Xorg oddity.

[Anton Aylward]

I have a working and acceptable full-res X running, including all the
eye candy like rotating box,, fading windows etc etc etc.

but, but but ...

The logs tell me things like

<quote>
[145070.025] (II) intel(0): Output VGA1 has no monitor section
</quote>

Well of course it doesn't!
My monitor is a Flatron and I have the definition:

<quote>

Section "Monitor"
Identifier "LG1"
Vendorname "LG"
Modelname "LG 1953 S"
</quote>

and of course that ties in with "Screen' definition etc.

Oh, and it complains

<quote>
[145070.025] (II) intel(0): Output VIRTUAL1 has no monitor section
</quote>

Where did that come from?

Are these hard wired into the intel driver or something?
I make no mention of VGA or VIRTUAL in my xorg.conf.d

The EDID seems to be working OK.

[Felix Miata]

Post by Anton Aylward
I have a working and acceptable full-res X running, including all the
eye candy like rotating box,, fading windows etc etc etc.
but, but but ...
The logs tell me things like
<quote>
[145070.025] (II) intel(0): Output VGA1 has no monitor section
</quote>
Well of course it doesn't!
<quote>
Section "Monitor"
Identifier "LG1"
Vendorname "LG"
Modelname "LG 1953 S"
</quote>
and of course that ties in with "Screen' definition etc.
Oh, and it complains
<quote>
[145070.025] (II) intel(0): Output VIRTUAL1 has no monitor section
</quote>
Where did that come from?

Quite some months back I went looking for docs on why VIRTUAL1 exists and how
to make use of it, but never found anything useful.

Post by Anton Aylward
Are these hard wired into the intel driver or something?
I make no mention of VGA or VIRTUAL in my xorg.conf.d

I think outputs showing up in Xorg.0.log actually depend on BIOS
implementation and settings. You can see in
http://fm.no-ip.com/Tmp/Linux/Xorg/xorg.0.log-big41-os132-intel1280x1024 I
have even more "outputs" than you mentioned, but what I think is going on is
that we're seeing all the outputs that the gfxchip *could* support if each of
the possibles were physically implemented on the motherboard, plus VIRTUAL1.
To generate that log just a bit ago I originally had a Radeon installed in
the PCIe slot and used the BIOS to force use of the onboard Intel. In that
configuration, the only output in Xorg.0.log
http://fm.no-ip.com/Tmp/Linux/Xorg/xorg.0.log-big41-os132rc-intel1280x1024
was VGA1. To get the log containing multiple outputs I had to restore the
BIOS to default setting and remove the Radeon.

Post by Anton Aylward
The EDID seems to be working OK.

If you're vexed by X not playing nice (on your GX755?), and simply want your
display's native resolution (1280x1024?), the following should be all you
need in your xorg.conf:

Section "Device"
Identifier "Default Device"
Driver "intel"
Option "AccelMethod" "UXA"
EndSection

Section "Monitor"
Identifier "Default Monitor"
EndSection

Section "Screen"
Identifier "Default Screen"
Device "Default Device"
Monitor "Default Monitor"
EndSection

Identifiers can be any foo you want, but they must be consistent within the
config file.

If you've not already seen them, you might want to look at
http://bugzilla.opensuse.org/show_bug.cgi?id=901506 and the upstream bug.

Common displays shouldn't need any xorg.conf* at all to work correctly from
EDID, but until the fixed Intel driver shows up, I've been doing my manual
Intel config via xrandr in a startup script.

If that minimalist xorg.conf fails to work for you, and you're using KDE, you
might be fighting against KScreen. Try in kdedrc adding:

[Module-kscreen]
autoload=false

Kscreen at least used to when new ignore manual X configs and do whatever it
pleased.
--
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata *** http://fm.no-ip.com/

[Anton Aylward]

This post might be inappropriate. Click to display it.

[Felix Miata]

Post by Anton Aylward
I'll try taking out all the extra modelines I have put in that I worked
out carefully with xrandr and cvt.

I don't get why it seems to be as common as it is that people manually
generate modelines to put into xorg.conf. I can't remember ever solving a
problem doing so in the eons since Xorg replaced XFree86 (and most CRTs were
replaced by flat panels). Xorg does what cvt does automatically given either
valid EDID and/or HozizSync and VertRefresh.
--
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata *** http://fm.no-ip.com/

[jdd]

This post might be inappropriate. Click to display it.

[Anton Aylward]

Post by Felix Miata

Post by Anton Aylward
I'll try taking out all the extra modelines I have put in that I worked
out carefully with xrandr and cvt.

I don't get why it seems to be as common as it is that people manually
generate modelines to put into xorg.conf. I can't remember ever solving a
problem doing so in the eons since Xorg replaced XFree86 (and most CRTs were
replaced by flat panels). Xorg does what cvt does automatically given either
valid EDID and/or HozizSync and VertRefresh.

There are a class of situation where EDID does not work.
The one I have is that I'm using a KVM switch and it does not do the
EDID stuff properly as you switch back and forth. It is even worse if
one of the machines on the switch is Windows!

[Felix Miata]

Post by Anton Aylward

I don't get ... or HozizSync and VertRefresh.

There are a class of situation where EDID does not work.
The one I have is that I'm using a KVM switch and it does not do the
EDID stuff properly as you switch back and forth. It is even worse if
one of the machines on the switch is Windows!

--
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata *** http://fm.no-ip.com/