR')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":134
 *         try:
 *             global __nvrtcGetOptiXIR
 *             __nvrtcGetOptiXIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIR')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_win32api); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 134, __pyx_L117_error)
              __Pyx_GOTREF(__pyx_t_5);
              __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 134, __pyx_L117_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_5 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {
                __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_6);
                if (likely(__pyx_t_5)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);
                  __Pyx_INCREF(__pyx_t_5);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_6, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetOptiXIR};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 134, __pyx_L117_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcGetOptiXIR};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 134, __pyx_L117_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_7 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 134, __pyx_L117_error)
                __Pyx_GOTREF(__pyx_t_7);
                if (__pyx_t_5) {
                  __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetOptiXIR);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetOptiXIR);
                PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_9, __pyx_n_u_nvrtcGetOptiXIR);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_7, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 134, __pyx_L117_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              }
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 134, __pyx_L117_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIR = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":132
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetOptiXIR
 *             __nvrtcGetOptiXIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIR')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L122_try_end;
            __pyx_L117_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":135
 *             global __nvrtcGetOptiXIR
 *             __nvrtcGetOptiXIR = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetOptiXIR')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L118_exception_handled;
            }
            __pyx_L118_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L122_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":137
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetProgramLogSize
 *             __nvrtcGetProgramLogSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLogSize')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":139
 *         try:
 *             global __nvrtcGetProgramLogSize
 *             __nvrtcGetProgramLogSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLogSize')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 139, __pyx_L123_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 139, __pyx_L123_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_6)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_6);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetProgramLogSize};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 139, __pyx_L123_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetProgramLogSize};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 139, __pyx_L123_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_5 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 139, __pyx_L123_error)
                __Pyx_GOTREF(__pyx_t_5);
                if (__pyx_t_6) {
                  __Pyx_GIVEREF(__pyx_t_6); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_6); __pyx_t_6 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetProgramLogSize);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetProgramLogSize);
                PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_9, __pyx_n_u_nvrtcGetProgramLogSize);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_5, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 139, __pyx_L123_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 139, __pyx_L123_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLogSize = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":137
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetProgramLogSize
 *             __nvrtcGetProgramLogSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLogSize')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L128_try_end;
            __pyx_L123_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":140
 *             global __nvrtcGetProgramLogSize
 *             __nvrtcGetProgramLogSize = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLogSize')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L124_exception_handled;
            }
            __pyx_L124_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L128_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":142
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetProgramLog
 *             __nvrtcGetProgramLog = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLog')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":144
 *         try:
 *             global __nvrtcGetProgramLog
 *             __nvrtcGetProgramLog = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLog')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_7, __pyx_n_s_win32api); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 144, __pyx_L129_error)
              __Pyx_GOTREF(__pyx_t_7);
              __pyx_t_5 = __Pyx_PyObject_GetAttrStr(__pyx_t_7, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 144, __pyx_L129_error)
              __Pyx_GOTREF(__pyx_t_5);
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_7 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_5))) {
                __pyx_t_7 = PyMethod_GET_SELF(__pyx_t_5);
                if (likely(__pyx_t_7)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_5);
                  __Pyx_INCREF(__pyx_t_7);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_5, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetProgramLog};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 144, __pyx_L129_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_5)) {
                PyObject *__pyx_temp[3] = {__pyx_t_7, __pyx_v_handle, __pyx_n_u_nvrtcGetProgramLog};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_5, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 144, __pyx_L129_error)
                __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_6 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 144, __pyx_L129_error)
                __Pyx_GOTREF(__pyx_t_6);
                if (__pyx_t_7) {
                  __Pyx_GIVEREF(__pyx_t_7); PyTuple_SET_ITEM(__pyx_t_6, 0, __pyx_t_7); __pyx_t_7 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_6, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetProgramLog);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetProgramLog);
                PyTuple_SET_ITEM(__pyx_t_6, 1+__pyx_t_9, __pyx_n_u_nvrtcGetProgramLog);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_5, __pyx_t_6, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 144, __pyx_L129_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              }
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 144, __pyx_L129_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLog = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":142
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetProgramLog
 *             __nvrtcGetProgramLog = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLog')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L134_try_end;
            __pyx_L129_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":145
 *             global __nvrtcGetProgramLog
 *             __nvrtcGetProgramLog = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetProgramLog')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L130_exception_handled;
            }
            __pyx_L130_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L134_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":147
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcAddNameExpression
 *             __nvrtcAddNameExpression = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcAddNameExpression')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_2, &__pyx_t_3, &__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_4);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":149
 *         try:
 *             global __nvrtcAddNameExpression
 *             __nvrtcAddNameExpression = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcAddNameExpression')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_5, __pyx_n_s_win32api); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 149, __pyx_L135_error)
              __Pyx_GOTREF(__pyx_t_5);
              __pyx_t_6 = __Pyx_PyObject_GetAttrStr(__pyx_t_5, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 149, __pyx_L135_error)
              __Pyx_GOTREF(__pyx_t_6);
              __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              __pyx_t_5 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_6))) {
                __pyx_t_5 = PyMethod_GET_SELF(__pyx_t_6);
                if (likely(__pyx_t_5)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_6);
                  __Pyx_INCREF(__pyx_t_5);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_6, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcAddNameExpression};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 149, __pyx_L135_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_6)) {
                PyObject *__pyx_temp[3] = {__pyx_t_5, __pyx_v_handle, __pyx_n_u_nvrtcAddNameExpression};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_6, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 149, __pyx_L135_error)
                __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_7 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 149, __pyx_L135_error)
                __Pyx_GOTREF(__pyx_t_7);
                if (__pyx_t_5) {
                  __Pyx_GIVEREF(__pyx_t_5); PyTuple_SET_ITEM(__pyx_t_7, 0, __pyx_t_5); __pyx_t_5 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_7, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcAddNameExpression);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcAddNameExpression);
                PyTuple_SET_ITEM(__pyx_t_7, 1+__pyx_t_9, __pyx_n_u_nvrtcAddNameExpression);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_6, __pyx_t_7, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 149, __pyx_L135_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              }
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 149, __pyx_L135_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcAddNameExpression = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":147
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcAddNameExpression
 *             __nvrtcAddNameExpression = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcAddNameExpression')
 */
            }
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            goto __pyx_L140_try_end;
            __pyx_L135_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":150
 *             global __nvrtcAddNameExpression
 *             __nvrtcAddNameExpression = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcAddNameExpression')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 *         try:
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L136_exception_handled;
            }
            __pyx_L136_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_ExceptionReset(__pyx_t_2, __pyx_t_3, __pyx_t_4);
            __pyx_L140_try_end:;
          }

          /* "cuda/_cuda/cnvrtc.pyx":152
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetLoweredName
 *             __nvrtcGetLoweredName = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLoweredName')
 */
          {
            __Pyx_PyThreadState_declare
            __Pyx_PyThreadState_assign
            __Pyx_ExceptionSave(&__pyx_t_4, &__pyx_t_3, &__pyx_t_2);
            __Pyx_XGOTREF(__pyx_t_4);
            __Pyx_XGOTREF(__pyx_t_3);
            __Pyx_XGOTREF(__pyx_t_2);
            /*try:*/ {

              /* "cuda/_cuda/cnvrtc.pyx":154
 *         try:
 *             global __nvrtcGetLoweredName
 *             __nvrtcGetLoweredName = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLoweredName')             # <<<<<<<<<<<<<<
 *         except:
 *             pass
 */
              __Pyx_GetModuleGlobalName(__pyx_t_6, __pyx_n_s_win32api); if (unlikely(!__pyx_t_6)) __PYX_ERR(0, 154, __pyx_L141_error)
              __Pyx_GOTREF(__pyx_t_6);
              __pyx_t_7 = __Pyx_PyObject_GetAttrStr(__pyx_t_6, __pyx_n_s_GetProcAddress); if (unlikely(!__pyx_t_7)) __PYX_ERR(0, 154, __pyx_L141_error)
              __Pyx_GOTREF(__pyx_t_7);
              __Pyx_DECREF(__pyx_t_6); __pyx_t_6 = 0;
              __pyx_t_6 = NULL;
              __pyx_t_9 = 0;
              if (CYTHON_UNPACK_METHODS && unlikely(PyMethod_Check(__pyx_t_7))) {
                __pyx_t_6 = PyMethod_GET_SELF(__pyx_t_7);
                if (likely(__pyx_t_6)) {
                  PyObject* function = PyMethod_GET_FUNCTION(__pyx_t_7);
                  __Pyx_INCREF(__pyx_t_6);
                  __Pyx_INCREF(function);
                  __Pyx_DECREF_SET(__pyx_t_7, function);
                  __pyx_t_9 = 1;
                }
              }
              #if CYTHON_FAST_PYCALL
              if (PyFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetLoweredName};
                __pyx_t_10 = __Pyx_PyFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 154, __pyx_L141_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              #if CYTHON_FAST_PYCCALL
              if (__Pyx_PyFastCFunction_Check(__pyx_t_7)) {
                PyObject *__pyx_temp[3] = {__pyx_t_6, __pyx_v_handle, __pyx_n_u_nvrtcGetLoweredName};
                __pyx_t_10 = __Pyx_PyCFunction_FastCall(__pyx_t_7, __pyx_temp+1-__pyx_t_9, 2+__pyx_t_9); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 154, __pyx_L141_error)
                __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
                __Pyx_GOTREF(__pyx_t_10);
              } else
              #endif
              {
                __pyx_t_5 = PyTuple_New(2+__pyx_t_9); if (unlikely(!__pyx_t_5)) __PYX_ERR(0, 154, __pyx_L141_error)
                __Pyx_GOTREF(__pyx_t_5);
                if (__pyx_t_6) {
                  __Pyx_GIVEREF(__pyx_t_6); PyTuple_SET_ITEM(__pyx_t_5, 0, __pyx_t_6); __pyx_t_6 = NULL;
                }
                __Pyx_INCREF(__pyx_v_handle);
                __Pyx_GIVEREF(__pyx_v_handle);
                PyTuple_SET_ITEM(__pyx_t_5, 0+__pyx_t_9, __pyx_v_handle);
                __Pyx_INCREF(__pyx_n_u_nvrtcGetLoweredName);
                __Pyx_GIVEREF(__pyx_n_u_nvrtcGetLoweredName);
                PyTuple_SET_ITEM(__pyx_t_5, 1+__pyx_t_9, __pyx_n_u_nvrtcGetLoweredName);
                __pyx_t_10 = __Pyx_PyObject_Call(__pyx_t_7, __pyx_t_5, NULL); if (unlikely(!__pyx_t_10)) __PYX_ERR(0, 154, __pyx_L141_error)
                __Pyx_GOTREF(__pyx_t_10);
                __Pyx_DECREF(__pyx_t_5); __pyx_t_5 = 0;
              }
              __Pyx_DECREF(__pyx_t_7); __pyx_t_7 = 0;
              __pyx_t_11 = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(__pyx_t_10); if (unlikely((__pyx_t_11 == (unsigned PY_LONG_LONG)-1) && PyErr_Occurred())) __PYX_ERR(0, 154, __pyx_L141_error)
              __Pyx_DECREF(__pyx_t_10); __pyx_t_10 = 0;
              __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLoweredName = ((void *)((unsigned PY_LONG_LONG)__pyx_t_11));

              /* "cuda/_cuda/cnvrtc.pyx":152
 *         except:
 *             pass
 *         try:             # <<<<<<<<<<<<<<
 *             global __nvrtcGetLoweredName
 *             __nvrtcGetLoweredName = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLoweredName')
 */
            }
            __Pyx_XDECREF(__pyx_t_4); __pyx_t_4 = 0;
            __Pyx_XDECREF(__pyx_t_3); __pyx_t_3 = 0;
            __Pyx_XDECREF(__pyx_t_2); __pyx_t_2 = 0;
            goto __pyx_L146_try_end;
            __pyx_L141_error:;
            __Pyx_XDECREF(__pyx_t_10); __pyx_t_10 = 0;
            __Pyx_XDECREF(__pyx_t_5); __pyx_t_5 = 0;
            __Pyx_XDECREF(__pyx_t_6); __pyx_t_6 = 0;
            __Pyx_XDECREF(__pyx_t_7); __pyx_t_7 = 0;
            __Pyx_XDECREF(__pyx_t_8); __pyx_t_8 = 0;

            /* "cuda/_cuda/cnvrtc.pyx":155
 *             global __nvrtcGetLoweredName
 *             __nvrtcGetLoweredName = <void*><unsigned long long>win32api.GetProcAddress(handle, 'nvrtcGetLoweredName')
 *         except:             # <<<<<<<<<<<<<<
 *             pass
 * 
 */
            /*except:*/ {
              __Pyx_ErrRestore(0,0,0);
              goto __pyx_L142_exception_handled;
            }
            __pyx_L142_exception_handled:;
            __Pyx_XGIVEREF(__pyx_t_4);
            __Pyx_XGIVEREF(__pyx_t_3);
            __Pyx_XGIVEREF(__pyx_t_2);
            __Pyx_ExceptionReset(__pyx_t_4, __pyx_t_3, __pyx_t_2);
            __pyx_L146_try_end:;
          }
        }

        /* "cuda/_cuda/cnvrtc.pyx":51
 * 
 *     # Load function
 *     with gil:             # <<<<<<<<<<<<<<
 *         try:
 *             global __nvrtcGetErrorString
 */
        /*finally:*/ {
          /*normal exit:*/{
            #ifdef WITH_THREAD
            __Pyx_PyGILState_Release(__pyx_gilstate_save);
            #endif
            goto __pyx_L20;
          }
          __pyx_L20:;
        }
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":36
 * 
 * cdef int cuPythonInit() except -1 nogil:
 *     global __cuPythonInit             # <<<<<<<<<<<<<<
 *     if __cuPythonInit:
 *         return 0
 */
  /*finally:*/ {
    /*normal exit:*/{
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L5;
    }
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
    __pyx_L5:;
  }

  /* "cuda/_cuda/cnvrtc.pyx":35
 * cdef void *__nvrtcGetLoweredName = NULL
 * 
 * cdef int cuPythonInit() except -1 nogil:             # <<<<<<<<<<<<<<
 *     global __cuPythonInit
 *     if __cuPythonInit:
 */

  /* function exit code */
  __pyx_r = 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_5);
  __Pyx_XDECREF(__pyx_t_6);
  __Pyx_XDECREF(__pyx_t_7);
  __Pyx_XDECREF(__pyx_t_8);
  __Pyx_XDECREF(__pyx_t_10);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc.cuPythonInit", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = -1;
  __pyx_L0:;
  __Pyx_XDECREF(__pyx_v_handle);
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":158
 *             pass
 * 
 * cdef const char* _nvrtcGetErrorString(nvrtcResult result) except ?NULL nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetErrorString
 *     cuPythonInit()
 */

static char const *__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetErrorString(nvrtcResult __pyx_v_result) {
  char const *__pyx_v_err;
  char const *__pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetErrorString", 1);
  __Pyx_TraceCall("_nvrtcGetErrorString", __pyx_f[0], 158, 1, __PYX_ERR(0, 158, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":159
 * 
 * cdef const char* _nvrtcGetErrorString(nvrtcResult result) except ?NULL nogil:
 *     global __nvrtcGetErrorString             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetErrorString == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":160
 * cdef const char* _nvrtcGetErrorString(nvrtcResult result) except ?NULL nogil:
 *     global __nvrtcGetErrorString
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetErrorString == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 160, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":161
 *     global __nvrtcGetErrorString
 *     cuPythonInit()
 *     if __nvrtcGetErrorString == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetErrorString == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":162
 *     cuPythonInit()
 *     if __nvrtcGetErrorString == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')
 *     err = (<const char* (*)(nvrtcResult) nogil> __nvrtcGetErrorString)(result)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":163
 *     if __nvrtcGetErrorString == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')             # <<<<<<<<<<<<<<
 *     err = (<const char* (*)(nvrtcResult) nogil> __nvrtcGetErrorString)(result)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__2, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 163, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 163, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":162
 *     cuPythonInit()
 *     if __nvrtcGetErrorString == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')
 *     err = (<const char* (*)(nvrtcResult) nogil> __nvrtcGetErrorString)(result)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":161
 *     global __nvrtcGetErrorString
 *     cuPythonInit()
 *     if __nvrtcGetErrorString == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":164
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')
 *     err = (<const char* (*)(nvrtcResult) nogil> __nvrtcGetErrorString)(result)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((char const *(*)(nvrtcResult))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetErrorString)(__pyx_v_result);

    /* "cuda/_cuda/cnvrtc.pyx":165
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')
 *     err = (<const char* (*)(nvrtcResult) nogil> __nvrtcGetErrorString)(result)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcVersion(int* major, int* minor) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":159
 * 
 * cdef const char* _nvrtcGetErrorString(nvrtcResult result) except ?NULL nogil:
 *     global __nvrtcGetErrorString             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetErrorString == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":158
 *             pass
 * 
 * cdef const char* _nvrtcGetErrorString(nvrtcResult result) except ?NULL nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetErrorString
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetErrorString", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NULL;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":167
 *     return err
 * 
 * cdef nvrtcResult _nvrtcVersion(int* major, int* minor) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcVersion
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcVersion(int *__pyx_v_major, int *__pyx_v_minor) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcVersion", 1);
  __Pyx_TraceCall("_nvrtcVersion", __pyx_f[0], 167, 1, __PYX_ERR(0, 167, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":168
 * 
 * cdef nvrtcResult _nvrtcVersion(int* major, int* minor) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcVersion             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcVersion == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":169
 * cdef nvrtcResult _nvrtcVersion(int* major, int* minor) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcVersion
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcVersion == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 169, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":170
 *     global __nvrtcVersion
 *     cuPythonInit()
 *     if __nvrtcVersion == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcVersion" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcVersion == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":171
 *     cuPythonInit()
 *     if __nvrtcVersion == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcVersion" not found')
 *     err = (<nvrtcResult (*)(int*, int*) nogil> __nvrtcVersion)(major, minor)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":172
 *     if __nvrtcVersion == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcVersion" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(int*, int*) nogil> __nvrtcVersion)(major, minor)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__3, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 172, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 172, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":171
 *     cuPythonInit()
 *     if __nvrtcVersion == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcVersion" not found')
 *     err = (<nvrtcResult (*)(int*, int*) nogil> __nvrtcVersion)(major, minor)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":170
 *     global __nvrtcVersion
 *     cuPythonInit()
 *     if __nvrtcVersion == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcVersion" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":173
 *         with gil:
 *             raise RuntimeError('Function "nvrtcVersion" not found')
 *     err = (<nvrtcResult (*)(int*, int*) nogil> __nvrtcVersion)(major, minor)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(int *, int *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcVersion)(__pyx_v_major, __pyx_v_minor);

    /* "cuda/_cuda/cnvrtc.pyx":174
 *             raise RuntimeError('Function "nvrtcVersion" not found')
 *     err = (<nvrtcResult (*)(int*, int*) nogil> __nvrtcVersion)(major, minor)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetNumSupportedArchs(int* numArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":168
 * 
 * cdef nvrtcResult _nvrtcVersion(int* major, int* minor) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcVersion             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcVersion == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":167
 *     return err
 * 
 * cdef nvrtcResult _nvrtcVersion(int* major, int* minor) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcVersion
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcVersion", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":176
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetNumSupportedArchs(int* numArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetNumSupportedArchs
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetNumSupportedArchs(int *__pyx_v_numArchs) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetNumSupportedArchs", 1);
  __Pyx_TraceCall("_nvrtcGetNumSupportedArchs", __pyx_f[0], 176, 1, __PYX_ERR(0, 176, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":177
 * 
 * cdef nvrtcResult _nvrtcGetNumSupportedArchs(int* numArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNumSupportedArchs             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetNumSupportedArchs == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":178
 * cdef nvrtcResult _nvrtcGetNumSupportedArchs(int* numArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNumSupportedArchs
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetNumSupportedArchs == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 178, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":179
 *     global __nvrtcGetNumSupportedArchs
 *     cuPythonInit()
 *     if __nvrtcGetNumSupportedArchs == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNumSupportedArchs == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":180
 *     cuPythonInit()
 *     if __nvrtcGetNumSupportedArchs == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetNumSupportedArchs)(numArchs)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":181
 *     if __nvrtcGetNumSupportedArchs == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetNumSupportedArchs)(numArchs)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__4, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 181, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 181, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":180
 *     cuPythonInit()
 *     if __nvrtcGetNumSupportedArchs == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetNumSupportedArchs)(numArchs)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":179
 *     global __nvrtcGetNumSupportedArchs
 *     cuPythonInit()
 *     if __nvrtcGetNumSupportedArchs == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":182
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetNumSupportedArchs)(numArchs)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(int *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNumSupportedArchs)(__pyx_v_numArchs);

    /* "cuda/_cuda/cnvrtc.pyx":183
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetNumSupportedArchs)(numArchs)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetSupportedArchs(int* supportedArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":177
 * 
 * cdef nvrtcResult _nvrtcGetNumSupportedArchs(int* numArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNumSupportedArchs             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetNumSupportedArchs == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":176
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetNumSupportedArchs(int* numArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetNumSupportedArchs
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetNumSupportedArchs", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":185
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetSupportedArchs(int* supportedArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetSupportedArchs
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetSupportedArchs(int *__pyx_v_supportedArchs) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetSupportedArchs", 1);
  __Pyx_TraceCall("_nvrtcGetSupportedArchs", __pyx_f[0], 185, 1, __PYX_ERR(0, 185, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":186
 * 
 * cdef nvrtcResult _nvrtcGetSupportedArchs(int* supportedArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetSupportedArchs             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetSupportedArchs == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":187
 * cdef nvrtcResult _nvrtcGetSupportedArchs(int* supportedArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetSupportedArchs
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetSupportedArchs == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 187, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":188
 *     global __nvrtcGetSupportedArchs
 *     cuPythonInit()
 *     if __nvrtcGetSupportedArchs == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetSupportedArchs == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":189
 *     cuPythonInit()
 *     if __nvrtcGetSupportedArchs == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetSupportedArchs)(supportedArchs)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":190
 *     if __nvrtcGetSupportedArchs == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetSupportedArchs)(supportedArchs)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__5, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 190, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 190, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":189
 *     cuPythonInit()
 *     if __nvrtcGetSupportedArchs == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetSupportedArchs)(supportedArchs)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":188
 *     global __nvrtcGetSupportedArchs
 *     cuPythonInit()
 *     if __nvrtcGetSupportedArchs == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":191
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetSupportedArchs)(supportedArchs)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(int *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetSupportedArchs)(__pyx_v_supportedArchs);

