In raw REPL ">" indicates the prompt.  We originally read this character
upon entering the raw REPL, and after reading the last bit of the
output.  This patch changes the logic so the ">" is read only just
before trying to send the next command.  To make this work (and as an
added feature) the input buffer is now flushed upon entering raw REPL.

The main reason for this change is so that pyboard.py recognises the EOF
when sys.exit() is called on the pyboard.  Ie, if you run pyboard.py
with a script that calls sys.exit(), then pyboard.py will exit after
the sys.exit() is called.

esp8266: Changed esp_scan to keep the current WiFi operating mode but throw an exception if WiFi is in AP only mode

Since the commit that moved those two functions failed to update
the documentation, this is a fix for that.

MicroPython "network" module interface requires it to contains classes
to instantiate. But as we have a static network interace, make WLAN()
"constructor" just return module itself, and just make all methods
module-global functions.

Fixes regression from a previous commit.

mpconfigport.mk contains configuration options which affect the way
MicroPython is linked. In this regard, it's "stronger" configuration
dependency than even mpconfigport.h, so if we rebuild everything on
mpconfigport.h change, we certianly should of that on mpconfigport.mk
change too.

This makes all common files "port-aware" using the .. only directive.

Timer.A and Timer.B can be ORed for this purpose.

This is so that the DMA can be shared by multiple peripherals.

Enabled simply by making the identifier lexing code 8-bit clean.

If heap allocation for the Python-stack of a function fails then we may
as well allocate the Python-stack on the C stack.  This will allow to
run more code without using the heap.

If IRQs are disabled then the USB CDC buffer will never be
drained/filled and the sys-tick timer will never increase, so we should
not busy wait in this case.