[wish-info] Problem with cm11a - Error opening /dev/ttyS1 (No such device or address)

Discussion:

Mark Mykkanen

2004-10-04 19:55:01 UTC

Hi,

I'm running into the following problem on a
linux-2.6.8-gentoo-r3 system. I patched my 2.6.8
kernel, compiled the driver as a module and loaded it
without any problems, but it doesn't want to read my
cm11a on ttyS1 or ttyS0. I'm using udev and the
/dev/x10/* files are all there. Oh, and the device is
plugged in. :)

Any ideas?

Thanks,
Mark

# ls -l /dev/ttyS[01]
crw-rw---- 1 root tty 4, 64 Jul 17 20:14 /dev/ttyS0
crw-rw---- 1 root tty 4, 65 Jul 17 20:14 /dev/ttyS1

Oct 1 21:38:58 shamu2 x10: X10 DEV module v2.0.0
(***@sprintmail.com)
Oct 1 21:38:58 shamu2 x10: $Id: dev.c,v 1.19
2004/07/05 03:13:13 whiles Exp whiles $
Oct 1 21:38:58 shamu2 x10: X10 driver successfully
loaded
Oct 1 21:39:19 shamu2 cm11ad[11194]: starting.
api-device=/dev/x10/.api, pidfile=/var/run/x10d.pid
Oct 1 21:39:19 shamu2 cm11ad[11194]: Successfully
opened X10 API device /dev/x10/.api
Oct 1 21:39:20 shamu2 cm11ad[11195]: Transmit thread
starting
Oct 1 21:39:20 shamu2 cm11ad[11195]: Error opening
/dev/ttyS1 (No such device or address)
Oct 1 21:39:20 shamu2 cm11ad[11195]: unable to start
transmitter, exiting
O

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

Scott Hiles

2004-10-05 03:10:01 UTC

Permalink

I doubt that udev has anything to do with it, but the drivers do not support
any of the dynamic device capabilities of udev. Since you can see the
/dev/x10/* directory, then you probably are just fine in the device driver
regard. Make sure that /dev/x10/.api exists and is writable by the user
that you are running the daemon under.

The error you are getting on /dev/ttyS1 (and probably /dev/ttyS0) appears to
be a problem with the daemon recognizing your tty device. I get the exact
same error on my system if I specify /dev/ttyS5 which doesn't exist. Even
though the /dev/ttyS0 and /dev/ttyS1 files exist, they may not have devices
attached. Run the "dmesg" command to look at your boot log and see if you
have a /dev/tty device found and what it named it. For example, here is
what my system says:

# dmesg | grep tty
ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A

Another way to do it is to look at /proc/tty/driver/serial and see if it
recognized a serial device on any of the known ports.

If you are absolutely sure that /dev/ttyS0 or /dev/ttyS1 exist as devices in
the kernel then it could be a problem communicating with the CM11A. The
error in the log doesn't indicate that it ever got to the point of even
being able to open the /dev/ttyS1 port. However, I have had a couple of
people that have noted that the drivers could not start the CM11A up and the
solution was to remove the batteries and let it sit for a while before
plugging in. This clears the memory and the drivers will often start
working. I have not figure out what is causing this behavior yet.

Scott

-----Original Message-----
Sent: Monday, October 04, 2004 3:55 PM
Subject: Problem with cm11a - Error opening /dev/ttyS1 (No
such device or address)
Hi,
I'm running into the following problem on a
linux-2.6.8-gentoo-r3 system. I patched my 2.6.8 kernel,
compiled the driver as a module and loaded it without any
problems, but it doesn't want to read my cm11a on ttyS1 or
ttyS0. I'm using udev and the
/dev/x10/* files are all there. Oh, and the device is
plugged in. :)
Any ideas?
Thanks,
Mark
# ls -l /dev/ttyS[01]
crw-rw---- 1 root tty 4, 64 Jul 17 20:14 /dev/ttyS0
crw-rw---- 1 root tty 4, 65 Jul 17 20:14 /dev/ttyS1
Oct 1 21:38:58 shamu2 x10: X10 DEV module v2.0.0
Oct 1 21:38:58 shamu2 x10: $Id: dev.c,v 1.19
2004/07/05 03:13:13 whiles Exp whiles $
Oct 1 21:38:58 shamu2 x10: X10 driver successfully
loaded
Oct 1 21:39:19 shamu2 cm11ad[11194]: starting.
api-device=/dev/x10/.api, pidfile=/var/run/x10d.pid Oct 1
21:39:19 shamu2 cm11ad[11194]: Successfully opened X10 API
Transmit thread starting Oct 1 21:39:20 shamu2
cm11ad[11195]: Error opening /dev/ttyS1 (No such device or
address) Oct 1 21:39:20 shamu2 cm11ad[11195]: unable to
start transmitter, exiting O
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com

Mark Mykkanen

2004-10-06 14:40:03 UTC

Permalink

# cm11ad -debug -device /dev/ttyS1
cm11ad, debug turned on
Argument 2: device
cm11ad, serial set to /dev/ttyS1
OK...let's start
cm11ad opening /dev/x10/.api for API
cm11ad starting
device=/dev/x10/.api
pidfile=/var/run/x10d.pid
logtag=cm11ad
X10 message: src=0x4, hc=0x0, uc=0x0, cmd=0x2, f=0x0

