Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 433742 - x11-drivers/xf86-video-nv makes X segfault after upgrading to x11-base/xorg-server-1.12.2
Summary: x11-drivers/xf86-video-nv makes X segfault after upgrading to x11-base/xorg-s...
Status: RESOLVED FIXED
Alias: None
Product: Gentoo Linux
Classification: Unclassified
Component: Current packages (show other bugs)
Hardware: All Linux
: Normal normal (vote)
Assignee: Gentoo X packagers
URL:
Whiteboard:
Keywords: REGRESSION
Depends on:
Blocks:
 
Reported: 2012-09-03 07:12 UTC by Paweł Hajdan, Jr. (RETIRED)
Modified: 2012-09-03 12:06 UTC (History)
0 users

See Also:
Package list:
Runtime testing required: ---


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Paweł Hajdan, Jr. (RETIRED) gentoo-dev 2012-09-03 07:12:23 UTC
I'm using xf86-video-nv on a stable amd64 system, and after an X11 update I started getting 100% reproducible crashes on X startup:

[ 33896.413] Backtrace:
[ 33896.413] 0: /usr/bin/Xorg (xorg_backtrace+0x36) [0x568536]
[ 33896.413] 1: /usr/bin/Xorg (0x400000+0x16c1d9) [0x56c1d9]
[ 33896.413] 2: /lib64/libpthread.so.0 (0x7f0b27d72000+0x10280) [0x7f0b27d82280]
[ 33896.413] 
[ 33896.413] Segmentation fault at address (nil)
[ 33896.413] 
Fatal server error:
[ 33896.413] Caught signal 11 (Segmentation fault). Server aborting

With this configuration:

Section "Device"
        Identifier "Nvidia"
        Driver "nv"
EndSection

Backporting upstream commit http://cgit.freedesktop.org/xorg/driver/xf86-video-nv/commit/?id=b3d61828837661c9b383378a7d6006752659108b to xf86-video-nv-2.1.18 or just upgrading to 2.1.20 reliably fixes the problem. Please get the relevant fixes to stable.

My previous version of xorg-server was x11-base/xorg-server-1.11.2-r2. Also see similar RedHat bug https://bugzilla.redhat.com/show_bug.cgi?id=782995 which fixes a problem in a different driver in a similar way.

gdb backtrace I obtained with the following:

gdb.txt:
file /usr/bin/Xorg
set confirm off
handle SIGUSR1 nostop
handle SIGUSR2 nostop
handle SIGPIPE nostop
run
bt full
cont
quit

gdb --quiet --command=gdb.txt 2>&1 >gdb.out

Program received signal SIGSEGV, Segmentation fault.
0x0000000000000000 in ?? ()
#0  0x0000000000000000 in ?? ()
No symbol table info available.
#1  0x00007f0b24a0b27a in vgaHWSaveColormap (scrninfp=0x21a3b50, save=0x21a5b88) at /var/tmp/portage/x1
1-base/xorg-server-1.12.2/work/xorg-server-1.12.2/hw/xfree86/vgahw/vgaHW.c:1124
        hwp = 0x21a5b40
        readError = 0
        i = <optimized out>
#2  0x00007f0b24a0c93d in vgaHWSave (scrninfp=0x21a3b50, save=0x21a5b88, flags=7) at /var/tmp/portage/x
11-base/xorg-server-1.12.2/work/xorg-server-1.12.2/hw/xfree86/vgahw/vgaHW.c:1155
No locals.
#3  0x00007f0b24e3c0c0 in NVDACSave () from /usr/lib64/xorg/modules/drivers/nv_drv.so
No symbol table info available.
#4  0x00007f0b24e3deb3 in NVScreenInit () from /usr/lib64/xorg/modules/drivers/nv_drv.so
No symbol table info available.
#5  0x0000000000436071 in AddScreen (pfnInit=<optimized out>, argc=1, argv=0x7fff1b0f3258) at /var/tmp/portage/x11-base/xorg-server-1.12.2/work/xorg-server-1.12.2/dix/dispatch.c:3797
Comment 1 Matt Turner gentoo-dev 2012-09-03 08:22:37 UTC
Why are you using -nv instead of nouveau or the proprietary driver?
Comment 2 Paweł Hajdan, Jr. (RETIRED) gentoo-dev 2012-09-03 11:45:40 UTC
(In reply to comment #1)
> Why are you using -nv instead of nouveau or the proprietary driver?

It's an old install, and -nv was just working fine. I don't like nvidia's binary blob, and nouveau seems quite experimental to me.

Anyway, I'm just reporting a bug so it can be fixed.
Comment 3 Chí-Thanh Christopher Nguyễn gentoo-dev 2012-09-03 12:06:19 UTC
I am marking this as fixed as 2.1.20 is in the tree. It will go stable in the next x11 stabilization round.