Bug 195079 - x11/nvidia-driver / x11-servers/xorg-server segfault on CURRENT
Summary: x11/nvidia-driver / x11-servers/xorg-server segfault on CURRENT
Status: Closed FIXED
Alias: None
Product: Ports & Packages
Classification: Unclassified
Component: Individual Port(s) (show other bugs)
Version: Latest
Hardware: Any Any
: --- Affects Only Me
Assignee: Alexey Dokuchaev
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-11-16 20:05 UTC by Kevin Bowling
Modified: 2014-12-06 08:24 UTC (History)
2 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Kevin Bowling freebsd_committer freebsd_triage 2014-11-16 20:05:06 UTC
I run 11-CURRENT (r274577 right now) with nvidia-driver.  It was very stable until about a month ago.  I was running the 340-series driver for a couple months before it was introduced into ports, so I don't think that was it.  I'm not sure if it's dri/mesa or something changing in -CURRENT.

I can reliably trigger this segfault immediately when logging in with KDE's kwin set to use OpenGL 2 or 3 for desktop effects.  The system is stable in "failsafe mode" that does not use desktop effects.  GL seems to function outside of this, testing with WebGL in Firefox and glxgears.

I've tried upgrading to nvidia-driver-346.16 and xorg-server-1.14.4 out of tree but see the same stacktrace.

Xorg.0.log:
[  9307.793] Segmentation fault at address 0x0
[  9307.793] 
Fatal server error:
[  9307.793] Caught signal 11 (Segmentation fault). Server aborting

Stack:
(gdb) bt
#0 0x0000000802059c9a in thr_kill () from /lib/libc.so.7
#1 0x0000000802130ef9 in abort () from /lib/libc.so.7
#2 0x000000000062e4be in OsAbort () at utils.c:1198
#3 0x00000000004a6880 in ddxGiveUp (error=EXIT_ERR_ABORT) at xf86Init.c:1009
#4 0x00000000004a6998 in AbortDDX (error=EXIT_ERR_ABORT) at xf86Init.c:1053
#5 0x0000000000637be6 in AbortServer () at log.c:476
#6 0x00000000006368fd in FatalError (
f=0x659496 "Caught signal %d (%s). Server aborting\n") at log.c:611
#7 0x000000000062b9d8 in OsSigHandler (signo=11, sip=0x7fffffffd2f0,
unused=0x7fffffffcf80) at osinit.c:146
#8 0x00000008010b84f6 in swapcontext () from /lib/libthr.so.3
#9 0x00000008010b80ec in sigaction () from /lib/libthr.so.3
#10 <signal handler called>
#11 0x00000008084af6a6 in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#12 0x000000080848dcc9 in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#13 0x0000000808143e5f in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#14 0x00000008081456f6 in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#15 0x0000000808146676 in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#16 0x000000080814810f in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#17 0x000000080813e1d2 in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#18 0x0000000808148331 in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#19 0x000000080814a2cb in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#20 0x000000080811fbd2 in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#21 0x00000008084783cd in nvidiaAddDrawableHandler ()
from /usr/local/lib/xorg/modules/drivers/nvidia_drv.so
#22 0x0000000803d7872b in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#23 0x0000000803c95247 in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#24 0x0000000803c95d58 in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#25 0x0000000803d7e185 in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#26 0x0000000805e51d9f in _nv013glcore () from /usr/local/lib/libnvidia-glcore.so.1
#27 0x0000000803c92aa5 in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#28 0x0000000803d78e2d in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#29 0x0000000805e533e4 in _nv013glcore () from /usr/local/lib/libnvidia-glcore.so.1
#30 0x0000000805e59dfa in _nv013glcore () from /usr/local/lib/libnvidia-glcore.so.1
#31 0x0000000803d7ba25 in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#32 0x0000000803d7792e in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#33 0x0000000803d7ac37 in .text ()
from /usr/local/lib/xorg/modules/extensions/libglx.so
#34 0x00000000004368b8 in Dispatch () at dispatch.c:429
#35 0x0000000000424f72 in main (argc=9, argv=0x7fffffffecc8, envp=0x7fffffffed18)
at main.c:288
Comment 1 Bugzilla Automation freebsd_committer freebsd_triage 2014-11-16 20:05:06 UTC
Maintainers CC'd
Comment 2 Alexey Dokuchaev freebsd_committer freebsd_triage 2014-11-17 02:56:31 UTC
> I'm not sure if it's dri/mesa or something changing in -CURRENT.

That is, KDE4 bits were not upgraded during this timeframe?  Does the problem occur in KDE4-free environment (you might also want to try some games in addition to WebGL and glxgears)?

Can you also try to build/install their next version (not in the ports yet):

  make DISTVERSION=340.58 -DNO_CHECKSUM install

and tell how/if it changes anything?

There was another report recently on x11@ list about nvidia-driver-340.46 causing kernel panics on 10.1-RELEASE/amd64, also with KDE4; this might or might not be related: http://lists.freebsd.org/pipermail/freebsd-x11/2014-November/015577.html
Comment 3 Kevin Bowling freebsd_committer freebsd_triage 2014-11-17 07:28:53 UTC
Same behavior with 340.58 and 346.16.

I'm not seeing any panics.

I think kwin is pretty stagnate code wise (http://www.freshports.org/x11/kde4-workspace/) but I will try reverting kdeworkspace.
Comment 4 Mark Linimon freebsd_committer freebsd_triage 2014-11-23 07:37:17 UTC
Over to maintainer.
Comment 5 Kevin Bowling freebsd_committer freebsd_triage 2014-12-02 05:56:25 UTC
r273250 in base -CURRENT is at fault here.  Thanks to bisecting by Peter Wemm!
Comment 6 Kevin Bowling freebsd_committer freebsd_triage 2014-12-06 08:24:29 UTC
This has been fixed in HEAD by r275514