[linux-dvb] Re: section demux rewritten

[Ralph Metzler <rjkm@metzlerbros.de>]

emard@softhome.net writes:
 > I planned to support dynamically alloc'ed buffer for purpose
 > of on-demand stretch or shrinnk buffer as section data creep in.
 > This is heading toward to allow network all pids reception by dvbnet -p 
 > 8192, which is now non-functional because for 8192 we would have
 > to alloc 8192 section structs each having 4096 bytes while only few of 
 > section structs would actuall be in use. 
 > 
 > So average_network_user wouldn't have to care which PID is there,
 > and as he szap's also restart network apps to change the dvb0_0
 > for dvb0_1, but instead just register dvbnet -p 8192 and have one
 > and only dvb0_0 for all network purposes 


So, there could be use for multi-PID section filters after all!

 
 > The comapre would be beneficial, as it (on my 7146 based card) it 
 > periodically (1 sec period) reports bursts of longer
 > out-of-section-syntax data for 1451, while on 451 it reports it
 > reports out-of-section data much rarely and data are short
 > (always only 1 byte our-of-section cca 10x a second) 
 > 
 > I wonder between 3 things:
 > 1) do I have phantom DMA issues
 > 2) did we need to implement some new specification standard
 > 3) are they in the TS simply wasting the bandwidth for
 >   PID 8192 bandwidth > 0.5 full TS bandwidth 

Hmm, we should do some tests with parallel recording of the same stream
on two different hardwares (e.g. SAA7146 and Fusion878 or SAA7134) 
to rule out or find DMA problems.


Ralph


-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe linux-dvb" as subject.