public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* 2.5.2-pre9 - HD performance degradation
@ 2002-01-07 21:07 Carl Scarfoglio
  2002-01-07 22:58 ` Matthias Hanisch
  0 siblings, 1 reply; 2+ messages in thread
From: Carl Scarfoglio @ 2002-01-07 21:07 UTC (permalink / raw)
  To: linux-kernel

I ran hdparm -t on my disks and discovered an abysmal drop in hard disk 
performance since 2.5.2-pre1. Yesterday I ran fsck on all disks and it 
took more than three times as much to complete under 2.5.2-pre9 than 
usual. I am running SuSE 6.3, MB is Asus A7v, controller ATA 33 + 
Promise PDC 20265.
These are the results for a disk, (ATA 100 IBM 307030),for the other 
disks they are pretty the same.
Resuslts for "hdpam -t /dev/hdg"
Kernel 2.5.2-pre1 - 35 MB/sec
Kernel 2.5.2-pre4 - 15 MB/sec
Kernel 2.5.2-pre9 - 10 MB/sec
For the rest, it seems pretty stable, but I still get kernel panic on 
cold boots from the AHA 2904 (AIC7850).
Cheers,
		Carlo Scarfoglio


^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: 2.5.2-pre9 - HD performance degradation
  2002-01-07 21:07 2.5.2-pre9 - HD performance degradation Carl Scarfoglio
@ 2002-01-07 22:58 ` Matthias Hanisch
  0 siblings, 0 replies; 2+ messages in thread
From: Matthias Hanisch @ 2002-01-07 22:58 UTC (permalink / raw)
  To: Carl Scarfoglio; +Cc: linux-kernel

On Mon, 7 Jan 2002, Carl Scarfoglio wrote:

> I ran hdparm -t on my disks and discovered an abysmal drop in hard disk 
> performance since 2.5.2-pre1. Yesterday I ran fsck on all disks and it 
> took more than three times as much to complete under 2.5.2-pre9 than 
> usual. I am running SuSE 6.3, MB is Asus A7v, controller ATA 33 + 
> Promise PDC 20265.
> These are the results for a disk, (ATA 100 IBM 307030),for the other 
> disks they are pretty the same.
> Resuslts for "hdpam -t /dev/hdg"
> Kernel 2.5.2-pre1 - 35 MB/sec
> Kernel 2.5.2-pre4 - 15 MB/sec
> Kernel 2.5.2-pre9 - 10 MB/sec
> For the rest, it seems pretty stable, but I still get kernel panic on 
> cold boots from the AHA 2904 (AIC7850).
> Cheers,
> 		Carlo Scarfoglio


Could be scheduler related. Please try the fresh fix from Davide:

> In sched.c::init_idle() :
>
> current->dyn_prio = -100;
> 
> Let me know.

For more info, see also the related thread here in LKML:

http://marc.theaimsgroup.com/?t=101019244200001&r=1&w=2&n=12

	Matze


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2002-01-07 23:06 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-01-07 21:07 2.5.2-pre9 - HD performance degradation Carl Scarfoglio
2002-01-07 22:58 ` Matthias Hanisch

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox