Old subject:
"Detection Problems with Latest Git (2008-04-11)"

The LCD doesn't give real DDC info on the LCD:
(II) RADEONHD(0): Connector[1] {RHD_CONNECTOR_PANEL, "LVDS LCD1", DDC_NONE, RHD_HPD_NONE, { RHD_OUTPUT_LVTMA, RHD_OUTPUT_NONE } }

The VESA BIOS often doesn't obtain DDC data thru I2C. Instead it's faked inside the BIOS. AtomBIOS may also contain a 'faked' DDC block and we try to retrieve it.
In your case this fails:
(II) RADEONHD(0): Query for AtomBIOS Get Panel EDID: failed
If you can provide a copy of your BIOS (which you can dump with 'rhd_conntest -d'

Looking at your connector table it shows 3 digital connectors:
PANEL, DVI and HDMI. This shouldn't be possible with this hardware.
The latter one doesn't seem to have any outputs. This could be the reason why it doesn't show in xrandr.

The black screen problem in 3 is under investigation.
> Also, if I unplug the VGA connection without first disabling it with xrandr,
> xrandr starts displaying bogus DVI-D_1 information (even if nothing has ever
> been plugged into the HDMI connection) until I re-plugin the VGA connection or
> run "xrandr --output VGA_1 --off" (which still goes through okay even after 
> the VGA connection unplugged):

Could you specify what is printed before you unplugged anything?

- If xrandr is ran before plugging in the VGA connector, the display always
shows a miore pattern in the colours.  This persists until Xorg is restarted
and the display is plugged in before running xrandr.

I don't really understand this:
Where do you see this displayed? On the panel or on the VGA?

> Looking at your connector table it shows 3 digital connectors:
> PANEL, DVI and HDMI. This shouldn't be possible with this hardware.
> The latter one doesn't seem to have any outputs. This could be the reason why
> it doesn't show in xrandr.

The laptop comes with an HDMI connection for sure (that is what I'm using with the HDMI to DVI cable), and, if I understand correctly, the docking bay also gives you a DVI output (perhaps that is just the HDMI output again though).

I'll attach to BIOS dump.

> > Also, if I unplug the VGA connection without first disabling it with xrandr,
> > xrandr starts displaying bogus DVI-D_1 information (even if nothing has ever
> > been plugged into the HDMI connection) until I re-plugin the VGA connection or
> > run "xrandr --output VGA_1 --off" (which still goes through okay even after 
> > the VGA connection unplugged):
> 
> Could you specify what is printed before you unplugged anything?

On plugin (this is a different VGA display than the original one):

Screen 0: minimum 320 x 200, current 1680 x 1050, maximum 1680 x 1680
VGA_1 connected
   1280x1024      60.0 +   75.0     59.9
   1152x864       75.0
   1024x768       84.9     75.1     70.1     60.0
   832x624        74.6
   800x600        84.7     72.2     75.0     60.3
   640x480        84.4     75.0     72.8
PANEL connected 1680x1050+0+0 444mm x 277mm
   1680x1050      60.1*+
DVI-D_1 disconnected

After 'xrandr --output VGA_1 --auto':

Screen 0: minimum 320 x 200, current 1680 x 1050, maximum 1680 x 1680
VGA_1 connected 1280x1024+0+0 360mm x 290mm
   1280x1024      60.0*+   75.0     59.9
   1152x864       75.0
   1024x768       84.9     75.1     70.1     60.0
   832x624        74.6
   800x600        84.7     72.2     75.0     60.3
   640x480        84.4     75.0     72.8
PANEL connected 1680x1050+0+0 444mm x 277mm
   1680x1050      60.1*+
DVI-D_1 disconnected

After unplugging (without 'xrandr --output VGA_1 --off'):

Screen 0: minimum 320 x 200, current 1680 x 1050, maximum 1680 x 1680
VGA_1 disconnected 1280x1024+0+0 360mm x 290mm
PANEL connected 1680x1050+0+0 444mm x 277mm
   1680x1050      60.1*+
DVI-D_1 disconnected
  1280x1024 (0x76)  108.0MHz
        h: width  1280 start 1328 end 1440 total 1688 skew    0 clock   64.0KHz
        v: height 1024 start 1025 end 1028 total 1066           clock   60.0Hz

