From owner-cvs-all  Tue Dec 29 23:24:44 1998
Return-Path: <owner-cvs-all@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id XAA13676
          for cvs-all-outgoing; Tue, 29 Dec 1998 23:24:44 -0800 (PST)
          (envelope-from owner-cvs-all@FreeBSD.ORG)
Received: from firewall.reed.wattle.id.au (darren2.lnk.telstra.net [139.130.53.33])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id XAA13670;
          Tue, 29 Dec 1998 23:24:40 -0800 (PST)
          (envelope-from darrenr@reed.wattle.id.au)
Received: (from root@localhost)
	by firewall.reed.wattle.id.au (8.9.1/8.8.7) id HAA26366;
	Wed, 30 Dec 1998 07:24:19 GMT
Received: from avalon.reed.wattle.id.au(192.168.1.1) by firewall.reed.wattle.id.au via smap (V1.3)
	id sma026364; Wed Dec 30 07:23:57 1998
Received: from percival.reed.wattle.id.au. (percival.reed.wattle.id.au [192.168.1.5])
	by avalon.reed.wattle.id.au (8.9.0.Beta3/8.9.0.Beta3) with SMTP id SAA16755;
	Wed, 30 Dec 1998 18:23:57 +1100 (EST)
From: Darren Reed <darrenr@reed.wattle.id.au>
Message-Id: <199812300723.SAA16755@avalon.reed.wattle.id.au>
Subject: Re: kvm_nlist emulation of n_type from kld symbol table at runtime.
To: peter@netplex.com.au (Peter Wemm)
Date: Wed, 30 Dec 1998 18:23:56 +1100 (EST)
Cc: phk@FreeBSD.ORG, committers@FreeBSD.ORG
In-Reply-To: <199812300707.PAA84781@spinner.netplex.com.au> from "Peter Wemm" at Dec 30, 98 03:07:43 pm
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text
Sender: owner-cvs-all@FreeBSD.ORG
Precedence: bulk

In some email I received from Peter Wemm, sie wrote:
> 
> Darren Reed wrote:
> > 
> > If you're exporting symbols through something like a /dev/ksyms where
> > you can do "nm /dev/ksyms" and get a human-readable listing of the in-kernel
> > table, then n_type is essential (for obvious reasons).
> 
> Yes, I thought about that a few times and I'm fairly convinced that doing 
> a /dev/ksyms will be Really Hard (tm) and/or resource expensive given the 
> way things are presently implemented.

I don't think a /dev/ksyms needs to be "fast", but if implemented well
could possibly be used with mmap().

> In a situation where there is a single loaded symbol table, this is easy.
> However, we have multiple symbol tables and in multiple formats.

Well, then you need to pick one which can be used as a superset of the two.
I'd expect they're all storing the same sort of information so this shouldn't
be too hard.

> At
> present, I think the majority of cases have an a.out kernel symbol table
> and ELF module symbols.  A coherent symbol table would have to be
> constructed, and that would take a large block of memory to do it in one
> chunk...

I'm confused about what you're saying here, the need for n_type shouldn't
go away because of a.out -> ELF.  Solaris has a.out & ELF support and at
least appears to still support nlist.h on both a.out and ELF binaries.

> I'm really not sure how to best get large synthetic blocks of data out to 
> the user.  I know sysctl(2) has enough trouble.

Try creating a pseudo chracacter device, with a buffer and apply read(2)
or mmap(2) where necessary.

Darren

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