generate the new warning only when we had bad code.

theano/tensor/opt.py

@@ -1709,8 +1709,6 @@ def local_sum_sum(node):
                     # do it all at once
                     return [T.Sum(None)(summed.owner.inputs[0])]
 
-                if theano.config.warn.sum_sum_bug:
-                    print "WARNING: Theano version between version 9923a40c7b7a and the w august 2010(fixed date), generated an error in that case. This happen when their is 2 consecutive sum in the graph, bad code was generated. To disable this warning, set the theano flags warn.sum_sum_bug to False."
                 newaxis=list(tuple(summed.owner.op.axis))
                 # figure out which dimensions of the original input are preserved
                 for i in node.op.axis:
@@ -1722,6 +1720,17 @@ def local_sum_sum(node):
                     newaxis.append(new_i)
                         
                 assert len(newaxis)==len(list(summed.owner.op.axis)+list(node.op.axis))
+
+                #The old bugged logic. We keep it their to generate a warning when we generated bad code.
+                alldims = range(summed.owner.inputs[0].type.ndim)
+                alldims = [d for i,d in enumerate(alldims) if i in summed.owner.op.axis]
+                alldims = [d for i,d in enumerate(alldims) if i in node.op.axis]
+                newaxis_old = [i for i in xrange(summed.owner.inputs[0].type.ndim)
+                               if i not in alldims]
+                
+                if theano.config.warn.sum_sum_bug and newaxis!=newaxis_old:
+                    print "WARNING: Theano version between version 9923a40c7b7a and the 2 august 2010(fixation date), generated an error in that case. This happen when their is 2 consecutive sum in the graph, bad code was generated. To disable this warning, set the theano flags warn.sum_sum_bug to False."
+
                 combined_sum = T.Sum(newaxis)
                 return [combined_sum(summed.owner.inputs[0])]