Harondel J. Sibble wrote:
> 
> On 24 Mar 2008 at 23:27, Tom Metro wrote:
> 
>> [60.710413] ivtv0: Unreasonably low latency timer, setting to 64 (was 32)
>> [61.096095] PCI: Setting latency timer of device 0000:00:06.0 to 64
>> [67.561088] ivtv1: Unreasonably low latency timer, setting to 64 (was 32)
> 
> Those are some of the errors, but those are fairly common on many linux 
> machines I've seen, can't seem to find how I came up with this, started with 
> some postings to the ivtv lists.
> 
> Anyhow the mythtv wiki has a good entry on this
> 
> http://www.mythtv.org/wiki/index.php/PCI_Latency
> 
>> I looked into this back when I first deployed the system in the Fall of 
>> 2006, and I don't recall what I found, but I'm pretty sure I never 
>> implemented any fixes.
> 
> It helps!  I was getting several lockups a week until I implemented this.
>  
>> It locks up some of the time when I transcode using mencoder, and once 
>> when using ffmpeg. I used mencoder to transcode only rarely, and stopped 
>> using it because of this.
> 
> Most of those things have 2 things in common, high cpu and high disk usage, 
> adjusting the pci latency helps with the latter.
> 
>>> ...and what exactly do the lockups look like?
>> Black screen or normal looking but frozen screen. Unresponsive to 
>> network connections or existing ssh connections.
> 
> Yup, that about sums up the problem, it was most noticable when multiple 
> concurrent recordings were set to go off.
>  
>>> My Sempron has a VIA chipset.
> 
> I'm also using a VIA based system
> 
> Anyhow, here's the line from my /etc/init.d/rc.local
> 
> ### added Sep 25/07 to allow more pci throughput from multiple tv tuner cards 
> to the hard drives
> setpci -v -s 00:07.1  latency_timer=[b0]
> 
> 

All of my latencies are set to 64, and I don't have any crashing of the type 
that you describe (I have 2-150's and 1-atsc 5500), and a pci 4 port sata card,
and I don't get the unreasonably low latency error.

I think those parameters are set by the bios (they are controllable in some 
bioses), so it is likely a bios bug causing it.

                              Roger

-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Mvpmc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mvpmc-users
mvpmc wiki: http://mvpmc.wikispaces.com/

Reply via email to