Re: [Lmuse-user] compiling error 0.7 for Gentoo -definitely not a 2.4 vs 2.6 issue

[Robert J. <rj...@sp...>]

More silly questions.

On Thursday 05 August 2004 18.08, Allan Klinbail wrote:
> Hi Robert
>
> I did everything as root, and other than the ladcca error and the jack
> stuff when it died, that was it.
>
> Just for the hell of it I reproduced it again...
>
> starting with default template
> cca_open_socket: could not connect to host 'localhost', service 'ladcca'
> cca_comm_connect_to_server: could not create server connection
> cca_init: could not connect to server 'localhost' - disabling ladcca
> JACK ERROR: zombified - calling shutdown handler
> JACK shutdown callback
>
> Here is the corresponding jack output - note the first xrun was when
> muse fired up, the second was when I hit the panic button and three midi
> ports were in use.
>
>
> JACK compiled with System V SHM support
> loading driver ..
> apparent rate = 44100
> creating alsa driver ...
> ice1712|ice1712|256|2|44100|0|0|nomon|swmeter|-|32bit
> configuring for 44100Hz, period = 256 frames, buffer = 2 periods

It seems it's no longer possible to tell from the output of jack whether it 
was started with -R or not. Hence I ask..

All in all this is very confusing.


To top it of I managed to get 2.6 running yesterday and my experiences are in 
no way conclusive with this....

It did work for me, I loaded a few songs that consume alot of CPU (though only 
using softsynths) and it ran along. I get xruns more often than with 2.4 but 
not unbearably so...

What was _really_ annoying though was that I could hardly touch the faders in 
the mixer without getting a ticking noise, something that never happens with 
2.4. 

I'd guess there is a rt-threading problem somewhere... really sad that we get 
so massively different experiences... hopefully they are related atleast.

/Robert



>
>
> **** alsa_pcm: xrun of at least 3.695 msecs
>
> subgraph starting at MusE-1 timed out (subgraph_wait_fd=11, status = 0,
> state = Running)
> client MusE-1 error: awake_at = 14465250630 state = 2 timed_out = 2
>
>
> **** alsa_pcm: xrun of at least 2.408 msecs
>
> On Thu, 2004-08-05 at 22:34, Robert Jonsson wrote:
> > On Thursday 05 August 2004 14.11, Allan Klinbail wrote:
> > > Hi All
> > >
> > > I bit the bullet and compiled the standard gentoo 2.4.26 kernel (after
> > > patching it for realtime capabilities). The gentoo patchset does
> > > include all the low latency stuff other than the capability hack.
> > >
> > > I found the results to be almost identical, on the 2.4 kernel as the 3
> > > different 2.6 kernels I have now tried. MusE creates a jack
> > > disconnection any time I run more than 2 midi ports simultaneously.
> > >
> > > Depending on how big the differences between muse 0.7.0 pre 2 and muse
> > > 0.7.0 are then this performance issue is only true of my gentoo
> > > installation and not of my mandrake 10 installation. This is odd
> > > considering I have to run everything else at much higher latency
> > > settings on Mandrake than I can get away with on Gentoo.
> >
> > Ok...troublesome...
> >
> > Did you try running as root? And did you get any printouts that indicate
> > an error? (apart from the obligatory ladcca error...)
> >
> > /Robert
> >
> > > cheers
> > >
> > >
> > > Allan
> > >
> > > On Thu, 2004-08-05 at 13:51, Allan Klinbail wrote:
> > > > > > Also here's a capture of the console output from Muse when I ran
> > > > > > it, imported a midi file to the project and resized the window
> > > > > > slightly which seemed to coincide with the disconnection:
> > > > > > starting with default template
> > > > > > midi thread 21695 _NOT_ running SCHED_FIFO
> > > > >
> > > > > Aha!
> > > > > This error message is a bad one. I bet this is the problem.
> > > > >
> > > > > I'm pretty sure this has to do with kernel 2.6 or NPTL.
> > > > > I'm not sure how to check if a system has NPTL, do you know if it
> > > > > you system uses it?
> > > >
> > > > Hi... While I also am experiencing similar problems on a Gentoo
> > > > system, I don't get the error above. I am also using a 2.6.5 kernel,
> > > > but i didn't manually patch it I grabbed it from thac's rpms for
> > > > mandrake 10 (using a little app call rpm2targz).
> > > >
> > > > I only get messages about "cannot create thread" when running as a
> > > > normal user.  Sean, just a point, for some reason I cannot explain,
> > > > the device permissions for /dev/snd/sequencer seem to default to my
> > > > normal user on startup, I have no idea how that happens (checked
> > > > devfsd.conf and it should be root) - could you check if you're gentoo
> > > > does the same thing?
> > > >
> > > > How can I check if I am running NPTL? Is it a choice in kernel
> > > > configuration or a library of some kind.
> > > >
> > > > It seems on my machine to be a performance issue, I cannot get
> > > > anymore than 3 ports of MIDI operational before the xruns start
> > > > flying in.
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > -------------------------------------------------------
> > > > This SF.Net email is sponsored by OSTG. Have you noticed the changes
> > > > on Linux.com, ITManagersJournal and NewsForge in the past few weeks?
> > > > Now, one more big change to announce. We are now OSTG- Open Source
> > > > Technology Group. Come see the changes on the new OSTG site.
> > > > www.ostg.com _______________________________________________
> > > > Lmuse-user mailing list
> > > > Lmu...@li...
> > > > https://lists.sourceforge.net/lists/listinfo/lmuse-user

-- 
http://spamatica.se/music/