| Summary: | xscreensaver gl crashes with nvidia on 2.4.x | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | david walker <azrail> |
| Component: | Current packages | Assignee: | Jonathan Smith (RETIRED) <smithj> |
| Status: | RESOLVED NEEDINFO | ||
| Severity: | major | CC: | azrail, desktop-misc, gehn, konstantin.sobolev, liquidx, pYrania, x11 |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
|
Description
david walker
2004-01-10 21:08:49 UTC
could you provide us with some kernel log right before it crashes? does xscreensaver dump core to anywhere? it sounds like its not dumping core but spinning, when it happens are you able to jump back to console? like CTRL+ALT+F1 ? or ssh into the box? No the system is totaly down, there is no abnormal kernel log (or panics) and there is no core (at least what I can find. I experienced the same problem with OpenGL screensavers and nvidia 1.0.5328 buggy drivers... Yea im using a GL lock, and i have a geforcefx 5200. my friend uses the gl with a geforce4 and has this problem from time to time, so that could very well be the case. well, don't know how to go about this, i'm tempted to mark this as upstream, especially since nvidia kernel is quite a complicated beast, and i don't have a nvidia card. Well, last check, do you have a VIA chipset on your motherboard... if that's the case nvidia drivers are the problem, a patch was issued (see bug #36742) for the initial problem, but bringing kernel freezes (at least for configuration). I stepped back to previous nvidia packages to solve the problem. I hope this will help... no my mb is not via. i already saw that bug. i am now running the old nvidia-kernel, and will check if it also borks out. so far it hasent. ok scratch my last comment, it borked again this afternoon. Im wondering if there is a log (from sysklogd) or some log i can set up to log this? There also seem to be no real cause, i mean like i have xmms running or it OOo. I seem to encounter this bug too. After about 1 hour of xscreensaver running the system crashes. The 'caps lock' and 'roll lock' LED are blinking at ca 2 Hz, the NIC cant be pinged but sshd does not respond. Config is Athlon XP 2000 with via chipset, VGA is nvidia GeForce4 MX440. I'm going to check for nvidia-kernel bugs this doesn't appear on 4496 drivers for me, but for any newer ones system crashes rather quickly when running xscreensaver this is totally out of my league. i don't have a nvidia card or able to reproduce it. i think it is a nvidia driver problem more than a problem with xscreensaver. Does this happen using the non-accelerated driver in XFree? nv? I have no idea what caused it, sice this posting i have migrated to the exp 2004.0 beta stage 1's and have reinstalled and am not getting any xscreensaver lock ups. So I have also but no idea. I have the same problem (system locks during some GL screensavers and I am using the tdfx module as I have a voodoo3. Time to rethink the NVidia blame game. well, xfree crashes usually is because of the driver. can everyone who is reporting this post their version of xfree, gl implementation and emerge info? I am having this problem using NVidia OpenGL 1.0.5332-r1 on a MSI FX5200: % X -version This is a pre-release version of XFree86, and is not supported in any way. Bugs may be reported to XFree86@XFree86.Org and patches submitted to fixes@XFree86.Org. Before reporting bugs in pre-release versions, please check the latest version in the XFree86 CVS repository (http://www.XFree86.Org/cvs). XFree86 Version 4.3.0.1 Release Date: 15 August 2003 X Protocol Version 11, Revision 0, Release 6.6 Build Operating System: Linux 2.6.3-rc2-gentoo x86_64 [ELF] Build Date: 16 February 2004 Before reporting problems, check http://www.XFree86.Org/ to make sure that you have the latest version. Module Loader present % emerge info Portage 2.0.50-r1 (default-amd64-2004.0, gcc-3.3.3, glibc-2.3.3_pre20040207-r0, 2.6.3-rc2-gentoo) ================================================================= System uname: 2.6.3-rc2-gentoo x86_64 4 Gentoo Base System version 1.4.3.13p1 Autoconf: sys-devel/autoconf-2.59-r3 Automake: sys-devel/automake-1.8.2 ACCEPT_KEYWORDS="amd64 ~amd64" AUTOCLEAN="yes" CFLAGS="-O2 -fPIC" CHOST="x86_64-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3.2/share/config /usr/kde/3/share/config /usr/lib/mozilla/defaults/pref /usr/share/config /usr/share/texmf/dvipdfm/config/ /usr/share/texmf/dvips/config/ /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /usr/share/texmf/xdvi/ /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/env.d" CXXFLAGS="-O2 -fPIC" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs ccache sandbox" GENTOO_MIRRORS="ftp://gentoo.ccccom.com http://mirrors.tds.net/gentoo