stv090x vs stv0900

Hi,

I was adding support for a non working version of DVBWorld HD 2104

It is listed on
http://www.linuxtv.org/wiki/index.php/DVBWorld_HD_2104_FTA_USB_Box as :

=====
for new solution : 2104B (Sharp0169 Tuner)

* STV6110A tuner
* ST0903 demod
* Cyrix CY7C68013A USB controller
=====

The 2104A is supposed to be working and also have ST0903 but uses stv0900, so I tried using it too but did not manage to get it working.

I now have some working code by using stv090x + stv6110x (copied config from budget) but I am wondering why do we have 2 drivers for stv0900, and is stv0900 supposed to handle stv0903 devices or is either the code or the wki wrong about 2104A?

Also, are they both maintained ? I wrote a patch to add get_frontend to stv090x but stv0900 also does not have it and I don't know which one should get new code.

And stv6110x seems to also handle stv6110 which also exists as a separate module...

Hi Pascal,

Hehe, you are not only one who is fooled by current situation regarding demods stv0900/stv0903 and plls stv6110.

Current status-quo is not good. Same question is asked here again and again. Interesting that it is against usual rule having only one driver for every chip in kernel, but this time that rather strong rule was not applied. Dunno why.

May be someone from "knowings" can answer that on wiki?

/Honza

PS: FYI I'm also using BOTH variants in 2 projects. For me both look very similar and work w/o probs. So I can't tell you which is better :)

But it working. I have the device and test it succesfully.

Code for stv0900 supposed to handle stv0903 devices as well. And it handles. Feel free to add for stv0900. I don't impede.

In time when I commit stv0900(stv6110 as well) there wasn't any driver for stv0900/stv0903. Now I'm wondering like you.

Look here:

root@useri:/etc/default# modinfo dvb-usb-dw2102
filename:       /lib/modules/2.6.34-rc6/kernel/drivers/media/dvb/dvb-usb/dvb-usb-dw2102.ko
license:        GPL
version:        0.1
description:    Driver for DVBWorld DVB-S 2101, 2102, DVB-S2 2104, DVB-C 3101 USB2.0, TeVii S600, 
S630, S650, S660 USB2.0, Prof 1100, 7500 USB2.0 devices
author:         Igor M. Liplianin (c) [email protected]
srcversion:     496C3974FA3791E61D28672
alias:          usb:v3034p7500d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v9022pD660d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v3011pB012d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v9022pD630d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v04B4p3101d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v0CCDp0064d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v9022pD650d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v04B4p2104d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v04B4p2101d*dc*dsc*dp*ic*isc*ip*
alias:          usb:v04B4p2102d*dc*dsc*dp*ic*isc*ip*
depends:        dvb-usb,i2c-core
vermagic:       2.6.34-rc6 SMP mod_unload PENTIUM4 
parm:           debug:set debugging level (1=info 2=xfer 4=rc(or-able)). (int)
parm:           keymap:set keymap 0=default 1=dvbworld 2=tevii 3=tbs  ... (int)
parm:           demod:demod to probe (1=cx24116 2=stv0903+stv6110 4=stv0903+stb6100(or-able)). 
(int)
parm:           adapter_nr:DVB adapter numbers (array of short)

So modprobe dvb-usb-dw2102 demod=2 brings DVBWorld 2104A to you on golden plate.

Best Regards

Igor M. Liplianin
Microsoft Windows Free Zone - Linux used for all Computing Tasks

2010/5/9 Igor M. Liplianin <[email protected]>:

Quite strange indeed..

http://www.linuxtv.org/pipermail/linux-dvb/2008-November/030492.html

http://www.mail-archive.com/[email protected]/msg13353.html

OK, but the B does not here while it seems to work with stv090x

All I get in log with femon is

stv0900_read_status: <7>DEMOD LOCK FAIL <6>stv0900_get_rf_level <6>stv0900_get_rf_level: RFLevel = -100 <6>stv0900_carr_get_quality

And with scandvb :

stv0900_read_status: <7>DEMOD LOCK FAIL
stv0900_read_status: <7>DEMOD LOCK FAIL
stv0900_read_status: <7>DEMOD LOCK FAIL
stv0900_read_status: <7>DEMOD LOCK FAIL <6>stv0900_search: <7>stv0900_set_tuner: Frequency=1939000
stv0900_set_tuner: Bandwidth=72000000 <6>stv0900_activate_s2_modcode
<6>Search Fail
stv0900_read_status: <7>DEMOD LOCK FAIL

