On Tue, Nov 15, 2011 at 11:12 PM, Kyle Strickland <k...@kyle.strickland.name> wrote: > Hi folks, > > I'm messing around trying to get my Kworld PC150-U ATSC hybrid tuner > card (17de:a134) to work, and I think I'm almost done. I've been able > to get all of the functionality to work on its own: remote control, > FM, NTSC, composite input, and ATSC, but for some reason, whenever I > use the ATSC decoder, the remote control stops working. > > Here's what I know about the card so far: > > Analog Decoder chip: SAA7134HL > Tuner: TDA8290+TDA18271 > ATSC/ClearQAM Decoder: Samsung S5H1411, TS using parallel input, chip > enabled by setting GPIO18 high > Remote control: I2C with GPIO8 going low signalling a key press, > similar to MSI TV@nywhere Plus, but with different GPIO pin used for > signalling. > Previous thread about this card found here: > http://article.gmane.org/gmane.linux.drivers.video-input-infrastructure/25680/ > > Looking at the pictures on newegg, the card looks to be the same as > the MSI Digi@nywhere A/D Plus, except with a green mask instead of > red, a different symbol on the remote control media button, and it > doesn't include batteries for the remote :) The EEPROM contents are > even the same except for the subdevice id at the beginning. > > So, whenever I run VLC to watch DTV, button presses stop showing up > with the GPIO, and if I force it to poll I2C rather than checking > GPIO8, I get a read error of either BUSY or ST_ERR or NO_DEVICE. For > that matter, I get a lot of BUSY and ST_ERR, but not NO_DEVICE, with > the S5H1411 in general, and hacked together a retry loop for reading > and writing registers to get past it. I would suspect just funkiness > on the I2C bus, but GPIO8 rarely registers the key press when watching > DTV. When it does register the key press, I am able to read the key > code over I2C. > > I'm completely new to working with tuner cards (just since I bought > this one a month ago), so I was wondering if any of you out there have > had a similar experience with other cards and might have some advice. > I'd appreciate any help I can get. > > I'll be happy to provide code, but I am not running the latest kernel > yet. I'm just making the modifications to the ubuntu source, to see > whether I could get the thing to work before I go all in building and > running the kernel from source control > > Distribution: Ubuntu 11.10 > uname -a: Linux home 3.0.0-12-generic #20 SMP Tue Nov 1 00:33:52 EDT > 2011 x86_64 x86_64 x86_64 GNU/Linux > Modifications made to s5h1411.ko, saa7134.ko, saa7134-alsa.ko, > saa7134-dvb.ko, tda8290.ko, tda18271.ko and new module > rc_kworld_pc150u.ko. > > Thanks, > Kyle Strickland >
In case anyone happens to be looking into this now or in the future ( see http://xkcd.com/979/ ), I found the problem, and it has to do with not reading enough bytes in s5h1411_readreg() in s5h1411.c. When I set .len to 3 (originally 2), it reads one more byte off of the i2c bus and leaves the bus in a working state for the IR device. I haven't figured out why it was set to 2 originally, or what the Right Way is to solve the problem so as not to break s5h1411 working already with other decoder chips. Any ideas? Thanks, Kyle Strickland -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
