Running make check in frysk-imports generates the following kernel message on my FC6 machine (Linux hermans.wildebeest.org 2.6.18-1.2868.fc6 #1 SMP Fri Dec 15 17:32:54 EST 2006 i686 i686 i386 GNU/Linux): Jan 2 10:43:48 hermans kernel: ------------[ cut here ]------------ Jan 2 10:43:48 hermans kernel: kernel BUG at kernel/utrace.c:201! Jan 2 10:43:48 hermans kernel: invalid opcode: 0000 [#1] Jan 2 10:43:48 hermans kernel: SMP Jan 2 10:43:48 hermans kernel: last sysfs file: /class/net/sit0/statistics/collisions Jan 2 10:43:48 hermans kernel: Modules linked in: i915 drm autofs4 cpufreq_ondemand video sbs ibm_acpi i2c_ec dock button battery asus_acpi ac ipv6 parport_pc lp parport snd_hda_intel snd_hda_codec joydev snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device sg snd_pcm_oss snd_mixer_oss snd_pcm snd_timer e1000 i2c_i801 serio_raw snd i2c_core ide_cd pcspkr cdrom soundcore snd_page_alloc dm_snapshot dm_zero dm_mirror dm_mod ahci libata sd_mod scsi_mod ext3 jbd ehci_hcd ohci_hcd uhci_hcd Jan 2 10:43:48 hermans kernel: CPU: 1 Jan 2 10:43:48 hermans kernel: EIP: 0060:[<c045164b>] Not tainted VLI Jan 2 10:43:48 hermans kernel: EFLAGS: 00210202 (2.6.18-1.2868.fc6 #1) Jan 2 10:43:48 hermans kernel: EIP is at check_dead_utrace+0x3b/0xe0 Jan 2 10:43:48 hermans kernel: eax: 00000020 ebx: f733f200 ecx: 00000000 edx: 00000008 Jan 2 10:43:48 hermans kernel: esi: f16d3440 edi: 00000000 ebp: f16d3440 esp: f1a6fe14 Jan 2 10:43:48 hermans kernel: ds: 007b es: 007b ss: 0068 Jan 2 10:43:48 hermans kernel: Process threadexec (pid: 2956, ti=f1a6f000 task=f733f200 task.ti=f1a6f000) Jan 2 10:43:48 hermans kernel: Stack: f16d3448 f16d3448 00000000 f16d3440 c0451741 f1a6fe5c f733f200 f16d3448 Jan 2 10:43:48 hermans kernel: 00000020 00000020 f733f200 c0452817 00000020 00000000 00200246 f1b21878 Jan 2 10:43:48 hermans kernel: f1b21840 f16d2440 f16d3440 00000010 f16d3440 f733f200 f6c283f0 c0427d19 Jan 2 10:43:48 hermans kernel: Call Trace: Jan 2 10:43:48 hermans kernel: [<c0451741>] remove_detached+0x51/0x74 Jan 2 10:43:48 hermans kernel: [<c0452817>] utrace_report_death+0x1d0/0x205 Jan 2 10:43:48 hermans kernel: [<c0427d19>] do_exit+0x6fc/0x776 Jan 2 10:43:48 hermans kernel: [<c0427e09>] sys_exit_group+0x0/0xd Jan 2 10:43:48 hermans kernel: [<f1a6ff94>] 0xf1a6ff94 Jan 2 10:43:48 hermans kernel: DWARF2 unwinder stuck at 0xf1a6ff94 Jan 2 10:43:48 hermans kernel: Leftover inexact backtrace: Jan 2 10:43:48 hermans kernel: [<c061542d>] do_page_fault+0x0/0x4db Jan 2 10:43:48 hermans kernel: [<c0430808>] get_signal_to_deliver+0x38b/0x3b3 Jan 2 10:43:48 hermans kernel: [<c0430005>] specific_send_sig_info+0x8e/0x99 Jan 2 10:43:48 hermans kernel: [<c061542d>] do_page_fault+0x0/0x4db Jan 2 10:43:48 hermans kernel: [<c0403626>] do_notify_resume+0x7e/0x6c1 Jan 2 10:43:48 hermans kernel: [<c041c9ca>] force_sig_info_fault+0x24/0x28 Jan 2 10:43:48 hermans kernel: [<c0614616>] _spin_unlock_irq+0x5/0x7 Jan 2 10:43:48 hermans kernel: [<c0451945>] utrace_quiescent+0xc2/0x1d5 Jan 2 10:43:48 hermans kernel: [<c0452ed1>] utrace_report_syscall+0x171/0x190 Jan 2 10:43:48 hermans kernel: [<c0615900>] do_page_fault+0x4d3/0x4db Jan 2 10:43:48 hermans kernel: [<c061542d>] do_page_fault+0x0/0x4db Jan 2 10:43:48 hermans kernel: [<c04040a2>] work_notifysig+0x13/0x19 Jan 2 10:43:48 hermans kernel: ======================= Jan 2 10:43:48 hermans kernel: Code: 7a 04 00 8b 93 a4 04 00 00 0f 45 f8 89 f8 83 e2 08 f7 d0 85 d0 74 69 8b 83 90 00 00 00 85 c0 74 5f 83 3e 00 75 5a 83 f8 10 74 08 <0f> 0b c9 00 70 7f 63 c0 c7 06 01 00 00 00 83 bb 98 00 00 00 ff Jan 2 10:43:48 hermans kernel: EIP: [<c045164b>] check_dead_utrace+0x3b/0xe0 SS:ESP 0068:f1a6fe14 Jan 2 10:43:48 hermans kernel: <1>Fixing recursive fault but reboot is needed!
