Re: X fails to start - Intel i810 (845) - etch + testing
Andrew: Just a quick feedback to say thanks a bunch for poring over
that lengthy post - really appreciate the details about what you tried
that worked and your insights into the verbose-yet-cryptic X error
log. Quick check verifies that there is just the builtin 82845G
chipset video - no extra video card - and certainly no second
monitor(!!). I will try to force the driver to use my attached
Hitachi CRT and post again.
Undocumented is an understatement - on the other hand, hope the X
gnomes out there are dreaming up code not docs at this point - aside
from maybe having a good spec....
Gordon
On 9/11/07, Andrew Sackville-West <andrew@farwestbilliards.com> wrote:
> On Mon, Sep 10, 2007 at 12:41:34PM -0500, Gordon Pedersen wrote:
> > On 9/9/07, Andrew Sackville-West <andrew@farwestbilliards.com> wrote:
>
> ...
> > >
> > > one note from the logs, it is selecting an initial resolution of
> > > 1920x1200. Does your monitor really support that? it looks like a
> > > hitachi from 1996, and I'd be surprised. I'm guessing that you're
> > > getting a bad mode setting that's not compatible with your monitor.
> >
> > See what you interpret from the new Xorg.0.log at bottom.
>
> comments below.
>
> >
> > If it is choosing an incorrect mode, how do I prevent that?
>
> its tricky with the intel driver as it really likes to do its own
> thing. You have to do some mojo to create a mode that appears in the
> list before any others. This is what I had to do for my wife's
> machine... and it was hard to suss out because of a lack of
> documentation. First, it appears to me that the intel driver uses the
> highest resolution it can come up with based on its probing of the
> devices. It will ignore anything else you tell it, mostly. And the
> listing of available resolutions comes out in the same order as you'd
> get using xrandr. So, for example, my wife's machine used to display
> the following (paraphrased) resolutions for her 1440x900 monitor:
>
> size Freq.
> 1440x900 59.8 60
> 1280x1024 blah blah blah
> 1024x768 blah blah blah
> ...
>
> and it would select 1440x900@59.8 which was an invalid spec for the
> monitor and would produce a stupidly off-center display.
>
> I had first tell the driver to use the monitor specifications in the
> xorg.conf, otherwise it will just ignore them. Do this in
>
> Section "Device"
> Identifier...
> Driver "intel"
> ...
> Option "Monitor-VGA" "monitor_id"
> EndSection
>
> this tells xorg to use the specifications for "monitor_id" instead of
> or alongside of the probed ones.
>
> Then I had to specify a mode in the monitor section
>
> Section "Monitor"
> Identifier "monitor_id"
> ...
> Modeline "1440x900_75" ...
> ...
> EndSection
>
> this mode appears before the others in the xrandr output and so
> becomes the one selected by xorg. Its confusing and undocumented and
> ymmv. And this doesn't deal with your real issue, I think, upon
> reading the logs below, only with the issue of resolution.
>
>
> ...
>
> snipped xorg.conf and a bunch of (EE)'s that I don't know diddly
> about.
>
> ...
>
> >
> > -------------
> > /var/log/Xorg.0.log
> > -------------
> >
> > X Window System Version 1.3.0
> > Release Date: 19 April 2007
> > X Protocol Version 11, Revision 0, Release 1.3
> > Build Operating System: Linux Debian (xorg-server 2:1.3.0.0.dfsg-12)
> > Current Operating System: Linux pan 2.6.18-5-686 #1 SMP Fri Jun 1 00:47:00 UTC 2
> > 007 i686
> > Build Date: 09 August 2007
> > Before reporting problems, check http://wiki.x.org
> > to make sure that you have the latest version.
> > Module Loader present
> > Markers: (--) probed, (**) from config file, (==) default setting,
> > (++) from command line, (!!) notice, (II) informational,
> > (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
> > (==) Log file: "/var/log/Xorg.0.log", Time: Mon Sep 10 11:23:21 2007
> > (==) Using config file: "/etc/X11/xorg.conf"
> > (==) ServerLayout "Default Layout"
> > (**) |-->Screen "Default Screen" (0)
> > (**) | |-->Monitor "Hitachi 203"
> > (**) | |-->Device "Intel 845 Brookdale"
> > (**) |-->Input Device "Generic Keyboard"
> > (**) |-->Input Device "Configured Mouse"
>
> okay, this all looks good.
> ...
>
> > (II) intel(0): Output VGA using monitor section Hitachi 203
>
> this part is real interesting because of this:
>
>
> > (II) intel(0): Output LVDS has no monitor section
>
> which shows up a bit later. You have two outputs on this video
> adapter? At least xorg thinks so.
>
>
> ... snipped a bunch of modeline stuff for the vga monitor. some of
> those are conflicting (for example, EDID says it goes upto 1600x1200
> but xorg shows some much higher res probed mode lines. that's all very
> interesting but not as interesting as this:
>
>
> > (II) intel(0): Output LVDS connected
>
> what? I thought that wasn't connected above...
>
>
> snip a bunch of modelines again...
>
> and again this shows up:
>
> > (II) intel(0): Output VGA connected
> > (II) intel(0): Output LVDS connected
> > (II) intel(0): Output VGA using initial mode 1920x1200
>
> probably thats a bad mode for your vga monitor.
>
> > (II) intel(0): Output LVDS using initial mode 1280x800
>
> is there another monitor attached?
>
>
> > (EE) intel(0): ivch: Unable to read register 0x01 from DVOI2C_B:04.
> > (EE) intel(0): Unable to write register 0x01 to DVOI2C_B:4.
> > (EE) intel(0): Unable to write register 0x40 to DVOI2C_B:4.
>
> so now we start getting all kinds of (EE)'s like these going
> forward. And here's where I think the real problem lies:
>
> > (II) intel(0): Output configuration:
> > (II) intel(0): Pipe A is on
> > (II) intel(0): Display plane A is now enabled and connected to pipe A.
> > (II) intel(0): Output VGA is connected to pipe none
> > (II) intel(0): Output LVDS is connected to pipe A
>
> see that? Its trying to use an output that so far as *I* know doesn't
> exist and isn't connected to anything. That's probably the source of a
> lot of the (EE)'s and ultimately causing the session to fail. That's
> my guess.
>
> I *think* that if you specify the
>
> Option "Monitor-VGA" "monitor_id"
>
> as I suggested above, that will force xorg to use that monitor as
> output and ignore this LVDS thing, but I can't be sure. This is over
> my head and hopefully some of the other xorg gurus can get involved.
>
> A
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.6 (GNU/Linux)
>
> iD8DBQFG5uJwaIeIEqwil4YRApktAKCNz1LgM2rqM2bC3yxvz9Q09hylygCgqaTN
> O9rUrUuYRNRlej7Nu/tHF7o=
> =Hlzy
> -----END PGP SIGNATURE-----
>
>
--
Gordon Pedersen gordon@visi.com
Reply to: