but for large loads it ties master, for many short loads it
is slightly faster.  if the code can be made clearer then
it's a win

running orm2010, the whole approach adds callcounts.
code is more complex, less reliable, pretty much a total
bust.

will try to see if create_row_processor can remain but
be called against either "result" or a context namespace
ahead of time, and if this somehow can be used to simplify
the need to call upon strategy lookup twice.

many-to-one ``!= None`` would fail if the source were an aliased
class, or if the query needed to apply special aliasing to the
expression due to aliased joins or polymorphic querying; also fixed
bug in the case where comparing a many-to-one to an object state
would fail if the query needed to apply special aliasing
due to aliased joins or polymorphic querying.
fixes #3310

an ``after_rollback()`` handler for a :class:`.Session` incorrectly
adds state to that :class:`.Session` within the handler, and the task
to warn and remove this state (established by :ticket:`2389`) attempts
to proceed.
fixes #3309

asserting.

in callcounts specific to the "expiration" of objects, as in
the "auto expire" feature of :meth:`.Session.commit` and
for :meth:`.Session.expire_all`, as well as in the "cleanup" step
which occurs when object states are garbage collected.
fixes #3307