Re: Latency data - 2.6.14-rc3-rt13

From: Steven Rostedt
Date: Tue Oct 11 2005 - 01:54:25 EST



On Mon, 10 Oct 2005, Mark Knecht wrote:

> On 10/10/05, Mark Knecht <markknecht@xxxxxxxxx> wrote:
> >
> > ( softirq-timer/0-3 |#0): new 3997 us maximum-latency critical section.
>
> So the root cause of this 4mS delay is the 250Hz timer. If I change
> the system to use the 1Khz timer then the time in this section drops,
> as expected, to 1mS.
>
> ( softirq-timer/0-3 |#0): new 998 us maximum-latency critical section.
> => started at timestamp 121040020: <acpi_processor_idle+0x20/0x379>
> => ended at timestamp 121041019: <thread_return+0xb5/0x11a>
>
> So, thinking very interesting here I think.
>
> Back to the drawing board as to my xruns.
>

Are your xruns showing the same type of latency? If you switch to the
1000Hz do you get only 1mS latency on your xruns too? This sounds like
the application might have gone to sleep and didn't wake up until
something scheduled it in. Or something else with the scheduler. I doubt
that this has to due with preemption or interrupts being off, but
something that uses jiffies to calculate.

-- Steve

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/