This moves the raw read/write access code from the recovery mode into
st3m, and gives us the ability to mount/unmount the FAT partition on
demand.

Only ctx.arc has a 6th integer argument, integer truncation was applied to
all 6-argument methods.

Tested by doing the following in REPL:

    >>> import hardware
    >>> ctx = hardware.get_ctx()
    >>> ctx.rgb(1, 0, 0)
    <ctx_type>
    >>> ctx.rectangle(-120, -120, 240, 240)
    <ctx_type>
    >>> ctx.fill()
    <ctx_type>
    >>> hardware.display_update(ctx)

Fixes #25.

Whoops.

This allows us to os.listdir("/").

We attempt to share st3m with the recovery firmware, and thus must
eliminate two components that can't live in the recovery firmware:

1. The startup function, which contains references to all matters of
   subfunctionality we don't want (audio, micropython, etc)
2. The FS code, which contains references to the tarball extraction
   logic (which we don't want to embed into recovery mode either).

We don't pre-generate a FAT partition anymore.

rip 2023-2023

This removes all hardware.* captouch calls from mpy and ports the entire
API to captouch.*.

Internally in C, we have unified all API calls under one
st3m_captouch_read call which lets the user access structured data in
any way they want.

This reimplements components/captouch into a separate BSP-based captouch
driver.

New features:

 1. Interrupt-driven, which seems to improve performance.
 2. Has ability to run multiple sequences one after another on the
    AD7147. This is needed to support 13 channels on the bottom captouch
    controller. However, this seems to be currently broken (see BUG).

Whoops, this shouldn't have happened. Sorry.

Apparently this is what we have to do to end up with correctly split
flags being passed into micropython qstr machinery...

We're going with spirals for the production model, and the code fork
doesn't make much sense anyway, as it all ended up being the same (so
far). Whenever the behaviour between spiral/spikes changes, we should
just drop support for spikes.

We have newer badges available for everyone to develop against, and p1
is so different from everything else that it's really not worth
maintaining support for it anymore.

We ripped that out when initially porting VfsPosix to work on ESP32.
This re-introduces it.

Fixes #10.

This makes sure we don't start claiming to be a ttyACM device before the
user actually enables USB.

st3m_io_badge_link_set would squash previous state.