Subject: Re: Various NetBSD kernel questions to help with port of FreeBSD "zaptel" drivers.
To: None <firstname.lastname@example.org>
From: Christos Zoulas <email@example.com>
Date: 11/08/2004 23:42:59
In article <B3C8E88A-31DA-11D9-ACC4-000A957650EC@shagadelic.org>,
Jason Thorpe <firstname.lastname@example.org> wrote:
>On Nov 8, 2004, at 3:00 PM, Hubert Feyrer wrote:
>> On Mon, 8 Nov 2004, Jason Thorpe wrote:
>>> That's Special Magick that other parts of the kernel interpret as
>>> "create a clone".
>> Why no ENEEDCLONE?
>I dunno... ask Christos .. I think he's the one who implemented the
>cloning stuff originally (long, long time ago).
>IMO, it's lame to add a special error code for this case anyway.
Actually it is not me. I wanted to get rid of the ENXIO hack...
The hack came actually from 4.4BSD [look at the history of vfs_syscalls.c],
and the credit goes to mycroft for tidying in the code. We should
have gotten rid of the ENODEV case since this is handled the same
way as ENXIO now. Both are used to pass fd's sideways through vn_open():
ENODEV /* XXX from fdopen or fdesc_open? */
ENXIO /* XXX from portal_open? */
as the comment showed at one time; now the comment is not there anymore.