Dual head operation on a PCI G450 fails. The first screen is not
initialized properly. The result is an incorrect video signal on the
first port. This likely has to do with mga driver not being able to read
the BIOS for the correct timing parameters. For a single head
configuration, the UseFBDev parameter can be used. This option causes
the mga driver to use the kernel fb device for mode switching and
mmapping of video memory and accelerator registers instead of directly
going to the hardware. For a dual head configuration this does not work
because the mga driver tries to resolve the fb /dev entry by only looking at
the PCI ID. As the PCI ID is the same for both screens, the mga driver
will use the same fb device for both screens which obviously results in
chaos :-). A working solution is to add an fbdev option which allows
specifying the fb /dev entry for each screen. It's probably adviseable
to always use the fb interface instead of trying to mess around with
mmapping the BIOS directly.

Hi,

About a year ago, you reported a bug to the Debian BTS regarding
dual-head failing on a MGA G450 board. Did you reproduce this problem
recently? With Xorg/Etch? with the latest driver in unstable? If not, I
will close this bug in the next weeks.

Thanks,
Brice

This problem is still unresolved. Did you actually look at it or do you
assume the problem magically disappeared ?

Cheers,

Peter.

p2@mind.be wrote:

I didn't look at it, I don't have any MGA board. The MGA driver you were
using at this time has changed a lot compared to the one in unstable
nowadays, so I hoped it would have been fixed in the meantime.

I don't see any report with your symptoms on bugzilla.freedesktop.org.
[1] reports problems on the second head, but does not seem to have any
problem with the first head. We might want to open a new bug report
there then.

Brice

[1] https://bugs.freedesktop.org/show_bug.cgi?id=2779

Hi,

I just uploaded xserver-xorg-video-mga 1.9.99.dfsg.1-1 to experimental.
It brings RandR 1.2 support, which improves mode detection and dynamic
configuration. Could you test it and see whether it helps for your
dual-head problem on G450. Note that you might need to update your
xorg.conf, see the pointers I inserted in
http://bgoglin.livejournal.com/12835.html

Thanks,
Brice

Did you try?

Brice

I just upgraded my (old) computer with its G450
to unstable and am unable to get a dual head
configuration to work.
The driver available under unstable is 1:1.4.9.dfsg-1
and in experimental 1:1.4.9.dfsg-2
(1.19.100.dfsg.1-1 for kfreebsd architectures).

I will try a more recent version if it could be uploaded
to experimental....

(I have RandR 1.2 dual head working on another computer
with xserver-xorg-video-radeon.)

Thank you
Alan

After upgrading to Squeeze from etch, I could no longer use the dual head display of my Matrox G450 PCI card.

I looked around on the web and a few attempts to fix this were made on the Ubuntu forums, but none of their recommendations or patches worked for me.

I would really like to have my two monitors working again. Also, I know several people who still use this card attached to two monitors and they will not upgrade to squeeze while this problem persists.

Will this bug be fixed soon?

Mateo.