Re: X fails to start - Intel i810 (845) - etch + testing
- From: Andrew Sackville-West <andrew@xxxxxxxxxxxxxxxxxxxx>
- Date: Tue, 11 Sep 2007 13:24:48 -0700
On Tue, Sep 11, 2007 at 02:11:25PM -0500, Gordon Pedersen wrote:
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.
happy to help, though its effectiveness remains to be seen. I reviewed
the Xorg logs on my wife's machine and confirmed that it does *NOT*
make mention (other than one passing reference, I think) to LVDS. I'm
sure that's your problem. And its surely worth a bug-report if that
ends up being the problem.
A
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@xxxxxxxxxxxxxxxxxxxx> wrote:
On Mon, Sep 10, 2007 at 12:41:34PM -0500, Gordon Pedersen wrote:
On 9/9/07, Andrew Sackville-West <andrew@xxxxxxxxxxxxxxxxxxxx> 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@xxxx 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@xxxxxxxx
--
To UNSUBSCRIBE, email to debian-user-REQUEST@xxxxxxxxxxxxxxxx
with a subject of "unsubscribe". Trouble? Contact listmaster@xxxxxxxxxxxxxxxx
Attachment:
signature.asc
Description: Digital signature
- Follow-Ups:
- Re: X fails to start - Intel i810 (845) - etch + testing
- From: Gordon Pedersen
- Re: X fails to start - Intel i810 (845) - etch + testing
- References:
- X fails to start - Intel i810 (845) - etch + testing
- From: Gordon Pedersen
- Re: X fails to start - Intel i810 (845) - etch + testing
- From: Andrew Sackville-West
- Re: X fails to start - Intel i810 (845) - etch + testing
- From: Gordon Pedersen
- Re: X fails to start - Intel i810 (845) - etch + testing
- From: Andrew Sackville-West
- Re: X fails to start - Intel i810 (845) - etch + testing
- From: Gordon Pedersen
- X fails to start - Intel i810 (845) - etch + testing
- Prev by Date: Re: modem works in sarge but not etch or lenny
- Next by Date: SOLVED: another script query (perl?)
- Previous by thread: Re: X fails to start - Intel i810 (845) - etch + testing
- Next by thread: Re: X fails to start - Intel i810 (845) - etch + testing
- Index(es):
Relevant Pages
|
