gpython: Optimize recursive call

[corona10]

Thanks to tengo and @ncw, we can compare the gpython with other interpreted which are implemented in Go.

https://github.com/d5/tengo#benchmark

fib(35) benchmark is slower than I expected, we can optimize this point later.
(tail call is faster than CPython, I can't believe it!)

[ncw]

:-) I suspect the times recorded for the tail recursive test are dominated by startup times rather than actually calling the functions.

Perhaps we should have some internal benchmarks which run python code using the go benchmark framework? That would give us a solid baseline to work with.

So we could have a benchmarks directory with python programs and a little driver to run them through the benchmark code.

[corona10]

@ncw Looks like we have to create a new issue for build benchmark framework for gpython

[drew-512]

recommend @sbinet close -- addressed in #50