[PlanetCCRMA] Re: Running JACK as non-root user

[Fernando Lopez-Lezcano]

On Thu, 2004-12-02 at 01:32, Artem Baguinski wrote:
> Kevin Ernste <kevinernste@gmail.com> writes:
> >> 23:00:29.563 Statistics reset.
> >> 23:00:29.570 JACK is starting...
> >> 23:00:29.571 /usr/bin/jackd -v -R -dalsa -dhw:0 -r48000 -p64 -n2
> >
> > I hope it's this easy, but are you using "jackstart" with qjackctl
> > (look in "Setup...-->Server Path" in qjackclt)?  I think this is
> > needed instead of "jackd" for the privelages you want.
> >
> > http://ccrma.stanford.edu/planetccrma/man/man1/jackstart.1.html
> 
> Just recently I read a thread on jack mailing list where they say that
> jackstart is 2.4 kernel specific, although should work (simply passing
> control to jackd) with 2.6 kernel. But that means jackd should work. 
> 
> Nevertheless if I try to use jackd I get the same result as
> Jean-Baptiste (can't lock down memory for jack).
> 
> And BTW from JACK FAQ:
> 
>    There is an allcaps=1 option, which duplicates the effect of
>    running Linux 2.4 with the capabilities patch. To use that you must
>    compile JACK with --enable-capabilites as you would for the 2.4
>    kernel. You will need this if you prefer to run jackstart -R rather
>    than jackd -R. If you only use Linux 2.6, you need not bother with
>    it.
> 
> Aha! that explain everything, no?

Pretty much. With allcaps=1 "jackstart -R" should work. With the group
stuff "jackd -R" should work fine. The "allcaps" option has been deleted
from the latest version of the lsm module (realcap) so we'll have to use
the group permissions (or all groups) in the future. 

> But I read further:
> 
>     Currently (as of 2.6.7) JACK has a serious problem creating
>     SCHED_FIFO threads for real-time processing. It is unclear whether
>     this is a bug in JACK, in the new Native PThreads Library (NPTL),
>     or in the 2.6 kernel. At the moment no one has a solution, but
>     there is a workaround: define LD_ASSUME_KERNEL=2.4.19 in the
>     environment of the jackd process and of every JACK client.
> 
> I don't have this variable set and I have no idea if JACK has these
> "serious problem". But I do experience some funny behaviour that I'm
> trying to explore further before reporting. I just don't know if it
> has anything to do with SCHED_FIFO threads.

It should not. The Planet CCRMA kernels along with the latest glibc in
fc2 should not have these problems (I'm not using or need to use
LD_ASSUME_KERNEL in my tests). 

> NB: they did say FAQ was abit out of date on jack mailing list. namely
> latest jackstart should work no matter what.

I don't think the Planet CCRMA version (latest official release) has
those patches incorporated. 

> > And since you're asking for techniques to make it work, I like to
> > create a group "audio" and then add myself and or other users to that
> > group (say it ends up being group 501).  Then when you load realcap do
> > it with:
> >
> > /sbin/modprobe allcaps=1 gid=501
> 
> you meant:
> 
> /sbin/modprobe realcap allcaps=1 gid=501
> 
> right? 
> 
> > You can even just edit the init script that Fernando mentions to make
> > this load this way at boot.
> >
> > http://ccrma.stanford.edu/planetccrma/root/etc/rc.d/init.d/realcap
> >
> > The spot where it says:
> >
> > REALCAP_OPTS="allcaps=1"
> >
> > Change to:
> >
> > REALCAP_OPTS="allcaps=1 gid=501"

Or create a file in /etc/sysconfig/realcap and add that environment
variable there...

-- Fernando