Yes but this one is 2104B

В сообщении от 10 мая 2010 14:44:08 автор Pascal Terjan написал:

OK, modprobe dvb-usb-dw2102 demod=2 brings DVBWorld 2104B to you on golden plate Look what I was tested already:
1. DVB-S2 USB 2104A (EARDA4B47, STB6100 + ST0903)
2. DVB-S2 USB 2104B (sharp0169, stv6110A + ST0903)
3. DVB-S2 USB 2104C (cx24116)
4. DVB-S2 USB 2104D (montage_ts2020 + ds3000)
5. DVB-S2 USB 2104E (SERIT2636, stv6110A + ST0903)

If someone interested:
md5sum dvb-usb-dw2104*.fw*

4cd4215e169c42f6d05cf23be2edd907  dvb-usb-dw2104.fw
4cd4215e169c42f6d05cf23be2edd907  dvb-usb-dw2104.fw.keep
9ccab99bdd4a0f252455d08dee189794  dvb-usb-dw2104d.fw
9ccab99bdd4a0f252455d08dee189794  dvb-usb-dw2104e.fw

BR
Igor

Igor M. Liplianin
Microsoft Windows Free Zone - Linux used for all Computing Tasks

Hello,

I have added get_frontend to stv0900 two months ago -
http://linuxtv.org/hg/v4l-dvb/rev/a3e28fbefdc3

I'm trying to describe my point of view about two drivers for stv6110 +stv0900.

History:
I have anounced our card on November 2008 -
http://www.linuxtv.org/pipermail/linux-dvb/2008-November/030439.html
As you can see I have mentioned that we developing code and will be publish it under GPL. All people in ML received this message. This should be prevent of duplicate work.
Also we have obtained permission (signed letter) from STM to publish resulting code under GPL. We have send pull request at Feb 2009 -
http://www.mail-archive.com/[email protected]/msg02180.html

(stv090x commit requested later - in May 2009 -
http://www.mail-archive.com/[email protected]/msg04978.html ).

Solution:
Ideally two drivers should be combined into one. stv0900 driver can be used as starting point. We (NetUP Inc.) can initiaite this job. But we need approval from Manu and all community who using stv090x. Manu what do you think about this ?
This is not trivial because some "features" may be broken when combining this two code.

Abylai Ospan <[email protected]>
NetUP Inc.

P.S.

engeneers designed "NetUP Dual DVB-S2 CI" card from "scratch". We know many nuances about this IC's. For example, we have tested 16APSK/32APSK -
http://linuxtv.org/wiki/index.php/STMicroelectronics_STV0900A_16APSK_32APSK

Also we developing new version of our "NetUP Dual DVB-S2 CI" based on FPGA (like our Dual DVB-T/C CI card -
http://linuxtv.org/wiki/index.php/NetUP_Dual_DVB_T_C_CI_RF ). This new version of card can proceed high bitrates from STV0900 (120Mbps and higher ). Also this card can receive raw frames from STV0900 ( not only TS ) for extra functionality ( GSE ).

The STV090x driver supports both the STV0900 and STV0903 broadcast demodulators very well and in multiple configurations on couple of bridges (SAA7146, nGene, SAA7160, some USB bridges, some others also in the process) and is quite a generic one. Currently, the STV090x driver handles a lot of quirks which has been documented as well as undocumented by STM, basically being actively supported by STM themselves and by a few card vendors as well, other than for this community to a very great extend.

Currently, since the driver has been fine tuned to this great extend and so many users of it, I have no plans of migrating to the STV0900 driver, which was developed later on and tested to a much lesser extend, by developers and vendors alike. So, I am much less inclined to go the STV0900 way.

At the behest of STM themselves, I have withheld from adding support for the AAB/AAC chips initially, due to the basic reasons that these advanced features are not really well tested in the field, due to the lack of that many users and the highly unlikely chance that the advanced stuff (ACM/VCM) is yet to be deployed for a home user segment. In the chance that there are more likely users for the advanced stuff, these can be added quite easily to the STV090x/STV6110x driver.

Also, interesting point to be noted is that there are no users for ACM/VCM in the Linux userspace for the driver to be supported for "any" broadcast purposes. The only case in where the Advanced chips are being sold (that too at a higher price) is where card manufacturers who cannot commit themselves to that volume, not that it adds any value to the general user as it is, at the this time. Simply adding in code and making it unmaintainable makes no sense to me at least.