[linux-dvb] HVR4000 Support

Darron Broad darron at kewl.org
Wed Mar 14 14:43:01 CET 2007


In message <45F77638.80508 at hauppauge.com>, Steven Toth wrote:
>Bob wrote:

hi guys

>> On Tuesday 13 March 2007 14:34, Steven Toth wrote:
>>   
>>> Hi,
>>>
>>> I've created a new tree based on the current mainline v4l/dvb tree,
>>> Manu's multiproto patches and the HVR400 specific patches. It can be
>>> found here http://linuxtv.org/hg/~stoth/hvr4000
>>>
>>> I don't have any immediate hardware available for test, so your mileage
>>> may vary.
>>>
>>> If you'd like to spend the time testing then I'll happy take
>>> feedback/bugs via this ML.
>>>
>>> Looking at the patch description:
>>>
>>> 1. DVB-T is not working, pending a GPIO change
>>> 2. DiSEqC is not working
>>>
>>> Regular DVB-S and DVB-S2 should work fine. Remember you'll need apps
>>> that support the multiproto API's to use the S2 functionality.
>>>
>>> Lastly, see my posting to the ML last week for instructions on obtaining
>>> the firmware.

I will test the above sometime later, yet in the meantime I retrofitted
the cx24116 demod driver into v4l-dvb hg minus the multiproto/dvb-s2 support and it
functions. A bug was discovered which is probably the cause of the issue
detailed below.

There is a tarball for if you are interested, plus a scan output and kaffeine
channel config for Astra-28.2E at http://dev.kewl.org/tmp/hvr4000/

>> Oh goody, we're getting there.
>>
>> When unloading the modules, I get:
>> isl6421                 6656  4294967295 
>> cx24116                17664  4294967295 
>>
>> Forcing there removal does not seem to harm the system.
>>
>> I'm not too sure whether it is handling the card correctly, although
>> it does see it
>>
>> Mar 13 20:31:49 eth5 kernel: cx2388x alsa driver version 0.0.6 loaded
>> Mar 13 20:31:49 eth5 kernel: PCI: Enabling device 0000:02:0a.1 (0110 -> 0112)
>> Mar 13 20:31:49 eth5 kernel: ACPI: PCI Interrupt 0000:02:0a.1[A] -> GSI 22 (level, low) -> IRQ 217
>> Mar 13 20:31:49 eth5 kernel: cx88[0]/1: CX88x/0: ALSA support for cx2388x boards
>> Mar 13 20:31:49 eth5 kernel: cx2388x dvb driver version 0.0.6 loaded
>> Mar 13 20:31:49 eth5 kernel: cx8802_register_driver() ->registering driver type=dvb access=shared
>> Mar 13 20:31:49 eth5 kernel: CORE cx88[0]: subsystem: 0070:6902, board: Hauppauge WinTV-HVR4000 DVB-S/S2/T/Hybrid [card=57]
>> Mar 13 20:31:49 eth5 kernel: cx88[0]/2: cx2388x based dvb card
>> Mar 13 20:31:49 eth5 kernel: cx24116: cx24116_attach
>> Mar 13 20:31:49 eth5 kernel: DVB: registering new adapter (cx88[0]).
>> Mar 13 20:31:49 eth5 kernel: DVB: registering frontend 1 (Conexant CX24116/CX24118)...
>> Mar 13 20:31:50 eth5 kernel: cx2388x blackbird driver version 0.0.6 loaded
>> Mar 13 20:31:50 eth5 kernel: cx8802_register_driver() ->registering driver type=blackbird access=shared
>> Mar 13 20:31:50 eth5 kernel: CORE cx88[0]: subsystem: 0070:6902, board: Hauppauge WinTV-HVR4000 DVB-S/S2/T/Hybrid [card=57]
>> Mar 13 20:31:50 eth5 kernel: cx8802_register_driver() ->probe failed err = -19
>>
>> Other than that, I still cannot get DVB-S or SVB-S2 to work.
>> That could be because my knowledge of about the sat stuff
>> is sadly lacking.
>>
>> Bob
>>
>>   
>Thanks for the feedback.
>
>Gregoire reported an issue via IRC today. It looks like the HVR4000 
>demod driver never receives the set_params call, to actually tune. I 
>think this is probably a bug in dvb_core - manu's patches.
>
>I plan to repro tomorrow, expect more progress in the next day or so.

There is a bug in set_fec which sets the local FEC value
to the value retrieved from the FEC array in error.

A fix for that is here: http://dev.kewl.org/tmp/hvr4000/stoth/cx24116.diff

This ought to solve the tuning issue, yet as I have not tested the multiproto
work, I cannot say for certain.

Bye

darron

--

 // /
{:)==={ Darron Broad <darron at kewl.org>
 \\ \ 




More information about the linux-dvb mailing list