The different realtime kernels

[Ralf Mardorf]

On Thu, 2010-09-30 at 19:53 +0200, Ralf Mardorf wrote:
> On Thu, 2010-09-30 at 12:58 -0400, Ronan Jouchet wrote:
> > On Thu, Sep 30, 2010 at 11:45 AM, Ralf Mardorf
> > <ralf.mardorf at alice-dsl.net> wrote:
> > > Why does Ubuntu Studio comes without PREEMPT RT, but just PREEMPT?!
> > > This is my intension.
> > >
> > > FWIW, I'm a professional audio and video engineer and did program oldish
> > > computers and I'm missing hard real-time for modern PCs. Even the
> > > kernel-rt isn't able to do hard real-time, so I don't understand why
> > > Ubuntu Studio does prefer a kernel without rt-patch. Today the rt-patch
> > > isn't good enough and any kernel without this patch is useless for
> > > multimedia production.
> > >
> > > So a misunderstanding ;)!
> > 
> > Hello Ralf,
> > 
> > You keep coming back to -rt/-realtime, but nobody ever questioned
> > their greatness. I trust you when you mention -rt is the ideal
> > solution in your heavy MIDI use case, and I would also *love* a
> > properly maintained -rt kernel in Studio.
> > 
> > Now, whether we want it or not, the facts are:
> > - Preparing -realtime (vanilla+rtpatch) or -rt
> > (vanilla+ubuntusauce+rtpatch) is a lot of work and there are no
> > resources for this
> > - On the contrary, -lowlatency (generic with some config tweaks) means
> > performance tradeoff, _but_ could happen in a PPA (maybe even in the
> > archives in Natty) because it is less of a maintenance hell
> > 
> > You mention you have some custom built kernels, so if you feel like
> > helping maintaining -realtime/-rt, by all means step in, talk to
> > Alessio and make it happen. But insisting again and again on -rt and
> > -realtime without considering the possibilities is only going to
> > discourage him from working on the feasible options.
> > 
> > He tries to propose sustainable options and the only echoes are
> > negative, without much questioning. What are the results of your own
> > tests with -lowlatency? What kind of performance drop have you met on
> > one of your heavy setups? How much latency lost, on which kind of
> > machine / firewire card?
> > 
> > Ronan
> 
> Hi Ronan :)
> 
> I marked your email and will come back to it ASAP. At the moment the
> influenza gained the upper hand.
> 
> I guess it's not that important what issues I had when using the PREEMPT
> kernel, there were issues and I didn't noticed that it was a PREEMPT,
> but a PREEMPT RT kernel, when I posted something, including uname -a at
> LAD or JACK mailing list. Somebody else noticed it. I might do another
> test, or search the archives, but I would prefer trying to compile and
> build a package for a kernel-rt again on my Ubuntu Studio and post the
> package for 64-bit or ask because of trouble, if the compiling should
> fail, resp. the startup when booting the kernel should fail.
> 
> Perhaps I could try to compile a kernel-rt at the weekend. I suspect
> issues for the startup regarding to X, but I'm not sure.
> 
> To be continued ... probably this weekend ... if I don't answer until
> the week after next, please remember me to compile a kernel-rt and to
> post the messages I get when booting the kernel-rt from the
> repositories.
> 
> Today I don't wish to test anything.
> 
> I've got two PCI Envy24 cards, Terratec EWX 24/96, a NVidia 7200 GS + an
> onboard Radeon X1250-based graphics. The mobo is a M2A-VM HDMI and
> 
> suse11-2:/home/spinymouse11.2 # hwinfo --cpu
> 01: None 00.0: 10103 CPU                                        
>   [Created at cpu.301]
>   Unique ID: rdCR.j8NaKXDZtZ6
>   Hardware Class: cpu
>   Arch: X86-64
>   Vendor: "AuthenticAMD"
>   Model: 15.107.2 "AMD Athlon(tm) X2 Dual Core Processor BE-2350"
>   Features:
> fpu,vme,de,pse,tsc,msr,pae,mce,cx8,apic,sep,mtrr,pge,mca,cmov,pat,pse36,clflush,mmx,fxsr,sse,sse2,ht,syscall,nx,mmxext,fxsr_opt,rdtscp,lm,3dnowext,3dnow,rep_good,extd_apicid,pni,cx16,lahf_lm,cmp_legacy,svm,extapic,cr8_legacy,3dnowprefetch
>   Clock: 1000 MHz

Oops, clock is higher ...

spinymouse11.2 at suse11-2:~> cpu-p
Password: 
spinymouse11.2 at suse11-2:~> su -c "hwinfo --cpu"
Password: 
01: None 00.0: 10103 CPU                                        
  [Created at cpu.301]
  Unique ID: rdCR.j8NaKXDZtZ6
  Hardware Class: cpu
  Arch: X86-64
  Vendor: "AuthenticAMD"
  Model: 15.107.2 "AMD Athlon(tm) X2 Dual Core Processor BE-2350"
  Features:
fpu,vme,de,pse,tsc,msr,pae,mce,cx8,apic,sep,mtrr,pge,mca,cmov,pat,pse36,clflush,mmx,fxsr,sse,sse2,ht,syscall,nx,mmxext,fxsr_opt,rdtscp,lm,3dnowext,3dnow,rep_good,extd_apicid,pni,cx16,lahf_lm,cmp_legacy,svm,extapic,cr8_legacy,3dnowprefetch
  Clock: 2100 MHz
[snip]

Frequency scaling was ondemand, but performance.