> On Thu, Mar 11, 1999 at 01:36:25PM +0200, Craig Schlenter wrote:
> > The only thing that works properly for me is either x11amp or amp-0.7.6.
> >
> > amp-0.7.6 was compiled to do realtime stuff and has to run as root.
>
> I used to use x11amp, and ran it as root for the real-time support.
> Had to switch to mpg123 when x11amp got unhappy with my BIIIIIIIG
> play-lists. So, I wrote a lil utility, which is against 2.0.x kernels,
> but may work straight off on 2.2.x too, to set a process to real-time
> priority. The source is attached to this email.
> Compile as 'gcc -o setrealtime setrealtime.c' install as that, make
> it setuid root, and a symlink to it called 'getrealtime'. Then you can
> set any process to real-time, and/or get the current scheduler and
> priority for any process. Yeah, it could do with some security checks
> on a multi-user box, but mine isn't.
>
> In the hope someone finds this useful,
Hi,
A while back, someone (name lost) posted a url to schedctl.c which
does the same thing but can start a job in whatever scheduler mode.
schedctl --rr --pri n command args
I use this tool to prevent skipping with both mpg123 and mpg123-nas
when stress testing. (used mostly for starting vmstat 1 and such)
I found that if I use the nas server and nas version of mpg123 both
set rr and set to the same priority, they interfere with each other
badly. The kernel gripes constantly with..
Sound: DMA (output) timed out - IRQ/DRQ config error?
If I set the nas server to priority x and mpg123.nas to priority y,
all is well. Is this normal rr behavior?
(in any case, it might be useful for folks to know that setting both
client and server to the same rr priority generates skippity-skip)
-Mike
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/