Holger Waechtler wrote:
Niklas Peinecke wrote:
Hi everybody,
I'm sorry, I'm a bit confused now (and partly I think this is my own fault). Could somebody please tell me what I should do now?
* As suggested by Wolfgang I could implement a separate ref counter for every pid.
one refcounter per hardware PID.
Ok, done.
The version that incorporates all changes by Wolfgang and me is "skystar2_hwfilter20031204_2.diff" against today's CVS from this morning 11.45.I think 8 bit should be really sufficient?* Where should I start? Wolfgang submitted two patches, one should work with actual CVS. If not, I'll be waiting for feedback and then try to incorporate into the CVS.
ok, can you then please clearly mark which patch you refer to? I'm getting kind of confused due to all the different patch versions...
btw, Roberto and Augusto should now have their CVS account, it's probably better when they commit b2c2 related changes - they have the hardware available for testing.
Probably right. :)
* If you really can be sure that there will be no request to remove a 0x2000 pid that was not added, then Wolfgang is right and we need only one ref counter for open_whole_bandwidth.
better check the source twice for this scenario to be perfectly sure - :)
If not it's trivial to re-add the second counter.
Holger
Niklas -- Info: To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe linux-dvb" as subject.