Most of printing infrastructure now uses streams, but mp_obj_print() used
libc's printf(), which led to weird buffering issues in output. So, switch
mp_obj_print() to streams too, even though it may make sense to move it to
a separate file, as it is purely a debugging function now.

Relative imports are based of a package, so we're currently at a module
within a package, we should get to package first.

Also, factor out path travsering operation, but this broke testing for
boundary errors with relative imports. TODO: reintroduce them, together
with proper tests.

So corresponding exception can be thrown even under tight memory conditions.

Traceback allocation for exception will now never lead to recursive
MemoryError exception - if there's no memory for traceback, it simply
won't be created.

They are directly cast to str object, skipping allocation of formatting
buffer.

Has to be enabled by manual editing, but at least it's there, as debugging
NLR issues may be weird.

Pushing same NLR record twice would lead to "infinite loop" in nlr_jump
(but more realistically, it will crash as soon as NLR record on stack is
overwritten).

Conceptually it is part of code state, so let it be allocated in the same way
as the rest of state.

Addresses issue #1117.

Addresses issue #1119.

Now all ports can use pybstdio.c to provide sys.stdin/stdout/stderr, so
long as they implement mp_hal_stdin_* and mp_hal_stdout_* functions.