X locks up after a random time

Mon Jan 5 09:31:32 UTC 2009

On Mon, Jan 5, 2009 at 2:06 AM, sangu <sangu.fedora at gmail.com> wrote:

> See Als :
> https://bugzilla.redhat.com/show_bug.cgi?id=464866 : Xorg lockup with
> i945: "EQ overflowing. The server is probably stuck in an infinite
> loop."
>
> https://bugs.freedesktop.org/show_bug.cgi?id=14633 X freezes with
> Composite and Mouse Wheel, mouse moves, no... (with Nvidia Proprietary
> Driver)
>
> 2009-01-03 (토), 01:36 +0100, Marko Vojinovic 쓰시길:
> > Ok, this is a Fujitsu Siemens Esprimo U9200, with Intel graphics, F10
> 64bit,
> > updated last week. KDE desktop, compiz and all, but I believe it is not
> > related.
> >
> > The symptom: system boots up regularly, I do regular work, and after a
> random
> > amount of time, X freezes. If compiz is on, all I can do is move the
> mouse
> > pointer. If compiz is off, even the mouse pointer is frozen. The OS
> itself is
> > not, I can ssh from a remote machine and read logs. This is not
> reproducible,
> > it happens too randomly and I can only sit and wait for it to happen.
> > However, it is extremely annoying, since it has a tendency to freeze when
> I'm
> > in the middle of some work in progress.
> >
> > If I log in remotely and kill -9 the X process, the whole machine locks
> up and
> > I can only power-cycle it. Don't know a way to restart X short of
> rebooting,
> > keyboard doesn't respond to ctrl-alt-backspace or anything else.
> >
> > So far, dmesg and /var/log/messages do not show anything related or even
> > suspicious (afaics), while /var/log/Xorg.0.log shows this at the end:
> >
> > =================
> > [snip most of the regular log of a working X server]
> >
> > (II) intel(0): EDID for output TMDS-1
> > (II) intel(0): EDID for output TV
> > [mi] EQ overflowing. The server is probably stuck in an infinite loop.
> >
> > Backtrace:
> > 0: /usr/bin/X(xorg_backtrace+0x26) [0x4e7a26]
> > 1: /usr/bin/X(mieqEnqueue+0x291) [0x4c8591]
> > 2: /usr/bin/X(xf86PostMotionEventP+0xc4) [0x491494]
> > 3: /usr/bin/X(xf86PostMotionEvent+0xa9) [0x491669]
> > 4: /usr/lib64/xorg/modules/input//evdev_drv.so [0x7feeeaf3e126]
> > 5: /usr/bin/X [0x47a765]
> > 6: /usr/bin/X [0x46b307]
> > 7: /lib64/libc.so.6 [0x3522232f90]
> > 8: /lib64/libc.so.6(ioctl+0x7) [0x35222de037]
> > 9: /usr/lib64/libdrm.so.2 [0x7feefd045023]
> > 10: /usr/lib64/libdrm.so.2(drmCommandNone+0x16) [0x7feefd045326]
> > 11: /usr/lib64/xorg/modules/drivers//intel_drv.so [0x7feefcdd0fc8]
> > 12: /usr/bin/X [0x51cdb9]
> > 13: /usr/bin/X [0x4f7311]
> > 14: /usr/bin/X(BlockHandler+0x50) [0x44a320]
> > 15: /usr/bin/X(WaitForSomething+0x161) [0x4e4eb1]
> > 16: /usr/bin/X(Dispatch+0x7f) [0x4465ef]
> > 17: /usr/bin/X(main+0x45d) [0x42cd1d]
> > 18: /lib64/libc.so.6(__libc_start_main+0xe6) [0x352221e576]
> > 19: /usr/bin/X [0x42c0f9]
> > [mi] mieqEnequeue: out-of-order valuator event; dropping.
> > [mi] EQ overflowing. The server is probably stuck in an infinite loop.
> >
> > [the last two lines repeat a lot of times]
> > =================
> >
> > So if the server is really in an infinite loop, then something so very
> Not
> > Good is happening. Now, given that these are Intel open-source drivers
> > driving my graphics card here, I have the following questions:
> >
> > 1) What does the [mi] stand for here? Can anyone decypher the backtrace
> for
> > me? (I don't mind reading a "too technical" answer)
> > 2) How can I further diagnose the problem and help fixing it?
> > 3) In order to be efficient in diagnosing, does anyone know how to
> trigger the
> > lockup on purpose?
> > 4) This has been happening since I installed F10 on this machine --- has
> > anyone spotted the problem already? Is there a known bugzilla report that
> I
> > can add to?
> > 5) Is there a workaround until a fix comes?
> > 6) Can I somehow restart the X server from a remote console (without
> rebooting
> > the whole machine) to recover clean from this?
> > 7) F10 is out for a while now, how come a bug such as this (X in infinite
> > loop) has not been fixed already? I mean, many people here bragg about
> the
> > open source drivers being better then closed source when it comes to
> fixing
> > broken things, so why is an issue such as this still present?
> > // ok, this is a rant, I know, but I really am frustrated by this //
> >
> > Oh yes, lspci -v says this:
> >
> > 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960
> > Integrated Graphics Controller (rev 03) (prog-if 00 [VGA controller])
> >         Subsystem: Fujitsu Siemens Computers Device 110f
> >         Flags: bus master, fast devsel, latency 0, IRQ 16
> >         Memory at fc000000 (64-bit, non-prefetchable) [size=1M]
> >         Memory at d0000000 (64-bit, prefetchable) [size=256M]
> >         I/O ports at 18f8 [size=8]
> >         Capabilities: [90] Message Signalled Interrupts: Mask- 64bit-
> > Count=1/1 Enable-
> >         Capabilities: [d0] Power Management version 3
> >         Kernel modules: intelfb
> >
> > 00:02.1 Display controller: Intel Corporation Mobile GM965/GL960
> Integrated
> > Graphics Controller (rev 03)
> >         Subsystem: Fujitsu Siemens Computers Device 110f
> >         Flags: bus master, fast devsel, latency 0
> >         Memory at fc100000 (64-bit, non-prefetchable) [disabled]
> [size=1M]
> >         Capabilities: [d0] Power Management version 3
> >
> > There is no /etc/X11/xorg.conf file, default stuff...
> >
> > Help please?
>

Impressive. It seems that everybody knows about this problem since before
the F10 release. Therefore, it is a plain case of "broken edge".

-- 
Paulo Roma Cavalcanti
LCG - UFRJ
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/fedora-list/attachments/20090105/b4c06864/attachment-0001.htm>