-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Thursday 2006-10-26 at 21:33 +0200, I wrote:
I'm trying to make another test with 10.1 when I find some time, perhaps in a hour or two. I will make sure no important programs are running, and I will open an ssh from another computer before hand. I want to make a log of the style the nvidia folks like.
Well... Mixed feelings. I'm writing this up lest I forget, and as a RFC, even if it is very late and I'm tired. I started X from a text console, from runlevel 3, with "startx gnome -- - -logverbose 5", as per nvidia recommendations. I also logged in externally via ssh, left top running, as root, as a safety net. I tried to crash the system, and I couldn't. Not exactly. The other day, resizing konq columns would crash it - not today. Perhaps the last YOU patches have something to do. I think there was something about qt, but it is not listed in my rpm database, which crashed the other day, anyway; the last updates listed are: INSTALLTIME BUILDTIME NAME VERSION RELEASE PACKAGER Fri Oct 20 2006 Fri Oct 20 2006 SimGear 0.3.10 1 checkinstall-1.6.0 Sat Oct 21 2006 Sat Oct 21 2006 rte 0.5.6 1 checkinstall-1.6.0 Sat Oct 21 2006 Sat Oct 21 2006 FlightGear 0.9.10 2 checkinstall-1.6.0 Sun Oct 22 2006 Sat Oct 21 2006 zapping 0.10cvs6 1 checkinstall-1.6.0 Wed Oct 25 2006 Thu Sep 21 2006 xorg-x11-driver-video 6.9.0 46.20 http://bugs.opensuse.org Thu Oct 26 2006 Tue Oct 24 2006 screen 4.0.2 62.5 http://bugs.opensuse.org Or I'm confused and the qt thing was in 9.3 - that's what happens with these late night testing and my memory O:-) I wrote my notes on the info you asked about the other day. glxgears turned very slowly, the mouse affected it, but the FPS were similar to what I got in 9.3. I played 3D games like "planet penguin racer" and "FlightGear". The former was more playable than the other day; there were few "jerks", but there were some. FlightGear did have some jerkiness, even if compiled with "-O3", but that is a very demanding program. Finally, after an hour or so, I decided to start zapping (gnome TV app, version 0.10cvs6), which previously crashed instantly the system. This time, what crashed (surprise) was the "/home" XFS partition, which become suddenly unreadable. Other partitions in the same disk seemed unaffected: Oct 27 01:03:53 nimrodel gconfd (cer-6078): Could not open saved state file '/home/cer/.gconfd/saved_state.tmp' for writing: Input/output error /var/log/kernel Oct 27 01:03:27 nimrodel kernel: xfs_iunlink_remove: xfs_itobp() returned an error 990 on hdd8. Returning error. Oct 27 01:03:27 nimrodel kernel: xfs_inactive: xfs_ifree() returned an error = 990 on hdd8 Oct 27 01:03:27 nimrodel kernel: xfs_force_shutdown(hdd8,0x1) called from line 1762 of file fs/xfs/xfs_vnodeops.c. Return address = 0xf92d9bcb Oct 27 01:03:27 nimrodel kernel: Filesystem "hdd8": I/O Error Detected. Shutting down filesystem: hdd8 Oct 27 01:03:27 nimrodel kernel: Please umount the filesystem, and rectify the problem(s) Oct 27 01:09:36 nimrodel kernel: xfs_force_shutdown(hdd8,0x1) called from line 338 of file fs/xfs/xfs_rw.c. Return address = 0xf92d9bcb Oct 27 01:09:43 nimrodel kernel: audit(1161904183.600:7): audit_pid=0 old=4155 by auid=4294967295 Oct 27 01:09:45 nimrodel kernel: pnp: Device 00:0d disabled. Oct 27 01:09:45 nimrodel kernel: gameport: kgameportd exiting Oct 27 01:09:47 nimrodel kernel: device eth0 left promiscuous mode Oct 27 01:09:50 nimrodel kernel: Kernel logging (proc) stopped. Oct 27 01:09:50 nimrodel kernel: Kernel log daemon terminating. I managed to halt and reboot the system, and no data seemed lost. Very weird. But being "zapping" a cvs version, even though it works well with the "nv" driver, I thought it could be something else. So I tried again. This time I went to "init 5", login via "wdm" into gnome. I tried playing again, no crash. Then I decided to try "kdetv", and... I got another weird crash. This time, any command I entered in an xterm would not return the prompt, not even from the external ssh session. In a remaining root xterm I fired "nvidia-bug-report.sh", and later I found the file correctly saved, fortunately. I could not log off the session. ctrl-alt-backspace closed the session, but I didn't get a prompt back. Tried "ctrl-alt-supr", the system reported halting, but it did not. Finally I had to power off the