Description of problem: While running the kernel tier one tests scrashme caused a kernel BUG Version-Release number of selected component (if applicable): 2.6.18-96.el5 debug How reproducible: Very intermittent Steps to Reproduce: 1. Run scrashme test Actual results: ------------[ cut here ]------------ kernel BUG at kernel/utrace.c:345! invalid opcode: 0000 [#1] SMP last sysfs file: /devices/pci0000:00/0000:00:10.0/0000:06:02.1/irq Modules linked in: autofs4 hidp rfcomm l2cap bluetooth sunrpc ipv6 xfrm_nalgo crypto_api cpufreq_ondemand dm_multipath video sbs backlight i2c_ec button battery asus_acpi ac parport_pc lp parport ide_cd sr_mod k8_edac i2c_nforce2 edac_mc cdrom serio_raw k8temp e1000 i2c_core hwmon sg pcspkr dm_snapshot dm_zero dm_mirror dm_mod usb_storage mptsas mptscsih mptbase scsi_transport_sas sd_mod scsi_mod ext3 jbd uhci_hcd ohci_hcd ehci_hcd CPU: 2 EIP: 0060:[<c045735d>] Not tainted VLI EFLAGS: 00010202 (2.6.18-96.el5debug #1) EIP is at check_dead_utrace+0x123/0x14d eax: 00000020 ebx: f083e794 ecx: c042fc59 edx: f1afb000 esi: de4a31e0 edi: 00000000 ebp: f083e78c esp: f1afbf44 ds: 007b es: 007b ss: 0068 Process scrashme (pid: 18353, ti=f1afb000 task=de4a31e0 task.ti=f1afb000) Stack: 00000020 f083e794 00000000 f083e794 f083e78c c04573e1 00000000 de4a31e0 de4a31e0 f083e78c de4a31e0 f1afbfa4 c0457415 00000000 00000010 c042910e f7c8eca0 de4a3294 00000000 00000002 00000000 f7167cf8 00000000 bfbd92b8 Call Trace: [<c04573e1>] remove_detached+0x5a/0x6d [<c0457415>] finish_report_death+0x21/0x24 [<c042910e>] do_exit+0x733/0x7b8 [<c0429209>] sys_exit_group+0x0/0xd [<c0404f7b>] syscall_call+0x7/0xb ======================= Code: 98 00 00 00 89 f0 e8 c7 87 fd ff 83 be 98 00 00 00 ff 75 1f b8 20 00 00 00 87 86 90 00 00 00 83 f8 10 c7 04 24 20 00 00 00 74 08 <0f> 0b 59 01 5a db 63 c0 b8 00 4a 73 c0 e8 c2 f8 1b 00 83 3c 24 EIP: [<c045735d>] check_dead_utrace+0x123/0x14d SS:ESP 0068:f1afbf44 Expected results: This should pass Additional info: The kexec tools were enabled during this test run. There is a vmcore file available for debugging.
vmcore is not available. no luck with reproducing. me stuck.
Perhaps this relates to BUG 466774? If the race do exists, do_wait() can win and set EXIT_DEAD before check_dead_utrace().
On RHTS Job 33958 Recipe 120752 system ibm-qs22-01.lab.bos.redhat.com. While running the /kernel/syscalls/scrashme/multiple. Using kernel 2.6.18-120.el5 kernel on issue on PPC. kernel BUG in check_dead_utrace at kernel/utrace.c:345! cpu 0x0: Vector: 700 (Program Check) at [c0000003ea41b920] pc: c0000000000b36f0: .check_dead_utrace+0x1d0/0x22c lr: c0000000000b36b8: .check_dead_utrace+0x198/0x22c sp: c0000003ea41bba0 msr: 9000000000029032 current = 0xc0000003fe19a470 paca = 0xc00000000052af00 pid = 23970, comm = scrashme kernel BUG in check_dead_utrace at kernel/utrace.c:345!
Now I am more or less sure this bug duplicates BUG 466774, see the test case I sent.
*** This bug has been marked as a duplicate of bug 466774 ***