Well it's been 2 more days and things have been going relatively well
with this firmware/ivtv combination..
I have recorded 70 programs with relative success, quite a few of them
being simultaneous recordings.
Now the problem appears to be "busy" device issues. There has been about
4 recordings that have simply failed to happen. MythTV thought it was
recording, but in actuality, nothing was being written to disk. I have
MythTV (SVN release as of last week) with verbose monitoring for
recording turned on, and I have attached what the log says for one of
these recordings below. Now I realize this is possibly a MythTV problem,
however, I have not had these issues on MythTV before with my PVR-250,
and as it's indicating a "device or resource busy" message, I thought it
was worth at least mentioning on the ivtv list. Does the ivtv "device"
need a longer delay between stopping one recording and starting another.
That is the most likely explanation for this that I can come up with. I
might investigate a patch for MythTV that implements a wait/retry for
when it receives these errors if that is the case, though I don't recall
seeing anyone else with this issue on either the ivtv or MythTV lists,
so I am a bit hessitant to try that.
Any help, debugging advise, etc. would be greatly appreciated.
- mike
mythbackend.log
-----
2005-10-03 22:59:48.123 scheduler: Started recording
2005-10-03 23:00:00.980 TVRec(1): Switching to overrecord for 15 more
seconds
2005-10-03 23:00:00.984 TVRec(1): ClearFlags(FinishRecording,) ->
RunMainLoop,AskAllowRecording,RecorderRunning,
2005-10-03 23:00:02.209 TVRec(1): ClearFlags(AskAllowRecording,) ->
RunMainLoop,RecorderRunning,
2005-10-03 23:00:02.221 TVRec(1): Changing from RecordingOnly to None
2005-10-03 23:00:02.280 TVRec(1):
ClearFlags(FrontendReadyCancelNextRecording,) ->
RunMainLoop,RecorderRunning,
2005-10-03 23:00:02.312 TVRec(1): SetFlags(AskAllowRecording,) ->
RunMainLoop,AskAllowRecording,RecorderRunning,
2005-10-03 23:00:02.327 TVRec(1): Request: Program(no) channel() input()
flags(CloseRec,)
2005-10-03 23:00:02.378 TVRec(1):
ClearFlags(WaitingForSignal,SignalMonitorRunning,) ->
RunMainLoop,AskAllowRecording,RecorderRunning,
2005-10-03 23:00:02.481 scheduler: Finished recording
2005-10-03 23:00:02.782 TVRec(1): ClearFlags(RecorderRunning,) ->
RunMainLoop,AskAllowRecording,
2005-10-03 23:00:02.785 TVRec(1): ClearFlags(PENDINGACTIONS,) ->
RunMainLoop,AskAllowRecording,
2005-10-03 23:00:02.787 TVRec(1): ClearFlags(CancelNextRecording,) ->
RunMainLoop,AskAllowRecording,
2005-10-03 23:00:02.789 TVRec(1): Changing from None to RecordingOnly
2005-10-03 23:00:02.790 TVRec(1):
ClearFlags(FrontendReadyCancelNextRecording,) ->
RunMainLoop,AskAllowRecording,
2005-10-03 23:00:02.792 TVRec(1): SetFlags(AskAllowRecording,) ->
RunMainLoop,AskAllowRecording,
2005-10-03 23:00:02.793 TVRec(1): Request: Program(yes) channel()
input() flags(Recording,)
2005-10-03 23:00:02.795 TVRec(1):
ClearFlags(WaitingForSignal,SignalMonitorRunning,) ->
RunMainLoop,AskAllowRecording,
2005-10-03 23:00:03.137 TVRec(1): SetFlags(NeedToStartRecorder,) ->
RunMainLoop,AskAllowRecording,NeedToStartRecorder,
2005-10-03 23:00:03.140 TVRec(1): Starting Recorder
2005-10-03 23:00:03.305 TVRec(1): Using profile 'High Quality' to record
2005-10-03 23:00:03.393 SetRecording(0x0x836f5f8)
2005-10-03 23:00:03.396 Prog title: Die Hard 2
Error setting format
VIDIOC_S_FMT:: Device or resource busy
2005-10-03 23:00:03.432 TVRec(1): SetFlags(RecorderRunning,) ->
RunMainLoop,AskAllowRecording,NeedToStartRecorder,RecorderRunning,
2005-10-03 23:00:03.555 TVRec(1): ClearFlags(NeedToStartRecorder,) ->
RunMainLoop,AskAllowRecording,RecorderRunning,
2005-10-03 23:00:03.569 scheduler: Started recording
QSocketNotifier: invalid socket 12 and type 'Read', disabling...
2005-10-03 23:00:10.130 scheduler: Scheduled items
[mpeg2video @ 0xb767d7d0]ac-tex damaged at 4 17
[mpeg2video @ 0xb767d7d0]Warning MVs not available
Mike Javorski wrote:
ivtv: Encoder revision: 0x02030207
Falcn: Version 2.05.032
As I understand it, to use the HcwFalcn.rom you need to symlink it to
ivtv-fw-enc.bin. Is that correct? It is not linked, its just there.
- mike
Hans Verkuil wrote:
On Saturday 01 October 2005 21:08, Mike Javorski wrote:
So it looks like firmware might be the cause of the trouble.
I tried r2693 with the r2704 changes, got the errors. Tried just r2693
(the one I had compiled earlier in the week), same error. I went and
looked at my notes and realized that I had upgraded the firmware to the
"recommended" version from the ivtv site, well I reverted back to the
version I had previously at 12:30am this morning, and I have recorded 10
programs with it since (using ivtv-svn-r2718 so no ghosties), and no
errors. It seemed to be crashing regularly within 1-2 recordings before
the firmware swap. Conveniently, this is the firmware I have had great
success with on my PVR-250, so that means it should be easy going down
the road :-)
Here is the checksums for the working firmware.
0feb4f6aef94b6fab843c5269d7f0848 /lib/modules/ivtv-fw-dec.bin
839fb0b71324fa2ef3c7c43a17a41396 /lib/modules/ivtv-fw-enc.bin
5f5fa240ada73c3565f5f7de4c7b5138 /lib/modules/HcwFalcn.rom
3a4803384f749d644ee1f1ca9dcb12fa /lib/modules/str
What is the firmware of the version you are using now as shown by the
dmesg output? And what is the version of /lib/modules/HcwFalcn.rom?
To get the last version just run 'strings HcwFalcn.rom|grep -i
version'. I ask this because the only firmware whose checksum I
recognize is the HcwMakoA.ROM.
Hans
_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
_______________________________________________
ivtv-devel mailing list
[email protected]
http://ivtvdriver.org/mailman/listinfo/ivtv-devel