PC, and after reboot, finally disable "nvidia" driver, and start typing this. I have seen a moment ago there was a kernel oops, two minutes before I run nvidia-bug-report.sh, therefore right on the crash: Oct 27 02:31:54 nimrodel kernel: printing eip: Oct 27 02:31:54 nimrodel kernel: c015d0b8 Oct 27 02:31:54 nimrodel kernel: *pde = 00000000 Oct 27 02:31:54 nimrodel kernel: Oops: 0000 [#1] Oct 27 02:31:54 nimrodel kernel: last sysfs file: /block/hda/hda5/stat Oct 27 02:31:54 nimrodel kernel: Modules linked in: xt_pkttype ipt_LOG ipt_recent af_packet adi joydev snd_pcm_oss snd_mixer_oss snd_seq_midi snd_seq_midi_event snd_seq ir_kbd_i2c button battery ac loop_fish2 ohci_hcd ehci_hcd ip6t_REJECT ip6t_LOG xt_limit xt_tcpudp ipt_REJECT xt_state iptable_mangle iptable_nat ip_nat iptable_filter ip6table_mangle ip_conntrack nfnetlink ip_tables ip6table_filter ip6_tables x_tables ipv6 apparmor aamatch_pcre xfs_quota xfs exportfs raid1 dm_mod twofish cryptoloop loop usbhid bt878 shpchp pci_hotplug tuner tvaudio bttv video_buf firmware_class compat_ioctl32 i2c_algo_bit v4l2_common btcx_risc ir_common tveeprom i2c_i801 videodev intel_agp snd_intel8x0 snd_ac97_codec snd_ac97_bus 8139too snd_pcm mii snd_timer uhci_hcd nvidia snd_page_alloc usbcore ns558 gameport agpgart i2c_core ide_cd cdrom snd_mpu401 snd_mpu401_uart snd_rawmidi snd_seq_device snd soundcore parport_pc lp parport reiserfs ext3 jbd fan thermal processor piix ide_disk ide_core Oct 27 02:31:54 nimrodel kernel: CPU: 0 Oct 27 02:31:54 nimrodel kernel: EIP: 0060:[<c015d0b8>] Tainted: P U VLI Oct 27 02:31:54 nimrodel kernel: EFLAGS: 00210286 (2.6.16.21-0.25-default #1) Oct 27 02:31:54 nimrodel kernel: EIP is at __d_lookup+0xa1/0xbf Oct 27 02:31:54 nimrodel kernel: eax: 84ef80ee ebx: c3f70114 ecx: 01843b47 edx: c1815180 Oct 27 02:31:54 nimrodel kernel: esi: d9ea0b64 edi: 84ef80ee ebp: d5efbe74 esp: d5efbe08 Oct 27 02:31:54 nimrodel kernel: ds: 007b es: 007b ss: 0068 Oct 27 02:31:54 nimrodel kernel: Process drkonqi (pid: 18556, threadinfo=d5efa000 task=d897aab0) Oct 27 02:31:54 nimrodel kernel: Stack: <0>d9ea3e14 00000004 01843b47 dfcb7027 01843b47 d9ea0b64 dfcb702c d5efbf48 Oct 27 02:31:54 nimrodel kernel: c0155292 d5efbe80 d5efbe74 d5efbf48 dfff4cc0 01843b47 d9ea0b64 dfcb702c Oct 27 02:31:54 nimrodel kernel: d5efbf48 c0156d4f dfcb702c 00000000 00000403 c01606d8 00000000 d5efbf10 Oct 27 02:31:54 nimrodel kernel: Call Trace: Oct 27 02:31:54 nimrodel kernel: [<c0155292>] do_lookup+0x24/0x135 Oct 27 02:31:54 nimrodel kernel: [<c0156d4f>] __link_path_walk+0x6da/0xaec Oct 27 02:31:54 nimrodel kernel: [<c01606d8>] mntput_no_expire+0x11/0x62 Oct 27 02:31:55 nimrodel kernel: [<c0157210>] link_path_walk+0xaf/0xb9 Oct 27 02:31:55 nimrodel kernel: [<c01571a8>] link_path_walk+0x47/0xb9 Oct 27 02:31:55 nimrodel kernel: [<c01522ec>] sys_stat64+0x1e/0x23 Oct 27 02:31:55 nimrodel kernel: [<c01574c9>] do_path_lookup+0x198/0x1e6 Oct 27 02:31:55 nimrodel kernel: [<c0157c0b>] __user_walk_fd+0x29/0x3a Oct 27 02:31:55 nimrodel kernel: [<c0149df4>] sys_faccessat+0x92/0x123 Oct 27 02:31:55 nimrodel kernel: [<c01522ec>] sys_stat64+0x1e/0x23 Oct 27 02:31:55 nimrodel kernel: [<c0149e94>] sys_access+0xf/0x13 Oct 27 02:31:55 nimrodel kernel: [<c010299b>] sysenter_past_esp+0x54/0x79 Oct 27 02:31:55 nimrodel kernel: Code: 39 42 04 75 20 8b 42 08 8b 4c 24 04 8b 54 24 0c e8 29 74 04 00 85 c0 75 0c f6 43 04 10 75 20 ff 03 89 d8 eb 1c 8b 3f 85 ff 74 14 <8b> 07 0f 18 00 90 8d 5f f4 8b 4c 24 08 39 4b 18 75 e8 eb 9b 31 Oct 27 02:32:00 nimrodel kernel: <1>Unable to handle kernel paging request at virtual address 84ef80ee Oct 27 02:32:00 nimrodel kernel: printing eip: Oct 27 02:32:00 nimrodel kernel: c015d0b8 Oct 27 02:32:00 nimrodel kernel: *pde = 00000000 Oct 27 02:32:00 nimrodel kernel: Oops: 0000 [#2] I have logged there 12 Oops in succession, for anyone wanting them, but as the kernel is tainted (nvidia driver) I don't think the kernel guys will even look at it. But I find very suspicious of something very wrong that two TV apps cause the system to crash this way. As a very wild guess, I'd say that something is writing in kernel space out of bounds. I'll fill a bugzilla report tomorrow, time permitting. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFFQWletTMYHG2NR9URAv8OAKCAXSCjwh83SuueXRiyqAeIYyGcgACghDXW 47v0JRuIi9dTMx3Tq/esc7E= =fNfG -----END PGP SIGNATURE-----