#1293 closed bug (invalid)
External LCD on a second DVI
Reported by: | modeenf | Owned by: | euan |
---|---|---|---|
Priority: | normal | Milestone: | Unscheduled |
Component: | Drivers/Graphics/radeon | Version: | R1/pre-alpha1 |
Keywords: | Cc: | ||
Blocked By: | Blocking: | ||
Platform: | x86 |
Description
I have a Radeon x300 in a Dell Latitude D610 and using an external LCD on DVI throug a D/Port. There are no output at all on .
If I remember right this are known problem but I coudn't find a ticket on this.
I'm using hrev21591 and have Zeta 1.5. The LCD are a Dell 1907FPV
Attachments (2)
Change History (14)
comment:1 by , 17 years ago
Owner: | changed from | to
---|
comment:2 by , 17 years ago
Status: | new → assigned |
---|
comment:3 by , 17 years ago
comment:4 by , 17 years ago
I now have a working prototype of the external DVI fix. There are a number of things that need resolved however:
The Radeon series cannot scale the secondary DVI. So it can only display at the native resolution. If the monitor supports internal scaling of DVI input, we would need some way of detecting this from the EDID data (needs investigated).
If the LVDS or Primary DVI is at it's native resolution we can re-assign the scaler to the secondary channel. messy, but in theory possible.
If the LVDS / Primary DVI has the same native resolution as the secondary DVI we can clone the primary output to secondary, and use the same scaler output to drive both channels.
It can't do centered timings (where it is at native res, but shows smaller resolutions as a centered box). This would be quite useful
I'll maybe be able to upload a patch for testing only on Thursday for those wanting to try it.
I noticed a couple of other issues with the driver. Selecting a higher resolution than the monitor results in the framebuffer wrapping on screen. Should Haiku not change to a virtual desktop? I can't test in in BeOS right now, so don't know if this is a Radeon bug, or appserver. Also selecting a combined desktop resolution lower than the native resolution displays on screen at native resoltution, but with the unused area of the screen filled with rubbish. Pretty, but useless. probably a radeon bug.
comment:5 by , 17 years ago
Here's a working beta patch.
Tested only on an X600 Dual DVI. Needs particular testing with laptop + DVI combination.
- Clone mode with resolutions below secondary DVI native resolution Second screen with not stretch to screen. Will display in top left corner, remaining unused screen space will be filled with garbage. Plan to center and blank surrounding area.
- Clone mode with resolutions above secondary DVI resolution screen will display decent looking screen except the extra area wraps the left hand side. it shouldn't allow this. Should probably change to scroll mode.
- Combine mode with resolution above or below secondary DVI native resolution Shouldn't allow it, will return "general system error" message box, or will change mode, but secondary screen shows same image as primary monitor.
patch is attached.
comment:6 by , 17 years ago
Can't test it right now, Will have to wait 2 weeks. But after that I can test it again :)
comment:8 by , 17 years ago
No doesn't work here on a Radeon x300
What happends are that the DVI port (LCD) are black and when I set's the desktop to use both screens the desktop are strech as it should but the LCD attatch to the DVI port are still black and inactive. Before it only used one screen and I hade to move the destop around to see different part of the desktop.
Do you need som mor information?
by , 17 years ago
Attachment: | syslog.zip added |
---|
the syslog. If evey bootup shows on a syslog then te second boot up are with 2 sceens
comment:10 by , 4 years ago
Milestone: | R1 → R1.1 |
---|
comment:11 by , 3 years ago
Resolution: | → invalid |
---|---|
Status: | in-progress → closed |
Closing these tickets due to the passing of time. Please consider opening new tickets with current syslogs if still using these graphics accelerators and seeing these issues.
comment:12 by , 3 years ago
Milestone: | R1.1 → Unscheduled |
---|
Move tickets closed as invalid/obsolete out of R1.1 milestone.
A log file would be useful to check the monitor routing (attach here). I have a patch in progress, is quite complex though, given the mapping and routing of the Radeon internals.