Merge pull request #4820 from nouiz/local_gpua_careduce

theano/gpuarray/opt.py

@@ -373,6 +373,14 @@ class GraphToGPU(NavigatorOptimizer):
 
             if new_ops:
                 node_created[lopt] += len(graph.ops([mapping[i] for i in node.inputs], outputs))
+                if any([getattr(old_o, 'dtype', None) != getattr(new_o, 'dtype', None)
+                        for old_o, new_o in zip(outputs, node.outputs)]):
+                    _logger.warning(
+                        "The optimization %s returned bad dtype. Skipping it."
+                        " Write to theano-dev mailing list about this." %
+                        str(lopt))
+                    newnode = node.clone_with_new_inputs([mapping.get(i) for i in node.inputs])
+                    outputs = newnode.outputs
 
             for new_o, old_o in zip(outputs, node.outputs):
                 assert len(outputs) == len(node.outputs)
@@ -1009,10 +1017,9 @@ def local_gpua_careduce(op, context_name, inputs, outputs):
         else:
             return False
         x, = inputs
-
         greduce = op2(
             op.scalar_op, axis=op.axis,
-            dtype=getattr(op, 'dtype', None),
+            dtype=getattr(op, 'dtype', outputs[0].dtype),
             acc_dtype=getattr(op, 'acc_dtype', None))
         gvar = greduce(x)
         # We need to have the make node called, otherwise the mask can
@@ -1051,7 +1058,7 @@ def local_gpua_careduce(op, context_name, inputs, outputs):
             greduce = op2(
                 op.scalar_op,
                 axis=new_axis, reduce_mask=new_mask,
-                dtype=getattr(op, 'dtype', None),
+                dtype=getattr(op, 'dtype', outputs[0].dtype),
                 acc_dtype=getattr(op, 'acc_dtype', None))
 
             reshaped_x = x.reshape(tensor.stack(new_in_shp))