LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* Re: [patch] Re: AMD viper chipset and UDMA100
@ 2001-10-05 21:26 Steven Newbury
  0 siblings, 0 replies; 2+ messages in thread
From: Steven Newbury @ 2001-10-05 21:26 UTC (permalink / raw)
  To: linux-kernel; +Cc: vojtech

On Sep 28, 2001 at 10:21:14, Vojetch Pavlik wrote:
>
> On Thu, Sep 27, 2001 at 04:37:50PM -0700, Sean
Swallow wrote:
> 
> > Vojtech,
> > I would be interested in trying out your code.
>
> Ok. It should make your ViperPlus work with UDMA100.
> The patch is
> against 2.4.10, but will probably work with any
similar kernel. It
> completely replaces Andre's driver with mine. Find
it > attached.
> 
> Good luck. 
[patch snipped]
I have a Tyan K7 Thunder and I have tried this patch
and the patch from Andre, synced with 2.4.10 by Jacob
Luna and have had the same problem with both.  Without
the patch I can only do UDMA33 but the hw supports
UDMA100.  I have a Quantum FireballPlus AS40 on my
primary IDE channel which I have tested on an MSI
K7Master and works fine in UDMA100.

When I reboot with the driver enabled, as soon as hda
is detected the kernel locks up with a DIVIDE (0)
error.  I will try it again and take notes of more
details if it helps.

Without the driver, using the generic code I am able
to use the drive, in I think UDMA100 by letting the
BIOS set the drive to UDMA100 then manually setting
DMA mode on with hdparm.  However if I try to set the
modes manually with hdparm I can not select above
UDMA66.

I have tried this with kernel 2.4.10 and 2.4.11pre1.
My system is a Dual K7, SMP kernel, 512MB RAM.

My root filesystem is on a RAID0 array across 4
Quantum AtlasV 9GB U160 SCSI disks, two on each
channel of the onboard Adaptec.

I have tried to use the Fireball in the array as well
but for some reason performance reduced!  110MB/s with
hdparm -t compared to 70MB/s. Could this be because of
the lack of chipset driver?

One other thing is I am running with the preemt patch,
though I tested it without as well as I remember.

Please CC any replys to steven.newbury1@ntlworld.com
as  I am not on the list.
--
Steven Newbury


____________________________________________________________
Do You Yahoo!?
Get your free @yahoo.co.uk address at http://mail.yahoo.co.uk
or your free @yahoo.ie address at http://mail.yahoo.ie

^ permalink raw reply	[flat|nested] 2+ messages in thread
* Re: AMD viper chipset and UDMA100
@ 2001-09-27 22:24 Vojtech Pavlik
  2001-09-27 23:37 ` Sean Swallow
  0 siblings, 1 reply; 2+ messages in thread
From: Vojtech Pavlik @ 2001-09-27 22:24 UTC (permalink / raw)
  To: Dan Hollis; +Cc: linux-kernel

On Thu, Sep 27, 2001 at 03:16:18PM -0700, Dan Hollis wrote:

> On Thu, 27 Sep 2001, Vojtech Pavlik wrote:
> > On Thu, Sep 27, 2001 at 02:26:50PM -0700, Sean Swallow wrote:
> > > I just got a tyan tiger w/ the AMD Viper chipset on it. For some reason
> > > Linux will only set the onboard (AMD viper) chains to UDMA33.
> > > I'm using linux 2.4.9, and I have also tried 2.4.10. Is there a limitation
> > > to the AMD Viper driver?
> > > PS. The cables (2) are BRAND new ata100 cables.
> > The Viper can do UDMA66 max. At least it's doing it for me quite well.
> 
> If he's got the Tyan Tiger i'm thinking of, it's the Tyan S2460 with
> AMD766 southbridge. Tyan says it will do U100:
> http://www.tyan.com/products/html/tigermp.html

This southbridge is called Viper Plus as far as I know. In the past,
I've written a driver for this, but it didn't make it into the kernel,
because Andre also had his version.

> AMD does too:
> http://www.amd.com/us-en/Processors/TechnicalResources/0,,30_182_873,00.html
> 
> And I've got the same problem with my S2460:
> 
> Uniform Multi-Platform E-IDE driver Revision: 6.31
> ide: Assuming 33MHz system bus speed for PIO modes; override with idebus=xx
> AMD7411: IDE controller on PCI bus 00 dev 39
> AMD7411: chipset revision 1
> AMD7411: not 100% native mode: will probe irqs later
> AMD7411: disabling single-word DMA support (revision < C4)
>     ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:pio
>     ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:pio
> hda: 20010816 sectors (10246 MB) w/2048KiB Cache, CHS=1245/255/63, UDMA(33)
> hdc: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(33)
> 
> But these drives will do U66/U100:
> 
> And yes, I *DO* have the proper cables. These same drives connected to a
> Promise 20267 or a VIA KT133 with the same cables will do U66/U100
> perfectly.

Perhaps I could dig up my code and you can give it a test. Would you be
interested?

-- 
Vojtech Pavlik
SuSE Labs

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

end of thread, other threads:[~2001-10-05 21:26 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-10-05 21:26 [patch] Re: AMD viper chipset and UDMA100 Steven Newbury
  -- strict thread matches above, loose matches on Subject: below --
2001-09-27 22:24 Vojtech Pavlik
2001-09-27 23:37 ` Sean Swallow
2001-09-28 10:21   ` [patch] " Vojtech Pavlik

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).