[Hardy Alpha-5] fglrx log noise if dualhead unless ForceMonitors=nocv

[Filippo Carletti]

Both /var/log/gdm/:0.log and /var/log/Xorg.0.log are flooded with lines like the following:

(EE) fglrx(1): [DRI] Locking deadlock.
 Already locked with context 137016756,
 trying to lock with context 2.
(EE) fglrx(1): [DRI] Unlocking inconsistency:
 Context 137016756 trying to unlock lock held by context 2

only when a second monitor is attached to the graphic card (01:00.0 VGA compatible controller: ATI Technologies Inc M24 1P [Radeon Mobility X600]).
Log size increases about 1G/day.

ii xorg-driver-fglrx 1:7.1.0-8-02+2.6.24.9-10.27

[boucman (jeremy-rosen)]

I can confirm that bug, when googling for the log line a little, it seems related to dual-head configurations...

[boucman (jeremy-rosen)]

ok, while digging around in the ATI bug list, I found the following related bugs

http://ati.cchtml.com/show_bug.cgi?id=941
http://ati.cchtml.com/show_bug.cgi?id=570
http://ati.cchtml.com/show_bug.cgi?id=902

with workarounds (bigdesktop, disabling DRI...)

hope this helps

[Filippo Carletti (filippo-carletti)]

I'm seeing bug 902. Even screensaver behaviour is the same.
But the workaround seems not working.

Are my modifications correct? Maybe my bigdesktop setup isn't correct.

--- /etc/X11/xorg.conf.good 2007-05-03 23:45:51.000000000 +0200
+++ /etc/X11/xorg.conf 2008-03-14 13:29:38.000000000 +0100
@@ -96,6 +96,7 @@
  Identifier "ATI Technologies Inc M24 1P [Radeon Mobility X600]"
  Driver "fglrx"
  BusID "PCI:1:0:0"
+ Option "ForceMonitors" "nocv"
 EndSection

 Section "Device"
@@ -103,6 +104,7 @@
  Driver "fglrx"
  BusID "PCI:1:0:0"
  Screen 1
+ Option "ForceMonitors" "nocv"
 EndSection

[boucman (jeremy-rosen)]

well, I've hadn't tested myself yet, I found the ATI bug at work, but I need to test at home.

I posted it here in case other people had the same problem, and help triaging the bug for upstream support

[Bryce Harrington (bryce)]

Filippo, your modifications for setting up dual screen don't look correct, but I'm not certain if that would be what causes the log errors. I would recommend referring to some setup guides for dual head on fglrx and generate a clean configuration. If the issue still exists even with a proper config, please feel free to reopen this bug report.

[boucman (jeremy-rosen)]

in what way is the xorg.conf incorrect ?

I might miss something obvious, but it seems OK to me...

the logs are flooded on my setup, and the display is correctly setup... it block on OGL and window redraw is very slow, but I have two screens with correct resolution working

[yota (yota-opensystems)]

Same problem here with Hardy beta on a asus M2A with x1250 on board.
I can see "(EE) fglrx(1): [DRI] Locking deadlock" on the logs and OGL is not working if dual head is enabled.

Only viable solution is to disable DRI with 'option "no_dri" "yes"' on xorg.conf, but that means no 2d acceleration and no 3d at all.

@Bryce: the xorg.conf above seems fine to me; moreover some searches starting from http://www.google.com/search?hl=en&q=fglrx+deadlock&btnG=Google+Search and in particular http://forums.gentoo.org/viewtopic-t-659846-view-next.html?sid=6927d2edb2b2367d4d9465ebac8910da lead me to believe that this is a problem with fglrx<=8.3 with Xorg 1.4 and dual head

Please unmark this issue as invalid and include a new driver from ati, when available, able to support dual head on Xorg 1.4

[Filippo Carletti (filippo-carletti)]

Log noise disappeared yesterday.

