Bug 18466 - [SDVO-VGA] First SDVO output reports failure to sync, no picture on 2nd monitor
Summary: [SDVO-VGA] First SDVO output reports failure to sync, no picture on 2nd monitor
Status: RESOLVED WONTFIX
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/intel (show other bugs)
Version: 7.3 (2007.09)
Hardware: x86 (IA32) Linux (All)
: medium major
Assignee: MaLing
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2008-11-10 08:38 UTC by Aidan Dixon
Modified: 2009-02-16 05:11 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Xorg-modedebug=on.0.log (279.10 KB, text/plain)
2008-11-10 08:40 UTC, Aidan Dixon
no flags Details
Xorg.conf (3.91 KB, application/octet-stream)
2008-11-10 08:45 UTC, Aidan Dixon
no flags Details
Xorg.conf (3.91 KB, text/plain)
2008-11-10 08:45 UTC, Aidan Dixon
no flags Details
latest xorg.conf (3.97 KB, text/plain)
2008-11-12 02:11 UTC, Aidan Dixon
no flags Details
output of xrandr -q --verbose (7.92 KB, text/plain)
2008-11-12 02:11 UTC, Aidan Dixon
no flags Details
Xorg.0.log with modedebug=on (124.53 KB, text/plain)
2008-11-12 02:20 UTC, Aidan Dixon
no flags Details

Description Aidan Dixon 2008-11-10 08:38:34 UTC
OS: Ubuntu Linux 8.04 (Hardy Heron)
Hardware: Intel GME945
Configuration: two monitors (flat panel) connected in VGA mode via on-board VGA and SDVO-VGA.
Intel driver: v2.2.1 (from Ubuntu) and v2.3.1 (from git)

Xorg.log, xorg.conf and output of xrandr -q --verbose all attached.
Comment 1 Aidan Dixon 2008-11-10 08:40:13 UTC
Created attachment 20189 [details]
Xorg-modedebug=on.0.log

Xorg.0.log from Ubuntu system with driver 2.3.1 and modedebug=on
Comment 2 Aidan Dixon 2008-11-10 08:45:35 UTC
Created attachment 20190 [details]
Xorg.conf
Comment 3 Aidan Dixon 2008-11-10 08:45:38 UTC
Created attachment 20191 [details]
Xorg.conf
Comment 4 Gordon Jin 2008-11-11 04:23:04 UTC
So what's the sympton? And you forgot to attach xrandr output.

The git driver version should be 2.5.x.

Comment 5 Aidan Dixon 2008-11-11 08:19:39 UTC
Regarding the git source, my tree is old and I cannot get anything 2.4.0 or newer to compile on my system without upgrading DRM also.

I will attach the missing xrandr -q output.

The symptom is that no picture is ever displayed on the second monitor.  The output of VGA is present, the output of VGA-1 not so.
Comment 6 Aidan Dixon 2008-11-12 02:11:04 UTC
Created attachment 20243 [details]
latest xorg.conf
Comment 7 Aidan Dixon 2008-11-12 02:11:37 UTC
Created attachment 20244 [details]
output of xrandr -q --verbose
Comment 8 Aidan Dixon 2008-11-12 02:20:29 UTC
Created attachment 20246 [details]
Xorg.0.log with modedebug=on
Comment 9 Aidan Dixon 2008-11-24 09:36:38 UTC
Some additional notes:

BIOS (Award) is configured for CRT+CRT video at boot and 2nd screen shows picture at boot and when at Linux console (and always under Windows with Intel driver).

I don't know if there's some confusion with the SDVO-DVI chip (Silicon Image Sil1362CLU) on the same SDVO channel as the SDVO-VGA chip (Chrontel CH7021A-TEF).

If the DVI output (TMDS-1) is enabled then the monitor gets synch but still not picture is displayed.  I will check a full SDVO-DVI configuration just to double check that works but SDVO-VGA is what I need to get working for my paying customer.
Comment 10 Aidan Dixon 2008-12-02 03:45:49 UTC
I have observed that if I start the computer up with a monitor connected only to the SDVO-VGA connector then the picture appears correctly on that output appears correctly.  It seems that on-board VGA alone or SDVO-VGA alone do work but both together does not work.
Comment 11 Wang Zhenyu 2008-12-07 21:25:25 UTC
Have you tried more recent drivers? How about removing "TMDS-1" ignore option first?
Comment 12 Michael Fu 2008-12-09 21:29:59 UTC
(In reply to comment #5)
> Regarding the git source, my tree is old and I cannot get anything 2.4.0 or
> newer to compile on my system without upgrading DRM also.
> 
> I will attach the missing xrandr -q output.
> 
> The symptom is that no picture is ever displayed on the second monitor.  The
> output of VGA is present, the output of VGA-1 not so.
> 

Aidan, 2.3.1 is too old for us to fix this bug. we do need your help to provide us latest status when using latest release or git repository. One hint is to use the latest LiveCD from various distributions. If you have a chance to try, please upload your xorg.log with modedebug turns on.. thanks.
Comment 13 Wang Zhenyu 2008-12-15 21:56:25 UTC
Aidan, could you give some info on your hardware? Looks there're two SDVO encoders on your board?
Comment 14 Aidan Dixon 2008-12-16 02:19:14 UTC
In reply to comment #13:

| Aidan, could you give some info on your hardware? Looks there're two SDVO
| encoders on your board?

