There appear to be some memory limits either in the kernel or in the nvidia-kernel driver. The entry below comes in the system log, and OpenGL applications start to behave quite strange afterwards. One example of strange in this context is to mirror the coordinate system. System log: Mar 3 14:06:28 dina Kernel BUG at pageattr:106 Mar 3 14:06:28 dina invalid operand: 0000 [1] Mar 3 14:06:28 dina CPU 0 Mar 3 14:06:28 dina Pid: 6082, comm: holoSeis.Linux_ Tainted: P Mar 3 14:06:28 dina RIP: 0010:[<ffffffff8011f4b5>] <ffffffff8011f4b5>{revert_page+85} Mar 3 14:06:28 dina RSP: 0018:000001003df05460 EFLAGS: 00010282 Mar 3 14:06:28 dina RAX: 0000000000000090 RBX: 0000010001005190 RCX: 0000010000105090 Mar 3 14:06:28 dina RDX: 0000010000103000 RSI: ffffffff8247d000 RDI: 0000010000000000 Mar 3 14:06:28 dina RBP: ffffffff8247d000 R08: 0000010000000000 R09: 8000000000000163 Mar 3 14:06:28 dina R10: 00000000000000cb R11: 0000000000000000 R12: 8000000000000163 Mar 3 14:06:28 dina R13: 0000010000105090 R14: 0000000000000001 R15: cccccccccccccccd Mar 3 14:06:28 dina FS: 0000002ab2687a40(0000) GS:ffffffff80431080(0000) knlGS:0000000000000000 Mar 3 14:06:28 dina CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 Mar 3 14:06:28 dina CR2: 0000002a95aec6b0 CR3: 0000000000101000 CR4: 00000000000006e0 Mar 3 14:06:28 dina Process holoSeis.Linux_ (pid: 6082, stackpage=1000097ea90) Mar 3 14:06:28 dina Stack: ffffffff8011f805 0000010001c5a9e0 00000100010b6710 0000000000000000 Mar 3 14:06:28 dina 0000000000000000 8000000000000163 ffffffff8011f8f1 0000000000000093 Mar 3 14:06:28 dina 00000100010b6710 000001000247d000 Mar 3 14:06:28 dina Call Trace:<ffffffff8011f805>{__change_page_attr+773} <ffffffff8011f8f1>{change_page_attr+209} Mar 3 14:06:28 dina <ffffffff802360d0>{agp_generic_destroy_page+176} <ffffffff80234ed6>{agp_free_memory+150} Mar 3 14:06:28 dina <ffffffffa002f91e>{:nvidia:KernFreeAGPPages+78} <ffffffffa002e901>{:nvidia:nv_free_pages+257} Mar 3 14:06:28 dina <ffffffffa004c2c6>{:nvidia:_nv000948rm+118} <ffffffffa00387f2>{:nvidia:_nv005342rm+150} Mar 3 14:06:28 dina <ffffffffa0038613>{:nvidia:_nv005346rm+187} <ffffffffa0037ce4>{:nvidia:_nv005355rm+298} Mar 3 14:06:28 dina <ffffffffa005108a>{:nvidia:_nv005319rm+1054} <ffffffff80130780>{thread_return+0} Mar 3 14:06:28 dina <ffffffff802413bf>{elv_merged_request+15} <ffffffff801318db>{io_schedule+43} Mar 3 14:06:28 dina <ffffffff8016cd9c>{__wait_on_buffer+204} <ffffffff801328b0>{autoremove_wake_function+0} Mar 3 14:06:28 dina <ffffffff80170ba4>{bio_alloc+276} <ffffffff8016e3cc>{__find_get_block+188} Mar 3 14:06:28 dina <ffffffff80151d5c>{buffered_rmqueue+300} <ffffffff80151e11>{__alloc_pages+161} Mar 3 14:06:28 dina <ffffffff8015567e>{cache_grow+686} <ffffffff8016e3cc>{__find_get_block+188} Mar 3 14:06:28 dina <ffffffff8016e3ff>{__getblk+31} <ffffffff8016e479>{__bread+9} Mar 3 14:06:28 dina <ffffffff801a3ac4>{ext3_get_branch+100} <ffffffff801a4212>{ext3_get_block_handle+354} Mar 3 14:06:28 dina <ffffffff80150c41>{mempool_alloc+161} <ffffffff801328b0>{autoremove_wake_function+0} Mar 3 14:06:28 dina <ffffffff801328b0>{autoremove_wake_function+0} <ffffffff80248b96>{as_update_arq+38} Mar 3 14:06:28 dina <ffffffff80244642>{__make_request+1330} <ffffffff8015dc2a>{handle_mm_fault+234} Mar 3 14:06:28 dina <ffffffff8011ca23>{pci_map_sg+691} <ffffffff801308b0>{__wake_up_common+64} Mar 3 14:06:28 dina <ffffffff8012e04e>{recalc_task_prio+430} <ffffffff802c37e8>{unix_write_space+72} Mar 3 14:06:28 dina <ffffffff802c60be>{unix_stream_recvmsg+1070} <ffffffff8014e3a7>{find_get_page+23} Mar 3 14:06:28 dina <ffffffff8014f2c9>{filemap_nopage+345} <ffffffff8015b7e6>{zap_pmd_range+198} Mar 3 14:06:28 dina <ffffffffa005088b>{:nvidia:rm_ioctl+9} <ffffffffa002da89>{:nvidia:nv_kern_ioctl+681} Mar 3 14:06:28 dina <ffffffff801673d3>{free_pages_and_swap_cache+19} <ffffffffa002cee9>{:nvidia:nv_kern_vma_release+89} Mar 3 14:06:28 dina <ffffffff8015fadc>{unmap_vma_list+28} <ffffffff8017dc5f>{sys_ioctl+511} Mar 3 14:06:28 dina <ffffffff80160065>{sys_munmap+85} <ffffffff8010ec84>{system_call+124} Mar 3 14:06:28 dina Mar 3 14:06:28 dina Mar 3 14:06:28 dina Code: 0f 0b 2c f2 31 80 ff ff ff ff 6a 00 48 b8 ff ff ff 7f ff ff Mar 3 14:06:28 dina RIP <ffffffff8011f4b5>{revert_page+85} RSP <000001003df05460> Reproducible: Always Steps to Reproduce: 1. Run any OpenGL application that uses a lot of memory for a while. 2. 3. Actual Results: The system log prints the above message, and the OpenGL application gets unpredictable behaviour. Expected Results: The OpenGL application should have continued to run as usual. Kernel 2.6.3 Nvidia kernel driver 5332 Nvidia FX5900 256MB card Dual Opteron 246 system Portage 2.0.50-r1 (default-amd64-2004.0, gcc-3.3.3, glibc-2.3.2-r9, 2.6.3-gentoo-r2) ================================================================= System uname: 2.6.3-gentoo-r2 x86_64 5 Gentoo Base System version 1.4.3.13 Autoconf: sys-devel/autoconf-2.58-r1 Automake: sys-devel/automake-1.7.8 ACCEPT_KEYWORDS="amd64" AUTOCLEAN="yes" CFLAGS="-O2" CHOST="x86_64-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 /usr/share/texmf/tex/generic/config/ /usr/share/texmf/tex/platex/config/ /var/qmail/control" CONFIG_PROTECT_MASK="/etc/gconf /etc/env.d" CXXFLAGS="-O2" DISTDIR="/usr/portage/distfiles" FEATURES="autoaddcvs ccache sandbox" GENTOO_MIRRORS="ftp://gentoo.linux.no/pub/gentoo/ http://212.219.247.20/sites/www.ibiblio.org/gentoo/ http://212.219.247.12/sites/www.ibiblio.org/gentoo/ http://194.83.57.7/sites/www.ibiblio.org/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 amd64 avi berkdb cdr crypt dvb dvd dvdr emacs encode esd foomaticdb gdbm gif gnome gpm gtk gtk2 gtkhtml imap imlib jpeg ldap libg++ libwww mikmod motif mozilla mpeg ncurses nls oggvorbis opengl pam pdflib perl png python quicktime readline sdl slang spell ssl tcltk tcpd tetex tiff truetype xinerama xml2 xmms xv zlib"
There have been reports of nvidia drivers being broken with recent 2.6.x kernels on lkml. I know that doesn't help much, but just to let you know you aren't alone and that people are probably looking into it
this problem is persistant and if you check the nvidia forums, it's existance is either denied, blamed on hardware problems, or considered invalid due to the minion.de patches. any way you look at it, however, it is unfortunately out of our power to fix this one. such is the pain of using binary drivers. *kix his nvidia card* :/
