[mythtv-commits] Ticket #5882: S2API support
MythTV
mythtv at cvs.mythtv.org
Tue Dec 2 18:56:28 UTC 2008
#5882: S2API support
-------------------------+--------------------------------------------------
Reporter: wagabunda | Owner: janne
Type: enhancement | Status: accepted
Priority: major | Milestone: 0.22
Component: dvb | Version: head
Severity: medium | Resolution:
Mlocked: 0 |
-------------------------+--------------------------------------------------
Comment(by mnix):
Replying to [comment:40 mnix]:
> ioctl(FE_READ_STATUS) returning Invalid Argument.
Finally got a fix on this one... YAY!
The issue stems from the channels.conf data from scan-s2 (specifically I
did not specify '-o zap' and so was feeding a vdr format channels.conf to
mythtv).
The parameters field has no modulation (M) value, so somewhere in the
translation myth defaulted to QAMAuto. When the scanner arrived at
dtvmultiplex_to_dtvproperties() it was unable to match up the tuner type
with the modulation, and did not fill in the tuning commands - leaving the
DTV_CLEAR command in-tact in the buffer.
DVBChannel::Tune() then sent this DTV_CLEAR when it would normally send
the tuning parameters, and as it did not cause an error, would wait around
for a signal lock.
As the S2API had never received tuning parameters, it had no idea what it
was supposed to be tuned to (DVB-S or DVB-S2) and returned an error on
FE_READ_STATUS.
Proposed fix: get dtvmultiplex_to_dtvproperties() to return false as well
as log an error. Also check the returned value and return false from
Tune(). Possibly also return false if cmds.num<2 (redundant, but may help
catch future coding deficiencies).
There were other issues with mythtv's parsing of the vdr format from
scan-s2 but those a are unrelated to the S2API code.
Unfortunately I don't know if I can watch the channels it finds yet as the
current revision of mythfrontend is having issues during startup. A job
for tomorrow I think.
--
Ticket URL: <http://svn.mythtv.org/trac/ticket/5882#comment:41>
MythTV <http://www.mythtv.org/>
MythTV
More information about the mythtv-commits
mailing list