[PlanetCCRMA] Reliable version of Fedora using rt-kernel?

[mark]

OK, i am sry... i had the information from this page: 
http://wiki.debianforum.de/Realtime_Kernel it is german unfortunately

regards

mark

On 14.11.2011 19:57, Fernando Lopez-Lezcano wrote:
> On 11/13/2011 11:09 PM, mark wrote:
>> Hello,
>>
>> as far as i know the Kernel 2.6.39.4 is already RT patched.
>
> No, that is incorrect.
>
> Parts of the rt patch set migrated to mainline in the 2.6.39.x cycle, 
> most importantly threaded irqs, but 2.6.39.x definitely does not have 
> all of the rt patch. Not only the kernel has to be built with the 
> right options but you have to include a kernel boot line parameter to 
> activate threaded irqs.
>
> The only (current) kernel versions to have rt patches available are 
> 2.6.33.x, 3.0.x and 3.2-rc1. The 3.1 kernel series will not have an rt 
> patch.
>
>> I have tried
>> to compile Kernel 3 with patch at debian 6 but it didn`t to run better.
>> The latency was worse than without the patch. Anyway don`t forget to
>> input those two lines for jackd into the limits.conf and add your user
>> to the desired group.
>>
>> i would suggest to use either Kernel 2.6.39.4 or just Ubuntu Studio
>> 11.04 (not 11.10!!!).
>
> As this list is about Planet CCRMA running on Fedora I would instead 
> suggest Fedora 14 and 2.6.33.x :-)
>
> So, the most tested, reliable and best performing rt kernel currently 
> is 2.6.33.x and that means Fedora 14 at most (the default startup 
> system in 15/16 needs newer versions of the kernel). Next down is 
> 3.0.9-rt25 which could be used in Fedora 15 or 16. Next is the default 
> Fedora kernel.
>
> -- Fernando
>
>
>> On 14.11.2011 00:23, Fernando Lopez-Lezcano wrote:
>>> On 11/13/2011 12:05 PM, John S. Dey wrote:
>>>> Hi Fernando,
>>>>
>>>> Thanks for getting back to me. No. You're presumption was incorrect. I
>>>> hadn't install rtirq. I have now but it didn't make much different.
>>> (it does not necessarily make a difference, basically it tries to give
>>> better priority to the soundcard hardware interrupts, to test you have
>>> to reboot or - as root - do a "/sbin/service rtirq start" from a
>>> terminal... just in case you did not)
>>>
>>> -- Fernando
>>>
>>>
>>>> On Nov 12, 2011, at 9:17 PM, Fernando Lopez-Lezcano wrote:
>>>>
>>>>> On 11/12/2011 02:06 PM, John S. Dey wrote:
>>>>>> Hi,
>>>>>>
>>>>>> I'm currently using 3.0.7-1.rt20.1.fc15.ccrma.x86_64.rt as my kernel
>>>>>> on fc15 running on a thinkpad T60p with ATI graphics. I haven't been
>>>>>> able to find settings for jackd that produce reliable real time
>>>>>> performance. In the past I could achieve latency as presented in
>>>>>> qjackctl of less than 10ms. Now I get xruns with latencies of 40ms
>>>>>> and more. The stock kernel is just as good.
>>>>> What soundcard do you have and what parameters did you use for jack?
>>>>> (just curious).
>>>>>
>>>> cat /proc/asound/cards
>>>>
>>>> 0 [Intel ]: HDA-Intel - HDA Intel
>>>> HDA Intel at 0xee400000 irq 49
>>>> 1 [Set ]: USB-Audio - C-Media USB Headphone Set
>>>> C-Media USB Headphone Set at usb-0000:00:1d.0-2, full speed
>>>> 29 [ThinkPadEC ]: ThinkPad EC - ThinkPad Console Audio Control
>>>> ThinkPad Console Audio Control at EC reg 0x30, fw 79HT50WW-1.07
>>>>
>>>> I have varied Frames/Period and Periods/Buffer (2,3 and 4) but haven't
>>>> been able to get
>>>> stable results better that what is shown. I ran with both sound cards
>>>> shown above. The rt kernel wasn't any better that the up to date stock
>>>> kernel.
>>>>
>>>> Thanks,
>>>>
>>>> John
>>> _______________________________________________
>>> PlanetCCRMA mailing list
>>> PlanetCCRMA at ccrma.stanford.edu
>>> http://ccrma-mail.stanford.edu/mailman/listinfo/planetccrma
>>>
>>>
>>
>> _______________________________________________
>> PlanetCCRMA mailing list
>> PlanetCCRMA at ccrma.stanford.edu
>> http://ccrma-mail.stanford.edu/mailman/listinfo/planetccrma
>
>