Solution of the error NMI watchdog: BUG: soft lockup – CPU#0 stuck for 23s!

I noticed several times that the server was crashing with Ubuntu Server installed on 14.04.01, after which only a hard reset with power outage helped.

I will quote some of the logs at this moment:

Jun 17 20:16:19 server-gw kernel: [3455576.132542] Code: be 00 02 00 00 e8 46 26 8c ff 5d c3 0f 1f 40 00 66 66 66 66 90 55 48 89 e5 53 48 89 f3 66 66 66 66 90 66 83 07 02 48 89 df 57 9d <66> 66 90 66 90 5b 5d c3 0f 1f 44 00 00 b8 02 00 00 00 f0 66 0f
Jun 17 20:16:47 server-gw kernel: [3455604.143698] NMI watchdog: BUG: soft lockup – CPU#0 stuck for 23s! [kworker/0:2:19206]
Jun 17 20:16:47 server-gw kernel: [3455604.146475] Modules linked in: xt_nat xt_tcpudp iptable_nat nf_conntrack_ipv4 nf_defrag_ipv4 nf_nat_ipv4 ipt_ISG(OE) iptable_filter ip_tables x_tables ipmi_ssif gpio_ich amdkfd amd_iommu_v2 intel_powerclamp coretemp radeon kvm_intel kvm crct10dif_pclmul crc32_pclmul ghash_clmulni_intel aesni_intel aes_x86_64 lrw gf128mul ttm glue_helper ablk_helper cryptd drm_kms_helper serio_raw drm lpc_ich joydev i2c_algo_bit i7core_edac hpilo edac_core shpchp ipmi_si 8250_fintek ipmi_msghandler mac_hid acpi_power_meter nf_nat_pptp nf_nat_proto_gre nf_conntrack_pptp nf_conntrack_proto_gre nf_nat nf_conntrack lp parport hid_generic usbhid hid psmouse be2net vxlan bnx2 pata_acpi ip6_udp_tunnel udp_tunnel hpsa
Jun 17 20:16:47 server-gw kernel: [3455604.146529] CPU: 0 PID: 19206 Comm: kworker/0:2 Tainted: G IOEL 3.19.0-39-generic #44~14.04.1-Ubuntu
Jun 17 20:16:47 server-gw kernel: [3455604.146531] Hardware name: HP ProLiant DL380 G7, BIOS P67 09/30/2010
Jun 17 20:16:47 server kernel: [3455604.146537] Workqueue: events inet_frag_worker

The problem was a bug in the Ubuntu kernel version – 3.19.0-39-generic, and naturally the solution to the problem was to update the system with commands:

sudo apt-get update
sudo apt-get upgrade

After that, the error disappeared.

The kernel version can be viewed by the command:

uname -r

Ubuntu version:

lsb_release -r

See also my article:
Solution BUG: Bad page state in process kworker

Leave a comment

Leave a Reply