Re: Fedora 14 & Hauppauge Nova-T & devinput

[Nigel J. <ni...@gr...>]

On 13/02/2011 10:36, Nigel Jewell wrote:
> On 13/02/2011 10:08, Nick Morrott wrote:
>> On 13 February 2011 08:52, Nigel Jewell<ni...@gr...>  wrote:
>>> I have a machine that has three Hauppauge Nova-T's in it.  With CentOS 5
>>> (x86_64) I was able to pick the right event file from /dev/input and
>>> then use a lirc built with the devinput driver to receive ir codes.
>>> Worked fine.
>> Were you using any udev rules to control/create the /dev/input devices
>> being created when these cards were modprobed?
>
> No ... but in /etc/rc.d/rc.local I have a couple of lines that do 
> something similar.  I'm 99% sure that I've got the right device as 
> when I "cat" the device and press buttons on the remote I see characters.
>
>>> Now I'm getting to point where I am stumped.  How can I try and work out
>>> what is going on here and why irw is not outputting anything?
>> Which startup options is lircd starting up with (look at
>> /etc/sysconfig/lirc on Fedora)? Older LIRC installs needed to know
>> which /dev/input/eventX device to connect to (LIRC_DEVICE), in
>> addition to the LIRC driver to use (LIRC_DRIVER). If either of these
>> are not configured correctly, update these values and restart lircd
>> (note that irw needs lircd running correctly in order to produce
>> output).
>
> I've tried running the daemon manually with:
>
> lircd -H devinput -d /dev/input/event5 -n /etc/lirc/lircd.conf
>
> and with the following /etc/sysconfig/lirc.conf
>
>     # Options to lircd(8).  Typically, this will be empty, as which
>     driver to use
>     # should be specified using the LIRC_DRIVER variable below.
>     LIRCD_OPTIONS=""
>
>     # The infrared receiver (and/or transmitter) driver to be used by
>     lircd(8),
>     # similar to passing "-H driver" to lircd(8).
>     # Run "/usr/sbin/lircd -H help" to get a listing of supported drivers.
>     LIRC_DRIVER="devinput"
>
>     # Which lirc device will be used by lircd(8).
>     # This is the same as passing "-d device" to lircd.
>     # An empty value will use the default /dev/lirc0 device.
>     LIRC_DEVICE="/dev/input/event5"
>
>     # If "yes", the init script will try to start lircmd(8) too.
>     ENABLE_LIRCMD="no"
>
>     # Options to lircmd(8).
>     LIRCMD_OPTIONS=""
>
>
> Regards,
> Nige.

Okay ... still not getting anywhere with this.

I've just downloaded lirc-0.8.7-1.fc14.src.rpm and rebuilt it with 
debugging enabled (./configure --enable-debug).

The device:

[root@pvr ~]# ls -al /dev/input/by-path/pci-0000:04:02.0-event-ir
lrwxrwxrwx. 1 root root 9 Feb 13 20:54 
/dev/input/by-path/pci-0000:04:02.0-event-ir -> ../event5

Should there be an entry in /dev/input/by-id?? There isn't if there should.

My symlink:

[root@pvr ~]# ls -al /dev/input/remote
lrwxrwxrwx. 1 root root 17 Feb 13 20:54 /dev/input/remote -> 
/dev/input/event5

Start lircd:

[root@pvr ~]# lircd -H devinput -d /dev/input/remote -n -D1 
/etc/lirc/lircd.conf > lircd.log 2>&1

Stopped with Ctrl-C after a while ...

Examine log:

[root@pvr ~]# head lircd.log
lircd-0.8.7[7105]: started server socket
lircd-0.8.7[7105]: parsing remote
lircd-0.8.7[7105]: parsing devinput remote
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: first signal is a space!

[root@pvr ~]# tail lircd.log
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: first signal is a space!
lircd-0.8.7[7105]: nothing to send
lircd-0.8.7[7105]: lengths: 132799 132799 132799 132799
lircd-0.8.7[7105]: config file read
lircd-0.8.7[7105]: lircd(devinput) ready, using /var/run/lirc/lircd
lircd-0.8.7[7105]: caught signal

The "first single is a space!" comment repeats ... a lot.  Is this normal??

Today I've read A LOT of messages on different lists and forums about 
fixable issues with lirc on Fedora 14.  I appreciate that this is due to 
Jarod's changes in 2.6.35+ and that I may need to make some config 
changes, but I've tried everything I can think of.

Clutching at straws I'm wondering if that fact that I have 3 cards with 
the same ir interface is causing some issue (can't think what!) even 
though I'm reading from the correct device for the card I'm using for ir.

Any other ideas?  I don't want to start looking at buying another ir 
device to solve this when I have three attached to the Nova-Ts that 
worked perfectly well with CentOS 4 and 5.

Regards,
Nige.