    /* "cuda/_cuda/cnvrtc.pyx":192
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetSupportedArchs)(supportedArchs)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcCreateProgram(nvrtcProgram* prog, const char* src, const char* name, int numHeaders, const char** headers, const char** includeNames) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":186
 * 
 * cdef nvrtcResult _nvrtcGetSupportedArchs(int* supportedArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetSupportedArchs             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetSupportedArchs == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":185
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetSupportedArchs(int* supportedArchs) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetSupportedArchs
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetSupportedArchs", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":194
 *     return err
 * 
 * cdef nvrtcResult _nvrtcCreateProgram(nvrtcProgram* prog, const char* src, const char* name, int numHeaders, const char** headers, const char** includeNames) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcCreateProgram
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcCreateProgram(nvrtcProgram *__pyx_v_prog, char const *__pyx_v_src, char const *__pyx_v_name, int __pyx_v_numHeaders, char const **__pyx_v_headers, char const **__pyx_v_includeNames) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcCreateProgram", 1);
  __Pyx_TraceCall("_nvrtcCreateProgram", __pyx_f[0], 194, 1, __PYX_ERR(0, 194, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":195
 * 
 * cdef nvrtcResult _nvrtcCreateProgram(nvrtcProgram* prog, const char* src, const char* name, int numHeaders, const char** headers, const char** includeNames) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcCreateProgram             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcCreateProgram == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":196
 * cdef nvrtcResult _nvrtcCreateProgram(nvrtcProgram* prog, const char* src, const char* name, int numHeaders, const char** headers, const char** includeNames) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcCreateProgram
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcCreateProgram == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 196, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":197
 *     global __nvrtcCreateProgram
 *     cuPythonInit()
 *     if __nvrtcCreateProgram == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCreateProgram == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":198
 *     cuPythonInit()
 *     if __nvrtcCreateProgram == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*, const char*, const char*, int, const char**, const char**) nogil> __nvrtcCreateProgram)(prog, src, name, numHeaders, headers, includeNames)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":199
 *     if __nvrtcCreateProgram == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram*, const char*, const char*, int, const char**, const char**) nogil> __nvrtcCreateProgram)(prog, src, name, numHeaders, headers, includeNames)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__6, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 199, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 199, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":198
 *     cuPythonInit()
 *     if __nvrtcCreateProgram == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*, const char*, const char*, int, const char**, const char**) nogil> __nvrtcCreateProgram)(prog, src, name, numHeaders, headers, includeNames)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":197
 *     global __nvrtcCreateProgram
 *     cuPythonInit()
 *     if __nvrtcCreateProgram == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":200
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*, const char*, const char*, int, const char**, const char**) nogil> __nvrtcCreateProgram)(prog, src, name, numHeaders, headers, includeNames)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram *, char const *, char const *, int, char const **, char const **))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCreateProgram)(__pyx_v_prog, __pyx_v_src, __pyx_v_name, __pyx_v_numHeaders, __pyx_v_headers, __pyx_v_includeNames);

    /* "cuda/_cuda/cnvrtc.pyx":201
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*, const char*, const char*, int, const char**, const char**) nogil> __nvrtcCreateProgram)(prog, src, name, numHeaders, headers, includeNames)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcDestroyProgram(nvrtcProgram* prog) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":195
 * 
 * cdef nvrtcResult _nvrtcCreateProgram(nvrtcProgram* prog, const char* src, const char* name, int numHeaders, const char** headers, const char** includeNames) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcCreateProgram             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcCreateProgram == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":194
 *     return err
 * 
 * cdef nvrtcResult _nvrtcCreateProgram(nvrtcProgram* prog, const char* src, const char* name, int numHeaders, const char** headers, const char** includeNames) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcCreateProgram
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcCreateProgram", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":203
 *     return err
 * 
 * cdef nvrtcResult _nvrtcDestroyProgram(nvrtcProgram* prog) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcDestroyProgram
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcDestroyProgram(nvrtcProgram *__pyx_v_prog) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcDestroyProgram", 1);
  __Pyx_TraceCall("_nvrtcDestroyProgram", __pyx_f[0], 203, 1, __PYX_ERR(0, 203, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":204
 * 
 * cdef nvrtcResult _nvrtcDestroyProgram(nvrtcProgram* prog) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcDestroyProgram             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcDestroyProgram == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":205
 * cdef nvrtcResult _nvrtcDestroyProgram(nvrtcProgram* prog) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcDestroyProgram
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcDestroyProgram == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 205, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":206
 *     global __nvrtcDestroyProgram
 *     cuPythonInit()
 *     if __nvrtcDestroyProgram == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcDestroyProgram == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":207
 *     cuPythonInit()
 *     if __nvrtcDestroyProgram == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*) nogil> __nvrtcDestroyProgram)(prog)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":208
 *     if __nvrtcDestroyProgram == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram*) nogil> __nvrtcDestroyProgram)(prog)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__7, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 208, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 208, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":207
 *     cuPythonInit()
 *     if __nvrtcDestroyProgram == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*) nogil> __nvrtcDestroyProgram)(prog)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":206
 *     global __nvrtcDestroyProgram
 *     cuPythonInit()
 *     if __nvrtcDestroyProgram == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":209
 *         with gil:
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*) nogil> __nvrtcDestroyProgram)(prog)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcDestroyProgram)(__pyx_v_prog);

    /* "cuda/_cuda/cnvrtc.pyx":210
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram*) nogil> __nvrtcDestroyProgram)(prog)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcCompileProgram(nvrtcProgram prog, int numOptions, const char** options) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":204
 * 
 * cdef nvrtcResult _nvrtcDestroyProgram(nvrtcProgram* prog) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcDestroyProgram             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcDestroyProgram == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":203
 *     return err
 * 
 * cdef nvrtcResult _nvrtcDestroyProgram(nvrtcProgram* prog) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcDestroyProgram
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcDestroyProgram", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":212
 *     return err
 * 
 * cdef nvrtcResult _nvrtcCompileProgram(nvrtcProgram prog, int numOptions, const char** options) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcCompileProgram
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcCompileProgram(nvrtcProgram __pyx_v_prog, int __pyx_v_numOptions, char const **__pyx_v_options) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcCompileProgram", 1);
  __Pyx_TraceCall("_nvrtcCompileProgram", __pyx_f[0], 212, 1, __PYX_ERR(0, 212, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":213
 * 
 * cdef nvrtcResult _nvrtcCompileProgram(nvrtcProgram prog, int numOptions, const char** options) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcCompileProgram             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcCompileProgram == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":214
 * cdef nvrtcResult _nvrtcCompileProgram(nvrtcProgram prog, int numOptions, const char** options) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcCompileProgram
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcCompileProgram == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 214, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":215
 *     global __nvrtcCompileProgram
 *     cuPythonInit()
 *     if __nvrtcCompileProgram == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCompileProgram == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":216
 *     cuPythonInit()
 *     if __nvrtcCompileProgram == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, int, const char**) nogil> __nvrtcCompileProgram)(prog, numOptions, options)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":217
 *     if __nvrtcCompileProgram == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, int, const char**) nogil> __nvrtcCompileProgram)(prog, numOptions, options)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__8, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 217, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 217, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":216
 *     cuPythonInit()
 *     if __nvrtcCompileProgram == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, int, const char**) nogil> __nvrtcCompileProgram)(prog, numOptions, options)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":215
 *     global __nvrtcCompileProgram
 *     cuPythonInit()
 *     if __nvrtcCompileProgram == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":218
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, int, const char**) nogil> __nvrtcCompileProgram)(prog, numOptions, options)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, int, char const **))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCompileProgram)(__pyx_v_prog, __pyx_v_numOptions, __pyx_v_options);

    /* "cuda/_cuda/cnvrtc.pyx":219
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, int, const char**) nogil> __nvrtcCompileProgram)(prog, numOptions, options)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetPTXSize(nvrtcProgram prog, size_t* ptxSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":213
 * 
 * cdef nvrtcResult _nvrtcCompileProgram(nvrtcProgram prog, int numOptions, const char** options) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcCompileProgram             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcCompileProgram == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":212
 *     return err
 * 
 * cdef nvrtcResult _nvrtcCompileProgram(nvrtcProgram prog, int numOptions, const char** options) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcCompileProgram
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcCompileProgram", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":221
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetPTXSize(nvrtcProgram prog, size_t* ptxSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetPTXSize
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetPTXSize(nvrtcProgram __pyx_v_prog, size_t *__pyx_v_ptxSizeRet) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetPTXSize", 1);
  __Pyx_TraceCall("_nvrtcGetPTXSize", __pyx_f[0], 221, 1, __PYX_ERR(0, 221, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":222
 * 
 * cdef nvrtcResult _nvrtcGetPTXSize(nvrtcProgram prog, size_t* ptxSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetPTXSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetPTXSize == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":223
 * cdef nvrtcResult _nvrtcGetPTXSize(nvrtcProgram prog, size_t* ptxSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetPTXSize
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetPTXSize == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 223, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":224
 *     global __nvrtcGetPTXSize
 *     cuPythonInit()
 *     if __nvrtcGetPTXSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTXSize == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":225
 *     cuPythonInit()
 *     if __nvrtcGetPTXSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetPTXSize)(prog, ptxSizeRet)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":226
 *     if __nvrtcGetPTXSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetPTXSize)(prog, ptxSizeRet)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__9, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 226, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 226, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":225
 *     cuPythonInit()
 *     if __nvrtcGetPTXSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetPTXSize)(prog, ptxSizeRet)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":224
 *     global __nvrtcGetPTXSize
 *     cuPythonInit()
 *     if __nvrtcGetPTXSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":227
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetPTXSize)(prog, ptxSizeRet)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, size_t *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTXSize)(__pyx_v_prog, __pyx_v_ptxSizeRet);

    /* "cuda/_cuda/cnvrtc.pyx":228
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetPTXSize)(prog, ptxSizeRet)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetPTX(nvrtcProgram prog, char* ptx) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":222
 * 
 * cdef nvrtcResult _nvrtcGetPTXSize(nvrtcProgram prog, size_t* ptxSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetPTXSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetPTXSize == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":221
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetPTXSize(nvrtcProgram prog, size_t* ptxSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetPTXSize
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetPTXSize", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":230
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetPTX(nvrtcProgram prog, char* ptx) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetPTX
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetPTX(nvrtcProgram __pyx_v_prog, char *__pyx_v_ptx) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetPTX", 1);
  __Pyx_TraceCall("_nvrtcGetPTX", __pyx_f[0], 230, 1, __PYX_ERR(0, 230, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":231
 * 
 * cdef nvrtcResult _nvrtcGetPTX(nvrtcProgram prog, char* ptx) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetPTX             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetPTX == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":232
 * cdef nvrtcResult _nvrtcGetPTX(nvrtcProgram prog, char* ptx) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetPTX
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetPTX == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 232, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":233
 *     global __nvrtcGetPTX
 *     cuPythonInit()
 *     if __nvrtcGetPTX == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTX == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":234
 *     cuPythonInit()
 *     if __nvrtcGetPTX == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetPTX)(prog, ptx)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":235
 *     if __nvrtcGetPTX == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetPTX)(prog, ptx)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__10, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 235, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 235, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":234
 *     cuPythonInit()
 *     if __nvrtcGetPTX == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetPTX)(prog, ptx)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":233
 *     global __nvrtcGetPTX
 *     cuPythonInit()
 *     if __nvrtcGetPTX == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":236
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetPTX)(prog, ptx)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTX)(__pyx_v_prog, __pyx_v_ptx);

    /* "cuda/_cuda/cnvrtc.pyx":237
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetPTX)(prog, ptx)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetCUBINSize(nvrtcProgram prog, size_t* cubinSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":231
 * 
 * cdef nvrtcResult _nvrtcGetPTX(nvrtcProgram prog, char* ptx) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetPTX             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetPTX == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":230
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetPTX(nvrtcProgram prog, char* ptx) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetPTX
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetPTX", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":239
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetCUBINSize(nvrtcProgram prog, size_t* cubinSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetCUBINSize
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetCUBINSize(nvrtcProgram __pyx_v_prog, size_t *__pyx_v_cubinSizeRet) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetCUBINSize", 1);
  __Pyx_TraceCall("_nvrtcGetCUBINSize", __pyx_f[0], 239, 1, __PYX_ERR(0, 239, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":240
 * 
 * cdef nvrtcResult _nvrtcGetCUBINSize(nvrtcProgram prog, size_t* cubinSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetCUBINSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetCUBINSize == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":241
 * cdef nvrtcResult _nvrtcGetCUBINSize(nvrtcProgram prog, size_t* cubinSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetCUBINSize
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetCUBINSize == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 241, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":242
 *     global __nvrtcGetCUBINSize
 *     cuPythonInit()
 *     if __nvrtcGetCUBINSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBINSize == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":243
 *     cuPythonInit()
 *     if __nvrtcGetCUBINSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetCUBINSize)(prog, cubinSizeRet)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":244
 *     if __nvrtcGetCUBINSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetCUBINSize)(prog, cubinSizeRet)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__11, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 244, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 244, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":243
 *     cuPythonInit()
 *     if __nvrtcGetCUBINSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetCUBINSize)(prog, cubinSizeRet)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":242
 *     global __nvrtcGetCUBINSize
 *     cuPythonInit()
 *     if __nvrtcGetCUBINSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":245
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetCUBINSize)(prog, cubinSizeRet)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, size_t *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBINSize)(__pyx_v_prog, __pyx_v_cubinSizeRet);

    /* "cuda/_cuda/cnvrtc.pyx":246
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetCUBINSize)(prog, cubinSizeRet)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetCUBIN(nvrtcProgram prog, char* cubin) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":240
 * 
 * cdef nvrtcResult _nvrtcGetCUBINSize(nvrtcProgram prog, size_t* cubinSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetCUBINSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetCUBINSize == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":239
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetCUBINSize(nvrtcProgram prog, size_t* cubinSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetCUBINSize
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetCUBINSize", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":248
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetCUBIN(nvrtcProgram prog, char* cubin) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetCUBIN
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetCUBIN(nvrtcProgram __pyx_v_prog, char *__pyx_v_cubin) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetCUBIN", 1);
  __Pyx_TraceCall("_nvrtcGetCUBIN", __pyx_f[0], 248, 1, __PYX_ERR(0, 248, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":249
 * 
 * cdef nvrtcResult _nvrtcGetCUBIN(nvrtcProgram prog, char* cubin) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetCUBIN             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetCUBIN == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":250
 * cdef nvrtcResult _nvrtcGetCUBIN(nvrtcProgram prog, char* cubin) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetCUBIN
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetCUBIN == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 250, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":251
 *     global __nvrtcGetCUBIN
 *     cuPythonInit()
 *     if __nvrtcGetCUBIN == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBIN == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":252
 *     cuPythonInit()
 *     if __nvrtcGetCUBIN == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetCUBIN)(prog, cubin)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":253
 *     if __nvrtcGetCUBIN == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetCUBIN)(prog, cubin)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__12, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 253, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 253, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":252
 *     cuPythonInit()
 *     if __nvrtcGetCUBIN == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetCUBIN)(prog, cubin)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":251
 *     global __nvrtcGetCUBIN
 *     cuPythonInit()
 *     if __nvrtcGetCUBIN == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":254
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetCUBIN)(prog, cubin)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBIN)(__pyx_v_prog, __pyx_v_cubin);

    /* "cuda/_cuda/cnvrtc.pyx":255
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetCUBIN)(prog, cubin)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetNVVMSize(nvrtcProgram prog, size_t* nvvmSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":249
 * 
 * cdef nvrtcResult _nvrtcGetCUBIN(nvrtcProgram prog, char* cubin) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetCUBIN             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetCUBIN == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":248
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetCUBIN(nvrtcProgram prog, char* cubin) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetCUBIN
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetCUBIN", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":257
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetNVVMSize(nvrtcProgram prog, size_t* nvvmSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetNVVMSize
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetNVVMSize(nvrtcProgram __pyx_v_prog, size_t *__pyx_v_nvvmSizeRet) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetNVVMSize", 1);
  __Pyx_TraceCall("_nvrtcGetNVVMSize", __pyx_f[0], 257, 1, __PYX_ERR(0, 257, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":258
 * 
 * cdef nvrtcResult _nvrtcGetNVVMSize(nvrtcProgram prog, size_t* nvvmSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNVVMSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetNVVMSize == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":259
 * cdef nvrtcResult _nvrtcGetNVVMSize(nvrtcProgram prog, size_t* nvvmSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNVVMSize
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetNVVMSize == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 259, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":260
 *     global __nvrtcGetNVVMSize
 *     cuPythonInit()
 *     if __nvrtcGetNVVMSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVMSize == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":261
 *     cuPythonInit()
 *     if __nvrtcGetNVVMSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetNVVMSize)(prog, nvvmSizeRet)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":262
 *     if __nvrtcGetNVVMSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetNVVMSize)(prog, nvvmSizeRet)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__13, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 262, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 262, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":261
 *     cuPythonInit()
 *     if __nvrtcGetNVVMSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetNVVMSize)(prog, nvvmSizeRet)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":260
 *     global __nvrtcGetNVVMSize
 *     cuPythonInit()
 *     if __nvrtcGetNVVMSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":263
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetNVVMSize)(prog, nvvmSizeRet)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, size_t *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVMSize)(__pyx_v_prog, __pyx_v_nvvmSizeRet);

    /* "cuda/_cuda/cnvrtc.pyx":264
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetNVVMSize)(prog, nvvmSizeRet)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetNVVM(nvrtcProgram prog, char* nvvm) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":258
 * 
 * cdef nvrtcResult _nvrtcGetNVVMSize(nvrtcProgram prog, size_t* nvvmSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNVVMSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetNVVMSize == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":257
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetNVVMSize(nvrtcProgram prog, size_t* nvvmSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetNVVMSize
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetNVVMSize", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":266
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetNVVM(nvrtcProgram prog, char* nvvm) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetNVVM
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetNVVM(nvrtcProgram __pyx_v_prog, char *__pyx_v_nvvm) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetNVVM", 1);
  __Pyx_TraceCall("_nvrtcGetNVVM", __pyx_f[0], 266, 1, __PYX_ERR(0, 266, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":267
 * 
 * cdef nvrtcResult _nvrtcGetNVVM(nvrtcProgram prog, char* nvvm) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNVVM             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetNVVM == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":268
 * cdef nvrtcResult _nvrtcGetNVVM(nvrtcProgram prog, char* nvvm) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNVVM
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetNVVM == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 268, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":269
 *     global __nvrtcGetNVVM
 *     cuPythonInit()
 *     if __nvrtcGetNVVM == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVM == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":270
 *     cuPythonInit()
 *     if __nvrtcGetNVVM == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetNVVM)(prog, nvvm)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":271
 *     if __nvrtcGetNVVM == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetNVVM)(prog, nvvm)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__14, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 271, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 271, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":270
 *     cuPythonInit()
 *     if __nvrtcGetNVVM == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetNVVM)(prog, nvvm)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":269
 *     global __nvrtcGetNVVM
 *     cuPythonInit()
 *     if __nvrtcGetNVVM == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":272
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetNVVM)(prog, nvvm)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVM)(__pyx_v_prog, __pyx_v_nvvm);

    /* "cuda/_cuda/cnvrtc.pyx":273
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetNVVM)(prog, nvvm)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetLTOIRSize(nvrtcProgram prog, size_t* LTOIRSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":267
 * 
 * cdef nvrtcResult _nvrtcGetNVVM(nvrtcProgram prog, char* nvvm) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetNVVM             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetNVVM == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":266
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetNVVM(nvrtcProgram prog, char* nvvm) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetNVVM
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetNVVM", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":275
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetLTOIRSize(nvrtcProgram prog, size_t* LTOIRSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetLTOIRSize
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetLTOIRSize(nvrtcProgram __pyx_v_prog, size_t *__pyx_v_LTOIRSizeRet) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetLTOIRSize", 1);
  __Pyx_TraceCall("_nvrtcGetLTOIRSize", __pyx_f[0], 275, 1, __PYX_ERR(0, 275, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":276
 * 
 * cdef nvrtcResult _nvrtcGetLTOIRSize(nvrtcProgram prog, size_t* LTOIRSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLTOIRSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetLTOIRSize == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":277
 * cdef nvrtcResult _nvrtcGetLTOIRSize(nvrtcProgram prog, size_t* LTOIRSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLTOIRSize
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetLTOIRSize == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 277, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":278
 *     global __nvrtcGetLTOIRSize
 *     cuPythonInit()
 *     if __nvrtcGetLTOIRSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIRSize == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":279
 *     cuPythonInit()
 *     if __nvrtcGetLTOIRSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetLTOIRSize)(prog, LTOIRSizeRet)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":280
 *     if __nvrtcGetLTOIRSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetLTOIRSize)(prog, LTOIRSizeRet)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__15, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 280, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 280, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":279
 *     cuPythonInit()
 *     if __nvrtcGetLTOIRSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetLTOIRSize)(prog, LTOIRSizeRet)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":278
 *     global __nvrtcGetLTOIRSize
 *     cuPythonInit()
 *     if __nvrtcGetLTOIRSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":281
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetLTOIRSize)(prog, LTOIRSizeRet)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, size_t *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIRSize)(__pyx_v_prog, __pyx_v_LTOIRSizeRet);

    /* "cuda/_cuda/cnvrtc.pyx":282
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetLTOIRSize)(prog, LTOIRSizeRet)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetLTOIR(nvrtcProgram prog, char* LTOIR) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":276
 * 
 * cdef nvrtcResult _nvrtcGetLTOIRSize(nvrtcProgram prog, size_t* LTOIRSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLTOIRSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetLTOIRSize == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":275
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetLTOIRSize(nvrtcProgram prog, size_t* LTOIRSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetLTOIRSize
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetLTOIRSize", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":284
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetLTOIR(nvrtcProgram prog, char* LTOIR) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetLTOIR
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetLTOIR(nvrtcProgram __pyx_v_prog, char *__pyx_v_LTOIR) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetLTOIR", 1);
  __Pyx_TraceCall("_nvrtcGetLTOIR", __pyx_f[0], 284, 1, __PYX_ERR(0, 284, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":285
 * 
 * cdef nvrtcResult _nvrtcGetLTOIR(nvrtcProgram prog, char* LTOIR) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLTOIR             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetLTOIR == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":286
 * cdef nvrtcResult _nvrtcGetLTOIR(nvrtcProgram prog, char* LTOIR) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLTOIR
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetLTOIR == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 286, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":287
 *     global __nvrtcGetLTOIR
 *     cuPythonInit()
 *     if __nvrtcGetLTOIR == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIR == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":288
 *     cuPythonInit()
 *     if __nvrtcGetLTOIR == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetLTOIR)(prog, LTOIR)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":289
 *     if __nvrtcGetLTOIR == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetLTOIR)(prog, LTOIR)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__16, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 289, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 289, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":288
 *     cuPythonInit()
 *     if __nvrtcGetLTOIR == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetLTOIR)(prog, LTOIR)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":287
 *     global __nvrtcGetLTOIR
 *     cuPythonInit()
 *     if __nvrtcGetLTOIR == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":290
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetLTOIR)(prog, LTOIR)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIR)(__pyx_v_prog, __pyx_v_LTOIR);

    /* "cuda/_cuda/cnvrtc.pyx":291
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetLTOIR)(prog, LTOIR)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIRSize(nvrtcProgram prog, size_t* optixirSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":285
 * 
 * cdef nvrtcResult _nvrtcGetLTOIR(nvrtcProgram prog, char* LTOIR) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLTOIR             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetLTOIR == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":284
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetLTOIR(nvrtcProgram prog, char* LTOIR) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetLTOIR
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetLTOIR", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":293
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIRSize(nvrtcProgram prog, size_t* optixirSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetOptiXIRSize
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetOptiXIRSize(nvrtcProgram __pyx_v_prog, size_t *__pyx_v_optixirSizeRet) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetOptiXIRSize", 1);
  __Pyx_TraceCall("_nvrtcGetOptiXIRSize", __pyx_f[0], 293, 1, __PYX_ERR(0, 293, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":294
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIRSize(nvrtcProgram prog, size_t* optixirSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetOptiXIRSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIRSize == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":295
 * cdef nvrtcResult _nvrtcGetOptiXIRSize(nvrtcProgram prog, size_t* optixirSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetOptiXIRSize
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetOptiXIRSize == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 295, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":296
 *     global __nvrtcGetOptiXIRSize
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIRSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIRSize == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":297
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIRSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetOptiXIRSize)(prog, optixirSizeRet)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":298
 *     if __nvrtcGetOptiXIRSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetOptiXIRSize)(prog, optixirSizeRet)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__17, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 298, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 298, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":297
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIRSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetOptiXIRSize)(prog, optixirSizeRet)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":296
 *     global __nvrtcGetOptiXIRSize
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIRSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":299
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetOptiXIRSize)(prog, optixirSizeRet)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, size_t *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIRSize)(__pyx_v_prog, __pyx_v_optixirSizeRet);

