Reboot on core2 machine

Severity

Reboot

Found by

perf_fuzzer

Fixed by

Test case no longer reproduces with 3577af70a2ce4853d58e57d832e687d739281479 perf: Fix a race condition in perf_remove_from_context() as appearing in 3.17-rc7

First Seen

3.17-rc4

Most recently Seen

3.17-rc4

Reproducible

yes

Found On

core2

Linux-kernel Mailing List Report

8 September 2014: perf: perf_fuzzer triggers instant reboot

Analysis

Kernel Splat

  1. [ 1103.184050] NOHZ: local_softirq_pending 100
    [ 1103.606673] NOHZ: local_softirq_pending 100
    [ 1103.851721] NOHZ: local_softirq_pending 100
    [ 1105.457132] NOHZ: local_softirq_pending 100
    [    0.000000] do_IRQ: 0.145 No irq handler for vector (irq -30719)
    
  2. [  222.457213] NOHZ: local_softirq_pending 100
    [  222.712448] NOHZ: local_softirq_pending 100
    [  222.947573] NOHZ: local_softirq_pending 100
    [  229.168047] BUG: unable to handle kernel NULL pointer dereference at 0000000000000087
    [  229.172003] IP: [] scheduler_tick+0x56/0xbb
    [  229.172003] PGD c4b72067 PUD c52df067 PMD 0 
    [  229.172003] Thread overran stack, or stack corrupted
    [  229.172003] Oops: 0000 [#1] SMP 
    [  229.172003] Modules linked in: cpufreq_userspace cpufreq_stats cpufreq_powersave cpufreq_conservative f71882fg mcs7830 usbnet coretemp evdev psmouse pcspkr ohci_pci ohci_hcd  ^PA!^A<88>hA!^A<88>
    ^P}7(OCE) B<81>p驁 <90>B<81>
    ^P驁 A<81>
     <80>A<81>
    
  3. [  316.040001] general protection fault: 0000 [#1] 
    
  4. [  178.024220] perf_fuzzer[2788]: segfault at ffff88011fa7bef8 ip 00007f8c0e088c2f sp 00007fff003315e0 error 5
    
  5. [   78.018145] NOHZ: local_softirq_pending 100
    [   83.012661] BUG: Bad page map in process perf_fuzzer  pte:ffff8800cafc5190 pmd:119d00067
    [   83.020863] addr:000000000120c000 vm_flags:00100073 anon_vma:ffff8800cae19dc0 mapping:          (null) index:120c
    
  6. *** perf_fuzzer 0.29-pre *** by Vince Weaver
    
            Linux version 3.17.0-rc4+ x86_64
            Processor: Intel 6/23/10
    
            Seeding random number generator with 1410197388
            /proc/sys/kernel/perf_event_max_sample_rate currently: 25000/s
            /proc/sys/kernel/perf_event_paranoid currently: 1
            Logging perf_event_open() failures: no
            Running fsync after every syscall: no
            To reproduce, try: ./perf_fuzzer -t OCIRMQWPFpAi -s 50000 -r 1410197388
    
    Pid=22414, sleeping 1s
    ==================================================
    Fuzzing the following syscalls:
            mmap perf_event_open close read write ioctl fork prctl poll 
    *NOT* Fuzzing the following syscalls:
            
    Also attempting the following:
            busy-instruction-loop accessing-perf-proc-and-sys-files trashing-the-mmap-page 
    *NOT* attempting the following:
            signal-handler-on-overflow 
    ==================================================
    Iteration 10000
            Open attempts: 203475  Successful: 915
                    EPERM : 24
                    ENOENT : 1749
                    E2BIG : 22124
                    EBADF : 8060
                    EACCES : 353
                    EINVAL : 169538
                    ENOSPC : 191
                    EOPNOTSUPP : 521
                    Type (Hardware 227/26831)(software 348/32100)(tracepoint 97/31834)(Cache 50/25664)(cpu 168/31670)(breakpoint 25/31929)(#6 0/23)(#7 0/27)(#8 0/90)(#9 0/26)(#10 0/34)(#11 0/20)(#12 0/34)(#13 0/19)(#14 0/16)(>14 0/23158)
            Close attempts: 864  Successful: 864
            Read attempts: 842  Successful: 731
            Write attempts: 881  Successful: 0
            Ioctl attempts: 871  Successful: 416
            Mmap attempts: 915  Successful: 329
            Prctl attempts: 924  Successful: 924
            Fork attempts: 467  Successful: 467
            Poll attempts: 881  Successful: 0
            Access attempts: 915  Successful: 460
            Trash mmap attempts: 891  Successful: 891
            Overflows: 0
            SIGIOs due to RT signal queue full: 0
    [ 2616.535995] Kernel panic - not syncing: Lost physical address for unconsumed uncorrectable error
    
