* hard drive performance in 2.6.x
@ 2003-10-20 23:49 Karol Czachorowski
0 siblings, 0 replies; only message in thread
From: Karol Czachorowski @ 2003-10-20 23:49 UTC (permalink / raw)
To: linux-kernel
I have very poor ide disk (and not only ide) preformance (file operations
aren't very slow, but load of system is very big - it's not even possible
to listen music; everything stops during copying of large files). I've
tried all 2.6.0-testX kernels. When I'm copying files, 100% CPU is used
(top says it is i/o wait) even if I'm using only floppy or CD-ROM. I have
Duron 850/256MB SDRAM on motherboard with VIA686B chipset (K7VAT+). It
looks like it's not DMA related problem (it is turned on). And I'm almost
sure, that on 2.6.0-test1 everything was ok. But now, even on test1 (I've
compiled it again) it's still bad.
hdparm settings look like in 2.4.x and they should be right. The only
difference that I've found is many errors in /proc/interrupts.
$ cat /proc/interrupts
CPU0
0: 36063353 XT-PIC timer
1: 61126 XT-PIC i8042
2: 0 XT-PIC cascade
5: 4 XT-PIC acpi
8: 183751 XT-PIC rtc
10: 3101021 XT-PIC nvidia
11: 1396448 XT-PIC uhci_hcd, uhci_hcd, eth0, eth1
12: 106607 XT-PIC EMU10K1
14: 164923 XT-PIC ide0
15: 1 XT-PIC ide1
NMI: 0
LOC: 36065300
ERR: 3075635
MIS: 0
About 3 milions errors in 10 hours. Is it possible, that all of this is
related to irq errors? What should I do to resolve the problem? 2.4.*
kernels work fine for me.
I'm using Debian/unstable.
dmesg, lspci are here (if you need more information, please let me
know):
http://syjon.fantastyka.net/~narel/2.6/
Best regards,
Karol
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2003-10-21 0:00 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-10-20 23:49 hard drive performance in 2.6.x Karol Czachorowski
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.