Re: USB-RS232 adapters vs. machines w/real consoles.

["John D. Baker" <jdbaker%mylinuxisp.com@localhost>]

On Fri, 25 Nov 2016, AGC wrote:

> On 2016-11-25 17:27, John D. Baker wrote:
> 
> > The Linux Mint system has so-far thwarted my attempts to operate any
> > USB-RS232 adapter without dbus and its minions interfering.  (I managed
> > to boot it single-user and start 'udevd', but 'cu' claims "Permission
> > denied" and that the port is in use, but it doesn't appear in 'lsof'
> > output.)
> 
> Check systemd and the gettys.  You may need to prevent systemd from
> handling the getty system.  It tends to be "helpful" in unhelpful ways.

The Linux Mint 17.3 system seems to use UpStart masquerading as systemd
or some strange chimera thereof.  (I was pleased that system logs were
human-readable.)  There are obvious 'getty' instances running on the
usual virtual terminals, but they don't show up in the output of
'ps -ax'.  Must be running as threads under 'logind'.  'logind.conf'
seems pretty ordinary in that regard--only PC virtual terminals,
not serial ports of any kind.

(I've been away from any serious penguin-OS use for long enough that
I'm just about a newbie again...)

Playing with it further, in its default (graphical) runlevel, I can
open "/dev/ttyUSB0" with 'cu' just fine (after putting my user in the
"dialout" group).

Connected to another system running 'cu' on a PCI-bus-resident UART
(NetBSD/amd64-7.0_STABLE), reveals interesting behavior.

The Belkin/MCT USB device on the Linux Mint system can receive just
fine.  It won't transmit but queues all keystrokes and transmits them
in a burst when I close the connection with "~.".

The Prolific devices receive just fine.  Attempted transmission seems
to do nothing at all.  No keytrokes appear to be queued, or if they
are, the queue is discarded on close rather than flushed (as appears
to happen on the Belkin/MCT device above).

The KeySpan USA-19HS also receives fine.  It doesn't transmit, but if
one types more than four characters, including carriage-return, 'cu' is
locked up so that one cannot exit with "~.".  Unplugging the device is
the only way to terminate the session.

So, it's still not clear what's getting in the way of these devices
transmitting under Linux Mint 17.3.

-- 
|/"\ John D. Baker, KN5UKS               NetBSD     Darwin/MacOS X
|\ / jdbaker[snail]mylinuxisp[flyspeck]com    OpenBSD            FreeBSD
| X  No HTML/proprietary data in email.   BSD just sits there and works!
|/ \ GPGkeyID:  D703 4A7E 479F 63F8 D3F4  BD99 9572 8F23 E4AD 1645