Move atmel-samd to tinyusb and support nRF flash.

[tannewt]

This started while adding USB MIDI support (and descriptor support is
in this change.) When seeing that I'd have to implement the MIDI class
logic twice, once for atmel-samd and once for nrf, I decided to refactor
the USB stack so its shared across ports. This has led to a number of
changes that remove items from the ports folder and move them into
supervisor.

Furthermore, we had external SPI flash support for nrf pending so I
factored out the connection between the usb stack and the flash API as
well. This PR also includes the QSPI support for nRF.

I'm using my own fork of tinyusb until I can land the changes upstream.

[siddacious]

Awesome work! 👍

[ladyada]

i dont think i'd be good for code reviewing this opus, but i could do hardware testing for sure :) just let me know!

[dhalbert]

Very impressive refactoring! In some ways, there are fewer changes than I would have expected due to tinyusb. There are many more cleanup changes, which are really welcome!

[tannewt]

Thanks for the quick review! I'll get back to you tomorrow. Trying to run down the last few things to get the build green now.

[tannewt]

@dhalbert This is ready for another look. Thanks!

[dhalbert]

Fantastic! This is a huge step forward: tinyusb, and getting rid of a lot of technical debt.

[notro]

This PR is causing problems for me.
I use a tool similar to ampy to copy files through the REPL to the board.
It is now broken and I get timeout/disconnect after a few files are copied.
The commit before this PR is fine (I've tried the latest master, same problem).

The board is connected to a Raspberry Pi.
Sometimes the /dev device name stays the same and sometime I get a new one like here:

pi@cp:~/work/circuitpython/projects/fs.repl/fs $ python3 -u stdlib_copy.py -vvv /dev/ttyACM1 --delete-lib

board Board('/dev/ttyACM1')
uname: posix.uname_result(sysname='samd51', nodename='samd51', release='4.0.0', version='4.0.0-alpha.2-144-g50f5d27c4-dirty on 2018-11-15', machine='Adafruit Metro M4 Express with samd51j19')
board time time.struct_time(tm_year=2018, tm_mon=11, tm_mday=15, tm_hour=17, tm_min=14, tm_sec=38, tm_wday=3, tm_yday=319, tm_isdst=-1)
board time time.struct_time(tm_year=2018, tm_mon=11, tm_mday=15, tm_hour=17, tm_min=14, tm_sec=39, tm_wday=3, tm_yday=319, tm_isdst=-1)

<replfs '/dev/ttyACM1'>
Deleting /lib ...
Source: <osfs '/home/pi/work/circuitpython/tmp_CircuitPython_stdlib/lib'>
Destination: <replfs '/dev/ttyACM1'>/lib
new: /contextlib.mpy
new: /copy.mpy
new: /copyreg.mpy
new: /datetime.mpy
RETRY
Traceback (most recent call last):
  File "/home/pi/.local/lib/python3.5/site-packages/serial/serialposix.py", line 537, in write
    n = os.write(self.fd, d)
OSError: [Errno 5] Input/output error

<snip>

cpboard.CPboardError: read error

pi@cp:~/work/circuitpython/projects/fs.repl/fs $ dmesg
[80063.522775] usb 1-1.4: reset full-speed USB device number 81 using dwc_otg
[80063.655651] cdc_acm 1-1.4:1.0: ttyACM2: USB ACM device
[80063.664710] sd 1:0:0:0: [sdc] 4089 512-byte logical blocks: (2.09 MB/2.00 MiB)
[80063.682308]  sdc: sdc1

The same problem is also experienced when trying out libraries in the REPL. Suddenly the connection closes down. In this case I can just reconnect, but loading a library and trying a few things, leads to disconnect again.
I've tried just keeping the REPL open without doing anything, and it stays connected.

Anything I can do to diagnose this further?

[tannewt]

@notro have you tried the tip of master? Some fixes went in yesterday.

[notro]

Yes, I've tried the latest master.

[tannewt]

@notro Please file a new issue with instructions to reproduce. For that issue:

• Beagle or wireshark traces are extra helpful.
• What is the status LED doing as it does this? One of the first things is to rule out a hardfault.
• Which board are you using?

[notro]

To be precise, this PR didn't work at all (looked like something was corrupted in the transfer), I needed the overrun fix in order to copy anything at all.