Ben Taylor writes:
I'm running Ubuntu 7.10 patched on a AMD2600+ system with latest radeonhd git.
This card has a been a real piece of work.
When the card is by itself, radeonhd driver only finds the VGA port, and not the DVI port. This is consistent with the rhd_conntest and it's complete non-recognition of the 02:00.1 device. That usualy results in the following message:
rhd_conntest: v1.2.0, git branch master, commit 9d131f90 + changes Unknown device: 0x1002:0x7166 (02:00.01).
Ben, This is the scondary device - which doesn't idenntify a graphics device on its own. please do as I suggested several times and use the PCI bus ID of the primary card: 2:0.1. Please also don't use this ID in any device section. Although your cards have more than one connector all the connectors are handled by the same PCI device.
If I boot with just the DVI connected, I see text until it tries to graphics, that fails and it results in a non-working X session. Booting with it connected to the VGA port, it works fine. (though there is some noticiable flickering on my Acer 22" FP).
When I've had the board in system, along with my 2600HD AGP (which radeonhd sorta recognizes) the DVI port is recognized in the Xorg.0.log file.
This is making me crazy. Is it possible that the PCI to PCI-Express bridge is causing this problem?
Please: 1. Only use one card in the system at a time (at least for nw). 2. Remove any device section that contains the PCI ID of the secondary card (bus 2:0.1). Only keep the device section referring to bus ID (2:0.0). 3. Obtain the log file of the Xserver in this configuration. 4. Run 'rhd_conntest 2:0.0' for a. a monitor connected to the VGA port. b. a monitor connected to the DVI port. (don't connect them both simulaniously as then we cannot destinguish the results any more). Thanks a lot! Cheers, Egbert. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Mon, Apr 14, 2008 at 8:12 AM, Egbert Eich
Ben Taylor writes:
I'm running Ubuntu 7.10 patched on a AMD2600+ system with latest radeonhd git.
This card has a been a real piece of work.
When the card is by itself, radeonhd driver only finds the VGA port, and not the DVI port. This is consistent with the rhd_conntest and it's complete non-recognition of the 02:00.1 device. That usualy results in the following message:
rhd_conntest: v1.2.0, git branch master, commit 9d131f90 + changes Unknown device: 0x1002:0x7166 (02:00.01).
Ben,
This is the scondary device - which doesn't idenntify a graphics device on its own. please do as I suggested several times and use the PCI bus ID of the primary card: 2:0.1.
You mean 02:00.0 ? I know I've been frustrated by this process.
Please also don't use this ID in any device section. Although your cards have more than one connector all the connectors are handled by the same PCI device.
Now I'm just using a stock generated xorg.conf file, and it only shows one device. If I need to hook up two monitors simulataneously, I can probably make that happen.
If I boot with just the DVI connected, I see text until it tries to graphics, that fails and it results in a non-working X session. Booting with it connected to the VGA port, it works fine. (though there is some noticiable flickering on my Acer 22" FP).
When I've had the board in system, along with my 2600HD AGP (which radeonhd sorta recognizes) the DVI port is recognized in the Xorg.0.log file.
This is making me crazy. Is it possible that the PCI to PCI-Express bridge is causing this problem?
Please: 1. Only use one card in the system at a time (at least for nw).
ok. I was just trying to figure out why it saw it when the R2600 pro was in there.
2. Remove any device section that contains the PCI ID of the secondary card (bus 2:0.1). Only keep the device section referring to bus ID (2:0.0).
Got it. This version is only the x1550 autogenerated xorg.conf file.
3. Obtain the log file of the Xserver in this configuration.
Got it. Xorg.0.log I started manually with -logverbose 7. Xorg.0.log.old was the previously autostarted gdm session. I will note that when I started X -logverbose 7, the screen is not particularly "clean". There's some distortion. I can probably get a picture if necessary now that I can recreate it.
4. Run 'rhd_conntest 2:0.0' for a. a monitor connected to the VGA port. b. a monitor connected to the DVI port. (don't connect them both simulaniously as then we cannot destinguish the results any more).
Ok. This was a bit confusing for me. I got it now. Just so you know, If I boot connected to the VGA port, I get an X11 screen. If I boot off the DVI it fails to start X. So I'm switching cables in text mode (gdm off to do these tests). Ben
participants (2)
-
Ben Taylor -
Egbert Eich