Arch Linux

cantab

Member

Registered: 2010-07-17

Posts: 6

[SOLVED] System hangs at random, suspect X is the culprit

For a bit over a week now, my system has been hanging seemingly at random. When watching youtube videos, when web browsing, when watching stuff in mplayer, probably more times too. The only way to get back to normal is hitting the physical reset button.

Using KDE (normal, not mod) on Arch 64. nVidia GeForce 210 graphics card, have had hangs with both the nvidia and nouveau drivers. I spent a couple days with the nv driver and no hang, but that driver lacks features I need (specifically, tvtime won't work, and I need that for my tv card). (And anyway, it's not certain the lack of hanging wasn't just luck).

Behaviour is as follows:

NumLock responds.
Mouse pointer moves normally.
If Amarok is playing music, the playing song continues but the next one does not start.

No GUI elements respond.
CapsLock does not work.
Ctrl+Alt+F1 does not work.
Alt+SysRq+REISUB does not work.

Have not tested pinging or sshing into the hung system.
Have not yet tried using my motherboard's onboard ATI graphics instead of the nVidia card.
Have not yet tried a different desktop environment.

Probably ruled out as causes:
Cooling issues. I took the case cover off and added a second fan, temps were notably lower than usual, but still got a hang.
Being hacked. chkrootkit run from a livecd showed up nothing at least.
Memory issues. memtest86+ passed twice in a row.
Xinerama, I explicitly disabled it in xorg.conf

I get the following in my Xorg.0.log after it hangs

[  5656.221] [mi] EQ overflowing. The server is probably stuck in an infinite loop.
[  5656.243] 
Backtrace:
[  5656.293] 0: /usr/bin/X (xorg_backtrace+0x28) [0x498588]
[  5656.293] 1: /usr/bin/X (mieqEnqueue+0x1f4) [0x4952f4]
[  5656.293] 2: /usr/bin/X (xf86PostMotionEventP+0xc4) [0x476504]
[  5656.293] 3: /usr/lib/xorg/modules/input/evdev_drv.so (0x7eff0d86f000+0x3ccf) [0x7eff0d872ccf]
[  5656.293] 4: /usr/bin/X (0x400000+0x707d7) [0x4707d7]
[  5656.294] 5: /usr/bin/X (0x400000+0x106513) [0x506513]
[  5656.294] 6: /lib/libpthread.so.0 (0x7eff13482000+0xf1c0) [0x7eff134911c0]
[  5656.294] 7: /lib/libc.so.6 (ioctl+0x7) [0x7eff124be5b7]
[  5656.302] 8: /usr/lib/libdrm.so.2 (drmIoctl+0x28) [0x7eff0ed0c568]
[  5656.302] 9: /usr/lib/libdrm.so.2 (drmCommandWrite+0x1b) [0x7eff0ed0e82b]
[  5656.302] 10: /usr/lib/libdrm_nouveau.so.1 (0x7eff0e6ce000+0x2d8d) [0x7eff0e6d0d8d]
[  5656.302] 11: /usr/lib/libdrm_nouveau.so.1 (nouveau_bo_map_range+0xfe) [0x7eff0e6d139e]
[  5656.302] 12: /usr/lib/libdrm_nouveau.so.1 (0x7eff0e6ce000+0x1f1d) [0x7eff0e6cff1d]
[  5656.302] 13: /usr/lib/libdrm_nouveau.so.1 (nouveau_pushbuf_flush+0x1af) [0x7eff0e6d047f]
[  5656.302] 14: /usr/lib/xorg/modules/drivers/nouveau_drv.so (0x7eff0e8d3000+0x21b04) [0x7eff0e8f4b04]
[  5656.302] 15: /usr/lib/xorg/modules/libexa.so (0x7eff0e06f000+0xcc96) [0x7eff0e07bc96]
[  5656.302] 16: /usr/lib/xorg/modules/libexa.so (0x7eff0e06f000+0xdcba) [0x7eff0e07ccba]
[  5656.302] 17: /usr/bin/X (0x400000+0xc604b) [0x4c604b]
[  5656.302] 18: /usr/bin/X (0x400000+0xbe41e) [0x4be41e]
[  5656.302] 19: /usr/bin/X (0x400000+0x2804c) [0x42804c]
[  5656.302] 20: /usr/bin/X (0x400000+0x2173a) [0x42173a]
[  5656.302] 21: /lib/libc.so.6 (__libc_start_main+0xfd) [0x7eff12415c4d]
[  5656.302] 22: /usr/bin/X (0x400000+0x212f9) [0x4212f9]

How can I resolve this issue?