After 'xrandr --output VGA_1 --off':

Screen 0: minimum 320 x 200, current 1680 x 1050, maximum 1680 x 1680
VGA_1 disconnected
PANEL connected 1680x1050+0+0 444mm x 277mm
   1680x1050      60.1*+
DVI-D_1 disconnected

> - If xrandr is ran before plugging in the VGA connector, the display always
> shows a miore pattern in the colours.  This persists until Xorg is restarted
> and the display is plugged in before running xrandr.
> 
> I don't really understand this:
> Where do you see this displayed? On the panel or on the VGA?

The VGA (I've now seen this with three different displays) looks like there is some sort of random colour map (i.e., white become orange, black becomes blue, gradients look like rainbows, etc.).  The colours on the panel remains fine.

I can take a picture when I have my digital camera with me if you would like.

I should also note that I've now seen this even when I haven't run xrandr before plugging the VGA in.  In fact, I believe the only time I haven't seen it for sure now is when I plug the VGA in before starting Xorg.

Thanks for looking into this!  : )

Created attachment 16076 [details]
ROM dump via 'rhd_conntest -d'

Tyson,

At least one of your issues should be fixed now, the broken VT return. Please try our latest git version and verify this, so that the remaining issues can be tackled

I just tried the latest git (2008-05-18 up to and including commit starting with 2ea7cdd7), and the following issues have all now been resolved:

1- Ones that were present only if an external monitor was plugged into the VGA port before and when XOrg was starting up:

a) I used to loose my ability to switch to a virtual terminal (ctrl+alt+F1 would just give a blank screen).

b) The latop display used to remain dark on any subsequent restarts of Xorg (without rebooting the latop) unless the external monitor was re-plugged into the VGA port before the restart.


2- The colours on a monitor plugged into the VGA port used to sometimes look like they had all been randomly mapped around unless the display was plugged into the VGA port before and when XOrg was started (note -- this didn't always happen, so it still might not be fixed, but I believe it is as I haven't seen this once since upgrading).


Still remaining is:

1- If I unplug an external monitor from the VGA port without first disabling it with xrandr, it then begans to report bogus DVI information:

Screen 0: minimum 320 x 200, current 1920 x 1200, maximum 1920 x 1920
VGA_1 disconnected 1920x1200+0+0 519mm x 324mm
PANEL connected 1680x1050+0+0 444mm x 277mm
   1680x1050      60.1*+
DVI-D_1 disconnected
  1920x1200 (0x46)  154.0MHz
        h: width  1920 start 1968 end 2000 total 2080 skew    0 clock   74.0KHz
        v: height 1200 start 1203 end 1209 total 1235           clock   60.0Hz

I can then get rid of by running 'xrandr --output VGA_1 --off':

Screen 0: minimum 320 x 200, current 1680 x 1050, maximum 1920 x 1920
VGA_1 disconnected
PANEL connected 1680x1050+0+0 444mm x 277mm
   1680x1050      60.1*+
DVI-D_1 disconnected


2- The VESA driver reports EDID information for the laptop display, however, the radeonhd driver is unable to obtain it.  I've attached a new xorg log file showing this again.


3- If I plug an external monitor into the HDMI port (through a HDMI -> DVI cable), either before or after starting Xorg, nothing happens on the computer end (xrandr report nothing and nothing shows up in the Xorg logs).  I've repeated the rhd_conntest thing:

a) With no external monitors plugged in to any of the ports:

rhd_conntest: v1.2.1, git branch master, commit 2ea7cdd7 + changes
Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
  Load Detection: RHD_OUTPUT_NONE
  HotPlug: RHD_HPD_NONE
  DDC: RHD_DDC_NONE

b) With only an external monitor plugged into the VGA port:

rhd_conntest: v1.2.1, git branch master, commit 2ea7cdd7 + changes
Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
  Load Detection: RHD_OUTPUT_DACA
  HotPlug: RHD_HPD_NONE
  DDC: RHD_DDC_0

c) With only an external monitor plugged into the HDMI port (HDMI -> DVI cable):

rhd_conntest: v1.2.1, git branch master, commit 2ea7cdd7 + changes
Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
  Load Detection: RHD_OUTPUT_NONE
  HotPlug: RHD_HPD_1
  DDC: RHD_DDC_1


Thanks again to you guys for all their hard work on this!

