[LTC-Test] - NMI watchdog Bug and call traces when trinity is executed.
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
linux (Ubuntu) |
Fix Released
|
High
|
Canonical Kernel Team | ||
Xenial |
Fix Released
|
Undecided
|
Tim Gardner | ||
Yakkety |
Fix Released
|
High
|
Canonical Kernel Team |
Bug Description
== Comment: #0 - Santhosh G ==
Problem Statement:
NMI watchdog bug and call traces occurs when trinity is executed.
Environment:
P8 PowerVM Lpar
uname o/p:
uname -a
Linux tuleta4u-lp5 4.4.0-11-generic #26-Ubuntu SMP Sat Mar 5 14:21:51 UTC 2016 ppc64le ppc64le ppc64le GNU/Linux
Steps to reproduce:
1) Install ubuntu 16.04 in a PowerVM LPAR.
2) Download trinity-1.5 and set up ./configure.
3)Execute trinity as
'./trinity --dangerous'
The test runs for more than one hour and trinity gets killed with call traces:
[19744.229979] NMI watchdog: BUG: soft lockup - CPU#3 stuck for 21s! [trinity-c3:26544]
[19744.229991] Modules linked in: hidp hid bnep rfcomm l2tp_ppp l2tp_netlink l2tp_core ip6_udp_tunnel udp_tunnel af_key mpls_router llc2 nfnetlink dn_rtmsg xfrm_user xfrm_algo can_raw crypto_user can_bcm cmtp kernelcapi scsi_transport_
[19744.230024] CPU: 3 PID: 26544 Comm: trinity-c3 Not tainted 4.4.0-11-generic #26-Ubuntu
[19744.230026] task: c00000000ae87e60 ti: c00000000ae24000 task.ti: c00000000ae24000
[19744.230028] NIP: c0000000003fac78 LR: c0000000003fabfc CTR: c00000000039ef10
[19744.230029] REGS: c00000000ae27980 TRAP: 0901 Not tainted (4.4.0-11-generic)
[19744.230030] MSR: 8000000000009033 <SF,EE,
[19744.230035] CFAR: c0000000003fae6c SOFTE: 1
[19744.230045] NIP [c0000000003fac78] ext4_es_
[19744.230047] LR [c0000000003fabfc] ext4_es_
[19744.230048] Call Trace:
[19744.230050] [c00000000ae27c00] [c0000000003fabfc] ext4_es_
[19744.230053] [c00000000ae27c50] [c0000000003a6f18] ext4_map_
[19744.230055] [c00000000ae27d10] [c00000000039f14c] ext4_llseek+
[19744.230057] [c00000000ae27de0] [c0000000002e02a8] SyS_lseek+
[19744.230060] [c00000000ae27e30] [c000000000009204] system_
[19744.230061] Instruction dump:
[19744.230062] 2fa90000 409effec e93e0028 3b800000 e9490458 e92a0440 39290001 f92a0440
[19744.230065] 7c2004ac 7d20d828 3129ffff 7d20d92d <40c2fff4> 60000000 7f83e378 38210050
== Comment: #8 - Santhosh G ==
Tried the scenario as given in https:/
-----
# Create a 624GiB file; Mostly filled with holes though
$ dd if=/dev/zero of=file-0.bin bs=1M count=1 seek=598382
# Invoke lseek with SEEK_DATA option starting with file offset 0
while [ 1 ]; do xfs_io -f -c "seek -d 0" file-0.bin; done
----
and I was able to hit the issue in 16.04.1
kernel version:
4.4.0-28-generic
dmesg o/p:
[ 1197.994822] 40-...: (5249 ticks this GP) idle=975/
[ 1197.995071] (t=5251 jiffies g=29144 c=29143 q=3418)
[ 1197.995115] Task dump for CPU 40:
[ 1197.995117] xfs_io R running task 0 3601 3489 0x00040004
[ 1197.995121] Call Trace:
[ 1197.995126] [c000003c7c8675b0] [c0000000000fbc00] sched_show_
[ 1197.995131] [c000003c7c867620] [c00000000013eb74] rcu_dump_
[ 1197.995134] [c000003c7c867670] [c0000000001442a4] rcu_check_
[ 1197.995136] [c000003c7c8677a0] [c00000000014c108] update_
[ 1197.995140] [c000003c7c8677d0] [c000000000163818] tick_sched_
[ 1197.995143] [c000003c7c867810] [c000000000163914] tick_sched_
[ 1197.995146] [c000003c7c867850] [c00000000014cbd4] __hrtimer_
[ 1197.995148] [c000003c7c8678e0] [c00000000014dbfc] hrtimer_
[ 1197.995152] [c000003c7c8679a0] [c00000000001f5bc] __timer_
[ 1197.995154] [c000003c7c8679f0] [c00000000001f970] timer_interrupt
[ 1197.995157] [c000003c7c867a20] [c000000000002714] decrementer_
[ 1197.995163] --- interrupt: 901 at ext4_es_
[ 1197.995166] [c000003c7c867d10] [c0000000003a170c] ext4_llseek+
[ 1197.995170] [c000003c7c867de0] [c0000000002e1f08] SyS_lseek+
[ 1197.995173] [c000003c7c867e30] [c000000000009204] system_
=====
The call traces does not occur when tried with the kernel with patch.
affects: | kernel-package (Ubuntu) → linux (Ubuntu) |
Changed in linux (Ubuntu): | |
assignee: | Taco Screen team (taco-screen-team) → Canonical Kernel Team (canonical-kernel-team) |
importance: | Undecided → High |
status: | New → Triaged |
Changed in linux (Ubuntu Xenial): | |
status: | In Progress → Fix Committed |
tags: |
added: verification-done-xenial removed: verification-needed-xenial |
Default Comment by Bridge