[1245] in linux-scsi channel archive
Who is the SCSI maintainer? & Re: 2.0.27: aha1740[9]_mbxout wait!
daemon@ATHENA.MIT.EDU (Michael Weller)
Thu Jan 9 10:04:32 1997
Date: Thu, 9 Jan 1997 16:00:26 +0100 (MEZ)
From: Michael Weller <eowmob@exp-math.uni-essen.de>
To: Jon Lewis <jlewis@inorganic5.fdt.net>,
Barrie Spence <barrie@calvin.demon.co.uk>
Cc: linux-scsi@vger.rutgers.edu, linux-kernel@vger.rutgers.edu,
Andreas Arens <ari@av.rwth-aachen.de>
I had a look at the source. I was right, it just emits the warning, ensures
that the printk does not open the hole for a race condition and continues.
As I got no report of lockups, only of these warnings, I assume that
there is no problem, just the driver is to picky.
Before I make a patch, two major questions remain:
a) I'd prefer if I could loop for a given amount of time (say 100ms) before
logging a warning. This could be done by enabling interrupts and counting
jiffies. However, although I think it might be save there, it sounds like
playing with Pandora's box to me to enable irqs in there.
Just making <n> check loops for a well choosen <n> might work but take 10
minutes for a warning on a slow machine and be to fast on a fast one.
Making <n> loops with a given delay loop (BogoMips based) in there would
work, but slow things down when the card becomes ready long before the
tiny delay expires.
Suggestions from someone more uptodate with recent kernel code?
b) I got an aha174x.c patch from Andreas Arens ages ago. It does not change
this code but adds multiple aha174x.c card support and some cosmetic
stuff and extended translation support.
THIS PATCH IS IN NONE OF THE OFFICIAL KERNELS, EVEN NOT IN THE ALPHA
SERIES => WHY?
I know it was submitted to the SCSI stuff maintainer by Andreas months
ago. What happened? And, where to send the patches instead?
Michael.
(eowmob@exp-math.uni-essen.de or eowmob@pollux.exp-math.uni-essen.de
Please do not use my vm or de0hrz1a accounts anymore. In case of real
problems reaching me try mat42b@spi.power.uni-essen.de instead.)
