Bug 22719 - Dual-monitor with two different videoadapters sends image only to primary monitor
Summary: Dual-monitor with two different videoadapters sends image only to primary mon...
Status: RESOLVED INVALID
Alias: None
Product: xorg
Classification: Unclassified
Component: Server/General (show other bugs)
Version: 7.4 (2008.09)
Hardware: x86 (IA32) FreeBSD
: medium normal
Assignee: Xorg Project Team
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2009-07-10 22:14 UTC by Eugene Zykov
Modified: 2018-06-12 18:44 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
xorg.conf with s3 driver (2.52 KB, text/plain)
2009-08-15 10:48 UTC, Eugene Zykov
no flags Details
Xorg.0.log with s3 driver (26.63 KB, text/plain)
2009-08-15 10:49 UTC, Eugene Zykov
no flags Details
xorg.conf - ATI (ati driver) + S3 (vesa driver) (5.36 KB, text/plain)
2009-08-15 13:49 UTC, Eugene Zykov
no flags Details
Xorg log (ATI + ati driver and S3 + vesa driver) (49.58 KB, text/plain)
2009-08-15 13:50 UTC, Eugene Zykov
no flags Details

Description Eugene Zykov 2009-07-10 22:14:53 UTC
I didn't manage to get two different videoadapters running on Xorg.

My operating system and software versions:
FreeBSD 7.2-RELEASE
xorg-server-1.6.1,1
xf86-video-vesa-2.1.0_1
xf86-video-ati-6.12.2_1
I have two videocards: ATI Radeon 9250 on AGP slot and S3 Trio64V+ on PCI slot (S3 works only with "vesa" driver on Xorg).

Before applying dual-monitor configuration, I want to test monitors separately.
I take working single-monitor xorg.conf:

=====
Driver "vesa"
BusID   "PCI:1:0:0"
=====

Image goes to first monitor, ATI, this is normal.

I change configuration to:

=====
Driver "vesa"
BusID   "PCI:2:7:0"
=====

Image have to display to the second monitor, but it goes to first monitor anyway.
As I remember, on older X Server version (1.4.1) the text part is left on first screen, graphical part goes to second monitor.

In Xorg.0.log this text is written:

=====
(!!) More than one possible primary device found
(--) PCI: (0@1:0:0) ATI Technologies Inc RV280 [Radeon 9200 PRO] rev 1, Mem @ 0xd8000000/134217728, 0xf69f0000/65536, I/O @ 0x0000c000/256, BIOS @ 0x????????/65536
(--) PCI: (0@1:0:1) ATI Technologies Inc RV280 [Radeon 9200 PRO] (Secondary) rev 1, Mem @ 0xd0000000/134217728, 0xf69e0000/65536
(--) PCI: (0@2:7:0) S3 Inc. 86c764/765 [Trio32/64/64V+] rev 84, Mem @ 0xf8000000/67108864, BIOS @ 0x????????/65536
(II) System resource ranges:
	[0] -1	0	0x000f0000 - 0x000fffff (0x10000) MX[B]
	[1] -1	0	0x000c0000 - 0x000effff (0x30000) MX[B]
	[2] -1	0	0x00000000 - 0x0009ffff (0xa0000) MX[B]
	[3] -1	0	0x0000ffff - 0x0000ffff (0x1) IX[B]
	[4] -1	0	0x00000000 - 0x000000ff (0x100) IX[B]
=====

When I set AGP as primary adapter in BIOS, everything draws on first monitor, if PCI - on second monitor, regardless what BusID value (PCI:1:0:0 or PCI:2:7:0) is set in xorg.conf.
If I try dual-monitor xorg.conf, two desktops simply draws to the same screen - for example: pointer goes out from left edge and appears from the right side.
Comment 1 Evgeny Zubok 2009-08-14 08:26:32 UTC
>I have two videocards: ATI Radeon 9250 on AGP slot and S3 Trio64V+ on PCI slot (S3 works only with "vesa" driver on Xorg).

This is strange. What version of s3 driver you have? The s3 driver didn't support depth 24 until version 0.6.1. Only the depths 8/16 were supported. Starting with version 0.6.1, the 24-bit mode should work.

S3 Trio64V+ with s3 driver works on Xorg/Linux. I have two testers with such board.
Comment 2 Eugene Zykov 2009-08-14 12:37:03 UTC
s3 driver does not work at all - so I have to use vesa instead of it. It is very easy to test, because Microsoft Virtual PC, that use S3-card also fails on FreeBSD installation.
System version: FreeBSD 7.2-RELEASE
Driver version: xf86-video-s3-0.6.1_2
But the problem, actually is that the vesa driver, that works with single monitor doesn't work with dual monitor as expected.
Comment 3 Evgeny Zubok 2009-08-14 14:31:24 UTC
(In reply to comment #2)
> s3 driver does not work at all - so I have to use vesa instead of it. It is
> very easy to test, because Microsoft Virtual PC, that use S3-card also fails on
> FreeBSD installation.
> System version: FreeBSD 7.2-RELEASE
> Driver version: xf86-video-s3-0.6.1_2

Hmm, there is a report that all is ok with Virtual PC but with Linux:
https://bugs.freedesktop.org/show_bug.cgi?id=14999

> But the problem, actually is that the vesa driver, that works with single
> monitor doesn't work with dual monitor as expected.