I found the following lines in /var/log/Xorg.0.log:
(WW) fglrx(0): ***********************************************
(WW) fglrx(0): * DRI initialization failed! *
(WW) fglrx(0): * (maybe driver kernel module missing or bad) *
(WW) fglrx(0): * 2D acceleraton available (MMIO) *
(WW) fglrx(0): * no 3D acceleration available *
(WW) fglrx(0): ********************************************* *

[Jørgen H. Seland (jorgen-fabeljet)]

Have installed ATI 8.4 drivers, and running Xinerama-based dual head on ATI Radeon Mobility X1300, hardy beta 5, AMD 64.

I now have the "log noise" problem; keeps repeating:

...
(EE) fglrx(1): [DRI] Unlocking inconsistency:
        Context 9124404 trying to unlock lock held by context 2
(EE) fglrx(1): [DRI] Locking deadlock.
        Already locked with context 9124404,
        trying to lock with context 2.
...

And: OpenGL blocks. For instance, if I run glxinfo, it stops after printing the first line of output:
jorgen@brain:~$ glxinfo
name of display: :0.0

The screensaver setup utility also stops on launch. (I have an OpenGL screensaver selected.)

[konny (s-launchpad-gadzicki-de)]

Pretty much the same here after upgrading to Hardy. Dual-head xinerama setup with ATI Radeon Mobility X700 was working with Gutsy normaly, but with Hardy I get the log noise mentioned above. My Xorg log grows to 500MB after 1h.

OpenGL related applications don't work.

[garydion (gdion)]

I too can confirm this problem. I'm running a Gigabyte GA-MA78GM-S2H motherboard (780G chipset) with the built-in graphics adapter in dual-head mode and have all the problems described above. If I walk away from my computer longer than ten minutes or so, the screen will comepletely lock up. Brand new install of 8.04 just today. Used aticonfig to set up dual screens.

[boucman (jeremy-rosen)]

the lock-up is due to open-gl screensavers trying to stop. You'd better disable screensavers entirely, or at least filter out open-gl ones

on a related note, new users will have compiz as their default window manager... you need to tweak with gconf to change back to a non open-gl one...

[Schmallon (matthias-kleine)]

I can confirm this bug: When configuring for dual head my Xorg.0.log grows with the same error messages. Manually installing the most recent official fglrx driver (8.476) does not fix it nor does adding ForceMonitors=nocv.

[Ken Collins (ken-confirmative)]

I have the same problem. I have a dual head Radeon X1300, and it was working well with the 8.3 driver and Gutsy. After upgrading to Hardy, I see the deadlock errors in Xorg.0.log, fglrxinfo hangs, the gnome screensaver application hangs, and locking the screen causes X to hang. I tried setting ForceMonitors to "nocv", but it didn't help. I had to disable DRI, which is less than ideal.

[pankostechka (pan-kostechka)]

Same bug on Macbook pro C2D, Hardy (fresh install), ATI Mobility Radeon X1600, in dual-head mode. No workaround found.

[Arvind Menon (amenon81)]

I can confirm the bug. I'm running this on a Thinkpad T60 with an ATI X1400 in dual-head mode. I had to completely disable my screensavers, or else my computer would lockup. The log files fill up my disk pretty quick and slows down the computer.

[Patrik Kristiansson (patrik-kristiansson)]

Same bug on ASUS x1950pro system but works me with
aticonfig --dtop=horizontal

[Greg (greg-edvenson)]

Seeing this bug on Macbook Pro C2D (X1600), Hardy upgrade from Gutsy with dual-head.

[Brian May (brian-microcomaustralia)]

I added the following to the end of http://ati.cchtml.com/show_bug.cgi?id=941

I just got a Dell Optiplex 755 with a twinhead X1300 card running Ubuntu Hardy
with the latest fglrx drivers from ATIs website, and have encountered what
looks like the same problem.

Apparently these bugs are the same:

https://bugs.launchpad.net/ubuntu/+source/linux-restricted-modules-2.6.24/+bug/196617
http://ati.cchtml.com/show_bug.cgi?id=570
http://ati.cchtml.com/show_bug.cgi?id=902

However in my case the only work around that works is to disable dri.

