I managed to do some testing of Christopher Pascoe's driver this evening with interesting results. I used the CVS version of linuxtv DVB along with Christopher's mt352 front end (the current one in the CVS didn't get nearly as far as Christopher's mt352.c from http://www.itee.uq.edu.au/~chrisp/DVICO-Linux/mt352.c). I also patched up the cx88-cards.c to reflect the addition of the Kworld/V-Stream card (attached) - the GPIO values replicate those used for the DVICO card, so it's not known if they're even correct. Using his startup script to load/unload modules, I get the following output in the kernel logs, indicating some activity: -- cx88_register_dvb_module Linux video capture interface: v1.00 cx2388x v4l2 driver version 0.0.4 loaded cx8800[0]: found at 0000:01:08.0, rev: 5, irq: 18, latency: 32, mmio: 0xde000000 cx8800[0]: subsystem: 17de:08a6, board: KWorld/VStream XPert DVB-t [card=11,autodetected] cx8800[0]: i2c register ok cx8800[0]: registered device video0 [v4l2] cx8800[0]: registered device vbi0 cx8800_dvb: Unknown symbol cx88_register_dvb_module cx8800[0]: couldn't register DVB module cx8800: Initialised TS code, frames_per_irq: 512, frames_to_buffer: 10240 cx88_register_dvb_module Trying DVB attach on card c4104000 DVB: registering new adapter (cx8800[0]). DVB attached to dev c4104000, dvb: cb1abc00 DVB: registering frontend 0:0 (Unknown Zarlink MT352-based DVB-T card)... -- Thereafter, I ran the "scan" program in an attempt to see if the "Unkown" MT352 handling actually worked for the Samsung tuner we're trying to interface with. Given the output (below), it doesn't seem to work yet: -- WARNING: >>> tuning failed!!! >>> tune to: 191625000:INVERSION_AUTO:BANDWIDTH_7_MHZ:FEC_3_4:FEC_NONE:QAM_64:TRANSMISSION_MODE_8K:GUARD_INTERVAL_1_16:HIERARCHY_NONE WARNING: >>> tuning failed!!! >>> tune to: 191625000:INVERSION_AUTO:BANDWIDTH_7_MHZ:FEC_3_4:FEC_NONE:QAM_64:TRANSMISSION_MODE_8K:GUARD_INTERVAL_1_16:HIERARCHY_NONE (tuning failed) WARNING: >>> tuning failed!!! >>> tune to: 219500000:INVERSION_AUTO:BANDWIDTH_7_MHZ:FEC_3_4:FEC_NONE:QAM_64:TRANSMISSION_MODE_8K:GUARD_INTERVAL_1_16:HIERARCHY_NONE WARNING: >>> tuning failed!!! >>> tune to: 219500000:INVERSION_AUTO:BANDWIDTH_7_MHZ:FEC_3_4:FEC_NONE:QAM_64:TRANSMISSION_MODE_8K:GUARD_INTERVAL_1_16:HIERARCHY_NONE (tuning failed) WARNING: >>> tuning failed!!! >>> tune to: 571500000:INVERSION_AUTO:BANDWIDTH_7_MHZ:FEC_2_3:FEC_NONE:QAM_64:TRANSMISSION_MODE_8K:GUARD_INTERVAL_1_8:HIERARCHY_NONE WARNING: >>> tuning failed!!! >>> tune to: 571500000:INVERSION_AUTO:BANDWIDTH_7_MHZ:FEC_2_3:FEC_NONE:QAM_64:TRANSMISSION_MODE_8K:GUARD_INTERVAL_1_8:HIERARCHY_NONE (tuning failed) WARNING: >>> tuning failed!!! ERROR: initial tuning failed dumping lists (0 services) Done. -- At the same time, there is some kernel log activity as scan is being run: -- buf9,10: 05 02 buf9,10: 05 02 buf9,10: 05 02 buf9,10: 05 02 buf9,10: 05 56 buf9,10: 05 56 buf9,10: 05 56 buf9,10: 05 56 buf9,10: 05 56 buf9,10: 05 56 buf9,10: 05 fe buf9,10: 05 fe buf9,10: 05 fe buf9,10: 05 fe buf9,10: 05 fe buf9,10: 05 fe buf9,10: 0e 3e buf9,10: 0e 3e buf9,10: 0e 3e buf9,10: 0e 3e buf9,10: 0e 3e buf9,10: 0e 3e -- It seems that we're quite close to getting this working, however the front end for the Samsung tuner still needs attention. Someone knowledgeable will need to find out the addresses used and any additional programming parameters and update the mt352 to suit. Is anyone working on this? Feedback would be appreciated. Thanks, Jonathan On Fri, 2 Jul 2004 01:39:08 +0100 (BST), Soyeb Aswat <soyeb@btinternet.com> wrote: > > --- Jonathan Thorpe <jlthorpe@gmail.com> wrote: > > Perhaps someone more knowledgeable can comment on > > this, but at this > > stage, we may still need addressing and programming > > information on the > > KWorld card to proceed. There are other cards around > > which use the > > same Samsung tuner, but with the 878 - would the > > same set of addresses > > and programming be used with the cx23883 when > > interfaced with the > > Samsung tuner? > > Dunno. I'm currently trying to get Chris Pascoes cx88 > driver to work but I'm not having much luck - this is > all very new to me. They build after lots of fiddling > with the kernel source tree, but then I get "disagrees > about version of symbol" errors when I try to load > them. > > > I noticed that you attempted to get in contact with > > KWorld with > > regards to obtaining progamming information for the > > card - did you > > receive the information you requested? > > Sadly I haven't had a reply (or I've had a reply and > my spam filter has eaten it). > > > If you have difficulty in getting this information > > from KWorld, and > > the addresses/programming of the tuner is different > > to that from other > > cards using the bt878, a bus analyser may be > > necessary. > > I sure hope not! > > Let me try and get Chris' driver working. > > > > Regards, > > Soyeb > > ===== > email: soyeb@btinternet.com > www: http://soyeb.homeip.net > work: http://www.tertio.com > ) > >
Attachment:
dv88_dvb_vstream.diff
Description: Binary data