Thomas Pundt wrote: > Hi, > > I'd really like to see some support for RDS/RBDS in v4l2; > there is a flag V4L2_CAP_RDS_CAPTURE, but AFAIK right now > nothing uses it. It isn't complete yet. The idea to do that is to have some struct v4l2_rds_format with the format specification. One more buffer type for RDS, put v4l2_rds_format into the struct v4l2_format union. Use VIDIO_S_FMT to ask the device driver to return RDS data in the specified format. Then simply use read() to receive the actual data. One format I'd like to see is just the raw data (i.e. the tree-byte-group blknr, byte1, byte2). Decoding in userspace. Maybe we also need other formats for devices which can't hand out raw data. > My question: I'm wondering how and where RDS support can be > added to v4l2? See above .. > For the SAA7134 driver, would it make sense to add > another flag to "struct saa7134_board"? rds should be i2c module, it can be used with other devices too then. It is possible to build bt878-based boards with the same rds chip ... the saa6588 should implement the v4l2 calls via i2c_command, so all the saa7134 driver has to do is to pass the ioctls to the driver (like it is done with the v4l1 ioctls for the tuner and the i2c audio modules). Gerd -- You can't please everybody. And usually if you _try_ to please everybody, the end result is one big mess. -- Linus Torvalds, 2002-04-20
