Add more DEBUG_printf statements to trace gc behaviour

This adds a hook to get/set pyb_uart_global_debug from Python, using
pyb.repl_uart().  You can set it to an arbitrary UART object, and then
the REPL (in and out) is repeated on this UART object (as well as on USB
CDC).

Ultimately, this will be replaced with a proper Pythonic interface to
set sys.stdin and sys.stdout.

Also make stdout_print_strn static (ultimately this function needs to be
merged with stdout_tx_strn).

Still some method names to iron out, and funtionality to add, but this
will do for the first, basic version.

char can be signedness, and using signedness types is dangerous - it can
lead to negative offsets when doing table lookups. We apparently should just
ban char usage.

Add __assert_func

* issue #692

In preparation for unicode support.

We still have that char vs byte dichotomy, but majority of string operations
now use byte.

This will allow roughly the same behavior as Python3 for non-ASCII strings,
for example, print("<phrase in non-Latin script>".split()) will print list
of words, not weird hex dump (like Python2 behaves). (Of course, that it
will print list of words, if there're "words" in that phrase at all, separated
by ASCII-compatible whitespace; that surely won't apply to every human
language in existence).

Getting a length of string may be expensive, depending on the underlying
implementation.

void* is of unknown size

Functionality we provide in builtin io module is fairly minimal. Some
code, including CPython stdlib, depends on more functionality. So, there's
a choice to either implement it in C, or move it _io, and let implement other
functionality in Python. 2nd choice is pursued. This setup matches CPython
too (_io is builtin, io is Python-level).

It needs to be reused for generator functions, too.

Also now returns correct POSIX errno when an IO operation fails.

Addresses issues #516 and #676.

Remove unnecessary bounds check from mp_seq_get_fast_slice_indexes.

At this point, start will be >= 0, so checking if stop < 0 is redundant with
checking if start > stop a few lines later.

This is getting more and more tangled, but that's old news.

As before, instead of pushing constant values on stack again and again, just
pass around pointer to a structure.

See PR #618.