    /* "cuda/_cuda/cnvrtc.pyx":300
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetOptiXIRSize)(prog, optixirSizeRet)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIR(nvrtcProgram prog, char* optixir) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":294
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIRSize(nvrtcProgram prog, size_t* optixirSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetOptiXIRSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIRSize == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":293
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIRSize(nvrtcProgram prog, size_t* optixirSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetOptiXIRSize
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetOptiXIRSize", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":302
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIR(nvrtcProgram prog, char* optixir) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetOptiXIR
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetOptiXIR(nvrtcProgram __pyx_v_prog, char *__pyx_v_optixir) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetOptiXIR", 1);
  __Pyx_TraceCall("_nvrtcGetOptiXIR", __pyx_f[0], 302, 1, __PYX_ERR(0, 302, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":303
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIR(nvrtcProgram prog, char* optixir) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetOptiXIR             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIR == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":304
 * cdef nvrtcResult _nvrtcGetOptiXIR(nvrtcProgram prog, char* optixir) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetOptiXIR
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetOptiXIR == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 304, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":305
 *     global __nvrtcGetOptiXIR
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIR == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIR == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":306
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIR == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetOptiXIR)(prog, optixir)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":307
 *     if __nvrtcGetOptiXIR == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetOptiXIR)(prog, optixir)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__18, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 307, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 307, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":306
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIR == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetOptiXIR)(prog, optixir)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":305
 *     global __nvrtcGetOptiXIR
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIR == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":308
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetOptiXIR)(prog, optixir)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIR)(__pyx_v_prog, __pyx_v_optixir);

    /* "cuda/_cuda/cnvrtc.pyx":309
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetOptiXIR)(prog, optixir)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetProgramLogSize(nvrtcProgram prog, size_t* logSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":303
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIR(nvrtcProgram prog, char* optixir) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetOptiXIR             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetOptiXIR == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":302
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetOptiXIR(nvrtcProgram prog, char* optixir) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetOptiXIR
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetOptiXIR", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":311
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetProgramLogSize(nvrtcProgram prog, size_t* logSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetProgramLogSize
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetProgramLogSize(nvrtcProgram __pyx_v_prog, size_t *__pyx_v_logSizeRet) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetProgramLogSize", 1);
  __Pyx_TraceCall("_nvrtcGetProgramLogSize", __pyx_f[0], 311, 1, __PYX_ERR(0, 311, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":312
 * 
 * cdef nvrtcResult _nvrtcGetProgramLogSize(nvrtcProgram prog, size_t* logSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetProgramLogSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetProgramLogSize == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":313
 * cdef nvrtcResult _nvrtcGetProgramLogSize(nvrtcProgram prog, size_t* logSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetProgramLogSize
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetProgramLogSize == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 313, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":314
 *     global __nvrtcGetProgramLogSize
 *     cuPythonInit()
 *     if __nvrtcGetProgramLogSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLogSize == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":315
 *     cuPythonInit()
 *     if __nvrtcGetProgramLogSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetProgramLogSize)(prog, logSizeRet)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":316
 *     if __nvrtcGetProgramLogSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetProgramLogSize)(prog, logSizeRet)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__19, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 316, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 316, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":315
 *     cuPythonInit()
 *     if __nvrtcGetProgramLogSize == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetProgramLogSize)(prog, logSizeRet)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":314
 *     global __nvrtcGetProgramLogSize
 *     cuPythonInit()
 *     if __nvrtcGetProgramLogSize == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":317
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetProgramLogSize)(prog, logSizeRet)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, size_t *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLogSize)(__pyx_v_prog, __pyx_v_logSizeRet);

    /* "cuda/_cuda/cnvrtc.pyx":318
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetProgramLogSize)(prog, logSizeRet)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetProgramLog(nvrtcProgram prog, char* log) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":312
 * 
 * cdef nvrtcResult _nvrtcGetProgramLogSize(nvrtcProgram prog, size_t* logSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetProgramLogSize             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetProgramLogSize == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":311
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetProgramLogSize(nvrtcProgram prog, size_t* logSizeRet) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetProgramLogSize
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetProgramLogSize", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":320
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetProgramLog(nvrtcProgram prog, char* log) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetProgramLog
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetProgramLog(nvrtcProgram __pyx_v_prog, char *__pyx_v_log) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetProgramLog", 1);
  __Pyx_TraceCall("_nvrtcGetProgramLog", __pyx_f[0], 320, 1, __PYX_ERR(0, 320, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":321
 * 
 * cdef nvrtcResult _nvrtcGetProgramLog(nvrtcProgram prog, char* log) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetProgramLog             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetProgramLog == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":322
 * cdef nvrtcResult _nvrtcGetProgramLog(nvrtcProgram prog, char* log) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetProgramLog
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetProgramLog == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 322, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":323
 *     global __nvrtcGetProgramLog
 *     cuPythonInit()
 *     if __nvrtcGetProgramLog == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLog == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":324
 *     cuPythonInit()
 *     if __nvrtcGetProgramLog == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetProgramLog)(prog, log)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":325
 *     if __nvrtcGetProgramLog == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetProgramLog)(prog, log)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__20, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 325, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 325, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":324
 *     cuPythonInit()
 *     if __nvrtcGetProgramLog == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetProgramLog)(prog, log)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":323
 *     global __nvrtcGetProgramLog
 *     cuPythonInit()
 *     if __nvrtcGetProgramLog == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":326
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetProgramLog)(prog, log)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLog)(__pyx_v_prog, __pyx_v_log);

    /* "cuda/_cuda/cnvrtc.pyx":327
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetProgramLog)(prog, log)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcAddNameExpression(nvrtcProgram prog, const char* name_expression) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":321
 * 
 * cdef nvrtcResult _nvrtcGetProgramLog(nvrtcProgram prog, char* log) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetProgramLog             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetProgramLog == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":320
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetProgramLog(nvrtcProgram prog, char* log) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetProgramLog
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetProgramLog", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":329
 *     return err
 * 
 * cdef nvrtcResult _nvrtcAddNameExpression(nvrtcProgram prog, const char* name_expression) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcAddNameExpression
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcAddNameExpression(nvrtcProgram __pyx_v_prog, char const *__pyx_v_name_expression) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcAddNameExpression", 1);
  __Pyx_TraceCall("_nvrtcAddNameExpression", __pyx_f[0], 329, 1, __PYX_ERR(0, 329, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":330
 * 
 * cdef nvrtcResult _nvrtcAddNameExpression(nvrtcProgram prog, const char* name_expression) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcAddNameExpression             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcAddNameExpression == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":331
 * cdef nvrtcResult _nvrtcAddNameExpression(nvrtcProgram prog, const char* name_expression) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcAddNameExpression
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcAddNameExpression == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 331, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":332
 *     global __nvrtcAddNameExpression
 *     cuPythonInit()
 *     if __nvrtcAddNameExpression == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcAddNameExpression == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":333
 *     cuPythonInit()
 *     if __nvrtcAddNameExpression == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*) nogil> __nvrtcAddNameExpression)(prog, name_expression)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":334
 *     if __nvrtcAddNameExpression == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*) nogil> __nvrtcAddNameExpression)(prog, name_expression)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__21, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 334, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 334, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":333
 *     cuPythonInit()
 *     if __nvrtcAddNameExpression == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*) nogil> __nvrtcAddNameExpression)(prog, name_expression)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":332
 *     global __nvrtcAddNameExpression
 *     cuPythonInit()
 *     if __nvrtcAddNameExpression == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":335
 *         with gil:
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*) nogil> __nvrtcAddNameExpression)(prog, name_expression)             # <<<<<<<<<<<<<<
 *     return err
 * 
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char const *))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcAddNameExpression)(__pyx_v_prog, __pyx_v_name_expression);

    /* "cuda/_cuda/cnvrtc.pyx":336
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*) nogil> __nvrtcAddNameExpression)(prog, name_expression)
 *     return err             # <<<<<<<<<<<<<<
 * 
 * cdef nvrtcResult _nvrtcGetLoweredName(nvrtcProgram prog, const char* name_expression, const char** lowered_name) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":330
 * 
 * cdef nvrtcResult _nvrtcAddNameExpression(nvrtcProgram prog, const char* name_expression) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcAddNameExpression             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcAddNameExpression == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":329
 *     return err
 * 
 * cdef nvrtcResult _nvrtcAddNameExpression(nvrtcProgram prog, const char* name_expression) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcAddNameExpression
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcAddNameExpression", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

/* "cuda/_cuda/cnvrtc.pyx":338
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetLoweredName(nvrtcProgram prog, const char* name_expression, const char** lowered_name) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetLoweredName
 *     cuPythonInit()
 */

static nvrtcResult __pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetLoweredName(nvrtcProgram __pyx_v_prog, char const *__pyx_v_name_expression, char const **__pyx_v_lowered_name) {
  nvrtcResult __pyx_v_err;
  nvrtcResult __pyx_r;
  __Pyx_TraceDeclarations
  __Pyx_RefNannyDeclarations
  int __pyx_t_1;
  int __pyx_t_2;
  PyObject *__pyx_t_3 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  #ifdef WITH_THREAD
  PyGILState_STATE __pyx_gilstate_save;
  #endif
  __Pyx_RefNannySetupContext("_nvrtcGetLoweredName", 1);
  __Pyx_TraceCall("_nvrtcGetLoweredName", __pyx_f[0], 338, 1, __PYX_ERR(0, 338, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":339
 * 
 * cdef nvrtcResult _nvrtcGetLoweredName(nvrtcProgram prog, const char* name_expression, const char** lowered_name) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLoweredName             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetLoweredName == NULL:
 */
  /*try:*/ {

    /* "cuda/_cuda/cnvrtc.pyx":340
 * cdef nvrtcResult _nvrtcGetLoweredName(nvrtcProgram prog, const char* name_expression, const char** lowered_name) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLoweredName
 *     cuPythonInit()             # <<<<<<<<<<<<<<
 *     if __nvrtcGetLoweredName == NULL:
 *         with gil:
 */
    __pyx_t_1 = __pyx_f_4cuda_5_cuda_6cnvrtc_cuPythonInit(); if (unlikely(__pyx_t_1 == ((int)-1))) __PYX_ERR(0, 340, __pyx_L4_error)

    /* "cuda/_cuda/cnvrtc.pyx":341
 *     global __nvrtcGetLoweredName
 *     cuPythonInit()
 *     if __nvrtcGetLoweredName == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')
 */
    __pyx_t_2 = ((__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLoweredName == NULL) != 0);
    if (__pyx_t_2) {

      /* "cuda/_cuda/cnvrtc.pyx":342
 *     cuPythonInit()
 *     if __nvrtcGetLoweredName == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*, const char**) nogil> __nvrtcGetLoweredName)(prog, name_expression, lowered_name)
 */
      {
          #ifdef WITH_THREAD
          PyGILState_STATE __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
          #endif
          /*try:*/ {

            /* "cuda/_cuda/cnvrtc.pyx":343
 *     if __nvrtcGetLoweredName == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*, const char**) nogil> __nvrtcGetLoweredName)(prog, name_expression, lowered_name)
 *     return err
 */
            __pyx_t_3 = __Pyx_PyObject_Call(__pyx_builtin_RuntimeError, __pyx_tuple__22, NULL); if (unlikely(!__pyx_t_3)) __PYX_ERR(0, 343, __pyx_L8_error)
            __Pyx_GOTREF(__pyx_t_3);
            __Pyx_Raise(__pyx_t_3, 0, 0, 0);
            __Pyx_DECREF(__pyx_t_3); __pyx_t_3 = 0;
            __PYX_ERR(0, 343, __pyx_L8_error)
          }

          /* "cuda/_cuda/cnvrtc.pyx":342
 *     cuPythonInit()
 *     if __nvrtcGetLoweredName == NULL:
 *         with gil:             # <<<<<<<<<<<<<<
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*, const char**) nogil> __nvrtcGetLoweredName)(prog, name_expression, lowered_name)
 */
          /*finally:*/ {
            __pyx_L8_error: {
              #ifdef WITH_THREAD
              __Pyx_PyGILState_Release(__pyx_gilstate_save);
              #endif
              goto __pyx_L4_error;
            }
          }
      }

      /* "cuda/_cuda/cnvrtc.pyx":341
 *     global __nvrtcGetLoweredName
 *     cuPythonInit()
 *     if __nvrtcGetLoweredName == NULL:             # <<<<<<<<<<<<<<
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')
 */
    }

    /* "cuda/_cuda/cnvrtc.pyx":344
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*, const char**) nogil> __nvrtcGetLoweredName)(prog, name_expression, lowered_name)             # <<<<<<<<<<<<<<
 *     return err
 */
    __pyx_v_err = ((nvrtcResult (*)(nvrtcProgram, char const *, char const **))__pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLoweredName)(__pyx_v_prog, __pyx_v_name_expression, __pyx_v_lowered_name);

    /* "cuda/_cuda/cnvrtc.pyx":345
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*, const char**) nogil> __nvrtcGetLoweredName)(prog, name_expression, lowered_name)
 *     return err             # <<<<<<<<<<<<<<
 */
    __pyx_r = __pyx_v_err;
    goto __pyx_L3_return;
  }

  /* "cuda/_cuda/cnvrtc.pyx":339
 * 
 * cdef nvrtcResult _nvrtcGetLoweredName(nvrtcProgram prog, const char* name_expression, const char** lowered_name) except ?NVRTC_ERROR_INVALID_INPUT nogil:
 *     global __nvrtcGetLoweredName             # <<<<<<<<<<<<<<
 *     cuPythonInit()
 *     if __nvrtcGetLoweredName == NULL:
 */
  /*finally:*/ {
    __pyx_L3_return: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L0;
    }
    __pyx_L4_error: {
      #ifdef WITH_THREAD
      __pyx_gilstate_save = __Pyx_PyGILState_Ensure();
      #endif
      goto __pyx_L1_error;
    }
  }

  /* "cuda/_cuda/cnvrtc.pyx":338
 *     return err
 * 
 * cdef nvrtcResult _nvrtcGetLoweredName(nvrtcProgram prog, const char* name_expression, const char** lowered_name) except ?NVRTC_ERROR_INVALID_INPUT nogil:             # <<<<<<<<<<<<<<
 *     global __nvrtcGetLoweredName
 *     cuPythonInit()
 */

  /* function exit code */
  __pyx_r = (nvrtcResult) 0;
  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_3);
  __Pyx_AddTraceback("cuda._cuda.cnvrtc._nvrtcGetLoweredName", __pyx_clineno, __pyx_lineno, __pyx_filename);
  __pyx_r = NVRTC_ERROR_INVALID_INPUT;
  __pyx_L0:;
  __Pyx_TraceReturn(Py_None, 1);
  #ifdef WITH_THREAD
  __Pyx_PyGILState_Release(__pyx_gilstate_save);
  #endif
  return __pyx_r;
}

static PyMethodDef __pyx_methods[] = {
  {0, 0, 0, 0}
};

#if PY_MAJOR_VERSION >= 3
#if CYTHON_PEP489_MULTI_PHASE_INIT
static PyObject* __pyx_pymod_create(PyObject *spec, PyModuleDef *def); /*proto*/
static int __pyx_pymod_exec_cnvrtc(PyObject* module); /*proto*/
static PyModuleDef_Slot __pyx_moduledef_slots[] = {
  {Py_mod_create, (void*)__pyx_pymod_create},
  {Py_mod_exec, (void*)__pyx_pymod_exec_cnvrtc},
  {0, NULL}
};
#endif

static struct PyModuleDef __pyx_moduledef = {
    PyModuleDef_HEAD_INIT,
    "cnvrtc",
    0, /* m_doc */
  #if CYTHON_PEP489_MULTI_PHASE_INIT
    0, /* m_size */
  #else
    -1, /* m_size */
  #endif
    __pyx_methods /* m_methods */,
  #if CYTHON_PEP489_MULTI_PHASE_INIT
    __pyx_moduledef_slots, /* m_slots */
  #else
    NULL, /* m_reload */
  #endif
    NULL, /* m_traverse */
    NULL, /* m_clear */
    NULL /* m_free */
};
#endif
#ifndef CYTHON_SMALL_CODE
#if defined(__clang__)
    #define CYTHON_SMALL_CODE
#elif defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3))
    #define CYTHON_SMALL_CODE __attribute__((cold))
#else
    #define CYTHON_SMALL_CODE
#endif
#endif

static __Pyx_StringTabEntry __pyx_string_tab[] = {
  {&__pyx_kp_u_Failed_to_LoadLibraryEx_nvrtc64, __pyx_k_Failed_to_LoadLibraryEx_nvrtc64, sizeof(__pyx_k_Failed_to_LoadLibraryEx_nvrtc64), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcAddNameExpression, __pyx_k_Function_nvrtcAddNameExpression, sizeof(__pyx_k_Function_nvrtcAddNameExpression), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcCompileProgram_not, __pyx_k_Function_nvrtcCompileProgram_not, sizeof(__pyx_k_Function_nvrtcCompileProgram_not), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcCreateProgram_not, __pyx_k_Function_nvrtcCreateProgram_not, sizeof(__pyx_k_Function_nvrtcCreateProgram_not), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcDestroyProgram_not, __pyx_k_Function_nvrtcDestroyProgram_not, sizeof(__pyx_k_Function_nvrtcDestroyProgram_not), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetCUBINSize_not_f, __pyx_k_Function_nvrtcGetCUBINSize_not_f, sizeof(__pyx_k_Function_nvrtcGetCUBINSize_not_f), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetCUBIN_not_found, __pyx_k_Function_nvrtcGetCUBIN_not_found, sizeof(__pyx_k_Function_nvrtcGetCUBIN_not_found), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetErrorString_not, __pyx_k_Function_nvrtcGetErrorString_not, sizeof(__pyx_k_Function_nvrtcGetErrorString_not), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetLTOIRSize_not_f, __pyx_k_Function_nvrtcGetLTOIRSize_not_f, sizeof(__pyx_k_Function_nvrtcGetLTOIRSize_not_f), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetLTOIR_not_found, __pyx_k_Function_nvrtcGetLTOIR_not_found, sizeof(__pyx_k_Function_nvrtcGetLTOIR_not_found), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetLoweredName_not, __pyx_k_Function_nvrtcGetLoweredName_not, sizeof(__pyx_k_Function_nvrtcGetLoweredName_not), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetNVVMSize_not_fo, __pyx_k_Function_nvrtcGetNVVMSize_not_fo, sizeof(__pyx_k_Function_nvrtcGetNVVMSize_not_fo), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetNVVM_not_found, __pyx_k_Function_nvrtcGetNVVM_not_found, sizeof(__pyx_k_Function_nvrtcGetNVVM_not_found), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetNumSupportedArc, __pyx_k_Function_nvrtcGetNumSupportedArc, sizeof(__pyx_k_Function_nvrtcGetNumSupportedArc), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetOptiXIRSize_not, __pyx_k_Function_nvrtcGetOptiXIRSize_not, sizeof(__pyx_k_Function_nvrtcGetOptiXIRSize_not), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetOptiXIR_not_fou, __pyx_k_Function_nvrtcGetOptiXIR_not_fou, sizeof(__pyx_k_Function_nvrtcGetOptiXIR_not_fou), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetPTXSize_not_fou, __pyx_k_Function_nvrtcGetPTXSize_not_fou, sizeof(__pyx_k_Function_nvrtcGetPTXSize_not_fou), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetPTX_not_found, __pyx_k_Function_nvrtcGetPTX_not_found, sizeof(__pyx_k_Function_nvrtcGetPTX_not_found), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetProgramLogSize, __pyx_k_Function_nvrtcGetProgramLogSize, sizeof(__pyx_k_Function_nvrtcGetProgramLogSize), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetProgramLog_not, __pyx_k_Function_nvrtcGetProgramLog_not, sizeof(__pyx_k_Function_nvrtcGetProgramLog_not), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcGetSupportedArchs, __pyx_k_Function_nvrtcGetSupportedArchs, sizeof(__pyx_k_Function_nvrtcGetSupportedArchs), 0, 1, 0, 0},
  {&__pyx_kp_u_Function_nvrtcVersion_not_found, __pyx_k_Function_nvrtcVersion_not_found, sizeof(__pyx_k_Function_nvrtcVersion_not_found), 0, 1, 0, 0},
  {&__pyx_n_s_GetProcAddress, __pyx_k_GetProcAddress, sizeof(__pyx_k_GetProcAddress), 0, 0, 1, 1},
  {&__pyx_n_s_LoadLibraryEx, __pyx_k_LoadLibraryEx, sizeof(__pyx_k_LoadLibraryEx), 0, 0, 1, 1},
  {&__pyx_n_s_RuntimeError, __pyx_k_RuntimeError, sizeof(__pyx_k_RuntimeError), 0, 0, 1, 1},
  {&__pyx_n_s_cline_in_traceback, __pyx_k_cline_in_traceback, sizeof(__pyx_k_cline_in_traceback), 0, 0, 1, 1},
  {&__pyx_n_s_error, __pyx_k_error, sizeof(__pyx_k_error), 0, 0, 1, 1},
  {&__pyx_n_s_import, __pyx_k_import, sizeof(__pyx_k_import), 0, 0, 1, 1},
  {&__pyx_n_s_main, __pyx_k_main, sizeof(__pyx_k_main), 0, 0, 1, 1},
  {&__pyx_n_s_name, __pyx_k_name, sizeof(__pyx_k_name), 0, 0, 1, 1},
  {&__pyx_kp_u_nvrtc64_120_0_dll, __pyx_k_nvrtc64_120_0_dll, sizeof(__pyx_k_nvrtc64_120_0_dll), 0, 1, 0, 0},
  {&__pyx_n_u_nvrtcAddNameExpression, __pyx_k_nvrtcAddNameExpression, sizeof(__pyx_k_nvrtcAddNameExpression), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcCompileProgram, __pyx_k_nvrtcCompileProgram, sizeof(__pyx_k_nvrtcCompileProgram), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcCreateProgram, __pyx_k_nvrtcCreateProgram, sizeof(__pyx_k_nvrtcCreateProgram), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcDestroyProgram, __pyx_k_nvrtcDestroyProgram, sizeof(__pyx_k_nvrtcDestroyProgram), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetCUBIN, __pyx_k_nvrtcGetCUBIN, sizeof(__pyx_k_nvrtcGetCUBIN), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetCUBINSize, __pyx_k_nvrtcGetCUBINSize, sizeof(__pyx_k_nvrtcGetCUBINSize), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetErrorString, __pyx_k_nvrtcGetErrorString, sizeof(__pyx_k_nvrtcGetErrorString), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetLTOIR, __pyx_k_nvrtcGetLTOIR, sizeof(__pyx_k_nvrtcGetLTOIR), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetLTOIRSize, __pyx_k_nvrtcGetLTOIRSize, sizeof(__pyx_k_nvrtcGetLTOIRSize), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetLoweredName, __pyx_k_nvrtcGetLoweredName, sizeof(__pyx_k_nvrtcGetLoweredName), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetNVVM, __pyx_k_nvrtcGetNVVM, sizeof(__pyx_k_nvrtcGetNVVM), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetNVVMSize, __pyx_k_nvrtcGetNVVMSize, sizeof(__pyx_k_nvrtcGetNVVMSize), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetNumSupportedArchs, __pyx_k_nvrtcGetNumSupportedArchs, sizeof(__pyx_k_nvrtcGetNumSupportedArchs), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetOptiXIR, __pyx_k_nvrtcGetOptiXIR, sizeof(__pyx_k_nvrtcGetOptiXIR), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetOptiXIRSize, __pyx_k_nvrtcGetOptiXIRSize, sizeof(__pyx_k_nvrtcGetOptiXIRSize), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetPTX, __pyx_k_nvrtcGetPTX, sizeof(__pyx_k_nvrtcGetPTX), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetPTXSize, __pyx_k_nvrtcGetPTXSize, sizeof(__pyx_k_nvrtcGetPTXSize), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetProgramLog, __pyx_k_nvrtcGetProgramLog, sizeof(__pyx_k_nvrtcGetProgramLog), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetProgramLogSize, __pyx_k_nvrtcGetProgramLogSize, sizeof(__pyx_k_nvrtcGetProgramLogSize), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcGetSupportedArchs, __pyx_k_nvrtcGetSupportedArchs, sizeof(__pyx_k_nvrtcGetSupportedArchs), 0, 1, 0, 1},
  {&__pyx_n_u_nvrtcVersion, __pyx_k_nvrtcVersion, sizeof(__pyx_k_nvrtcVersion), 0, 1, 0, 1},
  {&__pyx_n_s_pywintypes, __pyx_k_pywintypes, sizeof(__pyx_k_pywintypes), 0, 0, 1, 1},
  {&__pyx_n_s_struct, __pyx_k_struct, sizeof(__pyx_k_struct), 0, 0, 1, 1},
  {&__pyx_n_s_test, __pyx_k_test, sizeof(__pyx_k_test), 0, 0, 1, 1},
  {&__pyx_n_s_win32api, __pyx_k_win32api, sizeof(__pyx_k_win32api), 0, 0, 1, 1},
  {0, 0, 0, 0, 0, 0, 0}
};
static CYTHON_SMALL_CODE int __Pyx_InitCachedBuiltins(void) {
  __pyx_builtin_RuntimeError = __Pyx_GetBuiltinName(__pyx_n_s_RuntimeError); if (!__pyx_builtin_RuntimeError) __PYX_ERR(0, 47, __pyx_L1_error)
  return 0;
  __pyx_L1_error:;
  return -1;
}

static CYTHON_SMALL_CODE int __Pyx_InitCachedConstants(void) {
  __Pyx_RefNannyDeclarations
  __Pyx_RefNannySetupContext("__Pyx_InitCachedConstants", 0);

