FreeBSD 5.3b7and poor ata performance
John-Mark Gurney
gurney_j at resnet.uoregon.edu
Tue Oct 26 16:14:59 PDT 2004
fandino wrote this message on Tue, Oct 26, 2004 at 17:19 +0200:
> John-Mark Gurney wrote:
> >Yep, and I get decent performance:
>
> could repeat the test setting udma mode to UDMA4?
Sure, though there is only one drive in this machine, so there is no
problems with channel colissions...
> >dmesg clip:
> >atapci0: <VIA 82C686B UDMA100 controller> port
> >0xc000-0xc00f,0x376,0x170-0x177,0
> >x3f6,0x1f0-0x1f7 at device 7.1 on pci0
> >atapci0: Correcting VIA config for southbridge data corruption bug
>
> hhhmmm, my dmesg doesn't print the message for the 686B data corruption bug.
>
> >ata0: channel #0 on atapci0
> >ata1: channel #1 on atapci0
> >ad0: 43979MB <IBM-DTLA-307045/TX6OA60A> [89355/16/63] at ata0-master
> >UDMA100
> >Transfer rates:
> > outside: 102400 kbytes in 2.857371 sec = 35837
> > kbytes/sec
> > middle: 102400 kbytes in 3.561157 sec = 28755
> > kbytes/sec
> > inside: 102400 kbytes in 5.635770 sec = 18170
> > kbytes/sec
Well, looks like I get similar performance w/ UDMA4 aka UDMA66:
Transfer rates:
outside: 102400 kbytes in 2.828054 sec = 36209 kbytes/sec
middle: 102400 kbytes in 3.506356 sec = 29204 kbytes/sec
inside: 102400 kbytes in 5.641238 sec = 18152 kbytes/sec
Note, I didn't reboot just before switching to UDMA4...
--
John-Mark Gurney Voice: +1 415 225 5579
"All that I will do, has been done, All that I have, has not."
More information about the freebsd-current
mailing list