Speed up. get_scalar_constant_value and slow scan test.

This make less condition being tested. It speed up the slow scan tests.

The 2 subtensors tests change probably due to optimization order changes.
The 2->22 is that in the past, the Composite wasn't flattened, so the outer Composite was smaller.

We fix for atleast 158 loops of that problem.

with Composite.

This allow to pickle bigger Composite graph as it simplify the graph,
so we bust later the max recursion limit.

Flatten Composite.

This help gh-689 as it allow it to support up to 57 iteration of the
loop instead of just 43.

This could also help Theano cache in the case where graph could have
different fusion order, so different way the composite are in each
other, but do the same computation.

This make the printing of Composite more readable.

Faster optimizer

Make PatternSub use that feature to lower optimizer time by tracking
less frequent node. This allow to save 85s on 317s in optimizer time.

Mixed

Faster opt (and other unrelated things).

Cuda fortran order

Add other gpualloc tests. It currently fail as ger isn't implement on
the new back-end.

This make opt slower as it make the creation of GpuDimshuffle node slower.

Alloc cst fold

Remove condition that is always True as we check if it was "output" before and it is the only string supported.

Since we do constant folding earlier, this make a GpuAlloc test fail.
This could also speed some code as this allow more op to work inplace.

Faster opt

I review the code, we always put a tuple or a Theano variable in that
list. We only want to work with the tuple case. So just test for the
tuple. We don't want to cover the case where we have a Theano vector
with a know lenght.