From owner-freebsd-hackers  Wed Mar 11 08:01:17 1998
Return-Path: <owner-freebsd-hackers@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id IAA29904
          for freebsd-hackers-outgoing; Wed, 11 Mar 1998 08:01:17 -0800 (PST)
          (envelope-from owner-freebsd-hackers@FreeBSD.ORG)
Received: from ns1.yes.no (ns1.yes.no [195.119.24.10])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id IAA29893
          for <hackers@FreeBSD.ORG>; Wed, 11 Mar 1998 08:01:09 -0800 (PST)
          (envelope-from eivind@bitbox.follo.net)
Received: from bitbox.follo.net (bitbox.follo.net [194.198.43.36])
	by ns1.yes.no (8.8.7/8.8.7) with ESMTP id QAA19636;
	Wed, 11 Mar 1998 16:01:08 GMT
Received: (from eivind@localhost)
	by bitbox.follo.net (8.8.6/8.8.6) id RAA17876;
	Wed, 11 Mar 1998 17:01:07 +0100 (MET)
Message-ID: <19980311170107.44621@follo.net>
Date: Wed, 11 Mar 1998 17:01:07 +0100
From: Eivind Eklund <eivind@yes.no>
To: "Justin T. Gibbs" <gibbs@plutotech.com>
Cc: hackers@FreeBSD.ORG
Subject: Re: userconfig data -> linker set -> ELF segment
References: <19980311162228.43166@follo.net> <199803111530.IAA13420@pluto.plutotech.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
X-Mailer: Mutt 0.89.1i
In-Reply-To: <199803111530.IAA13420@pluto.plutotech.com>; from Justin T. Gibbs on Wed, Mar 11, 1998 at 08:27:13AM -0700
Sender: owner-freebsd-hackers@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

On Wed, Mar 11, 1998 at 08:27:13AM -0700, Justin T. Gibbs wrote:
> >	* Loading 'just the probe code' isn't possible using a.out
> >	  (AFAIK)
> >	* Depending on device drivers being LKMs lowers reliability (N
> >	  files that can fail, instead of just having a single kernel)
> 
> Then provide a mechanism to have the driver entry point also called when
> linked statically.

This sound like the following would be about right:

for device in (static set of devices)
	find driver for in (linker set of drivers)
		call driver callback with info on device
endfor

show userconfig with devices registered by register_device

#ifdef REALLY_CORRECT_SYSTEM_NOT_YET
for each bus
	call probes
endfor
show userconfig again
for each bus
	call attach
endfor
#else
for each bus
	call probes
	call attach
endfor
#endif

driver callback:
	call register_device for device, with suitable description
	etc.

For the LKM case, the load-function would call register_device, and
(possibly) force a re-probe.


Does that sound right to you?


The above actually sounds like it might be something that is
reasonably quick to implement, and it would both solve my problem
(creating an enabling technology for externally developed drivers in a
fairly short time horizont) and be an enabling technology for further
development (good handling of LKMs with a longer time horizon.)

Only tradeoff: Old device drivers/LKMs Will Not Work, as they don't
register themselves properly.  This might be solvable (e.g, by
scanning the present linker sets).

Eivind.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message