14 replies to this topic

[Dimitrij]

WanWizard

stbmodel=H9TwinSE
stbmodel=h9combose
kernelcmdline=root=/dev/mmcblk0p23 rootsubdir=linuxrootfs4 rootfstype=ext4 kernel=/dev/mmcblk0p22 console=ttyAMA0,115200 blkdevparts=mmcblk0:1M(fastboot),1M(bootargs),18M(recovery),2M(bootoptions),2M(deviceinfo),4M(baseparam),4M(pqparam),4M(logo),40M(fastplay),20M(kernel),20M(trustedcore),4M(misc),1024M(system),150M(vendor),306M(e2tv),256M(cache),50M(private),8M(securestore),16M(linuxkernel1),16M(linuxkernel2),16M(linuxkernel3),16M(linuxkernel4),-(userdata) hbcomp=/dev/block/mmcblk0p13 mem=937M mmz=ddr,0,0,60M vmalloc=500M
nimsockets=NIM Socket 0:
imageissue=openpli 9.1-release %h


REMOVE Peer h9combose
REMOVE Peer h9combose
REMOVE Peer h9combose
Traceback (most recent call last):
  File "/usr/lib/enigma2/python/Screens/Ci.py", line 351, in ciStateChanged
IndexError: list index out of range
[ePyObject] (CallObject(<bound method CiMessageHandler.ciStateChanged of <Screens.Ci.CiMessageHandler object at 0xb29ab580>>,(1,)) failed)

Problem in enigma2.info

 

stbmodel=H9TwinSE
stbmodel=h9combose

[WanWizard]

More info please.

 

What box is this, a Twin SE ( 2 x DVB-S2 ) or a Combo SE ( 1 x DVB-S2 + 1 x DVB-C/T2 ) ?

 

What exactly, which key, is wrong in /usr/lib/enigma.info, and what is the correct value?

 

Runtime model detection happens here: https://github.com/O...ma-info.bb#L503

 

The box model is determined here: https://github.com/O...ma-info.bb#L376

[Dimitrij]

BoxInfo.setItem("CommonInterface", model in ("h9combo", "h9combose", "h10", "pulse4kmini") and 1 or eDVBCIInterfaces.getInstance().getNumOfSlots())

 

[WanWizard]

BoxInfo.setItem("CommonInterface", model in ("h9combo", "h9combose", "h10", "pulse4kmini") and 1 or eDVBCIInterfaces.getInstance().getNumOfSlots())

That isn't an answer to my question.

[Dimitrij]

 

openpli-9.1-release-h9combose-20250103_emmc 

zgemma h9 twinse

 
 

-model = BoxInfo.getItem("machine")
+model = BoxInfo.getItem("model")

or

elif [ "$MACHINE" = "h9combose" ]; then
	if [ "$model" = "h9twinse" ]; then
		updateinfo "displaymodel" "H9 TWIN SE"
+               updateinfo "machine" "h9twinse"
      

or

Edited by Dimitrij, 6 January 2025 - 06:32.

[littlesat]

Does this box has one or two CI… hardware info indicates one hardcoded via a hack because it looks the box drivers itself seems not correctly identify the number for CI interfaces.

[ghuss]

Thanks

 

It has :  2 x DVB Common Interface slots

[Dimitrij]

Thanks

 

It has :  2 x DVB Common Interface slots

+

name H9 TWIN SE

[WanWizard]

This can never have worked?

 

That bit still looks like "old" ( = pre BoxInfo ) code, and the model name of the box in /proc/stb is "h9combose", so before introduction of BoxInfo, it would also have limited the number of slots to 1 ?

 

The use of "machine" here is very wrong, as that is OE build MACHINE, that says nothing about the actual box.

 

So  I think two things are needed here:

-model = BoxInfo.getItem("machine")
+model = BoxInfo.getItem("model")

and

+ updateinfo "model" "h9twinse"

or, alternatively, use

-model = BoxInfo.getItem("machine")
+model = BoxInfo.getItem("machinebuild")

as that should correctly reflect the actual machine ( can you double check your enigma.info? ).

 

If we go for the first solution, more boxes need their "model" fixing, as the old code had a massive if structure to deduct the "model" from all sorts of /proc data.

[littlesat]

Then the box does also not have the proper procs to detect the CI…as the setting to 1 was already for ever a work-a-round….

Edited by littlesat, 6 January 2025 - 13:54.

[WanWizard]

That is exactly why I remarked it could have never worked.

[Dimitrij]

WanWizard

This problem H9 TWIN SE/H9 TWIN

Maybe really use BoxInfo.getItem("model")

this

HardwareInfo.py

self.device_model = self.machine_name = BoxInfo.getItem("machine")

and

SystemInfo.py

and

RcModel.py

Edited by Dimitrij, Yesterday, 09:34.

[littlesat]

My brain explode when I read back this thread.... I understand specific Zgemma boxes have 2 CIs instead of one... they do not have the proper procs to detect them so we hardcoded the number of procs. We have a list of Zgemma boxes that worked around it.... Is this work-a-round not correct? Is the box model not correct detected? I'm going in circles and circles here so I cannot define any solution....

Edited by littlesat, Yesterday, 10:23.

[salim1965]

https://cdn.resimupl...50107232933.jpg

 

https://cdn.resimupl...50107232413.jpg

[salim1965]

zgemma h9 twinsSE software was given instead of combeSE software. Twinse is visible in 9.0, combese is visible in 9.1, twinse is visible in 8.3.