  /* "cuda/_cuda/cnvrtc.pyx":47
 *             handle = win32api.LoadLibraryEx("nvrtc64_120_0.dll", 0, LOAD_LIBRARY_SAFE_CURRENT_DIRS)
 *         except:
 *             raise RuntimeError('Failed to LoadLibraryEx nvrtc64_120_0.dll')             # <<<<<<<<<<<<<<
 * 
 * 
 */
  __pyx_tuple_ = PyTuple_Pack(1, __pyx_kp_u_Failed_to_LoadLibraryEx_nvrtc64); if (unlikely(!__pyx_tuple_)) __PYX_ERR(0, 47, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple_);
  __Pyx_GIVEREF(__pyx_tuple_);

  /* "cuda/_cuda/cnvrtc.pyx":163
 *     if __nvrtcGetErrorString == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetErrorString" not found')             # <<<<<<<<<<<<<<
 *     err = (<const char* (*)(nvrtcResult) nogil> __nvrtcGetErrorString)(result)
 *     return err
 */
  __pyx_tuple__2 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetErrorString_not); if (unlikely(!__pyx_tuple__2)) __PYX_ERR(0, 163, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__2);
  __Pyx_GIVEREF(__pyx_tuple__2);

  /* "cuda/_cuda/cnvrtc.pyx":172
 *     if __nvrtcVersion == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcVersion" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(int*, int*) nogil> __nvrtcVersion)(major, minor)
 *     return err
 */
  __pyx_tuple__3 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcVersion_not_found); if (unlikely(!__pyx_tuple__3)) __PYX_ERR(0, 172, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__3);
  __Pyx_GIVEREF(__pyx_tuple__3);

  /* "cuda/_cuda/cnvrtc.pyx":181
 *     if __nvrtcGetNumSupportedArchs == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNumSupportedArchs" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetNumSupportedArchs)(numArchs)
 *     return err
 */
  __pyx_tuple__4 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetNumSupportedArc); if (unlikely(!__pyx_tuple__4)) __PYX_ERR(0, 181, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__4);
  __Pyx_GIVEREF(__pyx_tuple__4);

  /* "cuda/_cuda/cnvrtc.pyx":190
 *     if __nvrtcGetSupportedArchs == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetSupportedArchs" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(int*) nogil> __nvrtcGetSupportedArchs)(supportedArchs)
 *     return err
 */
  __pyx_tuple__5 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetSupportedArchs); if (unlikely(!__pyx_tuple__5)) __PYX_ERR(0, 190, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__5);
  __Pyx_GIVEREF(__pyx_tuple__5);

  /* "cuda/_cuda/cnvrtc.pyx":199
 *     if __nvrtcCreateProgram == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCreateProgram" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram*, const char*, const char*, int, const char**, const char**) nogil> __nvrtcCreateProgram)(prog, src, name, numHeaders, headers, includeNames)
 *     return err
 */
  __pyx_tuple__6 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcCreateProgram_not); if (unlikely(!__pyx_tuple__6)) __PYX_ERR(0, 199, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__6);
  __Pyx_GIVEREF(__pyx_tuple__6);

  /* "cuda/_cuda/cnvrtc.pyx":208
 *     if __nvrtcDestroyProgram == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcDestroyProgram" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram*) nogil> __nvrtcDestroyProgram)(prog)
 *     return err
 */
  __pyx_tuple__7 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcDestroyProgram_not); if (unlikely(!__pyx_tuple__7)) __PYX_ERR(0, 208, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__7);
  __Pyx_GIVEREF(__pyx_tuple__7);

  /* "cuda/_cuda/cnvrtc.pyx":217
 *     if __nvrtcCompileProgram == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcCompileProgram" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, int, const char**) nogil> __nvrtcCompileProgram)(prog, numOptions, options)
 *     return err
 */
  __pyx_tuple__8 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcCompileProgram_not); if (unlikely(!__pyx_tuple__8)) __PYX_ERR(0, 217, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__8);
  __Pyx_GIVEREF(__pyx_tuple__8);

  /* "cuda/_cuda/cnvrtc.pyx":226
 *     if __nvrtcGetPTXSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTXSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetPTXSize)(prog, ptxSizeRet)
 *     return err
 */
  __pyx_tuple__9 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetPTXSize_not_fou); if (unlikely(!__pyx_tuple__9)) __PYX_ERR(0, 226, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__9);
  __Pyx_GIVEREF(__pyx_tuple__9);

  /* "cuda/_cuda/cnvrtc.pyx":235
 *     if __nvrtcGetPTX == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetPTX" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetPTX)(prog, ptx)
 *     return err
 */
  __pyx_tuple__10 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetPTX_not_found); if (unlikely(!__pyx_tuple__10)) __PYX_ERR(0, 235, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__10);
  __Pyx_GIVEREF(__pyx_tuple__10);

  /* "cuda/_cuda/cnvrtc.pyx":244
 *     if __nvrtcGetCUBINSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBINSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetCUBINSize)(prog, cubinSizeRet)
 *     return err
 */
  __pyx_tuple__11 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetCUBINSize_not_f); if (unlikely(!__pyx_tuple__11)) __PYX_ERR(0, 244, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__11);
  __Pyx_GIVEREF(__pyx_tuple__11);

  /* "cuda/_cuda/cnvrtc.pyx":253
 *     if __nvrtcGetCUBIN == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetCUBIN" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetCUBIN)(prog, cubin)
 *     return err
 */
  __pyx_tuple__12 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetCUBIN_not_found); if (unlikely(!__pyx_tuple__12)) __PYX_ERR(0, 253, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__12);
  __Pyx_GIVEREF(__pyx_tuple__12);

  /* "cuda/_cuda/cnvrtc.pyx":262
 *     if __nvrtcGetNVVMSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVMSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetNVVMSize)(prog, nvvmSizeRet)
 *     return err
 */
  __pyx_tuple__13 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetNVVMSize_not_fo); if (unlikely(!__pyx_tuple__13)) __PYX_ERR(0, 262, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__13);
  __Pyx_GIVEREF(__pyx_tuple__13);

  /* "cuda/_cuda/cnvrtc.pyx":271
 *     if __nvrtcGetNVVM == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetNVVM" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetNVVM)(prog, nvvm)
 *     return err
 */
  __pyx_tuple__14 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetNVVM_not_found); if (unlikely(!__pyx_tuple__14)) __PYX_ERR(0, 271, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__14);
  __Pyx_GIVEREF(__pyx_tuple__14);

  /* "cuda/_cuda/cnvrtc.pyx":280
 *     if __nvrtcGetLTOIRSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIRSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetLTOIRSize)(prog, LTOIRSizeRet)
 *     return err
 */
  __pyx_tuple__15 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetLTOIRSize_not_f); if (unlikely(!__pyx_tuple__15)) __PYX_ERR(0, 280, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__15);
  __Pyx_GIVEREF(__pyx_tuple__15);

  /* "cuda/_cuda/cnvrtc.pyx":289
 *     if __nvrtcGetLTOIR == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLTOIR" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetLTOIR)(prog, LTOIR)
 *     return err
 */
  __pyx_tuple__16 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetLTOIR_not_found); if (unlikely(!__pyx_tuple__16)) __PYX_ERR(0, 289, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__16);
  __Pyx_GIVEREF(__pyx_tuple__16);

  /* "cuda/_cuda/cnvrtc.pyx":298
 *     if __nvrtcGetOptiXIRSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIRSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetOptiXIRSize)(prog, optixirSizeRet)
 *     return err
 */
  __pyx_tuple__17 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetOptiXIRSize_not); if (unlikely(!__pyx_tuple__17)) __PYX_ERR(0, 298, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__17);
  __Pyx_GIVEREF(__pyx_tuple__17);

  /* "cuda/_cuda/cnvrtc.pyx":307
 *     if __nvrtcGetOptiXIR == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetOptiXIR" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetOptiXIR)(prog, optixir)
 *     return err
 */
  __pyx_tuple__18 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetOptiXIR_not_fou); if (unlikely(!__pyx_tuple__18)) __PYX_ERR(0, 307, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__18);
  __Pyx_GIVEREF(__pyx_tuple__18);

  /* "cuda/_cuda/cnvrtc.pyx":316
 *     if __nvrtcGetProgramLogSize == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLogSize" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, size_t*) nogil> __nvrtcGetProgramLogSize)(prog, logSizeRet)
 *     return err
 */
  __pyx_tuple__19 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetProgramLogSize); if (unlikely(!__pyx_tuple__19)) __PYX_ERR(0, 316, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__19);
  __Pyx_GIVEREF(__pyx_tuple__19);

  /* "cuda/_cuda/cnvrtc.pyx":325
 *     if __nvrtcGetProgramLog == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetProgramLog" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, char*) nogil> __nvrtcGetProgramLog)(prog, log)
 *     return err
 */
  __pyx_tuple__20 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetProgramLog_not); if (unlikely(!__pyx_tuple__20)) __PYX_ERR(0, 325, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__20);
  __Pyx_GIVEREF(__pyx_tuple__20);

  /* "cuda/_cuda/cnvrtc.pyx":334
 *     if __nvrtcAddNameExpression == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcAddNameExpression" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*) nogil> __nvrtcAddNameExpression)(prog, name_expression)
 *     return err
 */
  __pyx_tuple__21 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcAddNameExpression); if (unlikely(!__pyx_tuple__21)) __PYX_ERR(0, 334, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__21);
  __Pyx_GIVEREF(__pyx_tuple__21);

  /* "cuda/_cuda/cnvrtc.pyx":343
 *     if __nvrtcGetLoweredName == NULL:
 *         with gil:
 *             raise RuntimeError('Function "nvrtcGetLoweredName" not found')             # <<<<<<<<<<<<<<
 *     err = (<nvrtcResult (*)(nvrtcProgram, const char*, const char**) nogil> __nvrtcGetLoweredName)(prog, name_expression, lowered_name)
 *     return err
 */
  __pyx_tuple__22 = PyTuple_Pack(1, __pyx_kp_u_Function_nvrtcGetLoweredName_not); if (unlikely(!__pyx_tuple__22)) __PYX_ERR(0, 343, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_tuple__22);
  __Pyx_GIVEREF(__pyx_tuple__22);
  __Pyx_RefNannyFinishContext();
  return 0;
  __pyx_L1_error:;
  __Pyx_RefNannyFinishContext();
  return -1;
}

static CYTHON_SMALL_CODE int __Pyx_InitGlobals(void) {
  if (__Pyx_InitStrings(__pyx_string_tab) < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  __pyx_int_0 = PyInt_FromLong(0); if (unlikely(!__pyx_int_0)) __PYX_ERR(0, 1, __pyx_L1_error)
  return 0;
  __pyx_L1_error:;
  return -1;
}

static CYTHON_SMALL_CODE int __Pyx_modinit_global_init_code(void); /*proto*/
static CYTHON_SMALL_CODE int __Pyx_modinit_variable_export_code(void); /*proto*/
static CYTHON_SMALL_CODE int __Pyx_modinit_function_export_code(void); /*proto*/
static CYTHON_SMALL_CODE int __Pyx_modinit_type_init_code(void); /*proto*/
static CYTHON_SMALL_CODE int __Pyx_modinit_type_import_code(void); /*proto*/
static CYTHON_SMALL_CODE int __Pyx_modinit_variable_import_code(void); /*proto*/
static CYTHON_SMALL_CODE int __Pyx_modinit_function_import_code(void); /*proto*/

static int __Pyx_modinit_global_init_code(void) {
  __Pyx_RefNannyDeclarations
  __Pyx_RefNannySetupContext("__Pyx_modinit_global_init_code", 0);
  /*--- Global init code ---*/
  __Pyx_RefNannyFinishContext();
  return 0;
}

static int __Pyx_modinit_variable_export_code(void) {
  __Pyx_RefNannyDeclarations
  __Pyx_RefNannySetupContext("__Pyx_modinit_variable_export_code", 0);
  /*--- Variable export code ---*/
  __Pyx_RefNannyFinishContext();
  return 0;
}

static int __Pyx_modinit_function_export_code(void) {
  __Pyx_RefNannyDeclarations
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  __Pyx_RefNannySetupContext("__Pyx_modinit_function_export_code", 0);
  /*--- Function export code ---*/
  if (__Pyx_ExportFunction("_nvrtcGetErrorString", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetErrorString, "char const *(nvrtcResult)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcVersion", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcVersion, "nvrtcResult (int *, int *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetNumSupportedArchs", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetNumSupportedArchs, "nvrtcResult (int *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetSupportedArchs", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetSupportedArchs, "nvrtcResult (int *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcCreateProgram", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcCreateProgram, "nvrtcResult (nvrtcProgram *, char const *, char const *, int, char const **, char const **)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcDestroyProgram", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcDestroyProgram, "nvrtcResult (nvrtcProgram *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcCompileProgram", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcCompileProgram, "nvrtcResult (nvrtcProgram, int, char const **)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetPTXSize", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetPTXSize, "nvrtcResult (nvrtcProgram, size_t *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetPTX", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetPTX, "nvrtcResult (nvrtcProgram, char *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetCUBINSize", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetCUBINSize, "nvrtcResult (nvrtcProgram, size_t *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetCUBIN", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetCUBIN, "nvrtcResult (nvrtcProgram, char *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetNVVMSize", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetNVVMSize, "nvrtcResult (nvrtcProgram, size_t *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetNVVM", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetNVVM, "nvrtcResult (nvrtcProgram, char *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetLTOIRSize", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetLTOIRSize, "nvrtcResult (nvrtcProgram, size_t *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetLTOIR", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetLTOIR, "nvrtcResult (nvrtcProgram, char *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetOptiXIRSize", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetOptiXIRSize, "nvrtcResult (nvrtcProgram, size_t *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetOptiXIR", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetOptiXIR, "nvrtcResult (nvrtcProgram, char *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetProgramLogSize", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetProgramLogSize, "nvrtcResult (nvrtcProgram, size_t *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetProgramLog", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetProgramLog, "nvrtcResult (nvrtcProgram, char *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcAddNameExpression", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcAddNameExpression, "nvrtcResult (nvrtcProgram, char const *)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  if (__Pyx_ExportFunction("_nvrtcGetLoweredName", (void (*)(void))__pyx_f_4cuda_5_cuda_6cnvrtc__nvrtcGetLoweredName, "nvrtcResult (nvrtcProgram, char const *, char const **)") < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  __Pyx_RefNannyFinishContext();
  return 0;
  __pyx_L1_error:;
  __Pyx_RefNannyFinishContext();
  return -1;
}

static int __Pyx_modinit_type_init_code(void) {
  __Pyx_RefNannyDeclarations
  __Pyx_RefNannySetupContext("__Pyx_modinit_type_init_code", 0);
  /*--- Type init code ---*/
  __Pyx_RefNannyFinishContext();
  return 0;
}

static int __Pyx_modinit_type_import_code(void) {
  __Pyx_RefNannyDeclarations
  __Pyx_RefNannySetupContext("__Pyx_modinit_type_import_code", 0);
  /*--- Type import code ---*/
  __Pyx_RefNannyFinishContext();
  return 0;
}

static int __Pyx_modinit_variable_import_code(void) {
  __Pyx_RefNannyDeclarations
  __Pyx_RefNannySetupContext("__Pyx_modinit_variable_import_code", 0);
  /*--- Variable import code ---*/
  __Pyx_RefNannyFinishContext();
  return 0;
}

static int __Pyx_modinit_function_import_code(void) {
  __Pyx_RefNannyDeclarations
  __Pyx_RefNannySetupContext("__Pyx_modinit_function_import_code", 0);
  /*--- Function import code ---*/
  __Pyx_RefNannyFinishContext();
  return 0;
}


#ifndef CYTHON_NO_PYINIT_EXPORT
#define __Pyx_PyMODINIT_FUNC PyMODINIT_FUNC
#elif PY_MAJOR_VERSION < 3
#ifdef __cplusplus
#define __Pyx_PyMODINIT_FUNC extern "C" void
#else
#define __Pyx_PyMODINIT_FUNC void
#endif
#else
#ifdef __cplusplus
#define __Pyx_PyMODINIT_FUNC extern "C" PyObject *
#else
#define __Pyx_PyMODINIT_FUNC PyObject *
#endif
#endif


#if PY_MAJOR_VERSION < 3
__Pyx_PyMODINIT_FUNC initcnvrtc(void) CYTHON_SMALL_CODE; /*proto*/
__Pyx_PyMODINIT_FUNC initcnvrtc(void)
#else
__Pyx_PyMODINIT_FUNC PyInit_cnvrtc(void) CYTHON_SMALL_CODE; /*proto*/
__Pyx_PyMODINIT_FUNC PyInit_cnvrtc(void)
#if CYTHON_PEP489_MULTI_PHASE_INIT
{
  return PyModuleDef_Init(&__pyx_moduledef);
}
static CYTHON_SMALL_CODE int __Pyx_check_single_interpreter(void) {
    #if PY_VERSION_HEX >= 0x030700A1
    static PY_INT64_T main_interpreter_id = -1;
    PY_INT64_T current_id = PyInterpreterState_GetID(PyThreadState_Get()->interp);
    if (main_interpreter_id == -1) {
        main_interpreter_id = current_id;
        return (unlikely(current_id == -1)) ? -1 : 0;
    } else if (unlikely(main_interpreter_id != current_id))
    #else
    static PyInterpreterState *main_interpreter = NULL;
    PyInterpreterState *current_interpreter = PyThreadState_Get()->interp;
    if (!main_interpreter) {
        main_interpreter = current_interpreter;
    } else if (unlikely(main_interpreter != current_interpreter))
    #endif
    {
        PyErr_SetString(
            PyExc_ImportError,
            "Interpreter change detected - this module can only be loaded into one interpreter per process.");
        return -1;
    }
    return 0;
}
static CYTHON_SMALL_CODE int __Pyx_copy_spec_to_module(PyObject *spec, PyObject *moddict, const char* from_name, const char* to_name, int allow_none) {
    PyObject *value = PyObject_GetAttrString(spec, from_name);
    int result = 0;
    if (likely(value)) {
        if (allow_none || value != Py_None) {
            result = PyDict_SetItemString(moddict, to_name, value);
        }
        Py_DECREF(value);
    } else if (PyErr_ExceptionMatches(PyExc_AttributeError)) {
        PyErr_Clear();
    } else {
        result = -1;
    }
    return result;
}
static CYTHON_SMALL_CODE PyObject* __pyx_pymod_create(PyObject *spec, CYTHON_UNUSED PyModuleDef *def) {
    PyObject *module = NULL, *moddict, *modname;
    if (__Pyx_check_single_interpreter())
        return NULL;
    if (__pyx_m)
        return __Pyx_NewRef(__pyx_m);
    modname = PyObject_GetAttrString(spec, "name");
    if (unlikely(!modname)) goto bad;
    module = PyModule_NewObject(modname);
    Py_DECREF(modname);
    if (unlikely(!module)) goto bad;
    moddict = PyModule_GetDict(module);
    if (unlikely(!moddict)) goto bad;
    if (unlikely(__Pyx_copy_spec_to_module(spec, moddict, "loader", "__loader__", 1) < 0)) goto bad;
    if (unlikely(__Pyx_copy_spec_to_module(spec, moddict, "origin", "__file__", 1) < 0)) goto bad;
    if (unlikely(__Pyx_copy_spec_to_module(spec, moddict, "parent", "__package__", 1) < 0)) goto bad;
    if (unlikely(__Pyx_copy_spec_to_module(spec, moddict, "submodule_search_locations", "__path__", 0) < 0)) goto bad;
    return module;
bad:
    Py_XDECREF(module);
    return NULL;
}


