Support GpuCumsum on 3D array.

Slightly clean up registration of GPU convolution optimizers

Documentation for sliding windows 

Cache opt

Fix cuDNN detection

Add CUDA include dir so cuda.h is found; also add CUDA library dir to mimic configuration when compiling CUDA ops

Implemented grad for cudnn softmax.

Added test to ensure SoftmaxGrad to DnnSoftmaxGrad is not applied when cudnn is excluded from optimizations.

Added optimization that converts SoftmaxGrad to DnnSoftmaxGrad and associated test to make sure it is applied correctly.

allow cxx flag to be full path to compiler

Conflicts:
	theano/gof/cmodule.py

Dnn default and doc

Prevent computations in float16 in scalar and elemwise

Catch exception if open() fails as well.