Hi, Hermann, Mauro
hermann pitton schrieb:
> Hi,
>
> Am Samstag, den 26.04.2008, 20:19 -0300 schrieb Mauro Carvalho Chehab:
>> On Sun, 27 Apr 2008 00:10:21 +0200
>> hermann pitton <hermann-pitton@xxxxxxxx> wrote:
>>> Cool stuff!
>>>
>>> Works immediately for all tuners again. Analog TV, radio and DVB-T on
>>> that machine is tested.
>>>
>>> Reviewed-by: Hermann Pitton <hermann-pitton@xxxxxxxx>
>> Thanks. I'll add it to the patch.
>>
>>> Maybe Hartmut can help too, but I will test also on the triple stuff and
>>> the FMD1216ME/I MK3 hybrid tomorrow.
>> Thanks.
>>
>> It would be helpful if tda9887 conf could also be validated. I didn't touch at
>> the logic, but I saw some weird things:
>>
>> For example, SAA7134_BOARD_PHILIPS_EUROPA defines this:
>> .tda9887_conf = TDA9887_PRESENT | TDA9887_PORT1_ACTIVE
>>
>> And SAA7134_BOARD_PHILIPS_SNAKE keep the default values.
>>
>> However, there's an autodetection code that changes from EUROPA to SNAKE,
>> without cleaning tda9887_conf:
>>
>> case SAA7134_BOARD_PHILIPS_EUROPA:
>> if (dev->autodetected && (dev->eedata[0x41] == 0x1c)) {
>> /* Reconfigure board as Snake reference design */
>> dev->board = SAA7134_BOARD_PHILIPS_SNAKE;
>> dev->tuner_type = saa7134_boards[dev->board].tuner_type;
>> printk(KERN_INFO "%s: Reconfigured board as %s\n",
>> dev->name, saa7134_boards[dev->board].name);
>> break;
>>
>> I'm not sure if .tda9887_conf is missing at SNAKE board entry, or if the above
>> code should be doing, instead:
>>
>> dev->tda9887_conf = saa7134_boards[dev->board].tda9887_conf;
>>
>> If the right thing to do is to initialize SNAKE with the same tda9887
>> parameters as EUROPE, the better would be to add the .tda9887_conf to SNAKE
>> entry.
>>
>> Cheers,
>> Mauro
>
> Hartmut has the board and knows better, but it looks like it only has
> DVB-S and external analog video inputs. There is TUNER_ABSENT set, no
> analog tuner, no tda9887 and also no DVB-T, but it unfortunately shares
> the subsystem with the Philips Europa.
>
Hermann is right, SNAKE has no analog tuner. These boards indeed share the same PCI ID,
This code fragment reads the tuner ID from the eeprom to find out which board is there.
> I notice some unwanted behavior when testing md7134 FMD1216ME hybrid
> boards.
>
Aha! I modified my board that it no longer runs with the current driver. But i observed
something similar
> Unchanged is that the tda9887 is not up for analog after boot.
> Previously one did reload "tuner" just once and was done.
>
<snip>
Don't have the time today, but lets roll back history: Not absolutely sure but if
i remember correcly, the initialization sequence can be critical with hybrid tuners /
NIM modules. The tda9887 may only be visible on I2C after a certain bit in the MOPLL
is set (in byte4?)
Best regards
Hartmut
_______________________________________________
linux-dvb mailing list
linux-dvb@xxxxxxxxxxx
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
