I've debated this issue before here.
Althoug I've found no solution, I would like to try removing my seagate drive 
and make some tests, but I only have that drive, and I can't afford one now.
I've tried the patch from Tejun Heo and one patch from Silicon Image:
[http://marc.theaimsgroup.com/?l=linux-kernel&m=111173514717575&w=2]
[http://marc.theaimsgroup.com/?l=linux-ide&m=111260146423451&w=2]

and get better read performance, although I don't see any improvment in the 
write performance (I think Tejun said that).
You can try those patches and test them. I'm not sure, but I think the one 
from Silicon Image has been integrated in the latest kernel release.

When I can afford a new disk I will make some tests and then go to the store 
and claim for a new drive (even if I had to make some trick to get the drive 
faulty, while I have warranty).

By the way my controller is a Sil3112.

Regards,
Paulo Fidalgo

On Friday 22 July 2005 14:42, Roger Heflin wrote:

Hi Steve,

I saw the same thing with a tyan mb and a different Seagate disk,
removing from the blacklist sped things up a lot, and fixed some
bad kernel response issues (bad response was only with SMP,
not with single cpu).     I send an email about it several
weeks ago to this list detailing what I found.

                    Roger

quoted

-----Original Message-----
From: linux-ide-owner@vger.kernel.org
[mailto:linux-ide-owner@vger.kernel.org] On Behalf Of Steve Sutphen
Sent: Friday, July 22, 2005 6:55 AM
To: Jeff Garzik
Cc: linux-ide@vger.kernel.org
Subject: Silicon Image 3114 and Seagate disks


I have a system with a Tyan S2882 mother board that has a
Silicon Image
Si3114 SATA controller that I would like to use one of the
"blacklisted"
Seagate drives on.  As has been discussed in this mailing
list, when one uses the current driver performance is poor on
this combination (I see around 14MB/s using hdparm's buffered
read test--Bonnie++ confirms that).

I was hunting around to see if there might be a better
driver, and came across this note in the Release notes for the:

SiI SATARaid Linux device driver ver 1.0.0.20 Release

Both the Tyan support download driver and the one from
SiliconImage have the following "Limitation" listed:

3. Mod15Write fix for Seagate Drives (for chip-set versions
3112 1.21 or older)

It seems that this driver may treat the 3114 and the 3112 the
same way, and that at least in the 3112 code the Seagate
problem may not exist on newer versions of the SI chip.
Lspci thinks that I have a Rev.2 chip/firmware:
04:05.0 Unknown mass storage controller: Silicon Image, Inc.
SiI 3114 [SATALink/SATARaid] Serial ATA Controller (rev 02)

The SI firmware version that my SI chip is running is 5.0.48.

My question is--is it possible that my ST3120026AS Seagate
drive might work without the "blacklisting"?  Should I try
it?  If so, what is a good test--will it lock up, or corrupt data?

If not with the current driver, then is there a new driver
that is coming down the pipe that will fix this in the
reasonably near term?

Thank you,
	steve@cs.ualberta.ca
-
To unsubscribe from this list: send the line "unsubscribe
linux-ide" in the body of a message to
majordomo@vger.kernel.org More majordomo info at
http://vger.kernel.org/majordomo-info.html

-
To unsubscribe from this list: send the line "unsubscribe linux-ide" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html