Assert that the size of int is the same on cpu and gpu.

theano/sandbox/cuda/cuda_ndarray.cu

@@ -1886,12 +1886,13 @@ static PyTypeObject CudaNdarrayType =
 static __global__ void get_gpu_ptr_size(int* dst)
 {
   dst[0] = sizeof(float*);
+  dst[1] = sizeof(int);
 }
 
 PyObject *
 CudaNdarray_ptr_int_size(PyObject* _unused, PyObject* args)
 {
-  int *gpu_data = (int*)device_malloc(sizeof(int));
+  int *gpu_data = (int*)device_malloc(sizeof(int)*2);
   if(gpu_data == NULL){
     return PyErr_Format(PyExc_MemoryError,
                         "CudaNdarray_ptr_int_size: Can't allocate memory on the gpu.");
@@ -1903,14 +1904,17 @@ CudaNdarray_ptr_int_size(PyObject* _unused, PyObject* args)
     return PyErr_Format(PyExc_RuntimeError,
                         "CudaNdarray_ptr_int_size: error when calling the gpu code.");
   }
-  int gpu_ptr_size = -1;
-  cublasGetVector(1, sizeof(int), gpu_data, 1, &gpu_ptr_size, 1);
+
+  // Transfer the result to cpu
+  int gpu_sizes[] = {-1,-1};
+  cublasGetVector(2, sizeof(int), gpu_data, 1, gpu_sizes, 1);
   device_free(gpu_data);
+
   if (CUBLAS_STATUS_SUCCESS != cublasGetError()){
     PyErr_SetString(PyExc_RuntimeError, "error copying data to from memory");
     return NULL;
   }
-  return Py_BuildValue("iii", gpu_ptr_size, sizeof(float*), sizeof(int));
+  return Py_BuildValue("iiii", gpu_sizes[0], sizeof(float*), sizeof(int), gpu_sizes[1]);
 }
 
 // Initialize the gpu.

theano/sandbox/cuda/opt.py

@@ -852,6 +852,10 @@ def local_gpu_join(node):
 #               70.0, 'fast_run', 'inplace')
 
 def get_device_type_sizes():
+    """
+    :return:(gpu ptr size, cpu ptr size, int sizes(gpu and cpu))
+    :return type: tuple
+    """
     if hasattr(get_device_type_sizes, 'rval'):
         return get_device_type_sizes.rval
     gpu_ptr_size = 8
@@ -859,9 +863,10 @@ def get_device_type_sizes():
     int_size = 8
     try:
 
-        #RETURN (gpu ptr size, cpu ptr size, int sizes)
         t = cuda_ndarray.cuda_ndarray.ptr_int_size()
-        gpu_ptr_size, cpu_ptr_size, int_size = t
+        gpu_ptr_size, cpu_ptr_size, int_size, gpu_int_size = t
+        assert int_size == gpu_int_size
+        del gpu_int_size
         del t
     except Exception, e:
         _logger.warning(("OPTIMIZATION WARNING: "