[mythtv-users] recordings cut off early especially when watching recoding in progress

Jon Vanderwoude jvanderw at gmail.com
Thu Nov 9 23:21:44 UTC 2006

On 10/13/06, Nick <knowledgejunkie at gmail.com> wrote:
> On 12/10/06, McNally <mcnallychris at pobox.com> wrote:
> > Hi,
> >
> > I'm using mythtv 0.20.  When I watch recording in progress, such as a
> > baseball game, I begin watching about 40 minutes after the start, fast
> > forward through commercials and other junk,  and find that the actual
> > recording seems to have stopped soon after i started watching. This has
> > happened twice already to me. I see from the logs that there seems to be
> > a DMA error and it may be IVTV related. Other recordings have ended much
> > sooner than scheduled and there is an error in the Myth log.
> >
> > I am using an ASUS A8N-VM CSM mainboard with an AMD Sempron  which has
> > NVIDIA nforce and gforce chipsets. The backend and frontend are the same
> > box. I am using ext3. The CPU from top is around 90% when watching
> > recordings. I have sections of the logs below.  Thanks in advance.
> >
> > ivtv version
> > ivtv:  version 0.7.0 (tagged release) loading
> > ivtv:  Linux version: 2.6.17-1.2142_FC4 mod_unload 686 REGPARM 4KSTACKS
> > gcc-4.0
> Try upgrading to ivtv 0.7.1 which has a fix for DMA problems like this
> - from the ivtv changelog:
> IVTV changes:
> 0.7.1 stable release
>         - Add getopt support to test/vbi. Add new timeout option.
>         - Fix South African broadcast frequency table.
>         - Detect and workaround DMA offsets that can be introduced by
>           a DMA error.
>         - Sliced VBI always uses PIO, never DMA.
> Nick

I was having this issue even after installing the new firmware, turns
out the 3.3v line on my power supply was bad and the memory chip on
the PVR-250 wasn't getting the juice it needed to keep it going,
replacing the power supply resolved the issue.

More information about the mythtv-users mailing list