[2696] in linux-scsi channel archive
Re: ncr53c8xx experimental patch.
daemon@ATHENA.MIT.EDU (Gerard Roudier)
Tue Oct 28 18:31:34 1997
Date: Wed, 29 Oct 1997 00:25:24 +0100 (MET)
From: Gerard Roudier <groudier@club-internet.fr>
To: ghsong@eunhasu.kjist.ac.kr
cc: linux-scsi@vger.rutgers.edu, ncr53c810@colorado.edu
In-Reply-To: <Pine.LNX.3.95.971028143124.449A-100000@ardour.kjist.ac.kr>
On Tue, 28 Oct 1997, G. Hugh SONG wrote:
> I am extremely pleased with the following report that I am now making.
> Just like a magic, your patch worked for me like a charm. I have been
> having so much trouble with the Seagate Barracuda ST12550N 7200 rpm
> drive for a LONG time. After applying your patch, the Seagate now
> works.
It is a very good news.
> For someone who knows nothing about my problem, I once reported that
> my Seagate does not work with the ncr53c8xx driver (with 810, 815, 875
> cards) even though it works with 53c7,8xx driver with the 810,815 cards.
> The Seagate works fine under MS Windows driver with 810, 815, 875 cards.
> So, I have been saying that the problem is with ncr53c8xx driver for
> Linux. I now use ncr53c8xx Version 2.6b with kernel 2.1.60. I enabled
> multiple command queuing, 10Mbps, and all other SCSI-2 things. I
> don't really know whether my drive is actually faulty or not. Can you
> give me any possible explanation?
Not immediately, but I hope we will be able soon to understand which
change fixes your problem.
As I wrote in the announce, I think that this patch should not make
differences with flawless SCSI systems. But a SCSI system is something
very complex indeed, and 'flawless' perhaps does not apply to complex
systems. ;-)
There is no proof your drive is faulty, it only seems that one of my
changes has had the expected effect.
But there are several changes in this patch, especially in the SCRIPTS
code. This will make hard to find the real cause of the fix.
In order to have a chance to discover the good change, I need
more information, but just for now:
- driver messages about the failure with driver 2.5b or higher 2.5x.
- boot-up messages with 'verb:2' driver option for driver 2.6b.
(and 2.5x if possible)
- any (error?) message printed out by driver 2.6b after linux boot-up
and SCSI system initialisations.
> BTW, the ftp place does not have the patch for 2.5c to 2.6. So, it was
> quite a hassle to find the version 2.5b.
Sorry for the inconvenience.
Patches I make available are in the first place for lastest 2.0.X linux
versions, since, in my opinion, driver changes should be first tested
using a stable kernel version.
The same way, I think that kernel development and debugging is easier
using trustable drivers and trustable applications.
I will try to maintain a combination of patches that will allow to update
latest 2.1.X kernel too.
> I would like to express my sincere thank to Gerard.
I wish to share your thank with Richard Waltham for his tests that
helped me fixing bugs in the driver and with Stefan Esser for our
discussions about most of the changes included in the 2.6 driver.
Regards, Gerard.