Sorry that I'm asking about s3 and not about your problem. I'm a contributor to s3 driver, and this problem bothers me. Can you attach full Xorg log and xorg.conf when you use s3 driver?
Comment 4 Eugene Zykov 2009-08-15 10:48:31 UTC
Created attachment 28645 [details]
xorg.conf with s3 driver
Comment 5 Eugene Zykov 2009-08-15 10:49:29 UTC
Created attachment 28646 [details]
Xorg.0.log with s3 driver
Comment 6 Eugene Zykov 2009-08-15 11:00:28 UTC
I attached xorg.conf and Xorg.0.log.
Operating system - NetBSD 5.0.1 (guest), Emulator - Virtual PC 2007 (6.0.192.0)
This is almost "autogenerated" xorg.conf (I added 800x600 resolution for vesa driver - it does not set up 640x480 resolution).
The symptom in VirtualPC and real hardware is the same - display "hangs" and draws some red interlaced squares in the top of the screen.

I left all tries to get dual monitor setup working - just installed NVidia GeForce MX 440 with proprietary drivers and it worked on the first try.
Comment 7 Evgeny Zubok 2009-08-15 12:35:56 UTC
Please read at my explanation above. 24-bit colour mode was unavaible until version 0.6.1, and your log shows that NetBSD has version 0.5.0:


(II) Module s3: vendor="X.Org Foundation"
	compiled for 1.4.2, module version = 0.5.0
	Module class: X.Org Video Driver
	ABI class: X.Org Video Driver, version 2.0
...
(II) S3: driver (version 0.5.0 for S3 chipset: 964-0, 964-1, 968,
	Trio32/64, Aurora64V+, Trio64UV+, Trio64V2/DX/GX


and depth 24 (by defaul):

(==) s3(0): Depth 24, (==) framebuffer bpp 32
(==) s3(0): RGB weight 888
(==) s3(0): Default visual is TrueColor


So the only way is to lower depth to 16. Just add "DefaultDepth 16" to your xorg.conf:

Section "Screen"
	Identifier "Screen0"
	Device     "Card0"
	Monitor    "Monitor0"
        DefaultDepth 16
	SubSection "Display"
		Viewport   0 0
		Depth     1
		Modes "800x600"
	EndSubSection
	SubSection "Display"
		Viewport   0 0
		Depth     4
		Modes "800x600"
	EndSubSection
	SubSection "Display"
		Viewport   0 0
		Depth     8
		Modes "800x600"
	EndSubSection
	SubSection "Display"
		Viewport   0 0
		Depth     15
		Modes "800x600"
	EndSubSection
	SubSection "Display"
		Viewport   0 0
		Depth     16
		Modes "800x600"
	EndSubSection
	SubSection "Display"
		Viewport   0 0
		Depth     24
		Modes "800x600"
	EndSubSection
EndSection


Please, inform me about the result. 

Also I recommend you to leave here full log, full xorg.conf and lspci -vv output if you want more qualified help with the your main problem with dual head. The exerpts don't make clear anything.
Comment 8 Eugene Zykov 2009-08-15 13:47:30 UTC
Thank you, This works (at least in Virtual PC).

I suspect that x.org doesn't work properly with any pair of videocards and placed the steps to reproduce the bug.
I cannot paste lspci output here, but I kept xorg.conf and Xorg.0.log from that time, if this can help.

The only configuration that worked on 1.4.1 - was vga+vga (on different adapters). But the vga driver is broken now on XOrg (why? This is, probably, the most documented and known driver)...
Comment 9 Eugene Zykov 2009-08-15 13:49:45 UTC
Created attachment 28651 [details]
xorg.conf - ATI (ati driver) + S3 (vesa driver)
Comment 10 Eugene Zykov 2009-08-15 13:50:50 UTC
Created attachment 28652 [details]
Xorg log (ATI + ati driver and S3 + vesa driver)
Comment 11 Evgeny Zubok 2009-08-16 00:35:08 UTC
> I suspect that x.org doesn't work properly with any pair of videocards and
> placed the steps to reproduce the bug.
> I cannot paste lspci output here, but I kept xorg.conf and Xorg.0.log from that
> time, if this can help.

Did you specify the right BusIDs? Maybe you need 1:0:1 and not 1:0:0? This message from the log is strange:

(!!) More than one possible primary device found
(--) PCI: (0@1:0:0) ATI Technologies Inc RV280 [Radeon 9200 PRO] rev 1, Mem @ 0xd8000000/134217728, 0xf69f0000/65536, I/O @ 0x0000c000/256, BIOS @ 0x????????/65536
(--) PCI: (0@1:0:1) ATI Technologies Inc RV280 [Radeon 9200 PRO] (Secondary) rev 1, Mem @ 0xd0000000/134217728, 0xf69e0000/65536
(--) PCI: (0@2:7:0) S3 Inc. 86c764/765 [Trio32/64/64V+] rev 84, Mem @ 0xf8000000/67108864, BIOS @ 0x????????/65536

You may learn the correct BusIDs for videoboards from lspci or lshw.
Comment 12 Adam Jackson 2018-06-12 18:44:01 UTC
Mass closure: This bug has been untouched for more than six years, and is not obviously still valid. Please file a new report if you continue to experience issues with a current server.


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.