Hi, following earlier patches [1], [2], here's another version. I've added the remote for the Cinergy C card, and simplified it even a bit. The remote of the Cinergy C appears to be the same physically as the one of the Cinergy DT USB XS Diversity. If you compare the key definitions to the one in dib0700_devices.c you can see some differences: 1) I've mapped one button differently since the dib driver has two physical keys mapped to the same keycode... 2) the codes received in the dib driver are the same, except they are bit-inverted (the first 6 bits) and then the dib driver will always use the seventh bit instead of the sixth. I don't know if that's because the remote is actually different or if one of the drivers interprets the values a bit wrong... As for simplification, the driver got out of the non-working auto-repeat business... So no longer any worker thread when we know there's no work to do (actually should probably use work_struct instead of delayed_work now). The trouble with the previous auto-repeat code was that at least for the Cinergy C remote it can't work. This remote has an initial delay (initial button to first repeat) of ~270ms, and time between further repeats is ~220ms. So the code using polling 10 times per second didn't work and always registered single keypresses anyway. It couldn't be easily fixed, since the remote's repeat delay is larger than the fixed delay for repeats of the input layer (250ms) - meaning you wouldn't be able to do single key presses any longer... (I don't know if this really worked for the vp-2033 remote - the driver will now print out the time between irqs if the verbose option is set). So repeats are a bit slow (in fact annoyingly slow imho, with not even 5 per seconds here), but even a proper auto-repeat solution can't really fix this - if you'd emulate more repeats per second you obviously would not be able to stop pressing a key precisely, which is just as annoying. btw enabling verbose is a bit annoying due to the cam slot polling code filling up your log file even when the driver is just sitting idle - are people really expected to be able to insert/remove smartcards 10 times a second? Roland [1] http://www.linuxtv.org/pipermail/linux-dvb/2007-April/017403.html [2] http://www.linuxtv.org/pipermail/linux-dvb/2008-March/024301.html
Attachment:
mantis-rc2.diff
Description: application/mbox
_______________________________________________ linux-dvb mailing list linux-dvb@xxxxxxxxxxx http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
