linux-hotplug.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Oliver Neukum <oliver@neukum.org>
To: linux-hotplug@vger.kernel.org
Subject: Re: Should driver changes cause hotplug event
Date: Sun, 29 Feb 2004 00:43:11 +0000	[thread overview]
Message-ID: <200402290143.12489.oliver@neukum.org> (raw)
In-Reply-To: <20040216052819.28584.qmail@web14916.mail.yahoo.com>

Am Donnerstag, 26. Februar 2004 18:57 schrieb Greg KH:
> On Tue, Feb 24, 2004 at 11:30:19PM +0100, Oliver Neukum wrote:
> > Am Montag, 16. Februar 2004 22:04 schrieb Greg KH:
> > > On Sun, Feb 15, 2004 at 09:28:19PM -0800, Jon Smirl wrote:
> > > > 1) I had radeonfb loaded. It claims my video card.
> > > > 2) Next I load radeon, it will also claim my video card, but radeonfb is
> > > > blocking it, so it loads and doesn't initialize.
> > > > 3) Now I rmmod radeonfb. 
> > > > 
> > > > Should the hotplug system notice this and attach radeon to the
> > > > hardware since it is already loaded? I tried and it didn't attach
> > > > radeon to the hardware.
> > > 
> > > This isn't a hotplug issue.  It's a driver core issue.  Currently the
> > > kernel does not try to match up drivers to any device that just had a
> > > driver removed from it.
> > 
> > Potentially dangerous. There needs to be a way to get a driver off
> > a device.
> 
> I don't understand.  That would work just the same way as before.  Just
> that when we remove a driver off of a device, we might want to check all
> other drivers that are currently present in the system to see if they
> want to bind to the device.

But, as David has convinced me, there needs to be a way to make sure
that a device is bound to no driver. Currently we have some halfassed way
of doing this for USB. It will work while you don't load new modules.

> Something like this will be necessary if we start to add "levels" of
> bindness to drivers (kind of like what microsoft has.)  People have been
> talking about this being one of the things we do for 2.7.

Yes, but think of user space drivers. If you really want to do this, you
need a way of claiming devices from user space. This opens a new can of
worms. You need hooks in exit() to give up such claims in case of unplanned
termination.

	Regards
		Oliver



-------------------------------------------------------
SF.Net is sponsored by: Speed Start Your Linux Apps Now.
Build and deploy apps & Web services for Linux with
a free DVD software kit from IBM. Click Now!
http://ads.osdn.com/?ad_id\x1356&alloc_id438&op=click
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

      parent reply	other threads:[~2004-02-29  0:43 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-02-16  5:28 Should driver changes cause hotplug event Jon Smirl
2004-02-16 21:04 ` Greg KH
2004-02-24 22:30 ` Oliver Neukum
2004-02-26 17:57 ` Greg KH
2004-02-29  0:43 ` Oliver Neukum [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=200402290143.12489.oliver@neukum.org \
    --to=oliver@neukum.org \
    --cc=linux-hotplug@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).