On 10.04.2017 10:14, KimJeongYeon wrote: > > > 2017. 4. 9. ì?¤ì ? 5:22ì?? "Georg Chini" <georg at chini.tk > <mailto:georg at chini.tk>>ë??ì?´ ì??ì?±: > > On 07.04.2017 15:41, KimJeongYeon wrote: > > For example, a normal stream tried to attach to filter sink(or > source), which > filter loaded and managed by filter-apply. But, the stream > become to attach to > the ***master sink(or source)*** of filter module due to > restoring operation. > It seems should to be attached to the filter sink(or source) > properly. > > I do not think this is the right fix for the problem. With your > patch, if > the property list of a stream without filter.apply set is changed, the > stream will still falsely be moved to the master sink. > I believe the right way to solve it would be to set the > filter.apply property > and module_name when the sink_input_put hook is called and the sink is > a filter sink. > > I also wonder why you use an enumeration when you only use one > of the values. A boolean is_put argument to process() would do. > > Regards > Georg > > > You are right. I'll submit patch again soon. Thinking again, it looks that my suggestion also does not cover all cases. When the stream is moved away from the filter sink, it will have filter.apply set which it did not have initially. This is obviously wrong, so we need an additional property to indicate that filter.apply was set by module-filter-apply and that it can be removed when the stream moves away from the filter. Also I have only been talking about the put hook. The same applies to the move_finished hook. But you have been working with the code for a while now, so maybe you can come up with a better solution. Also Arun might object to my idea. He asked me to summarize and wanted to reply on the list. So maybe you wait until he confirmed/rejected. -------------- next part -------------- An HTML attachment was scrubbed... URL: <https://lists.freedesktop.org/archives/pulseaudio-discuss/attachments/20170410/45f568c1/attachment.html>