For reference, my Xorg.conf

# nvidia-settings: X configuration file generated by nvidia-settings
# nvidia-settings:  version 1.0  (buildmeister@builder58)  Fri Mar 12 02:13:46 PST 2010

# nvidia-xconfig: X configuration file generated by nvidia-xconfig
# nvidia-xconfig:  version 1.0  (buildmeister@builder58)  Fri Mar 12 02:12:40 PST 2010

Section "ServerLayout"
    Identifier     "Layout0"
    Screen      0  "Screen0" 0 0
    InputDevice    "Keyboard0" "CoreKeyboard"
    InputDevice    "Mouse0" "CorePointer"
    Option         "Xinerama" "0"
EndSection

Section "Files"
EndSection

Section "InputDevice"

    # generated from default
    Identifier     "Mouse0"
    Driver         "mouse"
    Option         "Protocol" "auto"
    Option         "Device" "/dev/psaux"
    Option         "Emulate3Buttons" "no"
    Option         "ZAxisMapping" "4 5"
EndSection

Section "InputDevice"

    # generated from default
    Identifier     "Keyboard0"
    Driver         "kbd"
EndSection

Section "Monitor"
    Identifier     "Monitor0"
    VendorName     "Unknown"
    ModelName      "Sony CPD-100ES"
    HorizSync       30.0 - 70.0
    VertRefresh     50.0 - 120.0
    Option         "DPMS"
EndSection

Section "Device"
    Identifier     "Device0"
    Driver         "nouveau"
    VendorName     "NVIDIA Corporation"
    BoardName      "GeForce 210"
EndSection

Section "Screen"
    Identifier     "Screen0"
    Device         "Device0"
    Monitor        "Monitor0"
    DefaultDepth    24
    Option         "TwinView" "0"
    Option         "metamodes" "1280x960 +0+0"
    SubSection     "Display"
        Depth       24
    EndSubSection
EndSection

Section "ServerFlags"
    Option         "Xinerama" "0"
EndSection

Last edited by cantab (2010-07-20 23:00:47)

cantab

Member

Registered: 2010-07-17

Posts: 6

Re: [SOLVED] System hangs at random, suspect X is the culprit

Well, it hung under the Linux Mint LiveCD too. So it's not an Arch specific or my setup specific issue. I've now pulled the nvidia card and switched to the onboard graphics. Still running off LiveCDs, will see if it crashes again, if not I think I can blame the nvidia card in some way.

valium97582

Member

Registered: 2010-06-19

Posts: 126

Re: [SOLVED] System hangs at random, suspect X is the culprit

I also had those problems on my Arch i686 Desktop, and very very probably (in my case) is that the memory is not safely placed. Much times it won't boot complaining with "beep-bip-bip-bip" and then going quiet. AFAIK those beeps mean (according to the manual of my BIOS) "no RAM disposable" or something of the sort, but I also think of overheating.

(Even if, after I installed Arch, I had only one hang in one month and a half.)

Try underclocking your computer, if it does not matter for you to run (slightly) slower, or checking heatsink and fans.

Last edited by valium97582 (2010-07-18 16:24:11)

---

I'm also known as zmv on IRC.

Kosmonavt

Member

Registered: 2010-02-15

Posts: 100

Re: [SOLVED] System hangs at random, suspect X is the culprit

Can confirm this bug on Arch x86_64 with Intel GM45 video card (driver xf86-video-intel-newest from AUR). Log message is quite the same. I'm using laptop HP 6530b and have no possibility to clock memory or whatever in BIOS config. The bug appeared some time after upgrade to Xorg 1.8.

Temperature is fine. E.g., CPU (Intel Core 2 Duo 2.4GHz) is the hottest - it's 45 degrees at average, fans are quite clean and functional. Compared to Debian testing install with Xorg 1.7 - no such behavior. Also no BIOS or kernel errors on start (dmesg and errors.log say nothing).

These hangs are quite rare for me, however, sometimes it's quite annoying. IMO, it's an upstream problem in new Xorg, not in drivers or hardware.