Cheers!  -Tyson

Created attachment 16613 [details]
XOrg log for run with driver compiled on 2008-05-18 (up to and including commit starting with 2ea7cdd7)

Thanks for testing!

(In reply to comment #5)
> Still remaining is:
> 
> 1- If I unplug an external monitor from the VGA port without first disabling it
> with xrandr, it then begans to report bogus DVI information:

This sounds extremely weird.
OTOH, this *could* be a side effect of a bug Luc fixed yesterday in the RandR code. Could you please try again with git head?

> 2- The VESA driver reports EDID information for the laptop display, however,
> the radeonhd driver is unable to obtain it.  I've attached a new xorg log file
> showing this again.

We have this with multiple laptops. Egbert, would it be an option to the vesa EDID emulation if we don't get any information from AtomBIOS?

> 3- If I plug an external monitor into the HDMI port (through a HDMI -> DVI
> cable), either before or after starting Xorg, nothing happens on the computer
> end (xrandr report nothing and nothing shows up in the Xorg logs).  I've
> repeated the rhd_conntest thing:

Atombios provides us information about a HDMI connector, but nothing about the used outputs:
(II) RADEONHD(0): Connector[3] {RHD_CONNECTOR_DVI_SINGLE, "HDMI_TYPE_A DFP1", RHD_DDC_1, RHD_HPD_1, { RHD_OUTPUT_NONE, RHD_OUTPUT_NONE } }

Egbert, does this sound like parsing error, or like broken connector table?

> c) With only an external monitor plugged into the HDMI port (HDMI -> DVI
> cable):
> 
> rhd_conntest: v1.2.1, git branch master, commit 2ea7cdd7 + changes
> Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
>   Load Detection: RHD_OUTPUT_NONE
>   HotPlug: RHD_HPD_1
>   DDC: RHD_DDC_1

So the DDC and hotplug information in the table is ok... Sounds like parsing error.

(In reply to comment #7)
> Thanks for testing!
> 
> > 2- The VESA driver reports EDID information for the laptop display, however,
> > the radeonhd driver is unable to obtain it.  I've attached a new xorg log file
> > showing this again.
> 
> We have this with multiple laptops. Egbert, would it be an option to the vesa
> EDID emulation if we don't get any information from AtomBIOS?
> 
How do we know which EDID the VESA BIOS will report?
Is it always the EDID of the panel?
I would like to avoid going down this route.
It's possible that the routine collecting the EDID data out of AtomBIOS still has problems. All the laptops I've access to (which would be my HP) obtain the panel EDID directly from the panel using DDC. On those there is no EDID block in AtomBIOS. I have no access to our other laptops. Let's Luc look into this.

In any case we should have a mode line for the panel. Therefore this information is only needed to calculate the DPI. I'm not sure if we want to call the VESA BIOS when we already call AtomBIOS just to obtain EDID data to calculate the screen resolution.

> > 3- If I plug an external monitor into the HDMI port (through a HDMI -> DVI
> > cable), either before or after starting Xorg, nothing happens on the computer
> > end (xrandr report nothing and nothing shows up in the Xorg logs).  I've
> > repeated the rhd_conntest thing:
> 
> Atombios provides us information about a HDMI connector, but nothing about the
> used outputs:
> (II) RADEONHD(0): Connector[3] {RHD_CONNECTOR_DVI_SINGLE, "HDMI_TYPE_A DFP1",
> RHD_DDC_1, RHD_HPD_1, { RHD_OUTPUT_NONE, RHD_OUTPUT_NONE } }
> 
> Egbert, does this sound like parsing error, or like broken connector table?
> 

{RHD_CONNECTOR_DVI_SINGLE, "HDMI_TYPE_A DFP1", RHD_DDC_1, RHD_HPD_1, { RHD_OUTPUT_NONE, RHD_OUTPUT_NONE } }

This information seems to be correct. We've got 3 digital outputs: PANEL,  DVI
and HDMI. These don't match nicely on the 2 digital blocks.
Now the TMDSA and the LVTMA are used for dual link. So it's not possible to split any of those.
Therefore there are 2 options left:
1. The BIOS has an option to select between the DVI and the HDMI port.
2. There is a DVO chip on the system.
I would be able to tell if I had the BIOS. Tyson, please run rhd_conntest -d to dump the BIOS and send it to me in an email (eich_at_suse_dot_de).

