Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 16 Apr 2017 09:07:47 +0200
From:      "O. Hartmann" <ohartmann@walstatt.org>
To:        Kevin Oberman <rkoberman@gmail.com>
Cc:        "O. Hartmann" <ohartmann@walstatt.org>, Conrad Meyer <cem@freebsd.org>, FreeBSD CURRENT <freebsd-current@freebsd.org>, svn-src-head@freebsd.org
Subject:   Re: svn commit: r316977 - head/sys/dev/syscons
Message-ID:  <20170416090747.5154d7a5@thor.intern.walstatt.dynvpn.de>
In-Reply-To: <CAN6yY1s-_GhZ3xv_h7JHqcbXcmvxK9DRUO0hvBn8dkz_J=-7Lw@mail.gmail.com>
References:  <201704152003.v3FK3o3w002356@repo.freebsd.org> <20170415222136.6c58a00d@thor.intern.walstatt.dynvpn.de> <CAG6CVpWy5Y13oMra90nMkStt%2B8w85%2Byx7Qto3RCsg5-6gAY9tw@mail.gmail.com> <20170416075542.022a6902@thor.intern.walstatt.dynvpn.de> <CAN6yY1s-_GhZ3xv_h7JHqcbXcmvxK9DRUO0hvBn8dkz_J=-7Lw@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
--Sig_/EjWsezWUoSFHAzLKiVkrOOq
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Am Sat, 15 Apr 2017 23:47:19 -0700
Kevin Oberman <rkoberman@gmail.com> schrieb:

> On Sat, Apr 15, 2017 at 10:55 PM, O. Hartmann <ohartmann@walstatt.org>
> wrote:
>=20
> > Am Sat, 15 Apr 2017 18:00:18 -0700
> > Conrad Meyer <cem@freebsd.org> schrieb:
> > =20
> > > On Sat, Apr 15, 2017 at 1:21 PM, O. Hartmann <ohartmann@walstatt.org>=
 =20
> > wrote: =20
> > > > Am Sat, 15 Apr 2017 20:03:50 +0000 (UTC)
> > > > Bruce Evans <bde@FreeBSD.org> schrieb:
> > > > =20
> > > >> Author: bde
> > > >> Date: Sat Apr 15 20:03:50 2017
> > > >> New Revision: 316977
> > > >> URL: https://svnweb.freebsd.org/changeset/base/316977 =20
> > > >
> > > > There is a lot of development going on theses days for syscons. Wha=
t's =20
> > about vt()? =20
> > > > vt() is considered broken for x11/nvidia-driver and vt() is conside=
red =20
> > a requirement =20
> > > > when UEFI is boot scheme, isn't it?
> > > >
> > > > I'm just curious. =20
> > >
> > > Hi O.,
> > >
> > > Bruce uses syscons and cares enough to improve it.  He likely does not
> > > care about UEFI and definitely does not care about vt.  I don't think
> > > there's anything wrong with that.  We can't force volunteers to work
> > > on things they are not interested in.
> > >
> > > Best,
> > > Conrad =20
> >
> > Hello Conrad, happy easter.
> >
> > There is and was never intention to apply "force", it is a question as =
I'm
> > curious about
> > what's going on with vt() - no personally bound to B. Evans or anybody
> > else - I just took
> > the chance to comment/ask on that subject when I saw postings.
> >
> > Maybe not the right place to spread some thinkings.
> >
> > Regards,
> >
> > Oliver Hartmann
> > =20
>=20
> vt(4) is not a pleasant thing to look at. I am not implying that it is bad
> code or badly done. I am just saying that it is pretty gnarly and is not
> the sort of thing most enjoy dealing with. I got the distinct feeling that
> ray@ found the job much uglier than he anticipated  when he took the
> Foundation commission to write it. Since then it has been widely disparag=
ed
> for the things that it does not do, but I am not aware that anyone has
> gotten further than looking at what is needed and then running far
> away.Some day someone (or some company) will get sufficiently inspired to
> either re-write if or add the missing features. I have no idea when that
> might happen, though.
> --
> Kevin Oberman, Part time kid herder and retired Network Engineer
> E-mail: rkoberman@gmail.com
> PGP Fingerprint: D03FB98AFA78E3B78C1694B318AB39EF1B055683

Hello Kevin.

So, what you say is: vt(4) is a kind of unwanted child, not to say: "part t=
ime orphaned"?
As far as I know, vt(4) is a requirement for UEFI/EFI booting. We do so wer=
e we can and
so we rely on vt(4). Please correct me if I'm wrong.

We also use vt(4) with pleasure for most of our servers's consoles - due to=
 the great
improvements due to the higher resolution which makes it very easy to edit =
files/issue
commands/get results on the console. Especially in an experimental environm=
ent for
science, where a sophisticated infrastructure isn't available like graphica=
l terminals
and so on.

Personally, I use FreeBSD even as workstation, apart from others, equipted =
with the
only-left working GPU vendor, nVidia, for high-performance graphics (Intel =
iGPU is fine
for laptops, but ... ), and here we run into a serious problem: on all nVid=
ia driven
systems, with the newer drivers the console is garbage (on non-vt) when usi=
ng sc(4) on
legacy booting boxes, with UEFI, vt(4) and nvidia produce a blank console.=
=20

I've already
written a message to nVidia, but with no response until now for more than h=
alf a year and
the issue is present more than a year(!!!). As long as the nvidia kernel mo=
dule is loaded,
the console is unusable and the nvidia module somehow blocks rebooting some=
 of our
Fujitsu Celsius workstations. On a test box, I run 381.09 - the same proble=
ms.

By definition, vt(4) and nVidia driver has to be considered broken and this=
 should be
reflected then in some message from the driver - but this isn't a subject f=
or this list.

Thanks for your patience to read and answer.

Happy Easter,

Oliver Hartmann

--=20
O. Hartmann

Ich widerspreche der Nutzung oder =C3=9Cbermittlung meiner Daten f=C3=BCr
Werbezwecke oder f=C3=BCr die Markt- oder Meinungsforschung (=C2=A7 28 Abs.=
 4 BDSG).

--Sig_/EjWsezWUoSFHAzLKiVkrOOq
Content-Type: application/pgp-signature
Content-Description: OpenPGP digital signature

-----BEGIN PGP SIGNATURE-----

iLUEARMKAB0WIQQZVZMzAtwC2T/86TrS528fyFhYlAUCWPMYQwAKCRDS528fyFhY
lLGkAfwJS9bfXMUsXbph0PbOcCOdEcMOuPQdtrZ0AycE34GXgHywycHmbIjD22G4
9E8dkVsOZ3znS5ENnFu4MA5NKO2HAf449rp2rVf3ANte2cTCbSo25+luKbQjFsDT
kQdp5MdozsrxJEUDS8HdEWSS9TtHGh7oZFyBhILTK7QNjHFpByQ1
=GcB2
-----END PGP SIGNATURE-----

--Sig_/EjWsezWUoSFHAzLKiVkrOOq--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20170416090747.5154d7a5>