Also happens with 2.6.18-1.2869.fc6
With 2.6.19-1.2895.fc6 kernel gives: kernel: BUGging on (tsk->exit_state != EXIT_ZOMBIE) and then seems to hang the machine.
A second make check run (after reboot) with 2.6.19-1.2895.fc6 gave the following: Jan 19 12:53:45 hermans kernel: BUGging on (tsk->exit_state != EXIT_ZOMBIE) Jan 19 12:53:45 hermans kernel: ------------[ cut here ]------------ Jan 19 12:53:45 hermans kernel: kernel BUG at kernel/utrace.c:193! Jan 19 12:53:45 hermans kernel: invalid opcode: 0000 [#1] Jan 19 12:53:45 hermans kernel: SMP Jan 19 12:53:45 hermans kernel: last sysfs file: /class/net/eth1/statistics/collisions Jan 19 12:53:45 hermans kernel: Modules linked in: i915 drm autofs4 cpufreq_ondemand video sbs ibm_acpi i2c_ec dock button battery asus_acpi ac ipv6 parport_pc lp parport joydev snd_hda_intel snd_hda_codec snd_seq_dummy sg snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss snd_pcm snd_timer serio_raw ata_piix snd soundcore nsc_ircc e1000 snd_page_alloc irda crc_ccitt i2c_i801 iTCO_wdt i2c_core ide_cd cdrom pcspkr dm_snapshot dm_zero dm_mirror dm_mod ahci libata sd_mod scsi_mod ext3 jbd ehci_hcd ohci_hcd uhci_hcd Jan 19 12:53:45 hermans kernel: CPU: 0 Jan 19 12:53:45 hermans kernel: EIP: 0060:[<c0456467>] Not tainted VLI Jan 19 12:53:45 hermans kernel: EFLAGS: 00210296 (2.6.19-1.2895.fc6 #1) Jan 19 12:53:45 hermans kernel: EIP is at check_dead_utrace+0x55/0x110 Jan 19 12:53:45 hermans kernel: eax: 0000002f ebx: de725670 ecx: c0697ed0 edx: 00200082 Jan 19 12:53:45 hermans kernel: esi: 00000000 edi: d02e53a0 ebp: d02e53a0 esp: dd8dfdf8 Jan 19 12:53:45 hermans kernel: ds: 007b es: 007b ss: 0068 Jan 19 12:53:45 hermans kernel: Process threadexec (pid: 20057, ti=dd8df000 task=de725670 task.ti=dd8df000) Jan 19 12:53:45 hermans kernel: Stack: c063e591 c064c238 d02e53a8 00000000 d02e53a8 d02e53a0 c045657c 00000000 Jan 19 12:53:45 hermans kernel: de725670 d02e53a8 00000020 de725670 d02e53a0 c04578ee 00000000 00000000 Jan 19 12:53:45 hermans kernel: dabc7778 00200246 de725b18 dabc7740 e4c313a0 d02e53b0 00000010 d02e53a0 Jan 19 12:53:45 hermans kernel: Call Trace: Jan 19 12:53:45 hermans kernel: [<c045657c>] remove_detached+0x5a/0x6d Jan 19 12:53:45 hermans kernel: [<c04578ee>] utrace_report_death+0x221/0x229 Jan 19 12:53:45 hermans kernel: [<c042a65a>] do_exit+0x6e1/0x787 Jan 19 12:53:45 hermans kernel: [<c042a78d>] sys_exit_group+0x0/0xd Jan 19 12:53:45 hermans kernel: [<c0433185>] get_signal_to_deliver+0x38b/0x3b3 Jan 19 12:53:45 hermans kernel: [<c040365b>] do_notify_resume+0x83/0x6c6 Jan 19 12:53:45 hermans kernel: [<c04040da>] work_notifysig+0x13/0x19 Jan 19 12:53:45 hermans kernel: ======================= Jan 19 12:53:45 hermans kernel: Code: f0 f7 d0 83 e2 08 85 d0 0f 84 80 00 00 00 85 c9 74 7c 83 f9 10 74 1c c7 44 24 04 38 c2 64 c0 c7 04 24 91 e5 63 c0 e8 6a 16 fd ff <0f> 0b c1 00 0a c2 64 c0 83 bb 98 00 00 00 ff 75 33 b8 20 00 00 Jan 19 12:53:45 hermans kernel: EIP: [<c0456467>] check_dead_utrace+0x55/0x110 SS:ESP 0068:dd8dfdf8 Jan 19 12:53:45 hermans kernel: <1>Fixing recursive fault but reboot is needed!
Seeing this problem as well. As far as I can determine right now, it is test frysk2130/strace-clone-exec.sh causing it. It is mostly reproducible, though I have been able to do full 'make check' runs on my frysk build that do not trigger it, occasionally.
Quick follow-up - did a reboot and executed just the strace-clone-exec.sh test from the command line. This also triggered the kernel bug, on FC6 with the 2.6.19-1.2895.fc6 kernel.
Anyone out there with the latest RHEL5 beta installed who could verify whether this problem occurs there also?
This is actual a kernel bug (as stated in the original bug that prompted this testcase to be added - #2130). However, note that the testcase is *not* testing the kernel bug in the sense that the testcase passes happily while it still often (but not always) triggers the kernel bug. So, noteworthy info: - This is definitely a kernel bug - The test case passes even if the kernel bug is triggered - The kernel bug is not 100% reproducible - it sometimes takes multiple attempts
*** This bug has been marked as a duplicate of 2130 ***
Reopening after talking to Mark. We believe this is not actually the same problem as 2130, but rather a similar case (occuring in a slightly different place).
Opened a bug against the FC6 kernel: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=227952 Suspending this ticket to track progress of the FC bug.