Re: [PATCH] In-kernel module loader 1/7

From: Daniel Phillips <phillips@arcor.de>
To: Rob Landley <landley@trommello.org>, Greg KH <greg@kroah.com>,
	Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Roman Zippel <zippel@linux-m68k.org>,
	Rusty Russell <rusty@rustcorp.com.au>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] In-kernel module loader 1/7
Date: Fri, 4 Oct 2002 02:10:28 +0200	[thread overview]
Message-ID: <E17xG33-0001ji-00@starship> (raw)
In-Reply-To: <20021003235152.B41E3397@merlin.webofficenow.com>

On Thursday 03 October 2002 20:53, Rob Landley wrote:
> On Monday 30 September 2002 11:32 am, Daniel Phillips wrote:
> 
> > Not being able to unload LSM would suck enormously.  At last count, we
> > knew how to do this:
> >
> >   1) Unhook the function hooks (using a call table simplifies this)
> >   2) Schedule on each CPU to ensure all tasks are out of the module
> >   3) A schedule where the module count is incremented doesn't count
> >
> > and we rely on the rule that and module code that could sleep must be
> > bracketed by inc/dec of the module count.
> >
> > Did somebody come up with a reason why this will not work?
> 
> Preemption?

Turn it off.

> Scheduling doesn't guarantee making any specific amount of progress within 
> the kernel with preemption enabled.  I thought the preferred strategy was to 
> wait for the time slices to refill and then exhaust (since everybody has to 
> exhaust their time slices before anybody gets new ones.  Unless I've missed 
> something...?)

Probably ;-)

-- 
Daniel