Option "ForceMonitors" "nocv" -- same problem.

aticonfig --dtop=horizontal -- same problem.

Symptoms: The mouse pointer is corrupt, and then the computer hangs (requiring
cold reboot) or reboots.

Question: What does the Option "ForceMonitors" "nocv" option do? I tried looking for reference documentation on the options but couldnt.

[evan martin (m42martin)]

Last month I upgraded to Hardy via the upgrade. I'm almost certain that I did not have this problem until today when I reinstalled directly from the full installer disk.

In case this might be a shared symptom:
My computer is taking a longer time to boot, and just about every 10 minutes my laptop monitor goes black. The external monitor looks normal and I have control of the mouse - but both monitors are unresponsive and I have to ctrl-alt-backspace to restart X.

Other than that, I too have the DRI unlocking / locking bug and I have tried using the same xorg.conf file that was working for me with the Hardy upgrade and before since Fiesty.

(see xorg.conf attached)

I have tried several xorg.conf versions - all result in the 10 minute black out - and as far as I can tell, also result in the DRI unlocking error.

[Jyri Sillanpaa (ubuntu-amorco)]

Same problem here, the log is flooded and grows about a megabyte per minute.
The card:
02:00.0 VGA compatible controller: ATI Technologies Inc RV370 5B60 [Radeon X300 (PCIE)]
02:00.1 Display controller: ATI Technologies Inc RV370 [Radeon X300SE]

And I can't start xscreensaver-demo, the configuration window appears blank and just hangs.

[Jyri Sillanpaa (ubuntu-amorco)]

And no, Option "ForceMonitors" "nocv" doesn't fix the problem.

[evan martin (m42martin)]

Just wanted to add that I'm experiencing this problem with :

00:00.0 Host bridge: ATI Technologies Inc RS480 Host Bridge (rev 01)
00:01.0 PCI bridge: ATI Technologies Inc RS480 PCI Bridge
00:05.0 PCI bridge: ATI Technologies Inc RS480 PCI Bridge
00:13.0 USB Controller: ATI Technologies Inc IXP SB400 USB Host Controller
00:13.1 USB Controller: ATI Technologies Inc IXP SB400 USB Host Controller
00:13.2 USB Controller: ATI Technologies Inc IXP SB400 USB2 Host Controller
00:14.0 SMBus: ATI Technologies Inc IXP SB400 SMBus Controller (rev 11)
00:14.1 IDE interface: ATI Technologies Inc IXP SB400 IDE Controller
00:14.3 ISA bridge: ATI Technologies Inc IXP SB400 PCI-ISA Bridge
00:14.4 PCI bridge: ATI Technologies Inc IXP SB400 PCI-PCI Bridge
01:05.0 VGA compatible controller: ATI Technologies Inc Radeon XPRESS 200M 5955 (PCIE)

and ForceMonitors doesn't work for me either.

I have also tried commenting out DPMS in the monitor sections :

Section "Monitor"
 Identifier "Laptop Monitor"
# Option "DPMS"
EndSection

Section "Monitor"
 Identifier "Samsung Monitor"
# Option "DPMS"
EndSection

and adding the following to the ServerLayout

 Option "BlankTime" "0"
 Option "StandbyTime" "0"
 Option "SuspendTime" "0"
 Option "OffTime" "0"
...
none of which helps either the monitor blanking out and system freezing - or the Xorg.0.log file from bloating with errors.

Could there be a link between the Xorg log file growing in size to the system crashing?

Anyone else experiencing similar / related issues?

[yota (yota-opensystems)]

Latest fglrx from amd (version 8.5) is working for me, with DRI and OGL enabled on a dual-head configuration.

Xvideo and textured video are still not working, so 'Option "TexturedVideo" "off"' is needed on device section in xorg.conf; anyway it's a significant improvement: please include fglrx 8.5 in the official repositories.

