oscam (1.20-9286~r9059-ppa1~saucy) saucy; urgency=medium
* [r9059]
if a ecm request contains a prid and/or chid and a cache answer not
cwc runs in a unneeded learning state
->fixed by using only such metadata from request as cwc base
* [r9058]
dvbapi: do not add duplicate socket fd's to the poll
The current code for IPBOX boxtype should be applicable for all cases.
If a the same fd is added twice to the poll() command it will return
two POLLIN events, which then lead to unnecessary timeout reading data
from socket.
Optimises work for multiple services on one socket.
* [r9057]
cosmetics: fix showing negative inactivity ms value in trace log
* [r9056]
Typo in manual table freq for smartreader mhz
* [r9055]
change the selected auto value smargo smartreader
Seems that for some cards when 5.00 Mhz is the atr f, the smargo does not do well. Changed value to use 5.36 Mhz instead off the 4.80
* [r9054]
revert r9051
* [r9053]
Autospeed also for smargo protocol
Now the autospeed is for smartreader and smargo protocol.
Internal readers is also in process and should come in time
Initial card init set at 369 instead off 400
* [r9052]
Changed smargoautospeed into autospeed
Adapted parameter smargoautospeed into autospeed.
This parameter is very ussefull for all hw readers and very easy to use in reader sources.
At the same time very easy for enduser .
* [r9051]
added patch from http://www.streamboard.tv/wbb2/thread.php?postid=487688#post487688
Thx to theparasol
* [r9050]
Replaced smargoautospeed
Placed the smargoautospeed below mhz into web if By hw reader setup
It's currently only in classic web if think that modern needs to be adapted ass well.
* [r9049]
dvbapi: handle LIST_FIRST, LIST_MORE and LIST_LAST not only in pmt_mode=6
Above CA list management requests was only properly handled in
pmtmode=6, while it could be handled in other pmt modes as well.
As a result one single socket can be used to manage PMTs for
several services (eg. in pmt_mode=4)
* [r9048]
Smartreader now autospeed according to atr.
This is the new standard setting off smartreader protocol. The speed will be set automatically according off the detected atr. mhz will have no effect anymore. The cardmhz is standard on 400. If you have one off these cards which needs lower just set cardmhz = xxx. The mhz will be automatically set .
If You want to over or under clock you're card. You must add parameter
smargoautospeed = 0
Then it behaves like before this patch. Parameter can also be set trough web iff just uncheck the
smargoautospeed box and set the desired mhz then save.
Docs not addapted yet
* [r9047]
Fix setting (non-)blocking mode for descriptors.
Previous flags was ignored, add a set_nonblock()
function and use it where applicable.
* [r9046]
Relocate smartreader mhz table
Placed the mhz table in reader common.
Finally have had the confirmation that v2 and Triple speed range is equal to that off v1. Default card init clockspeed is now 4.00 Mhz,that is the best for smartreader in general. If you omit cardmhz = xxx 400 will be used. Some card's needs 369 then you just set cardmhz=369 .
* [r9045]
Removed some log ouputs
* [r9044]
- Tryfix sh4 internal readers
* [r9043]
smartreader and sci
Added some extra reader speed settings log (really needed cause whe need to find out why some cards are not working)
Bitbang now enabled for all modes in smartreader.
prepare already a new setting to work out a full automated speed setup based on atr with smartreader protocol. The full work out only can be done when the problem for some cards are solved, they are about the same for sci cards.
* [r9042]
- Work ETU was calculated wrong -> corrected
- BWT was calculated wrong -> corrected
* [r9041]
some fine tunings
Adapted setting to prevent double sr lock
added line to see which are the real settings used to enable the reader.
handy for trouble shooting on v2 and Triple
V2 and Triple have full freq range possible at this time
* [r9040]
removed delay start off P2 and P3 on tripple. Not needed anymore.
* [r9039]
Small correction card detect Triple
Reduced the card detect time two 6 seconds instead of 12, get rid off and orfull long integer in triple card detect polling
* [r9038]
Added new parameter 'cwcycle_usecwcfromce' to documentation.
* [r9037]
-some simplify on cwc stage4
-add possibility to use (cwc OK) info from cacheex sources
[cache]
cwcycle_usecwcfromce = 0/1 default: off
* [r9036]
- Cache answers where not taken into account -> moved nds cwswap to better point
- Reverted cyclecheck (my changesets could only caused additional trouble!)
-- Andrey Pavlenko <email address hidden> Tue, 26 Nov 2013 23:32:56 +0400