static CYTHON_SMALL_CODE int __pyx_pymod_exec_cnvrtc(PyObject *__pyx_pyinit_module)
#endif
#endif
{
  __Pyx_TraceDeclarations
  PyObject *__pyx_t_1 = NULL;
  PyObject *__pyx_t_2 = NULL;
  int __pyx_lineno = 0;
  const char *__pyx_filename = NULL;
  int __pyx_clineno = 0;
  __Pyx_RefNannyDeclarations
  #if CYTHON_PEP489_MULTI_PHASE_INIT
  if (__pyx_m) {
    if (__pyx_m == __pyx_pyinit_module) return 0;
    PyErr_SetString(PyExc_RuntimeError, "Module 'cnvrtc' has already been imported. Re-initialisation is not supported.");
    return -1;
  }
  #elif PY_MAJOR_VERSION >= 3
  if (__pyx_m) return __Pyx_NewRef(__pyx_m);
  #endif
  #if CYTHON_REFNANNY
__Pyx_RefNanny = __Pyx_RefNannyImportAPI("refnanny");
if (!__Pyx_RefNanny) {
  PyErr_Clear();
  __Pyx_RefNanny = __Pyx_RefNannyImportAPI("Cython.Runtime.refnanny");
  if (!__Pyx_RefNanny)
      Py_FatalError("failed to import 'refnanny' module");
}
#endif
  __Pyx_RefNannySetupContext("__Pyx_PyMODINIT_FUNC PyInit_cnvrtc(void)", 0);
  if (__Pyx_check_binary_version() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #ifdef __Pxy_PyFrame_Initialize_Offsets
  __Pxy_PyFrame_Initialize_Offsets();
  #endif
  __pyx_empty_tuple = PyTuple_New(0); if (unlikely(!__pyx_empty_tuple)) __PYX_ERR(0, 1, __pyx_L1_error)
  __pyx_empty_bytes = PyBytes_FromStringAndSize("", 0); if (unlikely(!__pyx_empty_bytes)) __PYX_ERR(0, 1, __pyx_L1_error)
  __pyx_empty_unicode = PyUnicode_FromStringAndSize("", 0); if (unlikely(!__pyx_empty_unicode)) __PYX_ERR(0, 1, __pyx_L1_error)
  #ifdef __Pyx_CyFunction_USED
  if (__pyx_CyFunction_init() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  #ifdef __Pyx_FusedFunction_USED
  if (__pyx_FusedFunction_init() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  #ifdef __Pyx_Coroutine_USED
  if (__pyx_Coroutine_init() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  #ifdef __Pyx_Generator_USED
  if (__pyx_Generator_init() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  #ifdef __Pyx_AsyncGen_USED
  if (__pyx_AsyncGen_init() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  #ifdef __Pyx_StopAsyncIteration_USED
  if (__pyx_StopAsyncIteration_init() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  /*--- Library function declarations ---*/
  /*--- Threads initialization code ---*/
  #if defined(WITH_THREAD) && PY_VERSION_HEX < 0x030700F0 && defined(__PYX_FORCE_INIT_THREADS) && __PYX_FORCE_INIT_THREADS
  PyEval_InitThreads();
  #endif
  /*--- Module creation code ---*/
  #if CYTHON_PEP489_MULTI_PHASE_INIT
  __pyx_m = __pyx_pyinit_module;
  Py_INCREF(__pyx_m);
  #else
  #if PY_MAJOR_VERSION < 3
  __pyx_m = Py_InitModule4("cnvrtc", __pyx_methods, 0, 0, PYTHON_API_VERSION); Py_XINCREF(__pyx_m);
  #else
  __pyx_m = PyModule_Create(&__pyx_moduledef);
  #endif
  if (unlikely(!__pyx_m)) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  __pyx_d = PyModule_GetDict(__pyx_m); if (unlikely(!__pyx_d)) __PYX_ERR(0, 1, __pyx_L1_error)
  Py_INCREF(__pyx_d);
  __pyx_b = PyImport_AddModule(__Pyx_BUILTIN_MODULE_NAME); if (unlikely(!__pyx_b)) __PYX_ERR(0, 1, __pyx_L1_error)
  Py_INCREF(__pyx_b);
  __pyx_cython_runtime = PyImport_AddModule((char *) "cython_runtime"); if (unlikely(!__pyx_cython_runtime)) __PYX_ERR(0, 1, __pyx_L1_error)
  Py_INCREF(__pyx_cython_runtime);
  if (PyObject_SetAttrString(__pyx_m, "__builtins__", __pyx_b) < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  /*--- Initialize various global constants etc. ---*/
  if (__Pyx_InitGlobals() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #if PY_MAJOR_VERSION < 3 && (__PYX_DEFAULT_STRING_ENCODING_IS_ASCII || __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT)
  if (__Pyx_init_sys_getdefaultencoding_params() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  if (__pyx_module_is_main_cuda___cuda__cnvrtc) {
    if (PyObject_SetAttr(__pyx_m, __pyx_n_s_name, __pyx_n_s_main) < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  }
  #if PY_MAJOR_VERSION >= 3
  {
    PyObject *modules = PyImport_GetModuleDict(); if (unlikely(!modules)) __PYX_ERR(0, 1, __pyx_L1_error)
    if (!PyDict_GetItemString(modules, "cuda._cuda.cnvrtc")) {
      if (unlikely(PyDict_SetItemString(modules, "cuda._cuda.cnvrtc", __pyx_m) < 0)) __PYX_ERR(0, 1, __pyx_L1_error)
    }
  }
  #endif
  /*--- Builtin init code ---*/
  if (__Pyx_InitCachedBuiltins() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  /*--- Constants init code ---*/
  if (__Pyx_InitCachedConstants() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  /*--- Global type/function init code ---*/
  (void)__Pyx_modinit_global_init_code();
  (void)__Pyx_modinit_variable_export_code();
  if (unlikely(__Pyx_modinit_function_export_code() < 0)) __PYX_ERR(0, 1, __pyx_L1_error)
  (void)__Pyx_modinit_type_init_code();
  (void)__Pyx_modinit_type_import_code();
  (void)__Pyx_modinit_variable_import_code();
  (void)__Pyx_modinit_function_import_code();
  /*--- Execution code ---*/
  #if defined(__Pyx_Generator_USED) || defined(__Pyx_Coroutine_USED)
  if (__Pyx_patch_abc() < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  #endif
  __Pyx_TraceCall("__Pyx_PyMODINIT_FUNC PyInit_cnvrtc(void)", __pyx_f[0], 1, 0, __PYX_ERR(0, 1, __pyx_L1_error));

  /* "cuda/_cuda/cnvrtc.pyx":8
 * # this software and related documentation outside the terms of the EULA
 * # is strictly prohibited.
 * import win32api             # <<<<<<<<<<<<<<
 * import struct
 * from pywintypes import error
 */
  __pyx_t_1 = __Pyx_Import(__pyx_n_s_win32api, 0, 0); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 8, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_t_1);
  if (PyDict_SetItem(__pyx_d, __pyx_n_s_win32api, __pyx_t_1) < 0) __PYX_ERR(0, 8, __pyx_L1_error)
  __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;

  /* "cuda/_cuda/cnvrtc.pyx":9
 * # is strictly prohibited.
 * import win32api
 * import struct             # <<<<<<<<<<<<<<
 * from pywintypes import error
 * 
 */
  __pyx_t_1 = __Pyx_Import(__pyx_n_s_struct, 0, 0); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 9, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_t_1);
  if (PyDict_SetItem(__pyx_d, __pyx_n_s_struct, __pyx_t_1) < 0) __PYX_ERR(0, 9, __pyx_L1_error)
  __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;

  /* "cuda/_cuda/cnvrtc.pyx":10
 * import win32api
 * import struct
 * from pywintypes import error             # <<<<<<<<<<<<<<
 * 
 * cdef bint __cuPythonInit = False
 */
  __pyx_t_1 = PyList_New(1); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 10, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_t_1);
  __Pyx_INCREF(__pyx_n_s_error);
  __Pyx_GIVEREF(__pyx_n_s_error);
  PyList_SET_ITEM(__pyx_t_1, 0, __pyx_n_s_error);
  __pyx_t_2 = __Pyx_Import(__pyx_n_s_pywintypes, __pyx_t_1, 0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 10, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_t_2);
  __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;
  __pyx_t_1 = __Pyx_ImportFrom(__pyx_t_2, __pyx_n_s_error); if (unlikely(!__pyx_t_1)) __PYX_ERR(0, 10, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_t_1);
  if (PyDict_SetItem(__pyx_d, __pyx_n_s_error, __pyx_t_1) < 0) __PYX_ERR(0, 10, __pyx_L1_error)
  __Pyx_DECREF(__pyx_t_1); __pyx_t_1 = 0;
  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;

  /* "cuda/_cuda/cnvrtc.pyx":12
 * from pywintypes import error
 * 
 * cdef bint __cuPythonInit = False             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetErrorString = NULL
 * cdef void *__nvrtcVersion = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___cuPythonInit = 0;

  /* "cuda/_cuda/cnvrtc.pyx":13
 * 
 * cdef bint __cuPythonInit = False
 * cdef void *__nvrtcGetErrorString = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcVersion = NULL
 * cdef void *__nvrtcGetNumSupportedArchs = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetErrorString = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":14
 * cdef bint __cuPythonInit = False
 * cdef void *__nvrtcGetErrorString = NULL
 * cdef void *__nvrtcVersion = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetNumSupportedArchs = NULL
 * cdef void *__nvrtcGetSupportedArchs = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcVersion = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":15
 * cdef void *__nvrtcGetErrorString = NULL
 * cdef void *__nvrtcVersion = NULL
 * cdef void *__nvrtcGetNumSupportedArchs = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetSupportedArchs = NULL
 * cdef void *__nvrtcCreateProgram = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNumSupportedArchs = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":16
 * cdef void *__nvrtcVersion = NULL
 * cdef void *__nvrtcGetNumSupportedArchs = NULL
 * cdef void *__nvrtcGetSupportedArchs = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcCreateProgram = NULL
 * cdef void *__nvrtcDestroyProgram = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetSupportedArchs = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":17
 * cdef void *__nvrtcGetNumSupportedArchs = NULL
 * cdef void *__nvrtcGetSupportedArchs = NULL
 * cdef void *__nvrtcCreateProgram = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcDestroyProgram = NULL
 * cdef void *__nvrtcCompileProgram = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCreateProgram = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":18
 * cdef void *__nvrtcGetSupportedArchs = NULL
 * cdef void *__nvrtcCreateProgram = NULL
 * cdef void *__nvrtcDestroyProgram = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcCompileProgram = NULL
 * cdef void *__nvrtcGetPTXSize = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcDestroyProgram = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":19
 * cdef void *__nvrtcCreateProgram = NULL
 * cdef void *__nvrtcDestroyProgram = NULL
 * cdef void *__nvrtcCompileProgram = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetPTXSize = NULL
 * cdef void *__nvrtcGetPTX = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcCompileProgram = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":20
 * cdef void *__nvrtcDestroyProgram = NULL
 * cdef void *__nvrtcCompileProgram = NULL
 * cdef void *__nvrtcGetPTXSize = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetPTX = NULL
 * cdef void *__nvrtcGetCUBINSize = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTXSize = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":21
 * cdef void *__nvrtcCompileProgram = NULL
 * cdef void *__nvrtcGetPTXSize = NULL
 * cdef void *__nvrtcGetPTX = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetCUBINSize = NULL
 * cdef void *__nvrtcGetCUBIN = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetPTX = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":22
 * cdef void *__nvrtcGetPTXSize = NULL
 * cdef void *__nvrtcGetPTX = NULL
 * cdef void *__nvrtcGetCUBINSize = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetCUBIN = NULL
 * cdef void *__nvrtcGetNVVMSize = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBINSize = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":23
 * cdef void *__nvrtcGetPTX = NULL
 * cdef void *__nvrtcGetCUBINSize = NULL
 * cdef void *__nvrtcGetCUBIN = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetNVVMSize = NULL
 * cdef void *__nvrtcGetNVVM = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetCUBIN = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":24
 * cdef void *__nvrtcGetCUBINSize = NULL
 * cdef void *__nvrtcGetCUBIN = NULL
 * cdef void *__nvrtcGetNVVMSize = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetNVVM = NULL
 * cdef void *__nvrtcGetLTOIRSize = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVMSize = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":25
 * cdef void *__nvrtcGetCUBIN = NULL
 * cdef void *__nvrtcGetNVVMSize = NULL
 * cdef void *__nvrtcGetNVVM = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetLTOIRSize = NULL
 * cdef void *__nvrtcGetLTOIR = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetNVVM = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":26
 * cdef void *__nvrtcGetNVVMSize = NULL
 * cdef void *__nvrtcGetNVVM = NULL
 * cdef void *__nvrtcGetLTOIRSize = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetLTOIR = NULL
 * cdef void *__nvrtcGetOptiXIRSize = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIRSize = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":27
 * cdef void *__nvrtcGetNVVM = NULL
 * cdef void *__nvrtcGetLTOIRSize = NULL
 * cdef void *__nvrtcGetLTOIR = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetOptiXIRSize = NULL
 * cdef void *__nvrtcGetOptiXIR = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLTOIR = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":28
 * cdef void *__nvrtcGetLTOIRSize = NULL
 * cdef void *__nvrtcGetLTOIR = NULL
 * cdef void *__nvrtcGetOptiXIRSize = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetOptiXIR = NULL
 * cdef void *__nvrtcGetProgramLogSize = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIRSize = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":29
 * cdef void *__nvrtcGetLTOIR = NULL
 * cdef void *__nvrtcGetOptiXIRSize = NULL
 * cdef void *__nvrtcGetOptiXIR = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetProgramLogSize = NULL
 * cdef void *__nvrtcGetProgramLog = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetOptiXIR = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":30
 * cdef void *__nvrtcGetOptiXIRSize = NULL
 * cdef void *__nvrtcGetOptiXIR = NULL
 * cdef void *__nvrtcGetProgramLogSize = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetProgramLog = NULL
 * cdef void *__nvrtcAddNameExpression = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLogSize = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":31
 * cdef void *__nvrtcGetOptiXIR = NULL
 * cdef void *__nvrtcGetProgramLogSize = NULL
 * cdef void *__nvrtcGetProgramLog = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcAddNameExpression = NULL
 * cdef void *__nvrtcGetLoweredName = NULL
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetProgramLog = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":32
 * cdef void *__nvrtcGetProgramLogSize = NULL
 * cdef void *__nvrtcGetProgramLog = NULL
 * cdef void *__nvrtcAddNameExpression = NULL             # <<<<<<<<<<<<<<
 * cdef void *__nvrtcGetLoweredName = NULL
 * 
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcAddNameExpression = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":33
 * cdef void *__nvrtcGetProgramLog = NULL
 * cdef void *__nvrtcAddNameExpression = NULL
 * cdef void *__nvrtcGetLoweredName = NULL             # <<<<<<<<<<<<<<
 * 
 * cdef int cuPythonInit() except -1 nogil:
 */
  __pyx_v_4cuda_5_cuda_6cnvrtc___nvrtcGetLoweredName = NULL;

  /* "cuda/_cuda/cnvrtc.pyx":1
 * # Copyright 2021-2023 NVIDIA Corporation.  All rights reserved.             # <<<<<<<<<<<<<<
 * #
 * # Please refer to the NVIDIA end user license agreement (EULA) associated
 */
  __pyx_t_2 = __Pyx_PyDict_NewPresized(0); if (unlikely(!__pyx_t_2)) __PYX_ERR(0, 1, __pyx_L1_error)
  __Pyx_GOTREF(__pyx_t_2);
  if (PyDict_SetItem(__pyx_d, __pyx_n_s_test, __pyx_t_2) < 0) __PYX_ERR(0, 1, __pyx_L1_error)
  __Pyx_DECREF(__pyx_t_2); __pyx_t_2 = 0;
  __Pyx_TraceReturn(Py_None, 0);

  /*--- Wrapped vars code ---*/

  goto __pyx_L0;
  __pyx_L1_error:;
  __Pyx_XDECREF(__pyx_t_1);
  __Pyx_XDECREF(__pyx_t_2);
  if (__pyx_m) {
    if (__pyx_d) {
      __Pyx_AddTraceback("init cuda._cuda.cnvrtc", __pyx_clineno, __pyx_lineno, __pyx_filename);
    }
    Py_CLEAR(__pyx_m);
  } else if (!PyErr_Occurred()) {
    PyErr_SetString(PyExc_ImportError, "init cuda._cuda.cnvrtc");
  }
  __pyx_L0:;
  __Pyx_RefNannyFinishContext();
  #if CYTHON_PEP489_MULTI_PHASE_INIT
  return (__pyx_m != NULL) ? 0 : -1;
  #elif PY_MAJOR_VERSION >= 3
  return __pyx_m;
  #else
  return;
  #endif
}

/* --- Runtime support code --- */
/* Refnanny */
#if CYTHON_REFNANNY
static __Pyx_RefNannyAPIStruct *__Pyx_RefNannyImportAPI(const char *modname) {
    PyObject *m = NULL, *p = NULL;
    void *r = NULL;
    m = PyImport_ImportModule(modname);
    if (!m) goto end;
    p = PyObject_GetAttrString(m, "RefNannyAPI");
    if (!p) goto end;
    r = PyLong_AsVoidPtr(p);
end:
    Py_XDECREF(p);
    Py_XDECREF(m);
    return (__Pyx_RefNannyAPIStruct *)r;
}
#endif

/* PyObjectGetAttrStr */
#if CYTHON_USE_TYPE_SLOTS
static CYTHON_INLINE PyObject* __Pyx_PyObject_GetAttrStr(PyObject* obj, PyObject* attr_name) {
    PyTypeObject* tp = Py_TYPE(obj);
    if (likely(tp->tp_getattro))
        return tp->tp_getattro(obj, attr_name);
#if PY_MAJOR_VERSION < 3
    if (likely(tp->tp_getattr))
        return tp->tp_getattr(obj, PyString_AS_STRING(attr_name));
#endif
    return PyObject_GetAttr(obj, attr_name);
}
#endif

/* GetBuiltinName */
static PyObject *__Pyx_GetBuiltinName(PyObject *name) {
    PyObject* result = __Pyx_PyObject_GetAttrStr(__pyx_b, name);
    if (unlikely(!result)) {
        PyErr_Format(PyExc_NameError,
#if PY_MAJOR_VERSION >= 3
            "name '%U' is not defined", name);
#else
            "name '%.200s' is not defined", PyString_AS_STRING(name));
#endif
    }
    return result;
}

/* PyErrFetchRestore */
#if CYTHON_FAST_THREAD_STATE
static CYTHON_INLINE void __Pyx_ErrRestoreInState(PyThreadState *tstate, PyObject *type, PyObject *value, PyObject *tb) {
    PyObject *tmp_type, *tmp_value, *tmp_tb;
    tmp_type = tstate->curexc_type;
    tmp_value = tstate->curexc_value;
    tmp_tb = tstate->curexc_traceback;
    tstate->curexc_type = type;
    tstate->curexc_value = value;
    tstate->curexc_traceback = tb;
    Py_XDECREF(tmp_type);
    Py_XDECREF(tmp_value);
    Py_XDECREF(tmp_tb);
}
static CYTHON_INLINE void __Pyx_ErrFetchInState(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb) {
    *type = tstate->curexc_type;
    *value = tstate->curexc_value;
    *tb = tstate->curexc_traceback;
    tstate->curexc_type = 0;
    tstate->curexc_value = 0;
    tstate->curexc_traceback = 0;
}
#endif

/* Profile */
#if CYTHON_PROFILE
static int __Pyx_TraceSetupAndCall(PyCodeObject** code,
                                   PyFrameObject** frame,
                                   PyThreadState* tstate,
                                   const char *funcname,
                                   const char *srcfile,
                                   int firstlineno) {
    PyObject *type, *value, *traceback;
    int retval;
    if (*frame == NULL || !CYTHON_PROFILE_REUSE_FRAME) {
        if (*code == NULL) {
            *code = __Pyx_createFrameCodeObject(funcname, srcfile, firstlineno);
            if (*code == NULL) return 0;
        }
        *frame = PyFrame_New(
            tstate,                          /*PyThreadState *tstate*/
            *code,                           /*PyCodeObject *code*/
            __pyx_d,                  /*PyObject *globals*/
            0                                /*PyObject *locals*/
        );
        if (*frame == NULL) return 0;
        if (CYTHON_TRACE && (*frame)->f_trace == NULL) {
            Py_INCREF(Py_None);
            (*frame)->f_trace = Py_None;
        }
#if PY_VERSION_HEX < 0x030400B1
    } else {
        (*frame)->f_tstate = tstate;
#endif
    }
    __Pyx_PyFrame_SetLineNumber(*frame, firstlineno);
    retval = 1;
    __Pyx_EnterTracing(tstate);
    __Pyx_ErrFetchInState(tstate, &type, &value, &traceback);
    #if CYTHON_TRACE
    if (tstate->c_tracefunc)
        retval = tstate->c_tracefunc(tstate->c_traceobj, *frame, PyTrace_CALL, NULL) == 0;
    if (retval && tstate->c_profilefunc)
    #endif
        retval = tstate->c_profilefunc(tstate->c_profileobj, *frame, PyTrace_CALL, NULL) == 0;
    __Pyx_LeaveTracing(tstate);
    if (retval) {
        __Pyx_ErrRestoreInState(tstate, type, value, traceback);
        return __Pyx_IsTracing(tstate, 0, 0) && retval;
    } else {
        Py_XDECREF(type);
        Py_XDECREF(value);
        Py_XDECREF(traceback);
        return -1;
    }
}
static PyCodeObject *__Pyx_createFrameCodeObject(const char *funcname, const char *srcfile, int firstlineno) {
    PyCodeObject *py_code = 0;
#if PY_MAJOR_VERSION >= 3
    py_code = PyCode_NewEmpty(srcfile, funcname, firstlineno);
    if (likely(py_code)) {
        py_code->co_flags |= CO_OPTIMIZED | CO_NEWLOCALS;
    }
#else
    PyObject *py_srcfile = 0;
    PyObject *py_funcname = 0;
    py_funcname = PyString_FromString(funcname);
    if (unlikely(!py_funcname)) goto bad;
    py_srcfile = PyString_FromString(srcfile);
    if (unlikely(!py_srcfile)) goto bad;
    py_code = PyCode_New(
        0,
        0,
        0,
        CO_OPTIMIZED | CO_NEWLOCALS,
        __pyx_empty_bytes,     /*PyObject *code,*/
        __pyx_empty_tuple,     /*PyObject *consts,*/
        __pyx_empty_tuple,     /*PyObject *names,*/
        __pyx_empty_tuple,     /*PyObject *varnames,*/
        __pyx_empty_tuple,     /*PyObject *freevars,*/
        __pyx_empty_tuple,     /*PyObject *cellvars,*/
        py_srcfile,       /*PyObject *filename,*/
        py_funcname,      /*PyObject *name,*/
        firstlineno,
        __pyx_empty_bytes      /*PyObject *lnotab*/
    );
bad:
    Py_XDECREF(py_srcfile);
    Py_XDECREF(py_funcname);
#endif
    return py_code;
}
#endif

/* PyDictVersioning */
#if CYTHON_USE_DICT_VERSIONS && CYTHON_USE_TYPE_SLOTS
static CYTHON_INLINE PY_UINT64_T __Pyx_get_tp_dict_version(PyObject *obj) {
    PyObject *dict = Py_TYPE(obj)->tp_dict;
    return likely(dict) ? __PYX_GET_DICT_VERSION(dict) : 0;
}
static CYTHON_INLINE PY_UINT64_T __Pyx_get_object_dict_version(PyObject *obj) {
    PyObject **dictptr = NULL;
    Py_ssize_t offset = Py_TYPE(obj)->tp_dictoffset;
    if (offset) {
#if CYTHON_COMPILING_IN_CPYTHON
        dictptr = (likely(offset > 0)) ? (PyObject **) ((char *)obj + offset) : _PyObject_GetDictPtr(obj);
#else
        dictptr = _PyObject_GetDictPtr(obj);
#endif
    }
    return (dictptr && *dictptr) ? __PYX_GET_DICT_VERSION(*dictptr) : 0;
}
static CYTHON_INLINE int __Pyx_object_dict_version_matches(PyObject* obj, PY_UINT64_T tp_dict_version, PY_UINT64_T obj_dict_version) {
    PyObject *dict = Py_TYPE(obj)->tp_dict;
    if (unlikely(!dict) || unlikely(tp_dict_version != __PYX_GET_DICT_VERSION(dict)))
        return 0;
    return obj_dict_version == __Pyx_get_object_dict_version(obj);
}
#endif

/* GetModuleGlobalName */
#if CYTHON_USE_DICT_VERSIONS
static PyObject *__Pyx__GetModuleGlobalName(PyObject *name, PY_UINT64_T *dict_version, PyObject **dict_cached_value)
#else
static CYTHON_INLINE PyObject *__Pyx__GetModuleGlobalName(PyObject *name)
#endif
{
    PyObject *result;
#if !CYTHON_AVOID_BORROWED_REFS
#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX >= 0x030500A1
    result = _PyDict_GetItem_KnownHash(__pyx_d, name, ((PyASCIIObject *) name)->hash);
    __PYX_UPDATE_DICT_CACHE(__pyx_d, result, *dict_cached_value, *dict_version)
    if (likely(result)) {
        return __Pyx_NewRef(result);
    } else if (unlikely(PyErr_Occurred())) {
        return NULL;
    }
#else
    result = PyDict_GetItem(__pyx_d, name);
    __PYX_UPDATE_DICT_CACHE(__pyx_d, result, *dict_cached_value, *dict_version)
    if (likely(result)) {
        return __Pyx_NewRef(result);
    }
#endif
#else
    result = PyObject_GetItem(__pyx_d, name);
    __PYX_UPDATE_DICT_CACHE(__pyx_d, result, *dict_cached_value, *dict_version)
    if (likely(result)) {
        return __Pyx_NewRef(result);
    }
    PyErr_Clear();
#endif
    return __Pyx_GetBuiltinName(name);
}

