ATI Fire GL V3100 and Dual display
W Bauske
wsb at paralleldata.com
Fri Jul 22 23:36:08 PDT 2005
Alex,
I checked the DVI-VGA adapter on one of my other systems and it no longer works. I
examined it and found the four pins around the flat blade had been pushed up into
the shell. Then I looked at the video cards I've been using for tests. The Asus AX300SE-X
ATI based card's DVI connector has no holes around the flat blade so when I pushed
the connector onto it I must have done the damage. Anyway, swapped it for a good
adapter and now both screens are fine. Also, comparing the Xorg.0.log file, I found
this difference:
OLD
(II) RADEON(0): Primary:
Monitor -- CRT
Connector -- DVI-I
DAC Type -- TVDAC/ExtDAC
TMDS Type -- Internal
DDC Type -- DVI_DDC
(II) RADEON(0): Secondary:
Monitor -- CRT
Connector -- VGA
DAC Type -- Primary
TMDS Type -- NONE
DDC Type -- VGA_DDC
NEW:
(II) RADEON(0): Primary:
Monitor -- CRT
Connector -- VGA
DAC Type -- TVDAC/ExtDAC
TMDS Type -- NONE
DDC Type -- NONE
(II) RADEON(0): Secondary:
Monitor -- CRT
Connector -- VGA
DAC Type -- Primary
TMDS Type -- NONE
DDC Type -- VGA_DDC
The primary now states it is a VGA connection.
Thanks for the hints on what to look for. Now I have to go back and try my Opteron and
AMD64 boards to see if they have the same problem or something else. Hopefully it's
just the DVI-VGA adapter being broken.
Wes
Alex Deucher wrote:
> On 7/22/05, Wes Bauske <wsb at paralleldata.com> wrote:
>
>>
>>Alex Deucher wrote:
>>
>>>>After looking at the log closely, I realized that the radeon driver is switching around
>>>>what is the primary vs secondary "port". The Sceptre monitor is really the primary
>>>>display that the BIOS initializes and the Linux console runs on. The EN7100e is on the
>>>>DVI->VGA adapter and is not initialized at boot. So, this appears to be the problem.
>>>>How to tell the driver to swap them around?
>>>>
>>>
>>>
>>>are you still having problems getting the screen to come up or are
>>>your heads just swapped? if it's just the heads that are swapped,
>>>switch the CRT2Position option to RightOf. If you are still having
>>>problems getting one of the displays to show an image add:
>>>Option "MonitorLayout" "CRT, CRT"
>>>to your device section. That will force both DACs on. You may also
>>>want to check your dvi-vga adapter. sometimes they can flake out.
>>>
>>>Alex
>>>
>>
>>The monitor on the DVI port is blank.
>>
>>I tried the MonitorLayout option and it didn't seem to make a difference other than
>>the radeon driver complained the 2nd value was incorrect.
>
>
> that's odd. it shouldn't. can you send me the message? what did you
> specify? the only valid options are:
>
> CRT - vga connectors, dvi->vga adapters, etc. anything driven by a DAC
> TMDS - digital DVI connections, digital only.
> LVDS - laptop flatpanels
>
>
>>I'll try the RightOf setting this afternoon to see if that helps.
>>
>>The DVI-VGA adpter was working last week on another system. I'll fire that system
>>up later today and make sure the DVI-VGA adapter is still working but I've seen
>>this behavior on multiple system combinations, with multiple DVI-VGA adapters.
>>
>>Just to clarify, you say this "forces" both DAC's on. Does that also mean the driver
>>will do a soft reset using the "int10" module on both ports? I still think the driver
>>is confused which port is primary vs secondary. Think I need to go look at the driver
>>source to see what ports are reset and under what conditions.
>
>
> it forces it on in that it turns the controller on and turns on the
> DAC to drive it regardless of whether there is a monitor detected on
> that port or not. There is no post done unless the card is not the
> primary, I think. Regardless, there is only one graphics chip so the
> second output is never posted, either the card is posted or it isn't.
> Also since everything is done via direct register writes, there's
> really no need for the bios once the card has been initially posted.
> If one output works, the other should too especially if you are
> forcing it on.
>
> Alex
>
>
>>Wes
>>
>>
>>>
>>>>Any help appreciated.
>>>>
>>>>Wes
>>>>
>>>
>>>
>>>
>
>
More information about the xorg
mailing list