【OLK-5.10】openvswitch触发软锁

**【标题描述】能够简要描述问题：说明什么场景下，做了什么操作，出现什么问题（尽量使用正向表达方式）**
在云计算场景中，使用ovs作为网络插件，在物理服务器上触发软锁问题

**一、缺陷信息**

**内核信息：**
5.10.0-136.37.0.113.oe2203sp1.aarch64

**缺陷归属组件：**
net/openvswitch/openvswitch.ko

**缺陷归属的版本：**
5.10.0-136.37.0.113

**缺陷简述：**
在物理服务器上触发软锁问题，通过查看内核日志，确认是openvswitch模块触发

**【环境信息】**
硬件信息
System Information
        Manufacturer: Inspur
        Product Name: CS5280F3
软件信息
- OS版本及分支信息：openEuler22.03LTS-SP1
- 内核信息：5.10.0-136.37.0.113.oe2203sp1.aarch64

**【问题复现步骤】**
在物理服务器上安装ovs，使用vxlan方式支持虚拟机网络，运行一周左右，出现openvswitch软锁

**【实际结果】**
由于出现软锁，系统运行效率下降，待所有cpu锁住后，系统卡死。

**【其他相关附件信息】**
内核堆栈信息如下：
[31455.992466] watchdog: BUG: soft lockup - CPU#41 stuck for 22s! [swapper/41:0]
[31455.992468] Modules linked in: dummy vhost_net vhost tun tap ip6table_nat ip6_tables ipt_rpfilter xt_set xt_multiport iptable_raw ip_set_hash_ip ip_set_hash_net ip_set ipip tunnel4 ip_tunnel xt_statistic xt_nat xt_recent xt_addrtype ipt_REJECT nf_reject_ipv4 iptable_mangle xt_physdev xt_conntrack xt_comment xt_mark veth iptable_filter nf_conntrack_netlink iptable_nat xt_MASQUERADE rbd ceph libceph dns_resolver overlay sch_ingress nfnetlink_cttimeout nfnetlink openvswitch nsh nf_conncount nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 bonding vfat fat dm_multipath ipmi_ssif aes_ce_blk crypto_simd cryptd aes_ce_cipher crct10dif_ce ghash_ce i40iw sha3_ce sm4_ce_cipher sm4_ce ses joydev sm3_ce sm4 sha3_generic sha512_ce ib_uverbs enclosure sha512_arm64 ib_core sha1_ce sbsa_gwdt ipmi_si ky_lpcmux ipmi_devintf
[31455.992513]  ipmi_msghandler spi_phytium_plat phytium_ce phyt_ps2364_ddr_pmu phyt_ps2364_pcie_pmu spi_phytium sch_fq_codel ip_tables lpfc nvmet_fc sha2_ce nvmet sha256_arm64 nvme_fc nvme nvme_fabrics nvme_core ast nvme_common scsi_transport_fc megaraid_sas i40e igb br_netfilter bridge stp llc
[31455.992530] CPU: 41 PID: 0 Comm: swapper/41 Kdump: loaded Tainted: G             L    4.19.90-89.1.v2401.fortest.ky10.aarch64 #1
[31455.992531] Source Version: ff12b2c7c575ef6431b3edf16ee0d8a0f64b1a2e
[31455.992532] Hardware name: Inspur CS5280F3/CS5280F3, BIOS 2.04.00 12/21/2023
[31455.992533] pstate: 80400009 (Nzcv daif +PAN -UAO)
[31455.992535] pc : netdev_pick_tx+0x1f4/0x2a0
[31455.992538] lr : netdev_core_pick_tx+0xf8/0x128
[31455.992539] sp : ffffa10effbe03f0
[31455.992540] x29: ffffa10effbe03f0 x28: 000000000000002c
[31455.992541] x27: ffffa00003f9bb00 x26: ffff81003a1b48e4
[31455.992542] x25: 0000000000000000 x24: 00000000ffffffff
[31455.992543] x23: fffff000724a6000 x22: 0000000000000000
[31455.992544] x21: ffffa00003f9bb00 x20: fffff000724a6000
[31455.992545] x19: 0000000000000000 x18: ffff000009c189c0
[31455.992546] x17: 0000000000000001 x16: 0000000000000004
[31455.992547] x15: fffff000203cfb38 x14: 0000000000000000
[31455.992548] x13: ffff000008dc0ac8 x12: 0000000000000001
[31455.992549] x11: ffff000008dc0aa0 x10: 0000000000000178
[31455.992550] x9 : ffffa10effbe0648 x8 : ffff81003ac2f910
[31455.992551] x7 : 0000000000000000 x6 : ffffa00003f9ad00
[31455.992553] x5 : ffffa10effbe2ecc x4 : 0000000000000000
[31455.992554] x3 : 0000000000000000 x2 : 0000000000000000
[31455.992555] x1 : 0000000000000027 x0 : 0000000000000000
[31455.992557] Call trace:
[31455.992558]  netdev_pick_tx+0x1f4/0x2a0
[31455.992560]  netdev_core_pick_tx+0xf8/0x128
[31455.992561]  __dev_queue_xmit+0x1a4/0xb80
[31455.992564]  dev_queue_xmit+0x24/0x30
[31455.992575]  ovs_vport_send+0xd0/0x1a0 [openvswitch]
[31455.992579]  do_output+0x64/0x168 [openvswitch]
[31455.992582]  do_execute_actions+0xda0/0xdb8 [openvswitch]
[31455.992586]  ovs_execute_actions+0x70/0x1a0 [openvswitch]
[31455.992589]  ovs_dp_process_packet+0x70/0x130 [openvswitch]
[31455.992593]  ovs_vport_receive+0x70/0xe0 [openvswitch]
[31455.992596]  netdev_frame_hook+0xd0/0x1e8 [openvswitch]
[31455.992598]  __netif_receive_skb_core+0x314/0xeb8
[31455.992600]  __netif_receive_skb_one_core+0x38/0x68
[31455.992601]  __netif_receive_skb+0x28/0x80
[31455.992603]  netif_receive_skb_internal+0x3c/0xb0
[31455.992605]  napi_gro_receive+0xf8/0x170
[31455.992614]  i40e_clean_rx_irq+0x288/0xa10 [i40e]
[31455.992622]  i40e_napi_poll+0x124/0x898 [i40e]
[31455.992624]  __napi_poll+0x40/0x180
[31455.992625]  net_rx_action+0x2dc/0x370
[31455.992629]  __do_softirq+0x114/0x32c
[31455.992632]  irq_exit+0x108/0x120
[31455.992636]  __handle_domain_irq+0x6c/0xc0
[31455.992637]  gic_handle_irq+0x78/0x180
[31455.992638]  el1_irq+0xb8/0x140
[31455.992640]  arch_cpu_idle+0x38/0x1d0
[31455.992644]  default_idle_call+0x24/0x58
[31455.992646]  do_idle+0x1a4/0x268
[31455.992648]  cpu_startup_entry+0x2c/0x68
[31455.992650]  secondary_start_kernel+0x230/0x278

GVP openEuler/kernel
关闭

内容风险标识

评论 (1)

GVPopenEuler/kernel关闭

内容风险标识