http://mirror.tucdemonic.org/gentoo/" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="" SYNC="rsync://rsync.namerica.gentoo.org/gentoo-portage" USE="X aalib acpi alsa amd amd64 apache2 arts artswrappersuid avi berkdb bonobo cdr cgi chroot crypt cscope cups dillo dvd dvdr emacs encode esd ethereal evo faad flac foomaticdb freetype fs gb gdbm gif gimpprint gnome gpm gstreamer gtk gtk2 idea imagemagick imlib java javascript jpeg kde libg++ libwww mad maildir mbox mikmod mldonkeypango motif mozcalendar mozilla mpeg mpeg4 mysql ncurses nls offensive oggvorbis opengl openssh oss pam parse-clocks pda pdflib perl php plotutils png ppds python qt quicktime readline samba sdl slang spell ssl tcltk tcpd tetex truetype xine xml2 xmms xv zlib" I seem to get this problem, with a Radeon 9500 Pro, ATI (fglrx) drivers. I'm using XScreenSaver 4.15, and this mainly seems to happen with the Really Slick Screensavers hacks (I included the --nice switch) When I return to the computer, either the image is frozen on the monitor, or it is simply black. Keyboard won't respond - can't ctrl-alt-backspace or ctrl-alt-F1. *lock keys won't respond. I can't find *anything* untoward in either the XF86 or System logs at the time of the crash. Incidentally, I'm using RedHat Linux - not gentoo. This is probably an nvidia driver bug. Please file a different one for ati-drivers. When i still used gentoo i found that installing off of the 2004.1 Live CD and using the 2004 stage one it did not have that problem. I have since switched to debian (could not stand all that compiling for everything) and still do not have that problem. I too am now receiving this error, seemingly since I've updated to nividia-kernel-1.0.5336-r4 and nvidia-glx-1.0.5336-r2 just recently. The computer will hang: ie, the computer won't respond to the keyboard nor mouse and cannot be pinged. The numlock, capslock, and scrolllock LEDs blink at about twice per second. Same symptoms as posted above. nforce2 AN35N-ultra Shuttle Motherboard, with onboard NIC Verto PNY GeForce FX 5200 AGP ___________ $X -version This is a pre-release version of XFree86, and is not supported in any way. Bugs may be reported to XFree86@XFree86.Org and patches submitted to fixes@XFree86.Org. Before reporting bugs in pre-release versions, please check the latest version in the XFree86 CVS repository (http://www.XFree86.Org/cvs). XFree86 Version 4.3.0.1 Release Date: 15 August 2003 X Protocol Version 11, Revision 0, Release 6.6 Build Operating System: Linux 2.4.25-gentoo-r4 i686 [ELF] Build Date: 04 July 2004 Before reporting problems, check http://www.XFree86.Org/ to make sure that you have the latest version. Module Loader present ____________ $emerge info Gentoo Base System version 1.4.16 Portage 2.0.50-r9 (default-x86-2004.0, gcc-3.3.3, glibc-2.3.3.20040420-r1, 2.4.26-gentoo-r9) ================================================================= System uname: 2.4.26-gentoo-r9 i686 AMD Athlon(tm) Autoconf: sys-devel/autoconf-2.59-r4 Automake: sys-devel/automake-1.8.3 ACCEPT_KEYWORDS="x86" AUTOCLEAN="yes" CFLAGS="-O2 -march=athlon-mp -pipe -fomit-frame-pointer" CHOST="i686-pc-linux-gnu" COMPILER="gcc3" CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3.2/share/config /usr/kde/3/share/config /usr/share/config /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d" CXXFLAGS="-O2 -march=athlon-mp -pipe -fomit-frame-pointer" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs ccache sandbox" GENTOO_MIRRORS="http://mirror.tucdemonic.org/gentoo/ ftp://ftp.ndlug.nd.edu/pub/gentoo/ ftp://gentoo.ccccom.com http://gentoo.mirrors.pair.com/ ftp://gentoo.mirrors.pair.com/" MAKEOPTS="-j2" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="3dnow X aalib aim alsa apache2 apm arts avi berkdb cdr crypt cups dga doc encode esd flash foomaticdb gd gdbm gif gphoto2 gpm gtk2 guile imagemagick imlib java jpeg kde libg++ libwww mad mikmod mmx motif mozilla mpeg mysql nas ncurses nls offensive oggvorbis opengl oscar oss pam pdflib perl png python qt quicktime readline sdl slang spell sse ssl svga tcltk tcpd tiff truetype unicode usb videos x86 xml xml2 xmms xv zlib" switched to xorg-x11, but still have the same problem. i noticed that the computer will also encounter the same freeze/hang upon exit of quake3 and glxgears. I'm having the same problems.. any GL (not just xscreensaver) will freeze the box after a random amount of time.
Nvidia versions: 5328, 5336, 6111
X -version:
XFree86 Version 4.3.0.1
Release Date: 15 August 2003
X Protocol Version 11, Revision 0, Release 6.6
Build Operating System: Linux 2.4.26-gentoo-r9 i686 [ELF]
Build Date: 19 August 2004
Before reporting problems, check http://www.XFree86.Org/
to make sure that you have the latest version.
