RE: 2.6.11, IDE: Strange scheduling behaviour: high-pri RT process not scheduled?

[kus Kusche Klaus]

> From: Bartlomiej Zolnierkiewicz [mailto:bzolnier@xxxxxxxxx] 
>
> On Wed, 30 Mar 2005 13:52:05 +0200, kus Kusche Klaus 
> <kus@xxxxxxxx> wrote:
> > However, things break seriously when exercising the CF card 
> in parallel
> > (e.g. with a dd if=/dev/hda of=/dev/null):
> > 
> > * The rtc *interrupt handler* is delayed for up to 250 
> *micro*seconds.
> > This is very bad for my purpose, but easy to explain: It is 
> roughly the
> > time needed to transfer 512 Bytes from a CF card which can 
> transfer 2
> > Mbyte/sec, and obviously, the CPU blocks all interrupts 
> while making pio
> > 
> > transfers. (Why? Is this really necessary?)
> >  
> > (I know because I instrumented the rtc irq handler with 
> rdtscl(), too)
> 
> hdparm -u1 /dev/hda
> 
> should help

Hmmm, thanks, that sounds very reasonable, and I didn't know that flag.

Unfortunately, it doesn't help. The bad timings stay the same (still
delays in the 30-300 ms range), the number of context switches stays the
same, ...

The only thing which changes is the CPU load shown by vmstat:
* With -u0, I have 1 % user, ~50 % sys, ~50 % wa
* With -u1, I have 1 % user, ~98 % sys, 1 % wa

P.S.: Apologies for my badly formatted mails. The company forces us to
use outlook, we may not even change the settings... 

Klaus Kusche
>Entwicklung Software - Steuerung
>Software Development - Control
>
>KEBA AG
>A-4041 Linz
>Gewerbepark Urfahr
>Tel +43 / 732 / 7090-3120
>Fax +43 / 732 / 7090-8919
>E-Mail: kus@xxxxxxxx
>www.keba.com
>
-
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/