Jump to content


Photo

HdmiCEC crash - RC 9.0


  • Please log in to reply
29 replies to this topic

Re: HdmiCEC crash - RC 9.0 #21 WanWizard

  • PLi® Core member
  • 68,796 posts

+1,743
Excellent

Posted 26 August 2023 - 17:25

CEC is an single electrical bus (altough it is logically represented as a tree), with a hub in the center, which is 99.99% of cases built into the TV.

 

The hub numbers the devices, and itself has address 0.0.0.0, the device connected to HDMI 1 will be 1.0.0.0, the device on HDMI 2 will be 2.0.0.0, and so on.

 

Technically, you can have a second hub connected to the TV, for example an HDMI enabled amp. If you connect that to HDMI 1, it (the hub in the amp) will get address 1.0.0.0, and the device on HDMI 1 of the amp will get 1.1.0.0.

 

This is fixed, so I have no idea why one would want to fix the address, I have never ever found a reason for it.


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Pro (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.


Re: HdmiCEC crash - RC 9.0 #22 littlesat

  • PLi® Core member
  • 56,417 posts

+692
Excellent

Posted 27 August 2023 - 12:35

This is fixed, so I have no idea why one would want to fix the address, I have never ever found a reason for it.

+1


WaveFrontier 28.2E | 23.5E | 19.2E | 16E | 13E | 10/9E | 7E | 5E | 1W | 4/5W | 15W


Re: HdmiCEC crash - RC 9.0 #23 ims

  • PLi® Core member
  • 13,626 posts

+212
Excellent

Posted 27 August 2023 - 12:40

Does it bother anyone?!

When you want connect box from one tv to second tv (where is connected more boxes yet) ,boxes have stored own address and 2 boxes have same ?


Kdo nic nedělá, nic nezkazí!

Re: HdmiCEC crash - RC 9.0 #24 WanWizard

  • PLi® Core member
  • 68,796 posts

+1,743
Excellent

Posted 27 August 2023 - 12:55

For example, but also if you connect the box to a different port of the same TV, a box with a fixed address will no longer work.


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Pro (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.


Re: HdmiCEC crash - RC 9.0 #25 Stan

  • Senior Member
  • 346 posts

0
Neutral

Posted 27 August 2023 - 13:40

Does it bother anyone?!

 

Yes. The physical address allocation takes place every time the box is connected to the TV-set.

Setting a fixed address is clearly a violation of the CEC specification.



Re: HdmiCEC crash - RC 9.0 #26 Stan

  • Senior Member
  • 346 posts

0
Neutral

Posted 31 August 2023 - 12:22

Is the PLi team working on this?

 

I found more issues with HdmiCEC and the HDMItest plugin.

 

Should I post them here or maybe send a PM to the developer? Sometimes things get overlooked in the forum...


Edited by Stan, 31 August 2023 - 12:23.


Re: HdmiCEC crash - RC 9.0 #27 Dimitrij

  • PLi® Core member
  • 10,044 posts

+339
Excellent

Posted 1 September 2023 - 07:33

Stan

please test /usr/lib/enigma2/python/Components/HdmiCec.py

 

 


GigaBlue UHD Quad 4K /Lunix3-4K/Solo 4K


Re: HdmiCEC crash - RC 9.0 #28 Stan

  • Senior Member
  • 346 posts

0
Neutral

Posted 1 September 2023 - 21:19

Thanks... Unfortunately, I won't be able to test much in the next couple of days. Busy elsewhere... B)

 

In the meantime, can I leave you with some other findings related to HdmiCec to look at or think about?

 

Firstly, about the HDMI logfile (hdmicec.log)... Some RXdata, which can be represented as plain text, is not readable because of non-printable characters.
I've included a sample below. (The data of command 32 should be displayed as "ger"). Could it be converted before writing to logfile?

Attached File  hdmicec.log   1.49KB   1 downloads

 

Also, some of the received commands are not logged with their name (like the message A0 in the attached logfiile).

It's not important, but would be nice and easier to read. Maybe a few more common command names can be added. (especially broadcast messages)

 

About the infamous fixed physical address - I stronly suggest to remove this non-compliant option. It has no added value, only creates problems.
Besides, it is a violation of the HDMI-CEC standard. As for testing with a spoofed CEC address, one can always use the HDMItest plugin.

Next... On the Hdmi CEC setings page, the physical address is hidden when CEC is set to Disable. In earlier days it was always visible.
Can this be reverted? It should always be visible. The physical address allocation takes place, regardless if CEC feature are used or not!

Moving on... When the box is booted, the various Receiver models have different concepts for dealing with the initial physical address.
Some boxes will start with physical address to F.F.F.F, which is the correct behaviour.

But some boxes keep their previous (!) address instead. This behaviour is most questionable.

Is there any way to control these boxes to make them behave compliant?

Can the address be set to F.F.F.F on reboot? Or is this something hardcoded in the drivers?

 

Almost done now... The screen "Information" - "About" also shows the CEC address. But strangely, only if a fixed address is used.
Can it be changed to show the actual physical address? And could additionally be mentionend there, if the physical address has been assigned through the allocation process? (as opposed to the (invalid) self-assigned address)

 

Finally, considering the above, the receiver should not send any CEC commands prior to registration on the CEC-bus. (aka physical address allocation)

 

Hopefully the issues will be discussed and addressed by the PLi team. :) I consider these basics.

Once in place, the next step could be about scenarios, options, presets, and how to fully take advantage of the CEC capabilities with different TV types, sound systems, etc.

 

And btw. Often referenced in the forum is HDMI Specification 1.3. I think, the better version to consider is 1.4b (2011).


Edited by Stan, 1 September 2023 - 21:29.


Re: HdmiCEC crash - RC 9.0 #29 Dimitrij

  • PLi® Core member
  • 10,044 posts

+339
Excellent

Posted 2 September 2023 - 06:40

[HdmiCec] sanity check decode/add some cmd


GigaBlue UHD Quad 4K /Lunix3-4K/Solo 4K


Re: HdmiCEC crash - RC 9.0 #30 WanWizard

  • PLi® Core member
  • 68,796 posts

+1,743
Excellent

Posted 2 September 2023 - 18:17

Merged.


Currently in use: VU+ Duo 4K (2xFBC S2), VU+ Solo 4K (1xFBC S2), uClan Usytm 4K Pro (S2+T2), Octagon SF8008 (S2+T2), Zgemma H9.2H (S2+T2)

Due to my bad health, I will not be very active at times and may be slow to respond. I will not read the forum or PM on a regular basis.

Many answers to your question can be found in our new and improved wiki.



1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users