On Mon, Aug 27, 2012 at 5:46 AM, Paolo Carlini <paolo.carl...@oracle.com> wrote: > One last observation from me: I think we are being a little inconsistent in > terms of inlining. I see some __generate_impl with a non-trivial body inline > whereas other, with a tad smaller body, in random.tcc.
For the __generate functions there is not much of an upside to having the functions inline. Unlike for the operator() definitions, the values are not immediately used which would enable folding in additional arithmetic operations with the computations for the distribution. I've moved all definitions of __generate_impl into random.tcc. Will run some more checks and check in the change after success.