I have a laptop with no serial ports, so getting debugging information out is difficult... With recent kernels (certainly 2.6.7) I've been having problems with oopses under load. Compiling a kernel or something large seems to trigger it, although everyday use (and compilation) is fine. I was recently playing with autofs, and managed to get it to cause an oops which didn't kill the machine (normally I have to resort of the Magic SysRq keys). I can confirm that this *is* the same oops that does normally kill the machine, though: Code: 8b 30 8b 04 b5 c0 e7 34 c0 89 44 24 40 89 f0 e8 38 62 00 00 <1>Unable to handle kernel NULL pointer dereference at virtual address 00000000 printing eip: c01d5af4 *pde = 00000000 Oops: 0000 [#14] Modules linked in: vfat fat nls_base autofs eeprom af_packet hci_usb sd_mod usb_storage scsi_mod ipw2100 firmware_class ieee80211 ieee80211_crypt yenta_socket pcmcia_core 8250_pci 8250 serial_core usblp ehci_hcd uhci_hcd intel_agp snd_intel8x0 snd_ac97_codec snd_mpu401_uart snd_rawmidi snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_pcm snd_page_alloc snd_timer snd_mixer_oss snd soundcore ipt_state ipt_REJECT iptable_filter iptable_mangle iptable_nat ip_conntrack ip_tables ide_cd cdrom eth1394 usbcore binfmt_misc ohci1394 ieee1394 rfcomm l2cap bluetooth i2c_i801 i2c_algo_pcf i2c_sensor i2c_dev e100 mii uinput psmouse radeon agpgart hw_random evdev nvram rtc sonypi msr microcode cpuid softdog speedstep_centrino button ac battery fan thermal processor unix CPU: 0 EIP: 0060:[<c01d5af4>] Not tainted EFLAGS: 00010296 (2.6.7-gentoo-r13) EIP is at vt_ioctl+0x14/0x1af0 eax: 00000000 ebx: 00005401 ecx: 00005401 edx: d9a8d5a8 esi: c155b95c edi: d97fb000 ebp: d9a8d5a8 esp: d8be5eb8 ds: 007b es: 007b ss: 0068 Process consolelog.sh (pid: 18206, threadinfo=d8be4000 task=dcb4b2b0) Stack: d8c73080 df712550 c145d870 c1646005 00000000 dfdcde10 080a17c0 df935dd8 d8be5ef8 00000000 c01a8792 00000000 00000000 00000000 c014f6dd d8be5f6c 0000000b 00000000 00000000 000016f2 000021b0 00000001 00000000 00000005 Call Trace: [<c01a8792>] copy_to_user+0x32/0x50 [<c014f6dd>] cp_new_stat64+0xed/0x110 [<c01d15e1>] tty_ioctl+0x3d1/0x4b0 [<c015697f>] sys_ioctl+0xbf/0x210 [<c0103d4d>] sysenter_past_esp+0x52/0x71 Code: 8b 30 8b 04 b5 c0 e7 34 c0 89 44 24 40 89 f0 e8 38 62 00 00 <1>Unable to handle kernel NULL pointer dereference at virtual address 00000000 printing eip: c01d5af4 *pde = 00000000 Oops: 0000 [#15] Modules linked in: vfat fat nls_base autofs eeprom af_packet hci_usb sd_mod usb_storage scsi_mod ipw2100 firmware_class ieee80211 ieee80211_crypt yenta_socket pcmcia_core 8250_pci 8250 serial_core usblp ehci_hcd uhci_hcd intel_agp snd_intel8x0 snd_ac97_codec snd_mpu401_uart snd_rawmidi snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_pcm snd_page_alloc snd_timer snd_mixer_oss snd soundcore ipt_state ipt_REJECT iptable_filter iptable_mangle iptable_nat ip_conntrack ip_tables ide_cd cdrom eth1394 usbcore binfmt_misc ohci1394 ieee1394 rfcomm l2cap bluetooth i2c_i801 i2c_algo_pcf i2c_sensor i2c_dev e100 mii uinput psmouse radeon agpgart hw_random evdev nvram rtc sonypi msr microcode cpuid softdog speedstep_centrino button ac battery fan thermal processor unix CPU: 0 EIP: 0060:[<c01d5af4>] Not tainted EFLAGS: 00010296 (2.6.7-gentoo-r13) EIP is at vt_ioctl+0x14/0x1af0 eax: 00000000 ebx: 00005401 ecx: 00005401 edx: d9a8d478 esi: c155b95c edi: d97fb000 ebp: d9a8d478 esp: d8be5eb8 ds: 007b es: 007b ss: 0068 Process consolelog.sh (pid: 18209, threadinfo=d8be4000 task=dcb4b2b0) Stack: d93c8080 df712550 c145d870 c1646005 00000000 dfdcde10 080a17c0 df8be02c d8be5ef8 00000000 c01a8792 00000000 00000000 00000000 c014f6dd d8be5f6c 0000000b 00000000 00000000 000016f2 000021b0 00000001 00000000 00000005 Call Trace: [<c01a8792>] copy_to_user+0x32/0x50 [<c014f6dd>] cp_new_stat64+0xed/0x110 [<c01d15e1>] tty_ioctl+0x3d1/0x4b0 [<c015697f>] sys_ioctl+0xbf/0x210 [<c0103d4d>] sysenter_past_esp+0x52/0x71 Code: 8b 30 8b 04 b5 c0 e7 34 c0 89 44 24 40 89 f0 e8 38 62 00 00 Since this error occurred on a Gentoo-customised kernel, I thought I'd ask here before in the LKML. Oh, and the consolelog.sh script is only: #!/bin/sh # # consolelog.sh # For metalog -- log to a console # DEVICE="/dev/vc/12" if [ -n "$DEVICE" -a -c "$DEVICE" -a -w "$DEVICE" ]; then echo -e "$1 [$2]\t$3" &>"$DEVICE" fi ... which I guess to indicate a problem tickled by Metalog or by bash :( (I'm wondering if the problem occurs when something tries to write to syslog *whilst* the machine is under load) Any help greatly appreciated!
Does the same error occur on a generic unpatched linux-2.6.7 kernel? Also look at emerging ksymoops to get a more detailed infomation about your crash.
The ksymoops package is not really needed for the 2.6 kernel anymore. I also would like to know if this happens on a clean 2.6.7 kernel from kernel.org.
Can you try gentoo-dev-sources-2.6.8 and see if you can still reproduce this issue?
Please reopen bug with requested tests/info.
Closed due to lack of info.
