- Messages
- 10,198
Changeset 11562
build juli13
Tryfix dvbapi error during parallel decoding of some HD+ channels
build juli13
Tryfix dvbapi error during parallel decoding of some HD+ channels
Changeset 11566
HD+ HD05 Card support (tnx pehedima for patch!)
Changeset 11565
More tryfix socketdata handling
Changeset 11568
Fix for descramblers used not 100% fixed but it is more accurate than before
Changeset 11567
More fix for dvbapi client v3
[webif ssl]
fix certificate chain loading
fix Oscam webif won't send out a complete cert chain to the browser when using SSL webif
patch provided by @delilah23
*** DES/3DES Overcrypt support
fix bug for wrong decoded classD3 CMD's
improved Sky DE/CH/AT V14/V15 payload log output
patch provided by kabeltod
*** DES/3DES Overcrypt support
fix bug for wrong decoded classD3 CMD's
improved Sky DE/CH/AT V14/V15 payload log output
patch provided by kabeltod
fix for BSkyB 0963
fix READER_VIDEOGUARD = CONFIG_LIB_DES=y
fix for BSkyB 0963
fix READER_VIDEOGUARD = CONFIG_LIB_DES=y
[streamserver] Synced descriptor parsing code with the dvbapi
This should fix no AC-3 audio on some PowerVu channels.
[gbx] - try fix ecm and cw messages
fix ecmfmt g,j parameters
[dvbapi] - add ecm.info gbox style
[gbx] - fix ccc->gbox reshare
[gbx] - add caidtab as reshare setting
Changeset 11576
[gbx] - more generalized routing info in cw msg
fix minor bugs
thanks @chimpie23
Changeset 11575
cak7 - fix caid 0000 and all other known issues
no config change required
patch provided by @pehedima
build juli13fix HD03/05 CMD Overflow
patch provided by @pehedima
should fix resync error with HD05 after nearly 36500 CMD
remove some DOS CRLFs in oscam.work.c
rework code ref gbox_maxecmsend
fix for gbox_max_ecm_send
change default to 5
clean up some ecm stuff
Revision 11393 introduced sharing instability side effect.
My fix is rather lame since it restores the old behavior in case the nodeid is lower than 0x7fffffffff
Tnx to ceco2000 for research, pinpointing to this revision and testing my fix.
Fix Using gbox protocol the last byte is missing from the odd CW in the ecm debug string
Incoming CWs are saved and it is checked whether the CW has already been received for another srvid (confused mappings ) & whether the CW is already older than "timediff_old_cw".
If one of the two conditions applies and the mode is set to 2, the CW is discarded and, accordingly, is not redistributed via CacheEx.
If some use the patch, the confused CWs should be reduced significantly.
The module CS_CACHEEX is required
Configuration:
@ oscam.conf -> [cache]
cw_cache_size = count of max. CWs for CW cache (default 0)
cw_cache_memory = MByte of max. memory used for CW cache (default 0) // one entry is about 63byte;>
cw_cache_settings = caid [& mask] [@ provid] [$ servid]: mode: timediff_old_cw [, n]
mode = 0 (disabled)
= 1 log only
= 2 drop known CWs for other srvid || drop incoming CWs if timediff_old_cw> now - first_seen
timediff_old_cw = time in ms for dropping old CWs
To enable this patch, you need to define cw_cache_size or cw_cache_memory, not both is required.
If cw_cache_size or cw_cache_memory has reached, 10% of the oldest cache entries are deleted.
Logs are @loglevel 8192 & config via webif is possible.
Try it and take a look in loglevel 8192 @ your incoming CWs from all sources with for example:
cw_cache_memory = 32
cw_cache_settings = 0: 1: 1000
fix:
webif
picon-foobar for cx-aio user/reader/status-page added/fixed
add:
if a physical reader returns E_CORRUPT => dont forward ECM to proxys and return E_INVALID
//-----
push all localgenerated-flagged CWs to defined group/s without source-group membership
if you plan to use this function, think about the following:
if a cx-peer(user/reader) doesn't run in one OSCam-instance(so there are differen NodeIDs) and you seperate them via groups, the localgenerated-flagged CWs would be pushed back to them, if you assign the choosen grp for this function
to prevent this, name them identical (user->username/reader->label) to avoid pushing thier localgenerated-flagged CWs back (cause there is a skip on same names for this function)
cfg:
oscam.conf => [cache]
cacheex_push_lg_groups = 1,n
//-----
use cacheex_localgenerated_only_in & cacheex_localgenerated_only_in_caid only for detected cx-aio peers:
cfg-global:
oscam.conf => [cache]
cacheex_lg_only_in_aio_only = 0|1 (default: 0)
cfg-user/reader:
oscam.user/server => [account]/[reader]
cacheex_lg_only_in_aio_only = 0|1 (default: 0)
//-----
service based localgenerated_only-exception (for srvids which are rare)
this exception works global for incoming and outgoing traffic, so if you define localgenerated_only/_in or localgenerated_only/_in_caids this service-exception is in higher priority
cfg:
[service_foo]
lg_only_exception = 1
caid = 1337
provid = 123456
srvid = CAFF,EBAB,E000
fix:
webif
segfault: disable cx-reader @ reader-page
move videoguard2-reader rc-check before badcw/cwc-check, so it doesnt break wanted behaivor
change:
drop incoming cws (via cacheex) if CW is known(cw_cache), a cw_cache-rule is matching and the CW is older than 15 secs // hardcoded!
log-lvl@8192+ : [late-15sec+ CW] cache: CAIDROVID:SRVID:CW | in: CAIDROVID:SRVID:CW | diff(now): XY ms > Z - cacheex-peer-label - hop 1 (lg), src-nodeid 1337133713371337X
if a videoguard2-reader detect a bad/wrong ECM:
dont forward ECM to proxys and return E_INVALID
no (NOK) LB-entry is written
extend logs for cacheex
for cw-diffs - log-lvl@512: (grep "Different CW" /your/path/oscamlogfile.log)
if last byte is xor 0xF0 add "[last byte xor 0xFO]" to log
show nodelist for cw-diff
change log-suffix for localgenerated hits
...(localgenerated)
to
...(lg)
add:
lg-only for "normal" clients - to deliver just lg-flagged CWs from cache (service-based exception like for cacheex also working)
define @ oscam.user just:
cacheex_localgenerated_only = 1
or
cacheex_localgenerated_only_caid = 1337,BEEF