Subject: Re: CVS commit: src
To: Iain Hibbert <>
From: Quentin Garnier <>
List: source-changes
Date: 08/20/2006 22:02:02
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Sun, Aug 20, 2006 at 08:13:38PM +0100, Iain Hibbert wrote:
> On Sun, 20 Aug 2006, Quentin Garnier wrote:
> > You know, you just made a very good point why we'd want to get rid of
> > hard-coded limits.
> I see what you say, but its not exactly the 640k limit - this is a
> configuration option that is easily changed by advanced users..  The
> GENERIC config cannot be everything to everybody

Yeah but the example you gave needs 5 devices :)

> > Not sure I follow your point.  Someone or something has to keep the
> > information anyway.  Reading the code, it's completely beyond me why
> > you'd need more than one btdev device anyway.  All children should
> > attach to one you create in btdevattach(), and then you can use the name
> > of the attached device as the identifier.  You just have to keep a list
> > of the children.
> Hmm yes, dv_xname..  Maybe I built a house of cards then. The reason for
> multiple access points and the minor number index is so that we can find
> configured devices in order to remove them later if need be. Using the
> name would cover that and the house of cards could blow away.
> I will have to think about this some though, because it would throw up
> other issues (currently, the stored configuration is also based around the
> btdevN and I would have to redesign it)

We have a very small window of opportunity to make things right before
4.0_BETA is too advanced so we could forget about binary compatibility.
Very soon we'd have to have compatibility support...

> PS The commit was actually to fix a mistake I made - for some reason I
> thought that 'pseudo-device btdev', would result in btdevattach(0) so I
> used that to signal the default value. Hence, the GENERIC kernel only
> ended up with 1 btdev which was definitely not enough.

Oh I wasn't commenting on the commit itself;  you're right on that.

Quentin Garnier - -
"When I find the controls, I'll go where I like, I'll know where I want
to be, but maybe for now I'll stay right here on a silent sea."
KT Tunstall, Silent Sea, Eye to the Telescope, 2004.

Content-Type: application/pgp-signature
Content-Disposition: inline

Version: GnuPG v1.4.3 (NetBSD)