Return to BSD News archive
Newsgroups: comp.unix.bsd
Path: sserve!manuel.anu.edu.au!munnari.oz.au!spool.mu.edu!sol.ctr.columbia.edu!hamblin.math.byu.edu!news.byu.edu!ux1!fcom.cc.utah.edu!cs.weber.edu!terry
From: terry@cs.weber.edu (A Wizard of Earth C)
Subject: Re: 386BSD: Problem in wd.c (?) - affects swap, non-atatched drive(s)
Message-ID: <1992Dec15.190748.5260@fcom.cc.utah.edu>
Sender: news@fcom.cc.utah.edu
Organization: Weber State University (Ogden, UT)
References: <1992Dec11.175951.54@uc.msc.edu> <1992Dec14.155504.5110@novatel.cuc.ab.ca>
Date: Tue, 15 Dec 92 19:07:48 GMT
Lines: 88
In article <1992Dec14.155504.5110@novatel.cuc.ab.ca> hpeyerl@novatel.cuc.ab.ca (Herb Peyerl) writes:
>This really SHOULD be in the FAQ since it affects ANYONE using MFM/RLL
>drives and having installed the patches from the patchkit.
[ ... ]
>I went through the exact same thing a few weeks ago and came up with the
>same symptoms... When I posted the same sort of message to the net that
>you did, someone sent me a message telling me exactly what I'm telling
>you now.
>
>If you use the Norway wd driver, you won't have this problem, you will
>also have proper bad144 support and proper multi-drive support. You
>Can get the Norway wd driver from agate.berkeley.edu in (I think) the
>unsupported directories. It might be under "barsoom" (which is where
>it originated). Anyways, the file I got was called "drivepatch.tar" and
>contains:
>
>rwxr-xr-x 268/20 0 Jul 31 04:27 1992 ./
>rw-r--r-- 268/20 3162 Jul 31 04:22 1992 fd.c.patch
>rw-r--r-- 268/20 396 Jul 31 04:22 1992 isa_device.h.patch
>rw-r--r-- 268/20 5098 Jul 31 04:21 1992 mkioconf.c.patch
>rw-r--r-- 268/20 3510 Jul 31 04:22 1992 wd.c.patch
>rw-r--r-- 268/20 991 Jul 31 04:23 1992 CONFIG.SAMPLE
>
>I sent Terry a message when I discovered this asking him to possibly
>remove the appropriate patches from his next release of the patchkit
>and replace them with these patches but I didn't receive a response
>which could indicate that he's either very busy (likely) or didn't
>receive my message. (or third possibility: thinks I'm a weenie and
>chose to ignore me. :-> )
>
>Anyways, this is definitely one for the FAQ.. (Terry; are you listening?)
It may be one for the FAQ, but then again, it's certainly not one for
inclusion as a patch.
While it's true that the "Norway" wd driver is vastly superior to the
existing driver, there are several issues which complicate matters
greatly:
1) The lack of true configuration management prevents the use of
"drop in" driver modules which conflict with existing devices;
in particular, it is not possible to choose between the two
with an "option" entry in the config file.
2) In general, there are "patches" and there are "complete drivers";
the "Norway" wd driver is a complete module. Generally, the
patchkit software is *not* installation software -- that is, it
handles complete modules relatively poorly.
3) There is an existing problem with the wd probe routine which is
carried over into the "Norway" driver; this has to do with the
recognition of "non-wd wd drives" -- in particular, a distinction
must be made between noraml wd controllers and IDE/WD7000-FASST2
controllers, or we will never come to closure on these (and
other) problems. I have this problem nearly solved in the
"standard" driver, but not in the "Norway" driver.
4) A modification needs to be made to the user's config; unfortunately,
not only is it not possible to identify which config file the
user will be using, it's not possible to patch multiple config
files in an appropriate manner. This has something to do with
the fact that kernel configuration is still rather primitive.
The fact that "drop in" dirvers are not supported in the kernel or in
kernel management, the probe routine, and the deficiencies in the patchkit
software all collide to make installation non-trivial for the casual
386BSD user. Thus, it was not included in the patchkit.
I have been trying to update the FAQ, but the sheer volume of information
is retarding the process; while the "Norway" driver bears mentioning, I
believe that it will in all liklihood be a standard part of 0.2; given
that this will reduce the utility of it being in the FAQ, whether or not
it's included will depend greatly on other issues being resolved first.
PS: You're not a weenie. 8-).
Terry Lambert
terry@icarus.weber.edu
terry_lambert@novell.com
---
Any opinions in this posting are my own and not those of my present
or previous employers.
--
-------------------------------------------------------------------------------
"I have an 8 user poetic license" - me
Get the 386bsd FAQ from agate.berkeley.edu:/pub/386BSD/386bsd-0.1/unofficial
-------------------------------------------------------------------------------