C code for GPUAdvancedIncSubtensor1.Handles one case to make it faster

theano/sandbox/cuda/basic_ops.py

@@ -2442,6 +2442,50 @@ class GpuAdvancedIncSubtensor1(tensor.AdvancedIncSubtensor1, GpuOp):
                     x[i] += y
         out[0] = x
 
+    def c_code(self, node, name, inputs, outputs, sub):
+        if (self.set_instead_of_inc) or \
+           (node.inputs[0].ndim != node.inputs[1].ndim):
+            raise NotImplementedError("This case does not have C code yet.")
+
+        x = inputs[0]
+        y = inputs[1]
+        ind = inputs[2]
+        out = outputs[0]
+        fail = sub['fail']
+        inplace = int(self.inplace)
+
+        return """
+        PyObject *x_obj, *y_obj, *row_x, *row_y;
+        PyObject *x_rowind_obj, *y_rowind_obj;
+        int *p_index;
+        int num_indices, j;
+        
+        Py_XDECREF(%(out)s);
+        if (!%(inplace)s) {
+            %(out)s = (CudaNdarray*)CudaNdarray_Copy(%(x)s);
+        } else {
+            %(out)s = %(x)s;
+            Py_XINCREF(%(out)s);
+        }
+
+        x_obj = (PyObject*)CudaNdarray_View(%(x)s);
+        y_obj = (PyObject*)CudaNdarray_View(%(y)s);
+        num_indices = PyArray_SIZE(%(ind)s);
+
+        for (j = 0;j < num_indices; j++) {
+
+             p_index = (int *)PyArray_GETPTR1(%(ind)s, j);
+             x_rowind_obj = PyInt_FromLong(*p_index);
+             y_rowind_obj = PyInt_FromLong(j);
+             row_x = CudaNdarray_Subscript(x_obj, x_rowind_obj);
+             row_y = CudaNdarray_Subscript(y_obj, y_rowind_obj);
+             CudaNdarray_inplace_add(row_x, row_y);
+        }
+   
+        if (!%(out)s) {
+            %(fail)s
+        }
+        """ %locals()
 
 class GpuIncSubtensor(tensor.IncSubtensor, GpuOp):
     """

theano/sandbox/cuda/cuda_ndarray.cu

@@ -1889,7 +1889,7 @@ CudaNdarray_len(PyObject * py_self)
 }
 
 // Will by called by __getitem__ in Python
-static PyObject *
+PyObject *
 CudaNdarray_Subscript(PyObject * py_self, PyObject * key)
 {
     int verbose = 0;

theano/sandbox/cuda/cuda_ndarray.cuh

@@ -479,7 +479,7 @@ int fprint_CudaNdarray(FILE * fd, const CudaNdarray *self);
 
 PyObject * CudaNdarray_View(const CudaNdarray * self);
 PyObject * CudaNdarray_inplace_add(PyObject* py_self, PyObject * py_other);
-
+PyObject * CudaNdarray_Subscript(PyObject * py_self, PyObject * key);
 
 
 // Ensures that *arr is a pointer to a contiguous ndarray of the specified