    
  7. *** perf_fuzzer 0.29-pre *** by Vince Weaver
    
            Linux version 3.17.0-rc4+ x86_64
            Processor: Intel 6/23/10
    
            Seeding random number generator with 1410197388
            /proc/sys/kernel/perf_event_max_sample_rate currently: 25000/s
            /proc/sys/kernel/perf_event_paranoid currently: 1
            Logging perf_event_open() failures: no
            Running fsync after every syscall: no
            To reproduce, try: ./perf_fuzzer -t OCIRMQWPFpAi -s 50000 -r 1410197388 
    
    Pid=4236, sleeping 1s
    ==================================================
    Fuzzing the following syscalls:
            mmap perf_event_open close read write ioctl fork prctl poll 
    *NOT* Fuzzing the following syscalls:
            
    Also attempting the following:
            busy-instruction-loop accessing-perf-proc-and-sys-files trashing-the-mmap-page 
    *NOT* attempting the following:
            signal-handler-on-overflow 
    ==================================================
    Iteration 10000
            Open attempts: 203475  Successful: 915
                    EPERM : 34
                    ENOENT : 1749
                    E2BIG : 22124
                    EBADF : 8060
                    EACCES : 356
                    EINVAL : 169716
                    EOPNOTSUPP : 521
                    Type (Hardware 227/26831)(software 348/32100)(tracepoint 97/31834)(Cache 50/25664)(cpu 168/31670)(breakpoint 25/31929)(#6 0/23)(#7 0/27)(#8 0/90)(#9 0/26)(#10 0/34)(#11 0/20)(#12 0/34)(#13 0/19)(#14 0/16)(>14 0/23158)
            Close attempts: 864  Successful: 864
            Read attempts: 842  Successful: 719
            Write attempts: 881  Successful: 0
            Ioctl attempts: 871  Successful: 416
            Mmap attempts: 915  Successful: 329
            Prctl attempts: 924  Successful: 924
            Fork attempts: 467  Successful: 467
            Poll attempts: 881  Successful: 0
            Access attempts: 915  Successful: 460
            Trash mmap attempts: 891  Successful: 891
            Overflows: 0
            SIGIOs due to RT signal queue full: 0
    Iteration 20000
            Open attempts: 209445  Successful: 887
                    EPERM : 36
                    ENOENT : 1929
                    E2BIG : 22675
                    EBADF : 7661
                    EACCES : 431
                    EINVAL : 175256
                    EOPNOTSUPP : 570
                    Type (Hardware 197/27672)(software 346/33126)(tracepoint 99/32718
    )(Cache 43/26109)(cpu 180/32778)(breakpoint 22/32923)(#6 0/41)(#7 0/36)(#8 0/106)
    (#9 0/41)(#10 0/25)(#11 0/19)(#12 0/39)(#13 0/27)(#14 0/22)(>14 0/23763)
            Close attempts: 888  Successful: 888
            Read attempts: 899  Successful: 774
            Write attempts: 924  Successful: 0
            Ioctl attempts: 861  Successful: 384
            Mmap attempts: 887  Successful: 206
            Prctl attempts: 898  Successful: 898
            Fork attempts: 464  Successful: 464
            Poll attempts: 944  Successful: 2
            Access attempts: 989  Successful: 482
            Trash mmap attempts: 910  Successful: 910
            Overflows: 0
            SIGIOs due to RT signal queue full: 0
    [  749.582103] NOHZ: local_softirq_pending 100
    [  750.051530] NOHZ: local_softirq_pending 100
    [  750.410629] NOHZ: local_softirq_pending 100
    [  751.656861] NOHZ: local_softirq_pending 100
    [  756.009271] traps: perf_fuzzer[4236] trap invalid opcode ip:4044c0 sp:7fffb0ed6f90 error:0
    [  756.017590] BUG: unable to handle kernel paging request
    

Back to perf_fuzzer bugs found