Correct.  The board is our custom DPX-117 which we sell to the casino gaming industry.  We use a mobile 945-based chipset.  We offer a range of video options.  Normally just on-board VGA as standard but there's a MXM/PEG slot that can take either ATI/Nvidia's MXM-format GPUs or an Intel SDVO-based solution.  The SDVO solution is what's the problem here: the card is both a SDVO-DVI chip (Silicon Image Sil1362CLU) on the same SDVO channel as the SDVO-VGA chip (Chrontel CH7021A-TEF).  This allows us to offer VGA+VGA or VGA+DVI.
Comment 15 MaLing 2008-12-30 22:42:00 UTC
(In reply to comment #14)
> In reply to comment #13:
> | Aidan, could you give some info on your hardware? Looks there're two SDVO
> | encoders on your board?
> Correct.  The board is our custom DPX-117 which we sell to the casino gaming
> industry.  We use a mobile 945-based chipset.  We offer a range of video
> options.  Normally just on-board VGA as standard but there's a MXM/PEG slot
> that can take either ATI/Nvidia's MXM-format GPUs or an Intel SDVO-based
> solution.  The SDVO solution is what's the problem here: the card is both a
> SDVO-DVI chip (Silicon Image Sil1362CLU) on the same SDVO channel as the
> SDVO-VGA chip (Chrontel CH7021A-TEF).  This allows us to offer VGA+VGA or
> VGA+DVI.

Thanks for your info, but I strongly suggest you use the latest version because we have appended new features and logs, which can accelerate our modification.

Thanks
Ma Ling
Comment 16 MaLing 2009-01-04 03:52:11 UTC
hi Aidan,

I find the sdvo-vga is not attached succesfully. by using latest version, could you paste log when it works fine --Comment #10.

Thanks
Ma Ling
Comment 17 MaLing 2009-01-07 18:07:55 UTC
ping Aidan ~
Comment 18 Aidan Dixon 2009-01-08 00:55:14 UTC
My apologies for delays replying.   I have been on vacation for a while and the equipment required to test the latest drivers will have to be reclaimed.  I have not stopped working on this however.
Comment 19 MaLing 2009-01-08 01:04:06 UTC
(In reply to comment #18)
> My apologies for delays replying.   I have been on vacation for a while and the
> equipment required to test the latest drivers will have to be reclaimed.  I
> have not stopped working on this however.

hi Aidan

If everying is ready, Please let me know .

Thanks
Ma Ling
Comment 20 Michael Fu 2009-01-21 01:13:45 UTC
ping Aidan Dixon ....
Comment 21 Aidan Dixon 2009-01-21 02:24:54 UTC
I have the equipment in place now.  I have driver 2.5 ready too.  Just need to update kernel DRM2 driver.
Comment 22 MaLing 2009-01-22 06:15:09 UTC
hi Aidan Dixon

from dump register in xorg log:
 
(II) intel(0):                SDVOB: 0x80480084 (enabled, pipe A, stall disabled, detected)
(II) intel(0):                SDVOC: 0xc0480080 (enabled, pipe B, stall disabled, not detected)
(II) intel(0):              SDVOUDI: 0x00000000
(II) intel(0):               DSPARB: 0x00001d9c
(II) intel(0):               DSPFW1: 0x00000000
(II) intel(0):               DSPFW2: 0x00000000
(II) intel(0):               DSPFW3: 0x00000000
(II) intel(0):                 ADPA: 0x80000010 (enabled, pipe A, -hsync, +vsync)

There should be a conflict between VGA and SDVO-VGA.

So after you update driver, start Xorg, then paste Xorg log file again.


Thanks 
Ma Ling
Comment 23 MaLing 2009-02-01 22:39:11 UTC
ping Aidan Dixon ....

Thanks
Ma Ling
Comment 24 Michael Fu 2009-02-12 00:27:12 UTC
ping... please attach a log of 2.5 or 2.6 driver, with modedebug=on.
Comment 25 Michael Fu 2009-02-15 21:52:06 UTC
(In reply to comment #10)
> I have observed that if I start the computer up with a monitor connected only
> to the SDVO-VGA connector then the picture appears correctly on that output
> appears correctly.  It seems that on-board VGA alone or SDVO-VGA alone do work
> but both together does not work.
> 

please also either attach a reg_dump or xorg log with modedebug turns on of this senario... thanks.
Comment 26 Aidan Dixon 2009-02-16 00:37:19 UTC
Hi Michael (and Team)

I have been unable to work on this bug due to illness.  I do not know when I will be able to continue working.  You may close the bug if you feel that necessary.  

Kind Regards,
Aidan
Comment 27 MaLing 2009-02-16 05:11:39 UTC
hi  Aidan 
we are sorry to hear that, hope you get better soon!
I close the issue now. If it still occurs after using our latest version, please reopen it.

Best Regards!
Ma Ling 


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.