opensuse 11.3: prevent "AIGLX: Suspending AIGLX clients for VT switch"
Ralf Fassel
Opensuse 11.1
X.Org X Server 1.5.2
Release Date: 10 October 2008
X Protocol Version 11, Revision 0
Build Operating System: openSUSE SUSE LINUX
Current Operating System: Linux gepard2 2.6.27.48-0.2-pae #1 SMP 2010-07-29 20:06:52 +0200 i686
Build Date: 15 April 2010 04:23:30PM
to
Opensuse 11.3
X.Org X Server 1.8.0
Release Date: 2010-04-02
X Protocol Version 11, Revision 0
Build Operating System: openSUSE SUSE LINUX
Current Operating System: Linux gepard2 2.6.34.7-0.3-desktop #1 SMP PREEMPT 2010-09-20 15:27:38 +0200 i686
Build Date: 19 August 2010 03:34:17PM
- we start a second X-Server and switch between the two via
Ctrl-Alt-F7/F8
Whenever a vt is switched, the X-logfiles show
[ 16677.712] (II) AIGLX: Suspending AIGLX clients for VT switch
...
[ 16680.538] (II) AIGLX: Resuming AIGLX clients after VT switch
and the clients running on the non-displayed VT are suspended so it
seems, at least they don't make any progress.
Since we would like to use the second VT for automatic tests, the
clients should continue running (and they did under Opensuse 11.1/X
1.5.2)
Is there a way to disable this suspending/resuming of clients when VTs
are switched?
Does AIGLX have anything to do with this?
If so, will disabling AIGLX stop this unwanted behaviour?
Any hints welcome...
R'
1jam
> Whenever a vt is switched, the X-logfiles show
> [ 16677.712] (II) AIGLX: Suspending AIGLX clients for VT switch
> ...
> [ 16680.538] (II) AIGLX: Resuming AIGLX clients after VT switch
>
> and the clients running on the non-displayed VT are suspended so it
> seems, at least they don't make any progress.
You must be running the proprietary ATI driver then? Hmm I just tried
switching to VT and back. I see similar messages in my XOrg.0.log.
I dunno...
Maybe try getting help on the linux wiki site:
http://wiki.cchtml.com/index.php/Main_Page
-johnm
Darklight
I think you should read the release notes that might point
you in the right direction.
Ralf Fassel
| > [ 16677.712] (II) AIGLX: Suspending AIGLX clients for VT switch
| > ...
| > [ 16680.538] (II) AIGLX: Resuming AIGLX clients after VT switch
| You must be running the proprietary ATI driver then?
Intel 945G on-board.
16: PCI 02.0: 0300 VGA compatible controller (VGA)
[Created at pci.318]
Unique ID: _Znp.kw4rvyJqLZ2
SysFS ID: /devices/pci0000:00/0000:00:02.0
SysFS BusID: 0000:00:02.0
Hardware Class: graphics card
Model: "Intel 945G"
Vendor: pci 0x8086 "Intel Corporation"
Device: pci 0x2772 "945G"
SubVendor: pci 0x1028 "Dell"
SubDevice: pci 0x01ad
Revision: 0x02
Driver: "i915"
Driver Modules: "drm"
Memory Range: 0xfeb00000-0xfeb7ffff (rw,non-prefetchable)
I/O Ports: 0xe898-0xe89f (rw)
Memory Range: 0xe0000000-0xefffffff (ro,non-prefetchable)
Memory Range: 0xfeac0000-0xfeafffff (rw,non-prefetchable)
IRQ: 16 (7877164 events)
I/O Ports: 0x3c0-0x3e7 (rw)
Module Alias: "pci:v00008086d00002772sv00001028sd000001ADbc03sc00i00"
Driver Info #0:
XFree86 v4 Server Module: intel
Driver Info #1:
XFree86 v4 Server Module: intel
3D Support: yes
Extensions: dri
Config Status: cfg=no, avail=yes, need=no, active=unknown
17: PCI 02.1: 0380 Display controller
[Created at pci.318]
Unique ID: ruGf.ocHjFWzbCK9
SysFS ID: /devices/pci0000:00/0000:00:02.1
SysFS BusID: 0000:00:02.1
Hardware Class: graphics card
Model: "Intel 82945G/GZ Integrated Graphics Controller"
Vendor: pci 0x8086 "Intel Corporation"
Device: pci 0x2776 "82945G/GZ Integrated Graphics Controller"
SubVendor: pci 0x1028 "Dell"
SubDevice: pci 0x01ad
Revision: 0x02
Memory Range: 0xfeb80000-0xfebfffff (rw,non-prefetchable)
Module Alias: "pci:v00008086d00002776sv00001028sd000001ADbc03sc80i00"
Config Status: cfg=no, avail=yes, need=no, active=unknown
R'
Ralf Fassel
| >> Whenever a vt is switched, the X-logfiles show
| >> [ 16677.712] (II) AIGLX: Suspending AIGLX clients for VT switch
| >> ...
| >> [ 16680.538] (II) AIGLX: Resuming AIGLX clients after VT switch
| >>
| I think you should read the release notes that might point
| you in the right direction.
Ok, thanks for pointing that out. KMS. Never would have guessed that.
Adding 'nomodeset' to the boot line keeps my tests running in the
background on the second display, which is fine.
But disabling a good new feature just to keep processes on the other
display running sounds like a misdesign in that feature to me...is there
really no way to tell KMS "don't tell them processes the display has
gone away"?
<rant>
And why does that 'nomodeset' think my monitor can only display
1600x1200? Leaving out 'nomodeset' detects the correct 1920x1200...
Back to editing xorg.conf manually, weelllllll...
</rant>
R'
Ralf Fassel
| And why does that 'nomodeset' think my monitor can only display
| 1600x1200? Leaving out 'nomodeset' detects the correct 1920x1200...
| Back to editing xorg.conf manually, weelllllll...
I can't get the system with 'nomodeset' to set the screen resolution to
1920x1200, I always end up with 1600x1200.
- ran X :0 -configure, stored /root/xorg.conf.new as /etc/X11/xorg.conf
-> file is recognized, but X does not start
- just took the 'Monitor' sections and edited into
/etc/X11/xorg.conf.d/50-monitor.conf, no change (1600x1200)
- edited the 11.1 xorg.conf into /etc/X11/xorg.conf or 50-monitor.conf,
no effect (1600x1200)
- set NO_KMS_IN_INITRD, no change (1600x1200)
- added mode lines via cvt to 50-monitor.conf, no change (1600x1200)
Right now I'm back to KMS and all X11 config files restored to
original. <grmbl> Let the tests run overnight...just like on Windows...</grmbl>
Maybe I have time later to sort this mess out.
Thanks anyway for the pointer...
R'
Darklight
With my nvidia card i installed/compiled the nvidia driver
and use the nvidia-settings gui to configure my graphics card.
does not your ati driver do the same. Owe from what i read
they should be bringing back a replacement for sax2
can't wait.
Ralf Fassel
| With my nvidia card i installed/compiled the nvidia driver
| and use the nvidia-settings gui to configure my graphics card.
On a different machine with Nvidia onboard graphics and the Nvidia
proprietary driver, the suspend/resume does not happen when I switch
virtual terminals (there is no mention in the logs, and the processes
keep running on the other VT). Dito with an Nvidia GTX 285 in a third
machine. Only my cheap onboard Intel has the problem....or is it meant
to be a feature?
| does not your ati driver do the same.
It's an Intel onboard-graphics, not ATI.
| from what i read they should be bringing back a replacement for sax2
| can't wait.
/me 2.
R'