[mythtv-users] Debugging failed DVB-T recordings

[James Fidell]

I think I'm down to the last of my problems with failed recordings now.
I'm running "mythbackend -v important,general,record,channel" and
comparing the output of a program which failed to record against one
that worked ok.  First, the program that worked:

2007-11-06 08:22:02.463 Started recording: Lunar Jim "Jim and Ted's Wild
Ride": channel 1071 on cardid 6, sourceid 1
2007-11-06 08:22:02.464 DVBSM(3)::UpdateValues: Signal Locked
2007-11-06 08:22:02.471 DVBSM(3)::UpdateValues: Waiting for table
monitor to start
2007-11-06 08:22:02.473 DVBSM(3)::UpdateValues: Table monitor started
2007-11-06 08:22:02.473 DVBSM(3): RunTableMonitorTS(): begin (# of pids 3)
2007-11-06 08:22:02.475 DVBSM(3)::AddPIDFilter(0x0):
2007-11-06 08:22:02.476 DVBSM(3)::AddPIDFilter(0x10):
2007-11-06 08:22:02.477 DVBSM(3)::AddPIDFilter(0x11):
2007-11-06 08:22:02.593 DVBRec(2): PID 0x100 discontinuity detected
2007-11-06 08:22:02.632 SM(3)::AddFlags: Seen(PAT,) Match() Wait()
2007-11-06 08:22:02.633 SM(3)::AddFlags: Seen() Match(PAT,) Wait()
2007-11-06 08:22:02.634 CreatePATSingleProgram()
2007-11-06 08:22:02.634 PAT in input stream


And the failing one:

2007-11-06 08:17:02.038 DVBSM(0)::UpdateValues: Signal Locked
2007-11-06 08:17:02.075 DVBSM(0)::UpdateValues: Waiting for table
monitor to start
2007-11-06 08:17:02.072 Started recording: Thomas & Friends "Thomas's
Frosty Friend": channel 1005 on cardid 3, sourceid 1
2007-11-06 08:17:02.078 DVBSM(0): RunTableMonitorTS(): begin (# of pids 3)
2007-11-06 08:17:02.078 DVBSM(0)::UpdateValues: Table monitor started
2007-11-06 08:17:02.081 DVBSM(0)::AddPIDFilter(0x0):
2007-11-06 08:17:02.083 DVBSM(0)::AddPIDFilter(0x10):
2007-11-06 08:17:02.087 DVBSM(0)::AddPIDFilter(0x11):
2007-11-06 08:17:02.493 Writing PAT & PMT @851657672 + 1590480
2007-11-06 08:17:02.984 Writing PAT & PMT @262656304 + 1272384
2007-11-06 08:17:03.214 Writing PAT & PMT @851987800 + 1166352
2007-11-06 08:17:03.958 Writing PAT & PMT @852315296 + 742224


I'm not sure that the last few "Writing PAT" lines are relevant to
what's happening.  I think they probably relate to somethign else
recording at the same time.

It appears however that the failed recording never gets to the
"AddFlags: Seen(PAT,) Match() Wait()" and perhaps that's the cause of
the failure.  Is there any reason why, given that it has a signal lock,
it shouldn't see the PAT?  Are there any other flags I can enable to
debug this further (or even changes I can make to the code to shed
further light)?

James