Module Loader present
emerge info:
Portage 2.0.50-r9 (default-x86-2004.0, gcc-3.3.3, glibc-2.3.3.20040420-r1, 2.4.26-gentoo-r9)
=================================================================
System uname: 2.4.26-gentoo-r9 i686 AMD Athlon(tm) MP 2400+
Gentoo Base System version 1.4.16
Autoconf: sys-devel/autoconf-2.59-r4
Automake: sys-devel/automake-1.8.3
ACCEPT_KEYWORDS="x86"
AUTOCLEAN="yes"
CFLAGS="-O2 -march=athlon-xp -fomit-frame-pointer"
CHOST="i686-pc-linux-gnu"
COMPILER="gcc3"
CONFIG_PROTECT="/etc /usr/X11R6/lib/X11/xkb /usr/kde/2/share/config /usr/kde/3/share/config /usr/lib/mozilla/defaults/pref /usr/share/config /var/qmail/control"CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d"
CXXFLAGS="-O2 -march=athlon-xp -fomit-frame-pointer"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoaddcvs ccache sandbox"
GENTOO_MIRRORS="http://gentoo.binarycompass.org http://mirror.datapipe.net/gentoo"
MAKEOPTS="-j3"
PKGDIR="/usr/portage/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
PORTDIR_OVERLAY=""
SYNC="rsync://rsync.gentoo.org/gentoo-portage"
USE="X acpi adns aim alsa apache2 avi berkdb bzlib calendar cdr crypt curl divx4linux doc encode exif flac flash foomaticdb ftp gdbm gif gnome gphoto2 gtk gtk2 gtkhtml imagemagick imap imlib jabber java joystick jpeg junit libg++ libwww mad mbox mikmod mime mmx mozilla mpeg mpi ncurses nls nocd offensive oggvorbis openal opengl oscar pam pcre pdflib perl pie png posix postgres python quicktime readline samba sdl shared slang soap sockets spell ssl svga tcpd tiff truetype unicode usb videos vim-with-x wxwindows x86 xml xml2 xmlrpc xmms xpm xv zlib"
OK, I've found out that the crash (actually a hard lock) is a gentoo-sources kernel panic regarding preemption and APM. I had reverted back to a lock-free kernel config I had a long time ago and slowly modified that config until I encountered the locking problem. The routine to check if the kernel config was problematic was: 1. Add an option to the kernel with "make menuconfig", compile, emerge modules, and reboot 2. Start up X with DPMS enabled and using "nvAGP 1" to use the nvidia driver module. Also, I set standbytime, suspendtime, and offtime to 2, 3, and 4, respectively, for this testing. 3. Start up glxgears and let it run for 3 or more minutes. 4. Press the "escape" key to exit out of glxgears. 5. Repeat steps 3 and 4 several times until the system locks (no keyboard or mouse input, cannot ctrl+alt+backspace out of X, cannot ctrl+alt+delete to reboot, a hard reboot is necessary) or until I think the config is safe so far. With this routine, I found out that the PREEMPTIVE KERNEL option caused the system to lock up. So, I rebooted, started X and glxgears again; this time, instead of exiting glxgears, I pressed ctrl+alt+backspace to exit out of X while glxgears was running. I received this output before another lock occurs: xinit: connection to X server lost kernel BUG at exit.c: 524! invalid operand: 000 CPU: 0 EIP 0010: [<c01af491>] Tainted: P EFLAGS: 00010292 ... <other output> ... <0>Kernel panic: Attempted to kill the idle task! In idle task - not syncing I believe bug #62343 is talking about the same problem discussed here because the Greg Tassone's kernel config has CONFIG_PREEMPT set to "y" as well; that is, he has the preemptive kernel option set. He does not experience the system hang with the vanilla sources because a preemptive kernel is not an option there. Here's a link that discusses this problem for posterity: http://forums.gentoo.org/viewtopic.php?p=1595798 That thread discusses it, but offers no real solution. Needless to say, I am having the same problems -- on 3 different machines now. Anyone know how to the Nvidia development team? Because I think all we can do is wait. (viva la closed-source). sorry, that was "how to contact"... not "how to". david, what's the latest on this? Re comment 14. I have also now switched to Ubuntu (could not stand the compiling anymore) and have no problem. I assume it was a buggy ACPI, or Xscreensaver build. But I do not know the code of the nVidia driver, the kernel, or Xscreensaver well enough to have an answer. Since I was/am not the only one to have/had the problem I would get in contact with nVidia (for whatever that may help) to see if there are updates or known internal issues with it. i'm taking over xscreensaver with permission from liquidx does this problem still exist on 4.22-r1? repoen if/when you ever notice this and can test some things |
