Reduces size of this port by about 3300 bytes, and demonstrates how to use
this feature.

Signed-off-by: Damien George <damien@micropython.org>

This commit simplifies and cleans up the bare-arm port, and adds just
enough system and library code to make it execute on an STM32F405 MCU.

The mpconfigport.h configuration is simplified to just specify those
configuration values that are different from the defaults.  And the
addition of -fdata-sections and -ffunction-sections means the final
firmware is smaller than it previously was, by about 4200 bytes.

A README is also added.

Signed-off-by: Damien George <damien@micropython.org>

To keep these ports as minimal as possible.

Enabled on: bare-arm, minimal, unix coverage/dev/minimal, stm32, esp32,
esp8266, cc3200, teensy, qemu-arm, nrf.  Not enabled on others to be able
to test the code when the feature is disabled (the default case).

Code size change for this commit:

   bare-arm:  -600 -0.906%
minimal x86:  -308 -0.208%
   unix x64:    +0 +0.000%
unix nanbox:    +0 +0.000%
      stm32: -3368 -0.869% PYBV10
     cc3200: -1024 -0.558%
    esp8266: -2512 -0.368% GENERIC
      esp32: -2876 -0.205% GENERIC[incl -3168(data)]
        nrf: -1708 -1.173% pca10040
       samd:    +0 +0.000% ADAFRUIT_ITSYBITSY_M4_EXPRESS

To enable lazy loading of submodules (among other things), which is very
useful for MicroPython libraries that want to have optional subcomponents.

Disabled explicitly on minimal ports.

This is run with uncrustify 0.70.1, and black 19.10b0.

This string is recognised by uncrustify, to disable formatting in the
region marked by these comments.  This is necessary in the qstrdef*.h files
to prevent modification of the strings within the Q(...).  In other places
it is used to prevent excessive reformatting that would make the code less
readable.

Many ports already allow overriding CROSS_COMPILE.  This modifies the
remaining ports to allow it as well.

This option affects py/vm.c and py/gc.c and using -Os gets them compiling a
bit smaller, and small firmware is the aim of these two ports.  Also,
having these files compiled with -Os on these ports, and -O3 as the default
on other ports, gives a better understanding of code-size changes when
making changes to these files.

mp_compile no longer takes an emit_opt argument, rather this setting is now
provided by the global default_emit_opt variable.

Now, when -X emit=native is passed as a command-line option, the emitter
will be set for all compiled modules (included imports), not just the
top-level script.

In the future there could be a way to also set this variable from a script.

Fixes issue #4267.