Re: UDMA problem with Maxtor 7200rpm disk

[GOTO Masanori]

At Wed, 19 Dec 2001 15:16:36 -0500,
Jean-Francois Levesque <[e-mail]> wrote:
> I have a problem regarding my new Asus A7V266 board with VIA KT266 chipset.  Byron Stanoszek told me to ask my problem to this list so here it is :
> 
> My hard drive is a Maxtor 5T030H3 ATA DISK drive (30 gig).  The problem is that I'm not able to read more than 7 MB/sec :
> 
> [root@xyz jfl]# /sbin/hdparm -t /dev/hda
> 
> /dev/hda:
>  Timing buffered disk reads:  64 MB in  9.18 seconds =  6.97 MB/sec
> 
> 
> [root@xyz jfl]# /sbin/hdparm -d1 -X66 /dev/hda
> 
> /dev/hda:
>  setting using_dma to 1 (on)
>  setting xfermode to 66 (UltraDMA mode2)
>  using_dma    =  1 (on)
> [root@xyz jfl]# /sbin/hdparm -t /dev/hda
> 
> /dev/hda:
>  Timing buffered disk reads:  64 MB in  9.70 seconds =  6.60 MB/sec
> 
> [root@xyz jfl]# /sbin/hdparm /dev/hda
> 
> /dev/hda:
>  multcount    = 16 (on)
>  I/O support  =  1 (32-bit)
>  unmaskirq    =  1 (on)
>  using_dma    =  0 (off)
>  keepsettings =  0 (off)
>  nowerr       =  0 (off)
>  readonly     =  0 (off)
>  readahead    =  8 (on)
>  geometry     = 3736/255/63, sectors = 60030432, start = 0
> [root@xyz jfl]#
> 
> I also have some idebus errors.
> 
> The problem seems to be the DMA (ATA100 compatible board and disk).

BIOS parameter of your machine is set as "DISK auto detection" ?  If
you so, would you try to change your BIOS setting as "this disk have
ultra DMA mode 4 (or 5, if your disk is ATA100)" explicitly ?

These behavior is very similar to my enviroment.
I have Asus KT133A board, and I had this problem, but I changed BIOS
setting, all problems were gone and became happy.
If my suggestion is totally wrong, I apologize...
(However, if my suggestion is relevant, then what's the problem?
 IDE driver, or hardware problem?)

> PS: My disk also "freeze" my system for a few seconds (from 1/2 to maybe 3sec) while checking some data.

IMHO, the reason is that writing disk with PIO mode 4 (IIRC, 6.60MB/s
is corresponding with PIO mode 4) invites CPU occupation.
Andrew Morton also fixed 'low latency read problem', which is already
resolved the latest 2.4.17-rc2.

-- gotom