> >   HotPlug: RHD_HPD_1
> >   DDC: RHD_DDC_1
> 
> So the DDC and hotplug information in the table is ok... Sounds like parsing
> error.

No, as this exactly what's in the connector table. See above.

(In reply to comment #8)
> > We have this with multiple laptops. Egbert, would it be an option to the vesa
> > EDID emulation if we don't get any information from AtomBIOS?
> > 
> How do we know which EDID the VESA BIOS will report?
> Is it always the EDID of the panel?
> I would like to avoid going down this route.

Ok, I read this as a no-go. Fine with me.


> I would be able to tell if I had the BIOS. Tyson, please run rhd_conntest -d to
> dump the BIOS and send it to me in an email (eich_at_suse_dot_de).

The BIOS is already available in attachment #16076 [details].

(In reply to comment #7)
> (In reply to comment #5)
> > Still remaining is:
> > 
> > 1- If I unplug an external monitor from the VGA port without first disabling it
> > with xrandr, it then begans to report bogus DVI information:
> 
> This sounds extremely weird.
> OTOH, this *could* be a side effect of a bug Luc fixed yesterday in the RandR
> code. Could you please try again with git head?

Okay.  I tried the latest git again (2008-05-22 up to and including the commit beginning with 566ba690), and it did not resolve this issue.

I did, however, discover some additional information while playing around, and that is the bogus information also disappears if I plug the display back into the VGA connector.

1) Unplug the display from the VGA port after having activated it and without first deactivating it:

Screen 0: minimum 320 x 200, current 1920 x 1200, maximum 1920 x 1920
VGA_1 disconnected 1920x1200+0+0 519mm x 324mm
PANEL connected
   1680x1050      60.1 +
DVI-D_1 disconnected
  1920x1200 (0x46)  154.0MHz
        h: width  1920 start 1968 end 2000 total 2080 skew    0 clock   74.0KHz
        v: height 1200 start 1203 end 1209 total 1235           clock   60.0Hz

2) Plug the display back into the VGA port without doing anything via xrandr:

Screen 0: minimum 320 x 200, current 1920 x 1200, maximum 1920 x 1920
VGA_1 connected 1920x1200+0+0 519mm x 324mm
   1920x1200      60.0*+   59.9
   ... <cut> ...
PANEL connected
   1680x1050      60.1 +
DVI-D_1 disconnected

Bugger.

If this thing is really what i think it is, then we're seriously had.

(II) RADEONHD(0): Connector[3] {RHD_CONNECTOR_DVI_SINGLE, "HDMI_TYPE_ADFP1",
RHD_DDC_1, RHD_HPD_1, { RHD_OUTPUT_NONE, RHD_OUTPUT_NONE } }

Apparently, there is an external TMDS device over the DVO bus involved here.

I need to go dig into this further, and i will probably need to acquire this hardware from somewhere, will ask around inside SUSE already.

In the meantime, can you give me the result of rhd_conntest -s (scanning DDC busses)

Here's the output of "rhd_conntest -s" with the various loads attached.

1) Nothing attached:

rhd_conntest: v1.2.1, git branch master, commit 566ba690 + changes
Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
  Load Detection: RHD_OUTPUT_NONE 
  HotPlug: RHD_HPD_NONE 
  DDC: RHD_DDC_NONE 

2) VGA display attached:

rhd_conntest: v1.2.1, git branch master, commit 566ba690 + changes
Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
  Load Detection: RHD_OUTPUT_DACA
  HotPlug: RHD_HPD_NONE 
  DDC: RHD_DDC_0
  DDC Line[0]: Slaves: 6e a0 