[Jyri Sillanpaa (ubuntu-amorco)]

Ok, after the driver upgrade everything seems to work again (ati-driver-installer-8-5-x86.x86_64.run). No noise in the log and xscreensaver works again.

[Brian May (brian-microcomaustralia)]

Upgrading to version 8.5 does not help me. Nor can I check the length (or contents) of the log file, after the computer freezes I have to cold reset it (ctrl+alt+backspace doesn't do anything), and the log file becomes length 0.

[benoit16 (benoit-schmid)]

Good morning,

I had the same problem with hardy on my PC.
I did not have the proble with old Ubuntu releases.
I have changed the config:
Old xorg.conf:

Section "Device"
        Identifier "radeondvi"
        Driver "fglrx"
        Option "DesktopSetup" "horizontal"
        Option "OverlayOnCRTC2" "1"
        BusID "PCI:1:0:0"
EndSection

Section "Device"
        Identifier "radeonvga"
        Driver "fglrx"
        BusID "PCI:1:0:0"
EndSection

New xorg .conf
Section "Device"
        Identifier "radeondvi"
        Driver "fglrx"
        Option "DesktopSetup" "horizontal"
        Option "OverlayOnCRTC2" "1"
        BusID "PCI:1:0:0"
EndSection

Section "Device"
        Identifier "radeonvga"
        Driver "fglrx"
        BusID "PCI:1:0:1"
EndSection

The problem has disappeared when I set the two different PCI ids.
Unfortunately I used to have two different desktops (one per screen).
Now I have a single desktop with both screens.

I have to rtfm to reconfigure my two desktops in my xorg.conf.

See you.

[evan martin (m42martin)]

can someone please post how to specify the new 8.5 version using
apt-get?

I tried Envy - which sort of worked - the first time, though not well, and the ati GUI utility
that Envy installed was no help in adjusting configuration.
I had to dump it, which created more problems...

this is getting old, Ubuntu.
I love solving problems, but start to feel bad if I have to keep resolving the same one.

thanks.

[Jyri Sillanpaa (ubuntu-amorco)]

Evan,

I don't think that the driver has reached the repositories yet, you need to download the driver from ati but it is easy to install, just download the package, then do
chmod a+x ati-driver-installer-8-5-x86.x86_64.run
sudo ./ati-driver-installer-8-5-x86.x86_64.run

Just answer yes to everything and you are done.

[evan martin (m42martin)]

Thanks Jyri.

Just for ref: here's the link on amd's site: http://ati.amd.com/support/drivers/linux/linux-radeon.html
it took me much too long to locate that page.

[Bryce Harrington (bryce)]

[This is an automated message]

As of Intrepid (8.10), we have a dedicated package 'fglrx-installer' for fglrx bugs, which now includes a process for upstreaming bugs to AMD.

  http://bugs.launchpad.net/ubuntu/+source/fglrx-installer

To transition your bug into the new fglrx-package, we need your help. Please do the following:

 a. Verify the bug occurs in Intrepid.
     (Intrepid ISOs: http://cdimage.ubuntu.com)
 b. If you haven't already, please include in the bug:
     * Your /var/log/Xorg.0.log
     * The output of `lspci -vvnn`
     * Steps to reproduce the issue
 c. Click 'Also affects distribution'
 d. Set 'Source Package Name' to 'fglrx-installer'
 e. Click Continue

Thank you. This will assist us in reviewing and upstreaming your fglrx bug, as appropriate.

[We'll expire the fglrx bugs in l-r-m-* in a month or so.]

[Filippo Carletti (filippo-carletti)]

On Intrepid, I'm now using radeon driver and Virtualdesktop option in xorg.conf.
A big Virtualdesktop is more useful for me as I can move windows between monitors.
I briefly tested fglrx and it doesn't log errors.

Log noise in Hardy disappeared after an update (see my previous comment).

I'd close this bug now.

[Jonathan Thomas (echidnaman)]

Closing. Thanks for the prompt response.