svn commit: r335004 - head/release/tools

[Edward Napierala]

On 0613T0730, Warner Losh wrote:
> On Wed, Jun 13, 2018 at 6:39 AM, Edward Napierala <trasz at freebsd.org> wrote:
> 
> > 2018-06-13 12:43 GMT+01:00 Emmanuel Vadot <manu at bidouilliste.com>:
> >
> >> On 2018-06-12 18:45, Edward Tomasz Napierala wrote:

[..]

> >> +       echo '# USB OTG virtual serial port' \
> >>> +               >> ${CHROOTDIR}/${DESTDIR}/etc/ttys
> >>> +       echo 'ttyU0     "/usr/libexec/getty 3wire"      vt100
> >>>  onifconsole  secure' \
> >>> +               >> ${CHROOTDIR}/${DESTDIR}/etc/ttys
> >>> +       echo 'ttyU1     "/usr/libexec/getty 3wire"      vt100
> >>>  onifconsole  secure' \
> >>> +               >> ${CHROOTDIR}/${DESTDIR}/etc/ttys
> >>>
> >>
> >>  If I have no OTG port and a usb<->uart plugged into my board that will
> >> give weird result no ?
> >>
> >
> > No, because that port won't be marked as console.  This only applies
> > to the "virtual" OTG serial ports.
> >
> 
> Right, and console is an overloaded term.  Here it just means 'tty marked
> by the kernel that gets a getty started on it automatically after it shows
> up' not 'the device that gets all the kernel I/O.'

Yup.  But again - this reuses the functionality that's already there
in init(8), while avoiding the renaming of device nodes.  And eventually,
those ports might become actual consoles, making things nicely aligned.

[..]