[9fans] [dolecek@ics.muni.cz: Re: LKMs]

[9fans] [dolecek@ics.muni.cz: Re: LKMs]

[Lucio De Re]

Interesting coincidence, this...

I think I agree with Jaromír that debugging is a lot simpler, and with
Presotto that a generic kernel matching the union of the needs of a
large number of workstations is unlikely to be optimal in any other
fashion.

----- Forwarded message from Jaromír Dolecek <dolecek@ics.muni.cz> -----

From: Jaromír Dolecek <dolecek@ics.muni.cz>
Subject: Re: LKMs
In-Reply-To: <Pine.NEB.4.21.0011250148110.20903-100000@love.warhead.org.uk>
 "from Al Snell at Nov 25, 2000 02:02:38 am"
To: Al Snell <alaric@alaric-snell.com>
Date: Sat, 25 Nov 2000 10:03:04 +0100 (CET)
CC: NetBSD Kernel list <tech-kern@netbsd.org>
X-Mailer: ELM [version 2.4ME+ PL78 (25)]
Precedence: list
Delivered-To: tech-kern@netbsd.org

Al Snell wrote:
> What is the intended future of LKMs? I see that there are changes afoot in
> -current to try to enable emulations to be loaded as LKMs, but are there
> any plans to try to actually move stuff out of the kernel and into LKMs
> instead?

There are some plans to revamp the thing, adding some advanced
features as clean interdependencies, less memory wasting, correct
symbol visibility etc. The linker would run in kernel, userland
would then only upload the raw module.o file to kernel. However,
the work is in very initial stage ATM.

The definite plan is surely to have ~all loadable. The most immediate
effect is that it eases debugging tremendously; obviously, it's
nice even for user not to have to recompile kernel :)

Jaromir
-- 
Jaromir Dolecek <jdolecek@NetBSD.org>      http://www.ics.muni.cz/~dolecek/
@@@@  Wanna a real operating system ? Go and get NetBSD, damn!  @@@@

----- End forwarded message -----