Oct 6 09:37:35 shamu2 cm11ad[19607]: starting.
api-device=/dev/x10/.api, pidfile=/var/run/x10d.pid
Oct 6 09:37:35 shamu2 cm11ad[19607]: Successfully
opened X10 API device /dev/x10/.api
Oct 6 09:37:36 shamu2 cm11ad[19608]: starting
transceiver (pthread_create)
Oct 6 09:37:36 shamu2 cm11ad[19608]: Transmit thread
starting
Oct 6 09:37:36 shamu2 cm11ad[19608]: pthread_crate
completed...transceiver started
Oct 6 09:37:36 shamu2 cm11ad[19608]: xmit_init:
received time request from CM11A
Oct 6 09:37:36 shamu2 cm11ad[19608]: xmit_init:
Wrote ffffff9b 24 61 04 17 ffffff88 00
Oct 6 09:37:37 shamu2 cm11ad[19608]: Unable to
establish communications with CM11A transceiver
Oct 6 09:37:37 shamu2 cm11ad[19608]: got connect
message
Oct 6 09:37:37 shamu2 cm11ad[19608]: unable to start
transmitter, exiting

Figured out that I had to compile the serial device
into the kernel. :)
Unfortunately, I'm still getting errors. I receive
the following when
executing: cm11ad -debug -device /dev/ttyS1
Oct 6 08:48:17 shamu2 x10: X10 driver unloaded
Oct 6 08:48:22 shamu2 x10: X10 DEV module v2.0.0
Oct 6 08:48:22 shamu2 x10: $Id: dev.c,v 1.19
2004/07/05 03:13:13
whiles Exp whiles $
Oct 6 08:48:22 shamu2 x10: X10 driver successfully
loaded
Oct 6 08:49:45 shamu2 cm11ad[19406]: starting.
api-device=/dev/x10/.api, pidfile=/var/run/x10d.pid
Oct 6 08:49:45 shamu2 cm11ad[19406]: Successfully
opened X10 API
device /dev/x10/.api
Oct 6 08:49:46 shamu2 cm11ad[19407]: starting
transceiver
(pthread_create)
Oct 6 08:49:46 shamu2 cm11ad[19407]: Transmit
thread
starting
Oct 6 08:49:46 shamu2 cm11ad[19407]: pthread_crate
completed...transceiver started
Oct 6 08:49:47 shamu2 cm11ad[19407]: Unable to
establish
communications with CM11A transceiver
Oct 6 08:49:47 shamu2 cm11ad[19407]: got connect
message
Oct 6 08:49:47 shamu2 cm11ad[19407]: unable to
start
transmitter,
exiting
I also tried unplugging and removing the batteries
(dead anyways?) for
60 seconds to no avail. Same results for ttyS0 and
ttyS1.
Mark
On Mon, Oct 04, 2004 at 11:10:01PM -0400, Scott
Hiles

Post by Scott Hiles
I doubt that udev has anything to do with it, but

the drivers do not
support

Post by Scott Hiles
any of the dynamic device capabilities of udev.

Since you can see
the

Post by Scott Hiles
/dev/x10/* directory, then you probably are just

fine in the device
driver

Post by Scott Hiles
regard. Make sure that /dev/x10/.api exists and

is
writable by the
user

Post by Scott Hiles
that you are running the daemon under.
The error you are getting on /dev/ttyS1 (and

probably /dev/ttyS0)
appears to

Post by Scott Hiles
be a problem with the daemon recognizing your tty

device. I get the
exact

Post by Scott Hiles
same error on my system if I specify /dev/ttyS5

which doesn't exist.
Even

Post by Scott Hiles
though the /dev/ttyS0 and /dev/ttyS1 files exist,

they may not have
devices

Post by Scott Hiles
attached. Run the "dmesg" command to look at your

boot log and see
if you

Post by Scott Hiles
have a /dev/tty device found and what it named it.

For example, here
is

Post by Scott Hiles
# dmesg | grep tty
ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A
Another way to do it is to look at

/proc/tty/driver/serial and see if
it

Post by Scott Hiles
recognized a serial device on any of the known

ports.

Post by Scott Hiles
If you are absolutely sure that /dev/ttyS0 or

/dev/ttyS1 exist as
devices in

Post by Scott Hiles
the kernel then it could be a problem

communicating
with the CM11A.
The

Post by Scott Hiles
error in the log doesn't indicate that it ever got

to the point of
even

Post by Scott Hiles
being able to open the /dev/ttyS1 port. However,

I
have had a couple
of

Post by Scott Hiles
people that have noted that the drivers could not

start the CM11A up
and the

Post by Scott Hiles
solution was to remove the batteries and let it

sit
for a while
before

Post by Scott Hiles
plugging in. This clears the memory and the

drivers
will often start

Post by Scott Hiles
working. I have not figure out what is causing

this
behavior yet.

Post by Scott Hiles
Scott

-----Original Message-----
From: Mark Mykkanen
Sent: Monday, October 04, 2004 3:55 PM
Subject: Problem with cm11a - Error opening

/dev/ttyS1 (No