[mythtv-users] application not reading fast enough?

[Rich West]

Kevin Kuphal wrote:
> Rich West wrote:
>   
>> I hadn't noticed these errors previously, but I cannot say for sure if 
>> they were or were not showing up...
>>
>> I'm seeing errors like :
>> ivtv2: All encoder MPEG stream buffers are full. Dropping data.
>> ivtv2: Cause: the application is not reading fast enough.
>> ivtv0: All encoder MPEG stream buffers are full. Dropping data.
>> ivtv0: Cause: the application is not reading fast enough.
>>
>> I wasn't (and, after some googling, I am still not) sure what to make of 
>> them.  The only thing that I came across that had a fair amount of 
>> relevance was:
>> http://malformed.org/2006/01/31/new-ivtv-error-from-mythtv-ivtv0-all-encoder-mpeg-stream-buffers-are-full-dropping-data/
>>
>> Which was on some blog from back in Jan 2006.  I figure that information 
>> is probably a bit old, and, from the info on the ivtv wiki, the options 
>> mentioned in the blog entry are probably out dated as well.
>>
>> So.. the question.. is this really ivtv related, or is it a mythtv 
>> issue?  Or, perhaps, is it a hardware issue (not being able to write to 
>> disk fast enough)?
>>   
>>     
> There was some discussion on this on the ivtv-devel list that Hans 
> indicated he was going to fix.  How recent of an ivtv driver release are 
> you using?  It may be fixed in the latest releases or trunk versions.

Thanks for the response.  I'm going to guess that this is an ivtv issue
then, and not a mythtv or other hardware issue..

I'm using the Axel's RPM ivtv-0.9.1-123.fc5.at.  I had checked out the
list archives for ivtv prior to posting here and I saw mention of it in
the changelog of one of the 0.10.1 rc releases....

Hrmm.. thinking of that, in looking at the latest stable release
(0.10.0) changelog, the 0.9.1 release isn't even mentioned:
> IVTV changes:
>
> 0.10.0 release
> 	Completely redesigned the interrupt/DMA handling of the driver and also much
> 	of the other code.
> 	Ian Armstrong contributed huge improvements to the framebuffer code.
> 	Too many improvements, fixes, etc. to list here. 
>
> 0.8.0 release

-Rich