[plug] hdparm troubles

[Craig Ringer]

On Sat, 2004-07-17 at 11:14, Steve Boak wrote:
> Hi all,
> 
> I'm trying to optimise the hd performance on my Via C3 500MHz system, and 
> wondering why I have such terrible performance with what I thought was a 
> fairly good configuration.
> 
> min:~# hdparm -X68 -c3 -m16 -d1 /dev/hda
> /dev/hda:
>  setting 32-bit IO_support flag to 3
>  setting multcount to 16
>  setting using_dma to 1 (on)
>  setting xfermode to 68 (UltraDMA mode4)
>  multcount    = 16 (on)
>  IO_support   =  3 (32-bit w/sync)
>  using_dma    =  1 (on)

Have you tried flipping unmasked interrupts (-u) on? I'd also try
enabling DMA without trying to set the speed (let the disk/interface
decide it) by omitting the -X68. I'd also try 16 bit mode .. just in
case there's some stupid driver or controller bug. On the same line of
thinking, I'd also run some tests with multicount disabled.

When you're doing the testing, what sort of CPU load do you see?

Have you tried a recent kernel? Perhaps the via IDE drivers have
improved for these boards. IIRC they're known to be a little eccentric
and have had Linux support issues in the past.

Do the boards have an IO-APIC? If yes, try booting with 'noapic' in case
it's braindead. Check /proc/interrupts to see if your IDE controller is
being forced to share interrupts with anything. Do you have anything
else in the system that might be generating heavy interrupt loads (TV
input cards, for example)?

> min:~# hdparm -Tt /dev/hda
> /dev/hda:
>  Timing buffer-cache reads:   192 MB in  2.03 seconds =  94.58 MB/sec
>  Timing buffered disk reads:   12 MB in  3.20 seconds =   3.75 MB/sec
> 
> 3.75 MB/sec ???

Wow.

--
Craig Ringer