Description of problem: anolisos8 x86_64执行ltp冒烟测试产出crash: general protection fault, probably for non-canonical address 0xff44c675ff48089d: 0000 [#1] SMP PTI CPU: 11 PID: 253900 Comm: setsockopt09 Kdump: loaded Tainted: G E 5.10.84-10.3.an8.x86_64 #1 RIP: 0010:queued_spin_lock_slowpath+0x176/0x1b0 vmcore-dmesg.txt部分日志: [ 2446.115446] Call Trace: [ 2446.115803] packet_set_ring+0x270/0x730 [ 2446.116349] ? kmem_cache_alloc+0x431/0x490 [ 2446.116925] packet_setsockopt+0x37d/0x1120 [ 2446.117503] ? alloc_empty_file+0x6f/0xc0 [ 2446.118058] ? alloc_file_pseudo+0x97/0xf0 [ 2446.118621] __sys_setsockopt+0xba/0x1a0 [ 2446.119163] __x64_sys_setsockopt+0x20/0x30 [ 2446.119742] do_syscall_64+0x33/0x40 [ 2446.120240] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 2446.120945] RIP: 0033:0x7f24c34fe41e # cat /etc/image-id image_name="Anolis OS 8.4 ANCK 64 bit" image_id="anolisos_8_4_x64_20G_anck_alibase_20220518.vhd" release_date="20220518111246" Version-Release number of selected component (if applicable): 5.10.84-10.3.an8.x86_64 How reproducible: Steps to Reproduce: 1.git clone https://github.com/linux-test-project/ltp yum install gcc-c++ gcc git libaio* kernel-debuginfo -y --skip-broken cd ltp make autotools && ./configure && make && make install mkdir /disk1 wipefs -a --force /dev/vdb mkfs -t ext4 -q -F /dev/vdb mount -t ext4 /dev/vdb /disk1 mkdir -p /disk1/tmpdir/ltp lsblk mount | grep vdb cd /opt/ltp vim load.sh #!/bin/bash echo 1 > /proc/sys/kernel/panic echo 1 > /proc/sys/kernel/hardlockup_panic echo 1 > /proc/sys/kernel/softlockup_panic echo 50 > /proc/sys/kernel/watchdog_thresh echo 1200 > /proc/sys/kernel/hung_task_timeout_secs echo 0 > /proc/sys/kernel/hung_task_panic nr_cpu=$(nproc) mem_kb=$(grep ^MemTotal /proc/meminfo | awk '{print $2}') ./runltp \ -c $((nr_cpu / 2)) \ -m $((nr_cpu / 4)),4,$((mem_kb / nr_cpu / 2 * 1024)),1 \ -D $((nr_cpu / 10)),1,0,1 \ -i 2 \ -B ext4 \ -R -p -q \ -t 24h \ -d /disk1/tmpdir/ltp chmod +x load.sh nohup ./load.sh > t1.log & 2. 3. Actual results: 产生crash Expected results: 正常运行24h,无crash,hung及已知问题 Additional info:
crash解析:http://vmcore.alibaba-inc.com/vmcore_detail/20220526115707_192.92.10.113/
实例ecs.g7a.32xlarge有相同的问题 镜像:anolisos_8_4_x64_20G_anck_alibase_20220518.vhd [ 2734.051959] general protection fault, probably for non-canonical address 0xff44c675ff48089d: 0000 [#1] SMP NOPTI [ 2734.053406] CPU: 6 PID: 87526 Comm: setsockopt09 Kdump: loaded Tainted: G E 5.10.84-10.3.an8.x86_64 #1 [ 2734.067205] Call Trace: [ 2734.067563] packet_set_ring+0x270/0x730 [ 2734.068098] ? kmem_cache_alloc+0x431/0x490 [ 2734.068682] packet_setsockopt+0x37d/0x1120 [ 2734.069253] ? alloc_empty_file+0x6f/0xc0 [ 2734.069794] ? alloc_file_pseudo+0x97/0xf0 [ 2734.070350] __sys_setsockopt+0xba/0x1a0 [ 2734.070884] __x64_sys_setsockopt+0x20/0x30 [ 2734.071451] do_syscall_64+0x33/0x40 [ 2734.071941] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 2734.072617] RIP: 0033:0x7f7e322fe41e [ 2734.073103] Code: d8 64 89 02 48 c7 c0 ff ff ff ff eb b1 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 f3 0f 1e fa 49 89 ca b8 36 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d 3a 0a 2c 00 f7 d8 64 89 01 48 # cat /etc/image-id image_name="Anolis OS 8.4 ANCK 64 bit" image_id="anolisos_8_4_x64_20G_anck_alibase_20220518.vhd" release_date="20220518111246" vmcore镜像:http://vmcore.alibaba-inc.com/vmcore_detail/20220607140938_192.108.10.6/ 其余信息同上