/* PyFunctionFastCall */
#if CYTHON_FAST_PYCALL
static PyObject* __Pyx_PyFunction_FastCallNoKw(PyCodeObject *co, PyObject **args, Py_ssize_t na,
                                               PyObject *globals) {
    PyFrameObject *f;
    PyThreadState *tstate = __Pyx_PyThreadState_Current;
    PyObject **fastlocals;
    Py_ssize_t i;
    PyObject *result;
    assert(globals != NULL);
    /* XXX Perhaps we should create a specialized
       PyFrame_New() that doesn't take locals, but does
       take builtins without sanity checking them.
       */
    assert(tstate != NULL);
    f = PyFrame_New(tstate, co, globals, NULL);
    if (f == NULL) {
        return NULL;
    }
    fastlocals = __Pyx_PyFrame_GetLocalsplus(f);
    for (i = 0; i < na; i++) {
        Py_INCREF(*args);
        fastlocals[i] = *args++;
    }
    result = PyEval_EvalFrameEx(f,0);
    ++tstate->recursion_depth;
    Py_DECREF(f);
    --tstate->recursion_depth;
    return result;
}
#if 1 || PY_VERSION_HEX < 0x030600B1
static PyObject *__Pyx_PyFunction_FastCallDict(PyObject *func, PyObject **args, Py_ssize_t nargs, PyObject *kwargs) {
    PyCodeObject *co = (PyCodeObject *)PyFunction_GET_CODE(func);
    PyObject *globals = PyFunction_GET_GLOBALS(func);
    PyObject *argdefs = PyFunction_GET_DEFAULTS(func);
    PyObject *closure;
#if PY_MAJOR_VERSION >= 3
    PyObject *kwdefs;
#endif
    PyObject *kwtuple, **k;
    PyObject **d;
    Py_ssize_t nd;
    Py_ssize_t nk;
    PyObject *result;
    assert(kwargs == NULL || PyDict_Check(kwargs));
    nk = kwargs ? PyDict_Size(kwargs) : 0;
    if (Py_EnterRecursiveCall((char*)" while calling a Python object")) {
        return NULL;
    }
    if (
#if PY_MAJOR_VERSION >= 3
            co->co_kwonlyargcount == 0 &&
#endif
            likely(kwargs == NULL || nk == 0) &&
            co->co_flags == (CO_OPTIMIZED | CO_NEWLOCALS | CO_NOFREE)) {
        if (argdefs == NULL && co->co_argcount == nargs) {
            result = __Pyx_PyFunction_FastCallNoKw(co, args, nargs, globals);
            goto done;
        }
        else if (nargs == 0 && argdefs != NULL
                 && co->co_argcount == Py_SIZE(argdefs)) {
            /* function called with no arguments, but all parameters have
               a default value: use default values as arguments .*/
            args = &PyTuple_GET_ITEM(argdefs, 0);
            result =__Pyx_PyFunction_FastCallNoKw(co, args, Py_SIZE(argdefs), globals);
            goto done;
        }
    }
    if (kwargs != NULL) {
        Py_ssize_t pos, i;
        kwtuple = PyTuple_New(2 * nk);
        if (kwtuple == NULL) {
            result = NULL;
            goto done;
        }
        k = &PyTuple_GET_ITEM(kwtuple, 0);
        pos = i = 0;
        while (PyDict_Next(kwargs, &pos, &k[i], &k[i+1])) {
            Py_INCREF(k[i]);
            Py_INCREF(k[i+1]);
            i += 2;
        }
        nk = i / 2;
    }
    else {
        kwtuple = NULL;
        k = NULL;
    }
    closure = PyFunction_GET_CLOSURE(func);
#if PY_MAJOR_VERSION >= 3
    kwdefs = PyFunction_GET_KW_DEFAULTS(func);
#endif
    if (argdefs != NULL) {
        d = &PyTuple_GET_ITEM(argdefs, 0);
        nd = Py_SIZE(argdefs);
    }
    else {
        d = NULL;
        nd = 0;
    }
#if PY_MAJOR_VERSION >= 3
    result = PyEval_EvalCodeEx((PyObject*)co, globals, (PyObject *)NULL,
                               args, (int)nargs,
                               k, (int)nk,
                               d, (int)nd, kwdefs, closure);
#else
    result = PyEval_EvalCodeEx(co, globals, (PyObject *)NULL,
                               args, (int)nargs,
                               k, (int)nk,
                               d, (int)nd, closure);
#endif
    Py_XDECREF(kwtuple);
done:
    Py_LeaveRecursiveCall();
    return result;
}
#endif
#endif

/* PyCFunctionFastCall */
#if CYTHON_FAST_PYCCALL
static CYTHON_INLINE PyObject * __Pyx_PyCFunction_FastCall(PyObject *func_obj, PyObject **args, Py_ssize_t nargs) {
    PyCFunctionObject *func = (PyCFunctionObject*)func_obj;
    PyCFunction meth = PyCFunction_GET_FUNCTION(func);
    PyObject *self = PyCFunction_GET_SELF(func);
    int flags = PyCFunction_GET_FLAGS(func);
    assert(PyCFunction_Check(func));
    assert(METH_FASTCALL == (flags & ~(METH_CLASS | METH_STATIC | METH_COEXIST | METH_KEYWORDS | METH_STACKLESS)));
    assert(nargs >= 0);
    assert(nargs == 0 || args != NULL);
    /* _PyCFunction_FastCallDict() must not be called with an exception set,
       because it may clear it (directly or indirectly) and so the
       caller loses its exception */
    assert(!PyErr_Occurred());
    if ((PY_VERSION_HEX < 0x030700A0) || unlikely(flags & METH_KEYWORDS)) {
        return (*((__Pyx_PyCFunctionFastWithKeywords)(void*)meth)) (self, args, nargs, NULL);
    } else {
        return (*((__Pyx_PyCFunctionFast)(void*)meth)) (self, args, nargs);
    }
}
#endif

/* PyObjectCall */
#if CYTHON_COMPILING_IN_CPYTHON
static CYTHON_INLINE PyObject* __Pyx_PyObject_Call(PyObject *func, PyObject *arg, PyObject *kw) {
    PyObject *result;
    ternaryfunc call = Py_TYPE(func)->tp_call;
    if (unlikely(!call))
        return PyObject_Call(func, arg, kw);
    if (unlikely(Py_EnterRecursiveCall((char*)" while calling a Python object")))
        return NULL;
    result = (*call)(func, arg, kw);
    Py_LeaveRecursiveCall();
    if (unlikely(!result) && unlikely(!PyErr_Occurred())) {
        PyErr_SetString(
            PyExc_SystemError,
            "NULL result without error in PyObject_Call");
    }
    return result;
}
#endif

/* GetTopmostException */
#if CYTHON_USE_EXC_INFO_STACK
static _PyErr_StackItem *
__Pyx_PyErr_GetTopmostException(PyThreadState *tstate)
{
    _PyErr_StackItem *exc_info = tstate->exc_info;
    while ((exc_info->exc_type == NULL || exc_info->exc_type == Py_None) &&
           exc_info->previous_item != NULL)
    {
        exc_info = exc_info->previous_item;
    }
    return exc_info;
}
#endif

/* SaveResetException */
#if CYTHON_FAST_THREAD_STATE
static CYTHON_INLINE void __Pyx__ExceptionSave(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb) {
    #if CYTHON_USE_EXC_INFO_STACK
    _PyErr_StackItem *exc_info = __Pyx_PyErr_GetTopmostException(tstate);
    *type = exc_info->exc_type;
    *value = exc_info->exc_value;
    *tb = exc_info->exc_traceback;
    #else
    *type = tstate->exc_type;
    *value = tstate->exc_value;
    *tb = tstate->exc_traceback;
    #endif
    Py_XINCREF(*type);
    Py_XINCREF(*value);
    Py_XINCREF(*tb);
}
static CYTHON_INLINE void __Pyx__ExceptionReset(PyThreadState *tstate, PyObject *type, PyObject *value, PyObject *tb) {
    PyObject *tmp_type, *tmp_value, *tmp_tb;
    #if CYTHON_USE_EXC_INFO_STACK
    _PyErr_StackItem *exc_info = tstate->exc_info;
    tmp_type = exc_info->exc_type;
    tmp_value = exc_info->exc_value;
    tmp_tb = exc_info->exc_traceback;
    exc_info->exc_type = type;
    exc_info->exc_value = value;
    exc_info->exc_traceback = tb;
    #else
    tmp_type = tstate->exc_type;
    tmp_value = tstate->exc_value;
    tmp_tb = tstate->exc_traceback;
    tstate->exc_type = type;
    tstate->exc_value = value;
    tstate->exc_traceback = tb;
    #endif
    Py_XDECREF(tmp_type);
    Py_XDECREF(tmp_value);
    Py_XDECREF(tmp_tb);
}
#endif

/* GetException */
#if CYTHON_FAST_THREAD_STATE
static int __Pyx__GetException(PyThreadState *tstate, PyObject **type, PyObject **value, PyObject **tb)
#else
static int __Pyx_GetException(PyObject **type, PyObject **value, PyObject **tb)
#endif
{
    PyObject *local_type, *local_value, *local_tb;
#if CYTHON_FAST_THREAD_STATE
    PyObject *tmp_type, *tmp_value, *tmp_tb;
    local_type = tstate->curexc_type;
    local_value = tstate->curexc_value;
    local_tb = tstate->curexc_traceback;
    tstate->curexc_type = 0;
    tstate->curexc_value = 0;
    tstate->curexc_traceback = 0;
#else
    PyErr_Fetch(&local_type, &local_value, &local_tb);
#endif
    PyErr_NormalizeException(&local_type, &local_value, &local_tb);
#if CYTHON_FAST_THREAD_STATE
    if (unlikely(tstate->curexc_type))
#else
    if (unlikely(PyErr_Occurred()))
#endif
        goto bad;
    #if PY_MAJOR_VERSION >= 3
    if (local_tb) {
        if (unlikely(PyException_SetTraceback(local_value, local_tb) < 0))
            goto bad;
    }
    #endif
    Py_XINCREF(local_tb);
    Py_XINCREF(local_type);
    Py_XINCREF(local_value);
    *type = local_type;
    *value = local_value;
    *tb = local_tb;
#if CYTHON_FAST_THREAD_STATE
    #if CYTHON_USE_EXC_INFO_STACK
    {
        _PyErr_StackItem *exc_info = tstate->exc_info;
        tmp_type = exc_info->exc_type;
        tmp_value = exc_info->exc_value;
        tmp_tb = exc_info->exc_traceback;
        exc_info->exc_type = local_type;
        exc_info->exc_value = local_value;
        exc_info->exc_traceback = local_tb;
    }
    #else
    tmp_type = tstate->exc_type;
    tmp_value = tstate->exc_value;
    tmp_tb = tstate->exc_traceback;
    tstate->exc_type = local_type;
    tstate->exc_value = local_value;
    tstate->exc_traceback = local_tb;
    #endif
    Py_XDECREF(tmp_type);
    Py_XDECREF(tmp_value);
    Py_XDECREF(tmp_tb);
#else
    PyErr_SetExcInfo(local_type, local_value, local_tb);
#endif
    return 0;
bad:
    *type = 0;
    *value = 0;
    *tb = 0;
    Py_XDECREF(local_type);
    Py_XDECREF(local_value);
    Py_XDECREF(local_tb);
    return -1;
}

/* RaiseException */
#if PY_MAJOR_VERSION < 3
static void __Pyx_Raise(PyObject *type, PyObject *value, PyObject *tb,
                        CYTHON_UNUSED PyObject *cause) {
    __Pyx_PyThreadState_declare
    Py_XINCREF(type);
    if (!value || value == Py_None)
        value = NULL;
    else
        Py_INCREF(value);
    if (!tb || tb == Py_None)
        tb = NULL;
    else {
        Py_INCREF(tb);
        if (!PyTraceBack_Check(tb)) {
            PyErr_SetString(PyExc_TypeError,
                "raise: arg 3 must be a traceback or None");
            goto raise_error;
        }
    }
    if (PyType_Check(type)) {
#if CYTHON_COMPILING_IN_PYPY
        if (!value) {
            Py_INCREF(Py_None);
            value = Py_None;
        }
#endif
        PyErr_NormalizeException(&type, &value, &tb);
    } else {
        if (value) {
            PyErr_SetString(PyExc_TypeError,
                "instance exception may not have a separate value");
            goto raise_error;
        }
        value = type;
        type = (PyObject*) Py_TYPE(type);
        Py_INCREF(type);
        if (!PyType_IsSubtype((PyTypeObject *)type, (PyTypeObject *)PyExc_BaseException)) {
            PyErr_SetString(PyExc_TypeError,
                "raise: exception class must be a subclass of BaseException");
            goto raise_error;
        }
    }
    __Pyx_PyThreadState_assign
    __Pyx_ErrRestore(type, value, tb);
    return;
raise_error:
    Py_XDECREF(value);
    Py_XDECREF(type);
    Py_XDECREF(tb);
    return;
}
#else
static void __Pyx_Raise(PyObject *type, PyObject *value, PyObject *tb, PyObject *cause) {
    PyObject* owned_instance = NULL;
    if (tb == Py_None) {
        tb = 0;
    } else if (tb && !PyTraceBack_Check(tb)) {
        PyErr_SetString(PyExc_TypeError,
            "raise: arg 3 must be a traceback or None");
        goto bad;
    }
    if (value == Py_None)
        value = 0;
    if (PyExceptionInstance_Check(type)) {
        if (value) {
            PyErr_SetString(PyExc_TypeError,
                "instance exception may not have a separate value");
            goto bad;
        }
        value = type;
        type = (PyObject*) Py_TYPE(value);
    } else if (PyExceptionClass_Check(type)) {
        PyObject *instance_class = NULL;
        if (value && PyExceptionInstance_Check(value)) {
            instance_class = (PyObject*) Py_TYPE(value);
            if (instance_class != type) {
                int is_subclass = PyObject_IsSubclass(instance_class, type);
                if (!is_subclass) {
                    instance_class = NULL;
                } else if (unlikely(is_subclass == -1)) {
                    goto bad;
                } else {
                    type = instance_class;
                }
            }
        }
        if (!instance_class) {
            PyObject *args;
            if (!value)
                args = PyTuple_New(0);
            else if (PyTuple_Check(value)) {
                Py_INCREF(value);
                args = value;
            } else
                args = PyTuple_Pack(1, value);
            if (!args)
                goto bad;
            owned_instance = PyObject_Call(type, args, NULL);
            Py_DECREF(args);
            if (!owned_instance)
                goto bad;
            value = owned_instance;
            if (!PyExceptionInstance_Check(value)) {
                PyErr_Format(PyExc_TypeError,
                             "calling %R should have returned an instance of "
                             "BaseException, not %R",
                             type, Py_TYPE(value));
                goto bad;
            }
        }
    } else {
        PyErr_SetString(PyExc_TypeError,
            "raise: exception class must be a subclass of BaseException");
        goto bad;
    }
    if (cause) {
        PyObject *fixed_cause;
        if (cause == Py_None) {
            fixed_cause = NULL;
        } else if (PyExceptionClass_Check(cause)) {
            fixed_cause = PyObject_CallObject(cause, NULL);
            if (fixed_cause == NULL)
                goto bad;
        } else if (PyExceptionInstance_Check(cause)) {
            fixed_cause = cause;
            Py_INCREF(fixed_cause);
        } else {
            PyErr_SetString(PyExc_TypeError,
                            "exception causes must derive from "
                            "BaseException");
            goto bad;
        }
        PyException_SetCause(value, fixed_cause);
    }
    PyErr_SetObject(type, value);
    if (tb) {
#if CYTHON_FAST_THREAD_STATE
        PyThreadState *tstate = __Pyx_PyThreadState_Current;
        PyObject* tmp_tb = tstate->curexc_traceback;
        if (tb != tmp_tb) {
            Py_INCREF(tb);
            tstate->curexc_traceback = tb;
            Py_XDECREF(tmp_tb);
        }
#else
        PyObject *tmp_type, *tmp_value, *tmp_tb;
        PyErr_Fetch(&tmp_type, &tmp_value, &tmp_tb);
        Py_INCREF(tb);
        PyErr_Restore(tmp_type, tmp_value, tb);
        Py_XDECREF(tmp_tb);
#endif
    }
bad:
    Py_XDECREF(owned_instance);
    return;
}
#endif

/* Import */
static PyObject *__Pyx_Import(PyObject *name, PyObject *from_list, int level) {
    PyObject *empty_list = 0;
    PyObject *module = 0;
    PyObject *global_dict = 0;
    PyObject *empty_dict = 0;
    PyObject *list;
    #if PY_MAJOR_VERSION < 3
    PyObject *py_import;
    py_import = __Pyx_PyObject_GetAttrStr(__pyx_b, __pyx_n_s_import);
    if (!py_import)
        goto bad;
    #endif
    if (from_list)
        list = from_list;
    else {
        empty_list = PyList_New(0);
        if (!empty_list)
            goto bad;
        list = empty_list;
    }
    global_dict = PyModule_GetDict(__pyx_m);
    if (!global_dict)
        goto bad;
    empty_dict = PyDict_New();
    if (!empty_dict)
        goto bad;
    {
        #if PY_MAJOR_VERSION >= 3
        if (level == -1) {
            if ((1) && (strchr(__Pyx_MODULE_NAME, '.'))) {
                module = PyImport_ImportModuleLevelObject(
                    name, global_dict, empty_dict, list, 1);
                if (!module) {
                    if (!PyErr_ExceptionMatches(PyExc_ImportError))
                        goto bad;
                    PyErr_Clear();
                }
            }
            level = 0;
        }
        #endif
        if (!module) {
            #if PY_MAJOR_VERSION < 3
            PyObject *py_level = PyInt_FromLong(level);
            if (!py_level)
                goto bad;
            module = PyObject_CallFunctionObjArgs(py_import,
                name, global_dict, empty_dict, list, py_level, (PyObject *)NULL);
            Py_DECREF(py_level);
            #else
            module = PyImport_ImportModuleLevelObject(
                name, global_dict, empty_dict, list, level);
            #endif
        }
    }
bad:
    #if PY_MAJOR_VERSION < 3
    Py_XDECREF(py_import);
    #endif
    Py_XDECREF(empty_list);
    Py_XDECREF(empty_dict);
    return module;
}

/* ImportFrom */
static PyObject* __Pyx_ImportFrom(PyObject* module, PyObject* name) {
    PyObject* value = __Pyx_PyObject_GetAttrStr(module, name);
    if (unlikely(!value) && PyErr_ExceptionMatches(PyExc_AttributeError)) {
        PyErr_Format(PyExc_ImportError,
        #if PY_MAJOR_VERSION < 3
            "cannot import name %.230s", PyString_AS_STRING(name));
        #else
            "cannot import name %S", name);
        #endif
    }
    return value;
}

/* CLineInTraceback */
#ifndef CYTHON_CLINE_IN_TRACEBACK
static int __Pyx_CLineForTraceback(CYTHON_UNUSED PyThreadState *tstate, int c_line) {
    PyObject *use_cline;
    PyObject *ptype, *pvalue, *ptraceback;
#if CYTHON_COMPILING_IN_CPYTHON
    PyObject **cython_runtime_dict;
#endif
    if (unlikely(!__pyx_cython_runtime)) {
        return c_line;
    }
    __Pyx_ErrFetchInState(tstate, &ptype, &pvalue, &ptraceback);
#if CYTHON_COMPILING_IN_CPYTHON
    cython_runtime_dict = _PyObject_GetDictPtr(__pyx_cython_runtime);
    if (likely(cython_runtime_dict)) {
        __PYX_PY_DICT_LOOKUP_IF_MODIFIED(
            use_cline, *cython_runtime_dict,
            __Pyx_PyDict_GetItemStr(*cython_runtime_dict, __pyx_n_s_cline_in_traceback))
    } else
#endif
    {
      PyObject *use_cline_obj = __Pyx_PyObject_GetAttrStr(__pyx_cython_runtime, __pyx_n_s_cline_in_traceback);
      if (use_cline_obj) {
        use_cline = PyObject_Not(use_cline_obj) ? Py_False : Py_True;
        Py_DECREF(use_cline_obj);
      } else {
        PyErr_Clear();
        use_cline = NULL;
      }
    }
    if (!use_cline) {
        c_line = 0;
        (void) PyObject_SetAttr(__pyx_cython_runtime, __pyx_n_s_cline_in_traceback, Py_False);
    }
    else if (use_cline == Py_False || (use_cline != Py_True && PyObject_Not(use_cline) != 0)) {
        c_line = 0;
    }
    __Pyx_ErrRestoreInState(tstate, ptype, pvalue, ptraceback);
    return c_line;
}
#endif

/* CodeObjectCache */
static int __pyx_bisect_code_objects(__Pyx_CodeObjectCacheEntry* entries, int count, int code_line) {
    int start = 0, mid = 0, end = count - 1;
    if (end >= 0 && code_line > entries[end].code_line) {
        return count;
    }
    while (start < end) {
        mid = start + (end - start) / 2;
        if (code_line < entries[mid].code_line) {
            end = mid;
        } else if (code_line > entries[mid].code_line) {
             start = mid + 1;
        } else {
            return mid;
        }
    }
    if (code_line <= entries[mid].code_line) {
        return mid;
    } else {
        return mid + 1;
    }
}
static PyCodeObject *__pyx_find_code_object(int code_line) {
    PyCodeObject* code_object;
    int pos;
    if (unlikely(!code_line) || unlikely(!__pyx_code_cache.entries)) {
        return NULL;
    }
    pos = __pyx_bisect_code_objects(__pyx_code_cache.entries, __pyx_code_cache.count, code_line);
    if (unlikely(pos >= __pyx_code_cache.count) || unlikely(__pyx_code_cache.entries[pos].code_line != code_line)) {
        return NULL;
    }
    code_object = __pyx_code_cache.entries[pos].code_object;
    Py_INCREF(code_object);
    return code_object;
}
static void __pyx_insert_code_object(int code_line, PyCodeObject* code_object) {
    int pos, i;
    __Pyx_CodeObjectCacheEntry* entries = __pyx_code_cache.entries;
    if (unlikely(!code_line)) {
        return;
    }
    if (unlikely(!entries)) {
        entries = (__Pyx_CodeObjectCacheEntry*)PyMem_Malloc(64*sizeof(__Pyx_CodeObjectCacheEntry));
        if (likely(entries)) {
            __pyx_code_cache.entries = entries;
            __pyx_code_cache.max_count = 64;
            __pyx_code_cache.count = 1;
            entries[0].code_line = code_line;
            entries[0].code_object = code_object;
            Py_INCREF(code_object);
        }
        return;
    }
    pos = __pyx_bisect_code_objects(__pyx_code_cache.entries, __pyx_code_cache.count, code_line);
    if ((pos < __pyx_code_cache.count) && unlikely(__pyx_code_cache.entries[pos].code_line == code_line)) {
        PyCodeObject* tmp = entries[pos].code_object;
        entries[pos].code_object = code_object;
        Py_DECREF(tmp);
        return;
    }
    if (__pyx_code_cache.count == __pyx_code_cache.max_count) {
        int new_max = __pyx_code_cache.max_count + 64;
        entries = (__Pyx_CodeObjectCacheEntry*)PyMem_Realloc(
            __pyx_code_cache.entries, ((size_t)new_max) * sizeof(__Pyx_CodeObjectCacheEntry));
        if (unlikely(!entries)) {
            return;
        }
        __pyx_code_cache.entries = entries;
        __pyx_code_cache.max_count = new_max;
    }
    for (i=__pyx_code_cache.count; i>pos; i--) {
        entries[i] = entries[i-1];
    }
    entries[pos].code_line = code_line;
    entries[pos].code_object = code_object;
    __pyx_code_cache.count++;
    Py_INCREF(code_object);
}