3) DVI display attached though HDMI port:

rhd_conntest: v1.2.1, git branch master, commit 566ba690 + changes
Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
  Load Detection: RHD_OUTPUT_NONE 
  HotPlug: RHD_HPD_1
  DDC: RHD_DDC_1
  DDC Line[1]: Slaves: a0 

4) VGA display and DVI display though HDMI attached (note that this is a single display with both its analogue and digital inputs plugged into the computer):

rhd_conntest: v1.2.1, git branch master, commit 566ba690 + changes
Checking connectors on 0x9581, 0x103C, 0x30DF  (@01:00:00):
  Load Detection: RHD_OUTPUT_DACA
  HotPlug: RHD_HPD_1
  DDC: RHD_DDC_0 RHD_DDC_1
  DDC Line[0]: Slaves: 6e a0 
  DDC Line[1]: Slaves: a0 

Thanks!  -Tyson

I just finished trying with the latest git (9b2e2693d) and this still doesn't work.  Something that might be helpful, however, is that I have discovered if the lid of the laptop is closed on boot up, the BIOS will enable the external DVI display plugged into the HDMI port.

Would it be useful to get a registers dump from the console (using rhd_dump)  in this state (i.e., after this BIOS has successfully enabled the external DVI display plugged into the HDMI port but before Xorg touches anything), and are there any registers in particular it would be good to know the state of?

Thanks!  -Tyson

I got it to work under the radeon driver (which would detect the display, but was unable to turn it on) by writing some of the registers manually after enabling it with xrandr.

Details (specifically the BIOS enabled/disabled register analysis) are at

http://bugs.freedesktop.org/show_bug.cgi?id=18564

Hope that is helpful.

Cheers!  -Tyson

PS:  I'm not sure how this got assigned to me, so I'm going to revert it back to the default as well.

Hi Tyson,


The term of "detect or Detected" is being used in both cases of seen and or
named.
By this the enduser can see the card but not it's name or type. ie I have
two MSI 4870 and neither of these cards are named or identified as "not
known" but are seen to be running ( whole or part) under X and or VGA std
mode.

Thus I think that this or to be left open until all of the "not known" type
of errors can be cleared.
Thanks and rgds SEanS



.

2009/1/2 <bugzilla-daemon@freedesktop.org>

> http://bugs.freedesktop.org/show_bug.cgi?id=15478
>
>
> Tyson Whitehead <twhitehead@gmail.com> changed:
>
>           What    |Removed                     |Added
>
> ----------------------------------------------------------------------------
>         AssignedTo|twhitehead@gmail.com        |lverhaegen@suse.de
>             Status|ASSIGNED                    |NEW
>
>
>
>
> --- Comment #14 from Tyson Whitehead <twhitehead@gmail.com>  2009-01-02
> 12:42:17 PST ---
> I got it to work under the radeon driver (which would detect the display,
> but
> was unable to turn it on) by writing some of the registers manually after
> enabling it with xrandr.
>
> Details (specifically the BIOS enabled/disabled register analysis) are at
>
> http://bugs.freedesktop.org/show_bug.cgi?id=18564
>
> Hope that is helpful.
>
> Cheers!  -Tyson
>
> PS:  I'm not sure how this got assigned to me, so I'm going to revert it
> back
> to the default as well.
>
>
> --
> Configure bugmail: http://bugs.freedesktop.org/userprefs.cgi?tab=email
> ------- You are receiving this mail because: -------
> You are the QA contact for the bug.
> _______________________________________________
> xorg-team mailing list
> xorg-team@lists.x.org
> http://lists.x.org/mailman/listinfo/xorg-team
> --
> To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org<radeonhd%2Bunsubscribe@opensuse.org>
> For additional commands, e-mail: radeonhd+help@opensuse.org<radeonhd%2Bhelp@opensuse.org>
>
>

Does this issue occur with the preferred ati driver (xf86-vide-ati)?  If so, please move this to the Driver/Radeon component.  

Development of radeonhd has pretty much halted and development focus is on the ati driver.  Please see http://www.x.org/wiki/radeonhd

If the issue does not exist in the ati driver (or if there is no response to this message), this bug will be closed as WONTFIX unless someone contributes a patch.

Closing due to lack of response.  Please reopen and move to the Driver/Radeon 
component if this issue persists with xf86-video-ati