Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 13 Feb 2012 02:17:46 -0800
From:      perryh@pluto.rain.com
To:        Alexander@Leidinger.net
Cc:        thierry@freebsd.org, stable@freebsd.org
Subject:   Re: Reducing the need to compile a custom kernel
Message-ID:  <4f38e34a.lZtNaNETBImp/XiD%perryh@pluto.rain.com>
In-Reply-To: <20120212120633.0000302d@unknown>
References:  <20120210145604.Horde.ewjpSpjmRSRPNSH0YRHxgAk@webmail.leidinger.net> <20120211124041.GF32360@graf.pompo.net> <20120211183308.00007579@unknown> <4f379cde.l6lDd9rduQzDU/xx%perryh@pluto.rain.com> <20120212120633.0000302d@unknown>

next in thread | previous in thread | raw e-mail | index | archive | help
Alexander Leidinger <Alexander@Leidinger.net> wrote:
> On Sun, 12 Feb 2012 03:05:02 -0800 perryh@pluto.rain.com wrote:
> > Alexander Leidinger <Alexander@Leidinger.net> wrote:
> > > On Sat, 11 Feb 2012 13:40:41 +0100 Thierry Thomas
> > > <thierry@FreeBSD.org> wrote:
> > > > is there another place to put options to atkbd and sc, like
> > > > these ones:
> > > >
> > > > options 	ATKBD_DFLT_KEYMAP	# specify the built-in
> > > > keymap makeoptions	ATKBD_DFLT_KEYMAP=fr.iso.acc
> > > > ...
> > >
> > > No, there is no other way to add the keymap to the kernel
> > > directly (if you want to have it working correctly in
> > > single-user mode) instead of loading it with rc.conf.
> >
> > Might it be feasible to make it into a sysctl, so it could be
> > set in loader.conf?
>
> There is already a way to configure this as soon as you have a
> working userland. What this setting is doing is to replace the
> compiled-in default keymap with a different one, so that you have
> the one which matches your keyboard even when you enter the very
> first keystrokes in single-user mode (root-pw, path to shell, ...).

My point is, if it were made into a sysctl, it could presumably
be set in loader.conf -- thereby providing the correct keymap for
those "very first keystrokes" without needing a custom kernel.

I know that's not how it works _now_, but is there some reason why
this approach is not feasible?  It seems like something that could
potentially go on the the list of projects to reduce the need for
custom kernels.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4f38e34a.lZtNaNETBImp/XiD%perryh>