/* AddTraceback */
#include "compile.h"
#include "frameobject.h"
#include "traceback.h"
#if PY_VERSION_HEX >= 0x030b00a6
  #ifndef Py_BUILD_CORE
    #define Py_BUILD_CORE 1
  #endif
  #include "internal/pycore_frame.h"
#endif
static PyCodeObject* __Pyx_CreateCodeObjectForTraceback(
            const char *funcname, int c_line,
            int py_line, const char *filename) {
    PyCodeObject *py_code = NULL;
    PyObject *py_funcname = NULL;
    #if PY_MAJOR_VERSION < 3
    PyObject *py_srcfile = NULL;
    py_srcfile = PyString_FromString(filename);
    if (!py_srcfile) goto bad;
    #endif
    if (c_line) {
        #if PY_MAJOR_VERSION < 3
        py_funcname = PyString_FromFormat( "%s (%s:%d)", funcname, __pyx_cfilenm, c_line);
        if (!py_funcname) goto bad;
        #else
        py_funcname = PyUnicode_FromFormat( "%s (%s:%d)", funcname, __pyx_cfilenm, c_line);
        if (!py_funcname) goto bad;
        funcname = PyUnicode_AsUTF8(py_funcname);
        if (!funcname) goto bad;
        #endif
    }
    else {
        #if PY_MAJOR_VERSION < 3
        py_funcname = PyString_FromString(funcname);
        if (!py_funcname) goto bad;
        #endif
    }
    #if PY_MAJOR_VERSION < 3
    py_code = __Pyx_PyCode_New(
        0,
        0,
        0,
        0,
        0,
        __pyx_empty_bytes, /*PyObject *code,*/
        __pyx_empty_tuple, /*PyObject *consts,*/
        __pyx_empty_tuple, /*PyObject *names,*/
        __pyx_empty_tuple, /*PyObject *varnames,*/
        __pyx_empty_tuple, /*PyObject *freevars,*/
        __pyx_empty_tuple, /*PyObject *cellvars,*/
        py_srcfile,   /*PyObject *filename,*/
        py_funcname,  /*PyObject *name,*/
        py_line,
        __pyx_empty_bytes  /*PyObject *lnotab*/
    );
    Py_DECREF(py_srcfile);
    #else
    py_code = PyCode_NewEmpty(filename, funcname, py_line);
    #endif
    Py_XDECREF(py_funcname);  // XDECREF since it's only set on Py3 if cline
    return py_code;
bad:
    Py_XDECREF(py_funcname);
    #if PY_MAJOR_VERSION < 3
    Py_XDECREF(py_srcfile);
    #endif
    return NULL;
}
static void __Pyx_AddTraceback(const char *funcname, int c_line,
                               int py_line, const char *filename) {
    PyCodeObject *py_code = 0;
    PyFrameObject *py_frame = 0;
    PyThreadState *tstate = __Pyx_PyThreadState_Current;
    PyObject *ptype, *pvalue, *ptraceback;
    if (c_line) {
        c_line = __Pyx_CLineForTraceback(tstate, c_line);
    }
    py_code = __pyx_find_code_object(c_line ? -c_line : py_line);
    if (!py_code) {
        __Pyx_ErrFetchInState(tstate, &ptype, &pvalue, &ptraceback);
        py_code = __Pyx_CreateCodeObjectForTraceback(
            funcname, c_line, py_line, filename);
        if (!py_code) {
            /* If the code object creation fails, then we should clear the
               fetched exception references and propagate the new exception */
            Py_XDECREF(ptype);
            Py_XDECREF(pvalue);
            Py_XDECREF(ptraceback);
            goto bad;
        }
        __Pyx_ErrRestoreInState(tstate, ptype, pvalue, ptraceback);
        __pyx_insert_code_object(c_line ? -c_line : py_line, py_code);
    }
    py_frame = PyFrame_New(
        tstate,            /*PyThreadState *tstate,*/
        py_code,           /*PyCodeObject *code,*/
        __pyx_d,    /*PyObject *globals,*/
        0                  /*PyObject *locals*/
    );
    if (!py_frame) goto bad;
    __Pyx_PyFrame_SetLineNumber(py_frame, py_line);
    PyTraceBack_Here(py_frame);
bad:
    Py_XDECREF(py_code);
    Py_XDECREF(py_frame);
}

/* CIntFromPyVerify */
#define __PYX_VERIFY_RETURN_INT(target_type, func_type, func_value)\
    __PYX__VERIFY_RETURN_INT(target_type, func_type, func_value, 0)
#define __PYX_VERIFY_RETURN_INT_EXC(target_type, func_type, func_value)\
    __PYX__VERIFY_RETURN_INT(target_type, func_type, func_value, 1)
#define __PYX__VERIFY_RETURN_INT(target_type, func_type, func_value, exc)\
    {\
        func_type value = func_value;\
        if (sizeof(target_type) < sizeof(func_type)) {\
            if (unlikely(value != (func_type) (target_type) value)) {\
                func_type zero = 0;\
                if (exc && unlikely(value == (func_type)-1 && PyErr_Occurred()))\
                    return (target_type) -1;\
                if (is_unsigned && unlikely(value < zero))\
                    goto raise_neg_overflow;\
                else\
                    goto raise_overflow;\
            }\
        }\
        return (target_type) value;\
    }

/* CIntToPy */
static CYTHON_INLINE PyObject* __Pyx_PyInt_From_long(long value) {
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wconversion"
#endif
    const long neg_one = (long) -1, const_zero = (long) 0;
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic pop
#endif
    const int is_unsigned = neg_one > const_zero;
    if (is_unsigned) {
        if (sizeof(long) < sizeof(long)) {
            return PyInt_FromLong((long) value);
        } else if (sizeof(long) <= sizeof(unsigned long)) {
            return PyLong_FromUnsignedLong((unsigned long) value);
#ifdef HAVE_LONG_LONG
        } else if (sizeof(long) <= sizeof(unsigned PY_LONG_LONG)) {
            return PyLong_FromUnsignedLongLong((unsigned PY_LONG_LONG) value);
#endif
        }
    } else {
        if (sizeof(long) <= sizeof(long)) {
            return PyInt_FromLong((long) value);
#ifdef HAVE_LONG_LONG
        } else if (sizeof(long) <= sizeof(PY_LONG_LONG)) {
            return PyLong_FromLongLong((PY_LONG_LONG) value);
#endif
        }
    }
    {
        int one = 1; int little = (int)*(unsigned char *)&one;
        unsigned char *bytes = (unsigned char *)&value;
        return _PyLong_FromByteArray(bytes, sizeof(long),
                                     little, !is_unsigned);
    }
}

/* CIntFromPy */
static CYTHON_INLINE unsigned PY_LONG_LONG __Pyx_PyInt_As_unsigned_PY_LONG_LONG(PyObject *x) {
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wconversion"
#endif
    const unsigned PY_LONG_LONG neg_one = (unsigned PY_LONG_LONG) -1, const_zero = (unsigned PY_LONG_LONG) 0;
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic pop
#endif
    const int is_unsigned = neg_one > const_zero;
#if PY_MAJOR_VERSION < 3
    if (likely(PyInt_Check(x))) {
        if (sizeof(unsigned PY_LONG_LONG) < sizeof(long)) {
            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, long, PyInt_AS_LONG(x))
        } else {
            long val = PyInt_AS_LONG(x);
            if (is_unsigned && unlikely(val < 0)) {
                goto raise_neg_overflow;
            }
            return (unsigned PY_LONG_LONG) val;
        }
    } else
#endif
    if (likely(PyLong_Check(x))) {
        if (is_unsigned) {
#if CYTHON_USE_PYLONG_INTERNALS
            const digit* digits = ((PyLongObject*)x)->ob_digit;
            switch (Py_SIZE(x)) {
                case  0: return (unsigned PY_LONG_LONG) 0;
                case  1: __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, digit, digits[0])
                case 2:
                    if (8 * sizeof(unsigned PY_LONG_LONG) > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, unsigned long, (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) >= 2 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) (((((unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0]));
                        }
                    }
                    break;
                case 3:
                    if (8 * sizeof(unsigned PY_LONG_LONG) > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, unsigned long, (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) >= 3 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) (((((((unsigned PY_LONG_LONG)digits[2]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0]));
                        }
                    }
                    break;
                case 4:
                    if (8 * sizeof(unsigned PY_LONG_LONG) > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, unsigned long, (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) >= 4 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) (((((((((unsigned PY_LONG_LONG)digits[3]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[2]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0]));
                        }
                    }
                    break;
            }
#endif
#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX < 0x030C00A7
            if (unlikely(Py_SIZE(x) < 0)) {
                goto raise_neg_overflow;
            }
#else
            {
                int result = PyObject_RichCompareBool(x, Py_False, Py_LT);
                if (unlikely(result < 0))
                    return (unsigned PY_LONG_LONG) -1;
                if (unlikely(result == 1))
                    goto raise_neg_overflow;
            }
#endif
            if (sizeof(unsigned PY_LONG_LONG) <= sizeof(unsigned long)) {
                __PYX_VERIFY_RETURN_INT_EXC(unsigned PY_LONG_LONG, unsigned long, PyLong_AsUnsignedLong(x))
#ifdef HAVE_LONG_LONG
            } else if (sizeof(unsigned PY_LONG_LONG) <= sizeof(unsigned PY_LONG_LONG)) {
                __PYX_VERIFY_RETURN_INT_EXC(unsigned PY_LONG_LONG, unsigned PY_LONG_LONG, PyLong_AsUnsignedLongLong(x))
#endif
            }
        } else {
#if CYTHON_USE_PYLONG_INTERNALS
            const digit* digits = ((PyLongObject*)x)->ob_digit;
            switch (Py_SIZE(x)) {
                case  0: return (unsigned PY_LONG_LONG) 0;
                case -1: __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, sdigit, (sdigit) (-(sdigit)digits[0]))
                case  1: __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG,  digit, +digits[0])
                case -2:
                    if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, long, -(long) (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 2 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) (((unsigned PY_LONG_LONG)-1)*(((((unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0])));
                        }
                    }
                    break;
                case 2:
                    if (8 * sizeof(unsigned PY_LONG_LONG) > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, unsigned long, (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 2 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) ((((((unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0])));
                        }
                    }
                    break;
                case -3:
                    if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, long, -(long) (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 3 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) (((unsigned PY_LONG_LONG)-1)*(((((((unsigned PY_LONG_LONG)digits[2]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0])));
                        }
                    }
                    break;
                case 3:
                    if (8 * sizeof(unsigned PY_LONG_LONG) > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, unsigned long, (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 3 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) ((((((((unsigned PY_LONG_LONG)digits[2]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0])));
                        }
                    }
                    break;
                case -4:
                    if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, long, -(long) (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 4 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) (((unsigned PY_LONG_LONG)-1)*(((((((((unsigned PY_LONG_LONG)digits[3]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[2]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0])));
                        }
                    }
                    break;
                case 4:
                    if (8 * sizeof(unsigned PY_LONG_LONG) > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(unsigned PY_LONG_LONG, unsigned long, (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(unsigned PY_LONG_LONG) - 1 > 4 * PyLong_SHIFT) {
                            return (unsigned PY_LONG_LONG) ((((((((((unsigned PY_LONG_LONG)digits[3]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[2]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[1]) << PyLong_SHIFT) | (unsigned PY_LONG_LONG)digits[0])));
                        }
                    }
                    break;
            }
#endif
            if (sizeof(unsigned PY_LONG_LONG) <= sizeof(long)) {
                __PYX_VERIFY_RETURN_INT_EXC(unsigned PY_LONG_LONG, long, PyLong_AsLong(x))
#ifdef HAVE_LONG_LONG
            } else if (sizeof(unsigned PY_LONG_LONG) <= sizeof(PY_LONG_LONG)) {
                __PYX_VERIFY_RETURN_INT_EXC(unsigned PY_LONG_LONG, PY_LONG_LONG, PyLong_AsLongLong(x))
#endif
            }
        }
        {
#if CYTHON_COMPILING_IN_PYPY && !defined(_PyLong_AsByteArray)
            PyErr_SetString(PyExc_RuntimeError,
                            "_PyLong_AsByteArray() not available in PyPy, cannot convert large numbers");
#else
            unsigned PY_LONG_LONG val;
            PyObject *v = __Pyx_PyNumber_IntOrLong(x);
 #if PY_MAJOR_VERSION < 3
            if (likely(v) && !PyLong_Check(v)) {
                PyObject *tmp = v;
                v = PyNumber_Long(tmp);
                Py_DECREF(tmp);
            }
 #endif
            if (likely(v)) {
                int one = 1; int is_little = (int)*(unsigned char *)&one;
                unsigned char *bytes = (unsigned char *)&val;
                int ret = _PyLong_AsByteArray((PyLongObject *)v,
                                              bytes, sizeof(val),
                                              is_little, !is_unsigned);
                Py_DECREF(v);
                if (likely(!ret))
                    return val;
            }
#endif
            return (unsigned PY_LONG_LONG) -1;
        }
    } else {
        unsigned PY_LONG_LONG val;
        PyObject *tmp = __Pyx_PyNumber_IntOrLong(x);
        if (!tmp) return (unsigned PY_LONG_LONG) -1;
        val = __Pyx_PyInt_As_unsigned_PY_LONG_LONG(tmp);
        Py_DECREF(tmp);
        return val;
    }
raise_overflow:
    PyErr_SetString(PyExc_OverflowError,
        "value too large to convert to unsigned PY_LONG_LONG");
    return (unsigned PY_LONG_LONG) -1;
raise_neg_overflow:
    PyErr_SetString(PyExc_OverflowError,
        "can't convert negative value to unsigned PY_LONG_LONG");
    return (unsigned PY_LONG_LONG) -1;
}

/* CIntFromPy */
static CYTHON_INLINE long __Pyx_PyInt_As_long(PyObject *x) {
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wconversion"
#endif
    const long neg_one = (long) -1, const_zero = (long) 0;
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic pop
#endif
    const int is_unsigned = neg_one > const_zero;
#if PY_MAJOR_VERSION < 3
    if (likely(PyInt_Check(x))) {
        if (sizeof(long) < sizeof(long)) {
            __PYX_VERIFY_RETURN_INT(long, long, PyInt_AS_LONG(x))
        } else {
            long val = PyInt_AS_LONG(x);
            if (is_unsigned && unlikely(val < 0)) {
                goto raise_neg_overflow;
            }
            return (long) val;
        }
    } else
#endif
    if (likely(PyLong_Check(x))) {
        if (is_unsigned) {
#if CYTHON_USE_PYLONG_INTERNALS
            const digit* digits = ((PyLongObject*)x)->ob_digit;
            switch (Py_SIZE(x)) {
                case  0: return (long) 0;
                case  1: __PYX_VERIFY_RETURN_INT(long, digit, digits[0])
                case 2:
                    if (8 * sizeof(long) > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, unsigned long, (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) >= 2 * PyLong_SHIFT) {
                            return (long) (((((long)digits[1]) << PyLong_SHIFT) | (long)digits[0]));
                        }
                    }
                    break;
                case 3:
                    if (8 * sizeof(long) > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, unsigned long, (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) >= 3 * PyLong_SHIFT) {
                            return (long) (((((((long)digits[2]) << PyLong_SHIFT) | (long)digits[1]) << PyLong_SHIFT) | (long)digits[0]));
                        }
                    }
                    break;
                case 4:
                    if (8 * sizeof(long) > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, unsigned long, (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) >= 4 * PyLong_SHIFT) {
                            return (long) (((((((((long)digits[3]) << PyLong_SHIFT) | (long)digits[2]) << PyLong_SHIFT) | (long)digits[1]) << PyLong_SHIFT) | (long)digits[0]));
                        }
                    }
                    break;
            }
#endif
#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX < 0x030C00A7
            if (unlikely(Py_SIZE(x) < 0)) {
                goto raise_neg_overflow;
            }
#else
            {
                int result = PyObject_RichCompareBool(x, Py_False, Py_LT);
                if (unlikely(result < 0))
                    return (long) -1;
                if (unlikely(result == 1))
                    goto raise_neg_overflow;
            }
#endif
            if (sizeof(long) <= sizeof(unsigned long)) {
                __PYX_VERIFY_RETURN_INT_EXC(long, unsigned long, PyLong_AsUnsignedLong(x))
#ifdef HAVE_LONG_LONG
            } else if (sizeof(long) <= sizeof(unsigned PY_LONG_LONG)) {
                __PYX_VERIFY_RETURN_INT_EXC(long, unsigned PY_LONG_LONG, PyLong_AsUnsignedLongLong(x))
#endif
            }
        } else {
#if CYTHON_USE_PYLONG_INTERNALS
            const digit* digits = ((PyLongObject*)x)->ob_digit;
            switch (Py_SIZE(x)) {
                case  0: return (long) 0;
                case -1: __PYX_VERIFY_RETURN_INT(long, sdigit, (sdigit) (-(sdigit)digits[0]))
                case  1: __PYX_VERIFY_RETURN_INT(long,  digit, +digits[0])
                case -2:
                    if (8 * sizeof(long) - 1 > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, long, -(long) (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) - 1 > 2 * PyLong_SHIFT) {
                            return (long) (((long)-1)*(((((long)digits[1]) << PyLong_SHIFT) | (long)digits[0])));
                        }
                    }
                    break;
                case 2:
                    if (8 * sizeof(long) > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, unsigned long, (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) - 1 > 2 * PyLong_SHIFT) {
                            return (long) ((((((long)digits[1]) << PyLong_SHIFT) | (long)digits[0])));
                        }
                    }
                    break;
                case -3:
                    if (8 * sizeof(long) - 1 > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, long, -(long) (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) - 1 > 3 * PyLong_SHIFT) {
                            return (long) (((long)-1)*(((((((long)digits[2]) << PyLong_SHIFT) | (long)digits[1]) << PyLong_SHIFT) | (long)digits[0])));
                        }
                    }
                    break;
                case 3:
                    if (8 * sizeof(long) > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, unsigned long, (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) - 1 > 3 * PyLong_SHIFT) {
                            return (long) ((((((((long)digits[2]) << PyLong_SHIFT) | (long)digits[1]) << PyLong_SHIFT) | (long)digits[0])));
                        }
                    }
                    break;
                case -4:
                    if (8 * sizeof(long) - 1 > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, long, -(long) (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) - 1 > 4 * PyLong_SHIFT) {
                            return (long) (((long)-1)*(((((((((long)digits[3]) << PyLong_SHIFT) | (long)digits[2]) << PyLong_SHIFT) | (long)digits[1]) << PyLong_SHIFT) | (long)digits[0])));
                        }
                    }
                    break;
                case 4:
                    if (8 * sizeof(long) > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(long, unsigned long, (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(long) - 1 > 4 * PyLong_SHIFT) {
                            return (long) ((((((((((long)digits[3]) << PyLong_SHIFT) | (long)digits[2]) << PyLong_SHIFT) | (long)digits[1]) << PyLong_SHIFT) | (long)digits[0])));
                        }
                    }
                    break;
            }
#endif
            if (sizeof(long) <= sizeof(long)) {
                __PYX_VERIFY_RETURN_INT_EXC(long, long, PyLong_AsLong(x))
#ifdef HAVE_LONG_LONG
            } else if (sizeof(long) <= sizeof(PY_LONG_LONG)) {
                __PYX_VERIFY_RETURN_INT_EXC(long, PY_LONG_LONG, PyLong_AsLongLong(x))
#endif
            }
        }
        {
#if CYTHON_COMPILING_IN_PYPY && !defined(_PyLong_AsByteArray)
            PyErr_SetString(PyExc_RuntimeError,
                            "_PyLong_AsByteArray() not available in PyPy, cannot convert large numbers");
#else
            long val;
            PyObject *v = __Pyx_PyNumber_IntOrLong(x);
 #if PY_MAJOR_VERSION < 3
            if (likely(v) && !PyLong_Check(v)) {
                PyObject *tmp = v;
                v = PyNumber_Long(tmp);
                Py_DECREF(tmp);
            }
 #endif
            if (likely(v)) {
                int one = 1; int is_little = (int)*(unsigned char *)&one;
                unsigned char *bytes = (unsigned char *)&val;
                int ret = _PyLong_AsByteArray((PyLongObject *)v,
                                              bytes, sizeof(val),
                                              is_little, !is_unsigned);
                Py_DECREF(v);
                if (likely(!ret))
                    return val;
            }
#endif
            return (long) -1;
        }
    } else {
        long val;
        PyObject *tmp = __Pyx_PyNumber_IntOrLong(x);
        if (!tmp) return (long) -1;
        val = __Pyx_PyInt_As_long(tmp);
        Py_DECREF(tmp);
        return val;
    }
raise_overflow:
    PyErr_SetString(PyExc_OverflowError,
        "value too large to convert to long");
    return (long) -1;
raise_neg_overflow:
    PyErr_SetString(PyExc_OverflowError,
        "can't convert negative value to long");
    return (long) -1;
}