P.S. Sometimes Ctrl+Alt+F1 is accessible (probability is about 33%). Then killing and restarting X helps.

crabmarkjellozest

Member

Registered: 2010-07-09

Posts: 20

Re: [SOLVED] System hangs at random, suspect X is the culprit

This is not a bug with just Arch, a cursory search yields:
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=578056
https://bugzilla.novell.com/show_bug.cgi?id=597078
https://bugzilla.redhat.com/show_bug.cgi?id=465884
https://qa.mandriva.com/show_bug.cgi?id=42833

The problem with diagnosing it is that the "EQ Overflowing" bit is completely ambiguous.  It seems to affect both Intel and Nvidia users so it's possible that the bug is actually not in the driver code.  Sorry I can't really be of any substantial help.

Last edited by crabmarkjellozest (2010-07-18 19:56:20)

mutlu_inek

Member

From: all over the place

Registered: 2006-11-18

Posts: 683

Re: [SOLVED] System hangs at random, suspect X is the culprit

These are X related buffer overflows. Please file bug reports on bugs.freedesktop.org. And please file separate report, since most of the time the bugs are driver (i.e. hardware) specific.

cantab

Member

Registered: 2010-07-17

Posts: 6

Re: [SOLVED] System hangs at random, suspect X is the culprit

Compared to Debian testing install with Xorg 1.7 - no such behavior.

Might it be worth me trying to rollback to Xorg 1.7?

mutlu_inek

Member

From: all over the place

Registered: 2006-11-18

Posts: 683

Re: [SOLVED] System hangs at random, suspect X is the culprit

Compared to Debian testing install with Xorg 1.7 - no such behavior.

Might it be worth me trying to rollback to Xorg 1.7?

Depends on your luck. I had a number of different EQ overflows over the past maybe 1.5 years. For me, it finally works. Might be different for you, though.

Oh, and btw., it is getting better. Linus even fixed that annoying freeze that sometimes happened when resuming (from ram): http://repo.or.cz/w/linux-2.6.git/commi … 2d05d27d86

Last edited by mutlu_inek (2010-07-19 23:26:45)

Kosmonavt

Member

Registered: 2010-02-15

Posts: 100

Re: [SOLVED] System hangs at random, suspect X is the culprit

Compared to Debian testing install with Xorg 1.7 - no such behavior.

Might it be worth me trying to rollback to Xorg 1.7?

Basically it may help. But there could be some issues with latest KDE (or whatever DE) interaction with Xorg, so the problem may repeat.

I haven't tried downgrading but looks like (it may be only local, though) when a GTK+ application is run for a long time over the latest KDE (e.g., I use Audacious music player a lot), probability of such thing is higher.

cantab

Member

Registered: 2010-07-17

Posts: 6

Re: [SOLVED] System hangs at random, suspect X is the culprit

Well, I downgraded and am now posting from Arch again. No hangups...yet.

The procedure was roughly as follows. Note that I also had to redo my X setup to use the ATI graphics.

Boot from a LiveCD.
Change Arch's /etc/pacman.conf to use the Arch Rollback Machine. The 20th June is the last date to have Xorg 1.7, though I used the 19th.
Change Arch's /etc/inittab to boot to runlevel 3.
If you normally use the network manager system tray applet, download the package for cnetworkmanager. I could NOT get my wireless working using iwconfig and ifconfig for some reason.
Reboot into Arch.
If needed, install cnetworkmanager and then use it to connect to the internet.
pacman -Syy, pacman -Suuf
edit xorg.conf if needed
init 5 (or maybe startx, but that fails if KMS is in use) to check things work
set inittab back to boot into runlevel 5

cantab

Member

Registered: 2010-07-17

Posts: 6

Re: [SOLVED] System hangs at random, suspect X is the culprit

One final note - to upgrade again after this, have this in pacman.conf

IgnorePkg = xorg-server
IgnoreGroup = xorg xorg-input-drivers xorg-video-drivers

It may be possible to just rollback those groups, and not everything else (firefox, KDE, and so on and so on). I'm getting an impression Xorg 1.8 is giving a lot of people various problems, so maybe we should look into a Wiki article or forum sticky on reverting to and staying with Xorg 1.7?

Last edited by cantab (2010-07-24 20:37:48)