Jump to content


Photo

ioctl commands for the hdmi-cec device on vu+ ultimo


  • Please log in to reply
4 replies to this topic

#1 sirwio

  • Member
  • 13 posts

0
Neutral

Posted 13 February 2013 - 20:38

I'm investigating some hdmi-cec issues I'm facing on my vu+ ultimo.

Anyone out there has doc's or samples of the ioctl calls one can make to communicate with the hdmi-cec device.

Of interest would be:
* Get/Set ack bit part of the header block of the CEC frame.
* How to get the firmware version
* How to get vendor id
* How to set information i.e. the reverse command of ::ioctl(hdmiFd, 1, &addressinfo) used to get information.
* and others

I realize that what I'm requesting may not even be available or not known.

Btw. The issue I'm facing is that I cannot get the ultimo to be registered by my samsung tv as a recording device even though one send the "Report Physical Address" 0x84 with the correct recording device type set.


- Magnus

Re: ioctl commands for the hdmi-cec device on vu+ ultimo #2 ims

  • PLi® Core member
  • 13,625 posts

+212
Excellent

Posted 13 February 2013 - 23:52

a little test you can make with http://openpli.org/f...883#entry317883
Kdo nic nedělá, nic nezkazí!

Re: ioctl commands for the hdmi-cec device on vu+ ultimo #3 sirwio

  • Member
  • 13 posts

0
Neutral

Posted 14 February 2013 - 23:41

No difference but its a nice plugin to monitor the messages sent/recevied between the cec devices. Sort of an expected since one would assume that the hdmitest plugin is using the same hdmi userspace driver as implemented in enigma2.

Samsung appears to be implementing the hdme-cec very much by the book. If a device announce itself to be of type tuner it will simply only make available buttons on the remote that make sense for a tuner device.

I suspect that one need to change the type and logical address of the cec controller before reporting the physical address to the tv since it appears the type and logical address are ignored/overridden when sending a "report physical" address message. The chip does when queried always report the type to be a tuner.

To set/change the type one may need to use ioctl calls. Problem is that I can't find any documentation on these. Or it may be that its not possible to set. In that case one would need to contact vu+ to have them change the type from tuner to recording device in the kernel module.

Re: ioctl commands for the hdmi-cec device on vu+ ultimo #4 jeffphil

  • Senior Member
  • 116 posts

+7
Neutral

Posted 15 February 2013 - 00:25

Even better would be if the device is capable of recording ie has a valid recording path then announce as a recording device but when it doesn't announce as a tuner.
Until The Wheels are off the ground there is still hope


DM500,DM600,DM800HD,VU+ Duo

Re: ioctl commands for the hdmi-cec device on vu+ ultimo #5 sirwio

  • Member
  • 13 posts

0
Neutral

Posted 15 February 2013 - 18:25

Hooked up my raspberry pi to the same tv to be able to analyse a bit in more details what was send on the line from the vu+ ultimo. A bit surprising was that a request sent from the pi <Give Physical Address> 0x83 to the ultimo did get the ultimo to send a proper <Report Physical Address> 0x84. I had patched the driver to send the type recording device (1) and not the type tuner reported by the cec device on the ultimo when queried using ioctl calls.

Something is fishy. I will monitor the startup sequence of the pi when it announces itself as a recording device to see if I can spot any differences.

I also hope to get some progress in my effort of porting the libcec library to the ultimo.


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users