/* CIntFromPy */
static CYTHON_INLINE int __Pyx_PyInt_As_int(PyObject *x) {
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wconversion"
#endif
    const int neg_one = (int) -1, const_zero = (int) 0;
#ifdef __Pyx_HAS_GCC_DIAGNOSTIC
#pragma GCC diagnostic pop
#endif
    const int is_unsigned = neg_one > const_zero;
#if PY_MAJOR_VERSION < 3
    if (likely(PyInt_Check(x))) {
        if (sizeof(int) < sizeof(long)) {
            __PYX_VERIFY_RETURN_INT(int, long, PyInt_AS_LONG(x))
        } else {
            long val = PyInt_AS_LONG(x);
            if (is_unsigned && unlikely(val < 0)) {
                goto raise_neg_overflow;
            }
            return (int) val;
        }
    } else
#endif
    if (likely(PyLong_Check(x))) {
        if (is_unsigned) {
#if CYTHON_USE_PYLONG_INTERNALS
            const digit* digits = ((PyLongObject*)x)->ob_digit;
            switch (Py_SIZE(x)) {
                case  0: return (int) 0;
                case  1: __PYX_VERIFY_RETURN_INT(int, digit, digits[0])
                case 2:
                    if (8 * sizeof(int) > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, unsigned long, (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) >= 2 * PyLong_SHIFT) {
                            return (int) (((((int)digits[1]) << PyLong_SHIFT) | (int)digits[0]));
                        }
                    }
                    break;
                case 3:
                    if (8 * sizeof(int) > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, unsigned long, (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) >= 3 * PyLong_SHIFT) {
                            return (int) (((((((int)digits[2]) << PyLong_SHIFT) | (int)digits[1]) << PyLong_SHIFT) | (int)digits[0]));
                        }
                    }
                    break;
                case 4:
                    if (8 * sizeof(int) > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, unsigned long, (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) >= 4 * PyLong_SHIFT) {
                            return (int) (((((((((int)digits[3]) << PyLong_SHIFT) | (int)digits[2]) << PyLong_SHIFT) | (int)digits[1]) << PyLong_SHIFT) | (int)digits[0]));
                        }
                    }
                    break;
            }
#endif
#if CYTHON_COMPILING_IN_CPYTHON && PY_VERSION_HEX < 0x030C00A7
            if (unlikely(Py_SIZE(x) < 0)) {
                goto raise_neg_overflow;
            }
#else
            {
                int result = PyObject_RichCompareBool(x, Py_False, Py_LT);
                if (unlikely(result < 0))
                    return (int) -1;
                if (unlikely(result == 1))
                    goto raise_neg_overflow;
            }
#endif
            if (sizeof(int) <= sizeof(unsigned long)) {
                __PYX_VERIFY_RETURN_INT_EXC(int, unsigned long, PyLong_AsUnsignedLong(x))
#ifdef HAVE_LONG_LONG
            } else if (sizeof(int) <= sizeof(unsigned PY_LONG_LONG)) {
                __PYX_VERIFY_RETURN_INT_EXC(int, unsigned PY_LONG_LONG, PyLong_AsUnsignedLongLong(x))
#endif
            }
        } else {
#if CYTHON_USE_PYLONG_INTERNALS
            const digit* digits = ((PyLongObject*)x)->ob_digit;
            switch (Py_SIZE(x)) {
                case  0: return (int) 0;
                case -1: __PYX_VERIFY_RETURN_INT(int, sdigit, (sdigit) (-(sdigit)digits[0]))
                case  1: __PYX_VERIFY_RETURN_INT(int,  digit, +digits[0])
                case -2:
                    if (8 * sizeof(int) - 1 > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, long, -(long) (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) - 1 > 2 * PyLong_SHIFT) {
                            return (int) (((int)-1)*(((((int)digits[1]) << PyLong_SHIFT) | (int)digits[0])));
                        }
                    }
                    break;
                case 2:
                    if (8 * sizeof(int) > 1 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 2 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, unsigned long, (((((unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) - 1 > 2 * PyLong_SHIFT) {
                            return (int) ((((((int)digits[1]) << PyLong_SHIFT) | (int)digits[0])));
                        }
                    }
                    break;
                case -3:
                    if (8 * sizeof(int) - 1 > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, long, -(long) (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) - 1 > 3 * PyLong_SHIFT) {
                            return (int) (((int)-1)*(((((((int)digits[2]) << PyLong_SHIFT) | (int)digits[1]) << PyLong_SHIFT) | (int)digits[0])));
                        }
                    }
                    break;
                case 3:
                    if (8 * sizeof(int) > 2 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 3 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, unsigned long, (((((((unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) - 1 > 3 * PyLong_SHIFT) {
                            return (int) ((((((((int)digits[2]) << PyLong_SHIFT) | (int)digits[1]) << PyLong_SHIFT) | (int)digits[0])));
                        }
                    }
                    break;
                case -4:
                    if (8 * sizeof(int) - 1 > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, long, -(long) (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) - 1 > 4 * PyLong_SHIFT) {
                            return (int) (((int)-1)*(((((((((int)digits[3]) << PyLong_SHIFT) | (int)digits[2]) << PyLong_SHIFT) | (int)digits[1]) << PyLong_SHIFT) | (int)digits[0])));
                        }
                    }
                    break;
                case 4:
                    if (8 * sizeof(int) > 3 * PyLong_SHIFT) {
                        if (8 * sizeof(unsigned long) > 4 * PyLong_SHIFT) {
                            __PYX_VERIFY_RETURN_INT(int, unsigned long, (((((((((unsigned long)digits[3]) << PyLong_SHIFT) | (unsigned long)digits[2]) << PyLong_SHIFT) | (unsigned long)digits[1]) << PyLong_SHIFT) | (unsigned long)digits[0])))
                        } else if (8 * sizeof(int) - 1 > 4 * PyLong_SHIFT) {
                            return (int) ((((((((((int)digits[3]) << PyLong_SHIFT) | (int)digits[2]) << PyLong_SHIFT) | (int)digits[1]) << PyLong_SHIFT) | (int)digits[0])));
                        }
                    }
                    break;
            }
#endif
            if (sizeof(int) <= sizeof(long)) {
                __PYX_VERIFY_RETURN_INT_EXC(int, long, PyLong_AsLong(x))
#ifdef HAVE_LONG_LONG
            } else if (sizeof(int) <= sizeof(PY_LONG_LONG)) {
                __PYX_VERIFY_RETURN_INT_EXC(int, PY_LONG_LONG, PyLong_AsLongLong(x))
#endif
            }
        }
        {
#if CYTHON_COMPILING_IN_PYPY && !defined(_PyLong_AsByteArray)
            PyErr_SetString(PyExc_RuntimeError,
                            "_PyLong_AsByteArray() not available in PyPy, cannot convert large numbers");
#else
            int val;
            PyObject *v = __Pyx_PyNumber_IntOrLong(x);
 #if PY_MAJOR_VERSION < 3
            if (likely(v) && !PyLong_Check(v)) {
                PyObject *tmp = v;
                v = PyNumber_Long(tmp);
                Py_DECREF(tmp);
            }
 #endif
            if (likely(v)) {
                int one = 1; int is_little = (int)*(unsigned char *)&one;
                unsigned char *bytes = (unsigned char *)&val;
                int ret = _PyLong_AsByteArray((PyLongObject *)v,
                                              bytes, sizeof(val),
                                              is_little, !is_unsigned);
                Py_DECREF(v);
                if (likely(!ret))
                    return val;
            }
#endif
            return (int) -1;
        }
    } else {
        int val;
        PyObject *tmp = __Pyx_PyNumber_IntOrLong(x);
        if (!tmp) return (int) -1;
        val = __Pyx_PyInt_As_int(tmp);
        Py_DECREF(tmp);
        return val;
    }
raise_overflow:
    PyErr_SetString(PyExc_OverflowError,
        "value too large to convert to int");
    return (int) -1;
raise_neg_overflow:
    PyErr_SetString(PyExc_OverflowError,
        "can't convert negative value to int");
    return (int) -1;
}

/* FastTypeChecks */
#if CYTHON_COMPILING_IN_CPYTHON
static int __Pyx_InBases(PyTypeObject *a, PyTypeObject *b) {
    while (a) {
        a = a->tp_base;
        if (a == b)
            return 1;
    }
    return b == &PyBaseObject_Type;
}
static CYTHON_INLINE int __Pyx_IsSubtype(PyTypeObject *a, PyTypeObject *b) {
    PyObject *mro;
    if (a == b) return 1;
    mro = a->tp_mro;
    if (likely(mro)) {
        Py_ssize_t i, n;
        n = PyTuple_GET_SIZE(mro);
        for (i = 0; i < n; i++) {
            if (PyTuple_GET_ITEM(mro, i) == (PyObject *)b)
                return 1;
        }
        return 0;
    }
    return __Pyx_InBases(a, b);
}
#if PY_MAJOR_VERSION == 2
static int __Pyx_inner_PyErr_GivenExceptionMatches2(PyObject *err, PyObject* exc_type1, PyObject* exc_type2) {
    PyObject *exception, *value, *tb;
    int res;
    __Pyx_PyThreadState_declare
    __Pyx_PyThreadState_assign
    __Pyx_ErrFetch(&exception, &value, &tb);
    res = exc_type1 ? PyObject_IsSubclass(err, exc_type1) : 0;
    if (unlikely(res == -1)) {
        PyErr_WriteUnraisable(err);
        res = 0;
    }
    if (!res) {
        res = PyObject_IsSubclass(err, exc_type2);
        if (unlikely(res == -1)) {
            PyErr_WriteUnraisable(err);
            res = 0;
        }
    }
    __Pyx_ErrRestore(exception, value, tb);
    return res;
}
#else
static CYTHON_INLINE int __Pyx_inner_PyErr_GivenExceptionMatches2(PyObject *err, PyObject* exc_type1, PyObject *exc_type2) {
    int res = exc_type1 ? __Pyx_IsSubtype((PyTypeObject*)err, (PyTypeObject*)exc_type1) : 0;
    if (!res) {
        res = __Pyx_IsSubtype((PyTypeObject*)err, (PyTypeObject*)exc_type2);
    }
    return res;
}
#endif
static int __Pyx_PyErr_GivenExceptionMatchesTuple(PyObject *exc_type, PyObject *tuple) {
    Py_ssize_t i, n;
    assert(PyExceptionClass_Check(exc_type));
    n = PyTuple_GET_SIZE(tuple);
#if PY_MAJOR_VERSION >= 3
    for (i=0; i<n; i++) {
        if (exc_type == PyTuple_GET_ITEM(tuple, i)) return 1;
    }
#endif
    for (i=0; i<n; i++) {
        PyObject *t = PyTuple_GET_ITEM(tuple, i);
        #if PY_MAJOR_VERSION < 3
        if (likely(exc_type == t)) return 1;
        #endif
        if (likely(PyExceptionClass_Check(t))) {
            if (__Pyx_inner_PyErr_GivenExceptionMatches2(exc_type, NULL, t)) return 1;
        } else {
        }
    }
    return 0;
}
static CYTHON_INLINE int __Pyx_PyErr_GivenExceptionMatches(PyObject *err, PyObject* exc_type) {
    if (likely(err == exc_type)) return 1;
    if (likely(PyExceptionClass_Check(err))) {
        if (likely(PyExceptionClass_Check(exc_type))) {
            return __Pyx_inner_PyErr_GivenExceptionMatches2(err, NULL, exc_type);
        } else if (likely(PyTuple_Check(exc_type))) {
            return __Pyx_PyErr_GivenExceptionMatchesTuple(err, exc_type);
        } else {
        }
    }
    return PyErr_GivenExceptionMatches(err, exc_type);
}
static CYTHON_INLINE int __Pyx_PyErr_GivenExceptionMatches2(PyObject *err, PyObject *exc_type1, PyObject *exc_type2) {
    assert(PyExceptionClass_Check(exc_type1));
    assert(PyExceptionClass_Check(exc_type2));
    if (likely(err == exc_type1 || err == exc_type2)) return 1;
    if (likely(PyExceptionClass_Check(err))) {
        return __Pyx_inner_PyErr_GivenExceptionMatches2(err, exc_type1, exc_type2);
    }
    return (PyErr_GivenExceptionMatches(err, exc_type1) || PyErr_GivenExceptionMatches(err, exc_type2));
}
#endif

/* CheckBinaryVersion */
static int __Pyx_check_binary_version(void) {
    char ctversion[5];
    int same=1, i, found_dot;
    const char* rt_from_call = Py_GetVersion();
    PyOS_snprintf(ctversion, 5, "%d.%d", PY_MAJOR_VERSION, PY_MINOR_VERSION);
    found_dot = 0;
    for (i = 0; i < 4; i++) {
        if (!ctversion[i]) {
            same = (rt_from_call[i] < '0' || rt_from_call[i] > '9');
            break;
        }
        if (rt_from_call[i] != ctversion[i]) {
            same = 0;
            break;
        }
    }
    if (!same) {
        char rtversion[5] = {'\0'};
        char message[200];
        for (i=0; i<4; ++i) {
            if (rt_from_call[i] == '.') {
                if (found_dot) break;
                found_dot = 1;
            } else if (rt_from_call[i] < '0' || rt_from_call[i] > '9') {
                break;
            }
            rtversion[i] = rt_from_call[i];
        }
        PyOS_snprintf(message, sizeof(message),
                      "compiletime version %s of module '%.100s' "
                      "does not match runtime version %s",
                      ctversion, __Pyx_MODULE_NAME, rtversion);
        return PyErr_WarnEx(NULL, message, 1);
    }
    return 0;
}

/* FunctionExport */
static int __Pyx_ExportFunction(const char *name, void (*f)(void), const char *sig) {
    PyObject *d = 0;
    PyObject *cobj = 0;
    union {
        void (*fp)(void);
        void *p;
    } tmp;
    d = PyObject_GetAttrString(__pyx_m, (char *)"__pyx_capi__");
    if (!d) {
        PyErr_Clear();
        d = PyDict_New();
        if (!d)
            goto bad;
        Py_INCREF(d);
        if (PyModule_AddObject(__pyx_m, (char *)"__pyx_capi__", d) < 0)
            goto bad;
    }
    tmp.fp = f;
#if PY_VERSION_HEX >= 0x02070000
    cobj = PyCapsule_New(tmp.p, sig, 0);
#else
    cobj = PyCObject_FromVoidPtrAndDesc(tmp.p, (void *)sig, 0);
#endif
    if (!cobj)
        goto bad;
    if (PyDict_SetItemString(d, name, cobj) < 0)
        goto bad;
    Py_DECREF(cobj);
    Py_DECREF(d);
    return 0;
bad:
    Py_XDECREF(cobj);
    Py_XDECREF(d);
    return -1;
}

/* InitStrings */
static int __Pyx_InitStrings(__Pyx_StringTabEntry *t) {
    while (t->p) {
        #if PY_MAJOR_VERSION < 3
        if (t->is_unicode) {
            *t->p = PyUnicode_DecodeUTF8(t->s, t->n - 1, NULL);
        } else if (t->intern) {
            *t->p = PyString_InternFromString(t->s);
        } else {
            *t->p = PyString_FromStringAndSize(t->s, t->n - 1);
        }
        #else
        if (t->is_unicode | t->is_str) {
            if (t->intern) {
                *t->p = PyUnicode_InternFromString(t->s);
            } else if (t->encoding) {
                *t->p = PyUnicode_Decode(t->s, t->n - 1, t->encoding, NULL);
            } else {
                *t->p = PyUnicode_FromStringAndSize(t->s, t->n - 1);
            }
        } else {
            *t->p = PyBytes_FromStringAndSize(t->s, t->n - 1);
        }
        #endif
        if (!*t->p)
            return -1;
        if (PyObject_Hash(*t->p) == -1)
            return -1;
        ++t;
    }
    return 0;
}

static CYTHON_INLINE PyObject* __Pyx_PyUnicode_FromString(const char* c_str) {
    return __Pyx_PyUnicode_FromStringAndSize(c_str, (Py_ssize_t)strlen(c_str));
}
static CYTHON_INLINE const char* __Pyx_PyObject_AsString(PyObject* o) {
    Py_ssize_t ignore;
    return __Pyx_PyObject_AsStringAndSize(o, &ignore);
}
#if __PYX_DEFAULT_STRING_ENCODING_IS_ASCII || __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT
#if !CYTHON_PEP393_ENABLED
static const char* __Pyx_PyUnicode_AsStringAndSize(PyObject* o, Py_ssize_t *length) {
    char* defenc_c;
    PyObject* defenc = _PyUnicode_AsDefaultEncodedString(o, NULL);
    if (!defenc) return NULL;
    defenc_c = PyBytes_AS_STRING(defenc);
#if __PYX_DEFAULT_STRING_ENCODING_IS_ASCII
    {
        char* end = defenc_c + PyBytes_GET_SIZE(defenc);
        char* c;
        for (c = defenc_c; c < end; c++) {
            if ((unsigned char) (*c) >= 128) {
                PyUnicode_AsASCIIString(o);
                return NULL;
            }
        }
    }
#endif
    *length = PyBytes_GET_SIZE(defenc);
    return defenc_c;
}
#else
static CYTHON_INLINE const char* __Pyx_PyUnicode_AsStringAndSize(PyObject* o, Py_ssize_t *length) {
    if (unlikely(__Pyx_PyUnicode_READY(o) == -1)) return NULL;
#if __PYX_DEFAULT_STRING_ENCODING_IS_ASCII
    if (likely(PyUnicode_IS_ASCII(o))) {
        *length = PyUnicode_GET_LENGTH(o);
        return PyUnicode_AsUTF8(o);
    } else {
        PyUnicode_AsASCIIString(o);
        return NULL;
    }
#else
    return PyUnicode_AsUTF8AndSize(o, length);
#endif
}
#endif
#endif
static CYTHON_INLINE const char* __Pyx_PyObject_AsStringAndSize(PyObject* o, Py_ssize_t *length) {
#if __PYX_DEFAULT_STRING_ENCODING_IS_ASCII || __PYX_DEFAULT_STRING_ENCODING_IS_DEFAULT
    if (
#if PY_MAJOR_VERSION < 3 && __PYX_DEFAULT_STRING_ENCODING_IS_ASCII
            __Pyx_sys_getdefaultencoding_not_ascii &&
#endif
            PyUnicode_Check(o)) {
        return __Pyx_PyUnicode_AsStringAndSize(o, length);
    } else
#endif
#if (!CYTHON_COMPILING_IN_PYPY) || (defined(PyByteArray_AS_STRING) && defined(PyByteArray_GET_SIZE))
    if (PyByteArray_Check(o)) {
        *length = PyByteArray_GET_SIZE(o);
        return PyByteArray_AS_STRING(o);
    } else
#endif
    {
        char* result;
        int r = PyBytes_AsStringAndSize(o, &result, length);
        if (unlikely(r < 0)) {
            return NULL;
        } else {
            return result;
        }
    }
}
static CYTHON_INLINE int __Pyx_PyObject_IsTrue(PyObject* x) {
   int is_true = x == Py_True;
   if (is_true | (x == Py_False) | (x == Py_None)) return is_true;
   else return PyObject_IsTrue(x);
}
static CYTHON_INLINE int __Pyx_PyObject_IsTrueAndDecref(PyObject* x) {
    int retval;
    if (unlikely(!x)) return -1;
    retval = __Pyx_PyObject_IsTrue(x);
    Py_DECREF(x);
    return retval;
}
static PyObject* __Pyx_PyNumber_IntOrLongWrongResultType(PyObject* result, const char* type_name) {
#if PY_MAJOR_VERSION >= 3
    if (PyLong_Check(result)) {
        if (PyErr_WarnFormat(PyExc_DeprecationWarning, 1,
                "__int__ returned non-int (type %.200s).  "
                "The ability to return an instance of a strict subclass of int "
                "is deprecated, and may be removed in a future version of Python.",
                Py_TYPE(result)->tp_name)) {
            Py_DECREF(result);
            return NULL;
        }
        return result;
    }
#endif
    PyErr_Format(PyExc_TypeError,
                 "__%.4s__ returned non-%.4s (type %.200s)",
                 type_name, type_name, Py_TYPE(result)->tp_name);
    Py_DECREF(result);
    return NULL;
}
static CYTHON_INLINE PyObject* __Pyx_PyNumber_IntOrLong(PyObject* x) {
#if CYTHON_USE_TYPE_SLOTS
  PyNumberMethods *m;
#endif
  const char *name = NULL;
  PyObject *res = NULL;
#if PY_MAJOR_VERSION < 3
  if (likely(PyInt_Check(x) || PyLong_Check(x)))
#else
  if (likely(PyLong_Check(x)))
#endif
    return __Pyx_NewRef(x);
#if CYTHON_USE_TYPE_SLOTS
  m = Py_TYPE(x)->tp_as_number;
  #if PY_MAJOR_VERSION < 3
  if (m && m->nb_int) {
    name = "int";
    res = m->nb_int(x);
  }
  else if (m && m->nb_long) {
    name = "long";
    res = m->nb_long(x);
  }
  #else
  if (likely(m && m->nb_int)) {
    name = "int";
    res = m->nb_int(x);
  }
  #endif
#else
  if (!PyBytes_CheckExact(x) && !PyUnicode_CheckExact(x)) {
    res = PyNumber_Int(x);
  }
#endif
  if (likely(res)) {
#if PY_MAJOR_VERSION < 3
    if (unlikely(!PyInt_Check(res) && !PyLong_Check(res))) {
#else
    if (unlikely(!PyLong_CheckExact(res))) {
#endif
        return __Pyx_PyNumber_IntOrLongWrongResultType(res, name);
    }
  }
  else if (!PyErr_Occurred()) {
    PyErr_SetString(PyExc_TypeError,
                    "an integer is required");
  }
  return res;
}
static CYTHON_INLINE Py_ssize_t __Pyx_PyIndex_AsSsize_t(PyObject* b) {
  Py_ssize_t ival;
  PyObject *x;
#if PY_MAJOR_VERSION < 3
  if (likely(PyInt_CheckExact(b))) {
    if (sizeof(Py_ssize_t) >= sizeof(long))
        return PyInt_AS_LONG(b);
    else
        return PyInt_AsSsize_t(b);
  }
#endif
  if (likely(PyLong_CheckExact(b))) {
    #if CYTHON_USE_PYLONG_INTERNALS
    const digit* digits = ((PyLongObject*)b)->ob_digit;
    const Py_ssize_t size = Py_SIZE(b);
    if (likely(__Pyx_sst_abs(size) <= 1)) {
        ival = likely(size) ? digits[0] : 0;
        if (size == -1) ival = -ival;
        return ival;
    } else {
      switch (size) {
         case 2:
           if (8 * sizeof(Py_ssize_t) > 2 * PyLong_SHIFT) {
             return (Py_ssize_t) (((((size_t)digits[1]) << PyLong_SHIFT) | (size_t)digits[0]));
           }
           break;
         case -2:
           if (8 * sizeof(Py_ssize_t) > 2 * PyLong_SHIFT) {
             return -(Py_ssize_t) (((((size_t)digits[1]) << PyLong_SHIFT) | (size_t)digits[0]));
           }
           break;
         case 3:
           if (8 * sizeof(Py_ssize_t) > 3 * PyLong_SHIFT) {
             return (Py_ssize_t) (((((((size_t)digits[2]) << PyLong_SHIFT) | (size_t)digits[1]) << PyLong_SHIFT) | (size_t)digits[0]));
           }
           break;
         case -3:
           if (8 * sizeof(Py_ssize_t) > 3 * PyLong_SHIFT) {
             return -(Py_ssize_t) (((((((size_t)digits[2]) << PyLong_SHIFT) | (size_t)digits[1]) << PyLong_SHIFT) | (size_t)digits[0]));
           }
           break;
         case 4:
           if (8 * sizeof(Py_ssize_t) > 4 * PyLong_SHIFT) {
             return (Py_ssize_t) (((((((((size_t)digits[3]) << PyLong_SHIFT) | (size_t)digits[2]) << PyLong_SHIFT) | (size_t)digits[1]) << PyLong_SHIFT) | (size_t)digits[0]));
           }
           break;
         case -4:
           if (8 * sizeof(Py_ssize_t) > 4 * PyLong_SHIFT) {
             return -(Py_ssize_t) (((((((((size_t)digits[3]) << PyLong_SHIFT) | (size_t)digits[2]) << PyLong_SHIFT) | (size_t)digits[1]) << PyLong_SHIFT) | (size_t)digits[0]));
           }
           break;
      }
    }
    #endif
    return PyLong_AsSsize_t(b);
  }
  x = PyNumber_Index(b);
  if (!x) return -1;
  ival = PyInt_AsSsize_t(x);
  Py_DECREF(x);
  return ival;
}
static CYTHON_INLINE Py_hash_t __Pyx_PyIndex_AsHash_t(PyObject* o) {
  if (sizeof(Py_hash_t) == sizeof(Py_ssize_t)) {
    return (Py_hash_t) __Pyx_PyIndex_AsSsize_t(o);
#if PY_MAJOR_VERSION < 3
  } else if (likely(PyInt_CheckExact(o))) {
    return PyInt_AS_LONG(o);
#endif
  } else {
    Py_ssize_t ival;
    PyObject *x;
    x = PyNumber_Index(o);
    if (!x) return -1;
    ival = PyInt_AsLong(x);
    Py_DECREF(x);
    return ival;
  }
}
static CYTHON_INLINE PyObject * __Pyx_PyBool_FromLong(long b) {
  return b ? __Pyx_NewRef(Py_True) : __Pyx_NewRef(Py_False);
}
static CYTHON_INLINE PyObject * __Pyx_PyInt_FromSize_t(size_t ival) {
    return PyInt_FromSize_t(ival);
}


#endif /* Py_PYTHON_H */
