Florian Schirmer wrote: > > >For point 4, the h/w that I am using does allow the capability to discard > >all packets until the PUSI flag is found in a TS header (after which all > >TS packets are matched). This feature just reduces the amount of packets to > >handle/copy/parse etc. For h/w that didn't implement it, it would > >probably be best to implement it in kernel space. In this way, a user > >space process that opened up a fd for a filter on a SI table would get > >back the complete 1024 or 4096 bytes of the section table without having > >to worry about any additional parsing. > > Just use that feature of you hw. It doesnt make any sense to deliver > incomplete data. If your hw cant handle it just route it to the dvb-core. It > will take care of it and discard the data. No need to worry about it ;-)
Yes, it's no problem to use this hardware feature in the section filter implementation. There's just no need to expose it in the API. Regards, Johannes -- Info: To unsubscribe send a mail to [EMAIL PROTECTED] with "unsubscribe linux-dvb" as subject.
