Re: 2.6.16.1 & D state processes

[Srihari Vijayaraghavan <sriharivijayaraghavan@yahoo.com.au>]

--- Mike Galbraith <efault@gmx.de> wrote:
> On Tue, 2006-04-18 at 19:35 +1000, Srihari
> Vijayaraghavan wrote:
> > --- Mike Galbraith <efault@gmx.de> wrote:
> > > On Tue, 2006-04-18 at 15:07 +1000, Srihari
> > > Vijayaraghavan wrote:
> > > > io scheduler cfq registered (default)
> > > ...
> > > Hmm.  Recovers [odd] but takes long time
> [odder]. 
> > > I'd try to eliminate
> > > io scheduler at this point.
> > 
> > Interesting. Considering the minimal .config,
> where I
> > haven't experienced this problem over a week
> uptime,
> > also having CFQ as the default elevator, do you
> still
> > believe CFQ is involved? (I guess if CFQ could be
> > influenced by other kernel configurations, then
> > perhaps another elevator might help. It's worth
> > trying.)
> 
> I don't know that CFQ is involved.  With it
> recovering though, the only
> thing I could think of was a request stucking in the
> io scheduler's
> gizzard for some reason.
> 
> It's just a suggestion, and one you can try without
> even rebooting.

Thanks for that. When it happened today on 2.6.16.7,
I've started using deadline now.

A couple of interesting things I've noticed after its
recovery:
1. Executing "time sleep 1", takes more than 1 second.
It reports real as 3 to 5 seconds, while my stop watch
measures it as closer to 50 seconds.
2. Pressing & holding a key at the console doesn't
produce repeating characters.

Thought they might shed some light on the problem. (I
ought to look at all the RTC & Time related settings
between my minimal .config & FC5's, for I believe
they're connected.)

I haven't rebooted it yet. I'm sure it'll be back to
normal after the reboot.

Thanks

Send instant messages to your online friends http://au.messenger.yahoo.com 
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/