Update chr to be compatible with CPython

Not all exposed to python code yet

This commit is pretty large because it contains two separate but interrelated changes:
- Rewrite the function argument handling code (callCompiledFunction and resolveCLFunc)
  into a single callFunc that does its own rewriting, and support the new features.
-- this required a change of data representations, so instead of having each function
   consist of variants with unrelated signatures, we can only have a single signature,
   but multiple type specializations of that signature
- To do that, had to rewrite all of the stdlib functions that used signature-variation
  (ex range1 + range2 + range3) to be a single function that took default arguments, and
  then took action appropriately.

Can emit add/sub/etc instructions with 32-bit operands in addition to 8-bit now
Can get the RSP and RBP in rewriter1, for accessing scratch space
Change some debugging output

Fix a gc bug: if an object gets new'd, and takes a parameter that gets new'd, the sequence is
1) object space gets allocated
2) parameter space gets allocated
3) parameter gets constructed
4) object gets constructed
The bug is that the object construction is what initializes the GC header,
so if step #3 causes a collection, it can see that the allocation from step #1
has an invalid header.
As a workaround, always zero out the header in allocation, and skip blocks with
zeroed headers.
The real solution is probably to have the GC manage the header itself rather than
expecting the user to; this would mean that gc_alloc would take the allocation kind,
put that into the header, and then return a pointer to the post-header data section
of the allocation.

Experimenting with vim+make integration
Changed installation instructions slightly
Added rlwrap support for repl

Fix crash: register string iterator class

Tried up to 210379 with the same segfaults.

Replace a simple "num_args" argument with a packed struct that takes
num_args and adds num_keywords, has_varargs, and has_kwargs.  Tried
to add asserts in all the places that don't allow
keywords/varargs/starargs

Started refactoring things; got to the point of attempting
argument->parameter shuffling, but it's tricky if we allow every
compilation to have a different signature (used by builtins).
Really they all have the same signatures but different
specializations; to get to that point, need to add defaults.

- Support parsing (but not running) **kw in function definitions
- more consistent assertion of what parts handle non-simple arguments or not

Added implementation of 'dir'

Implementation of __iter__ for tuple, added tests

Signed-off-by: Vinzenz Feenstra <evilissimo@gmail.com>

Signed-off-by: Vinzenz Feenstra <evilissimo@gmail.com>

- Added dir0 which would in theory return the local scope, but
  for just returns the content of the builtins module
- Added a unit test file for dir which won't print anything as the
  list sorting does not work as expected at the moment
  (missing full implementation) and the attributes are different
  as well (pyston has some CPython 2.7.5 does not have and vice versa)
- dir1 implements the lookup on a passed object
  - If the object has a __dir__ method, it will call it and return the result
    returned by this method
  - Otherwise all class and instance attributes and methods are returned
  - If the object has a __dict__ attribute, additional to the previous one
    it will return the keys from that one as well
Signed-off-by: Vinzenz Feenstra <evilissimo@gmail.com>

Signed-off-by: Vinzenz Feenstra <evilissimo@gmail.com>

Signed-off-by: Vinzenz Feenstra <evilissimo@gmail.com>

For the release build, was building both tools/ and tools/llc (and tools/opt),
which could potentially run at the same time and interfere with each other.

These seem to work fairly well as tests since
- they don't use any external libraries, and
- they have a very straightforward input-to-output model

Added some more runtime functions
gcj_2014_2d is currently crashing pyston

The presence of an allocated "elts" pointer is determined by capacity>0,
but the GC handler was incorrectly only choosing to visit it if size>0.
Not all code paths free the elts array if size goes down to zero, so
it could happen that capacity>0 and size=0, and then the elts array would
not be tracked even though it was expected to be.

Added a test that illustrates this.

Implement str.lstrip and str.rstrip