【openEuler-1.0-LTS】sched: Fix enqueue_task_fair() warning

【标题描述】被throttled的cfs_rq其nr_running可能大于1，无法再次加入leaf链表，导致父子链表关系不完整，触发warning
【环境信息】
硬件信息：
arm64/x86
软件信息：
1）	OS版本及分支：openEuler 1.0 LTS
2）	内核信息：NA
3）	发现问题的组件版本信息：NA
如果有特殊组网，请提供网络拓扑图
【问题复现步骤】
1.分别创建cpu和cpuset组，tg1, tg1/tg2, tg1/tg2/tg3, cpuset.cpus 都设置0-5，cfs quato都设置30000
2.运行3个 rt-app【6个实例，运行时间与周期：25000/100000】，分别绑定tg1,tg1/tg2,tg1/tg2/tg3,间隔4s运行。

具体如下：
mkdir -p /sys/fs/cgroup/cpuset/tg1/
mkdir -p /sys/fs/cgroup/cpuset/tg1/tg2
mkdir -p /sys/fs/cgroup/cpuset/tg1/tg2/tg3

mkdir -p /sys/fs/cgroup/cpu/tg1/
mkdir -p /sys/fs/cgroup/cpu/tg1/tg2
mkdir -p /sys/fs/cgroup/cpu/tg1/tg2/tg3
echo 30000 > /sys/fs/cgroup/cpu/tg1/cpu.cfs_quota_us
echo 30000 > /sys/fs/cgroup/cpu/tg1/tg2/cpu.cfs_quota_us
echo 30000 > /sys/fs/cgroup/cpu/tg1/tg2/tg3/cpu.cfs_quota_us

echo 0-5 > /sys/fs/cgroup/cpuset/tg1/cpuset.cpus
echo 0 > /sys/fs/cgroup/cpuset/tg1/cpuset.mems
echo 0-5 > /sys/fs/cgroup/cpuset/tg1/tg2/cpuset.cpus
echo 0 > /sys/fs/cgroup/cpuset/tg1/tg2/cpuset.mems
echo 0-5 > /sys/fs/cgroup/cpuset/tg1/tg2/tg3/cpuset.cpus
echo 0 > /sys/fs/cgroup/cpuset/tg1/tg2/tg3/cpuset.mems

cgexec -g cpuset:tg1 -g cpu:tg1 /home/tanghui/rt-app/src/rt-app /home/tanghui/rt-app/config/bw-warn.json >> logout-1.txt 2>&1 &
sleep 4
cgexec -g cpuset:tg1/tg2 -g cpu:tg1/tg2 /home/tanghui/rt-app/src/rt-app /home/tanghui/rt-app/config/bw-warn.json >> logout-1.txt 2>&1 &
sleep 4
cgexec -g cpuset:tg1/tg2/tg3 -g cpu:tg1/tg2/tg3 /home/tanghui/rt-app/src/rt-app /home/tanghui/rt-app/config/bw-warn.json >> logout-1.txt 2>&1 &

出现概率(是否必现，概率性错误)必现
【预期结果】
dmesg 日志不会出现warning
【实际结果】
May 12 12:05:01 euleros-pxe kernel: [ 4171.952084] ------------[ cut here ]------------
May 12 12:05:01 euleros-pxe kernel: [ 4171.952087] rq->tmp_alone_branch != &rq->leaf_cfs_rq_list
May 12 12:05:01 euleros-pxe kernel: [ 4171.952115] WARNING: CPU: 26 PID: 897674 at kernel/sched/fair.c:408 enqueue_task_fair+0x676/0x820
May 12 12:05:01 euleros-pxe kernel: [ 4171.952117] Modules linked in: fuse loop binfmt_misc bridge stp llc libcrc32c ip_set nfnetlink sunrpc intel_rapl sb_edac x86_pkg_temp_thermal kvm_intel kvm irqbypass crct10dif_pclmul crc32_pclmul ghash_clmulni_intel intel_rapl_perf intel_cstate intel_uncore ipmi_si ipmi_devintf ipmi_msghandler pcspkr ses enclosure scsi_transport_sas i2c_i801 joydev mei_me sg lpc_ich mei drm ext4 mbcache jbd2 sd_mod crc32c_intel ixgbe ahci libahci mdio dca libata tg3 megaraid_sas dm_mirror dm_region_hash dm_log dm_mod [last unloaded: ip_tables]
May 12 12:05:01 euleros-pxe kernel: [ 4171.952209] CPU: 26 PID: 897674 Comm: sh Kdump: loaded Not tainted 4.18.0-147.5.2.10.h906.kasan.eulerosv2r10.x86_64 #1
May 12 12:05:01 euleros-pxe kernel: [ 4171.952213] Hardware name: Huawei RH2288H V3/BC11HGSA0, BIOS 1.57 08/11/2015
May 12 12:05:01 euleros-pxe kernel: [ 4171.952220] RIP: 0010:enqueue_task_fair+0x676/0x820
May 12 12:05:01 euleros-pxe kernel: [ 4171.952227] Code: 1d 90 76 91 02 80 fb 01 0f 87 16 01 00 00 83 e3 01 0f 85 43 fb ff ff 48 c7 c7 60 f3 48 bb c6 05 70 76 91 02 01 e8 0a a5 f7 ff <0f> 0b e9 29 fb ff ff 49 8d bd 30 0a 00 00 e8 37 1d 36 00 48 c7 c7
May 12 12:05:01 euleros-pxe kernel: [ 4171.952232] RSP: 0018:ffff88e07f607ca8 EFLAGS: 00010086
May 12 12:05:01 euleros-pxe kernel: [ 4171.952237] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffffffffba01a9de
May 12 12:05:01 euleros-pxe kernel: [ 4171.952240] RDX: 0000000000010000 RSI: 0000000000000008 RDI: ffff88e07f61f9c0
May 12 12:05:01 euleros-pxe kernel: [ 4171.952244] RBP: 0000000000000049 R08: ffffed1c0fec3f39 R09: ffffed1c0fec3f39
May 12 12:05:01 euleros-pxe kernel: [ 4171.952255] R10: ffff88e07f607ca0 R11: ffffed1c0fec3f38 R12: ffff88e07f82c4c0
May 12 12:05:01 euleros-pxe kernel: [ 4171.952258] R13: ffff88e07f82c440 R14: ffff88b0e18681d8 R15: 0000000000000000
May 12 12:05:01 euleros-pxe kernel: [ 4171.952263] FS: 0000000000000000(0000) GS:ffff88e07f600000(0000) knlGS:0000000000000000
May 12 12:05:01 euleros-pxe kernel: [ 4171.952266] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
May 12 12:05:01 euleros-pxe kernel: [ 4171.952270] CR2: 000056271fb43f88 CR3: 000000314116a003 CR4: 00000000001606e0
May 12 12:05:01 euleros-pxe kernel: [ 4171.952272] Call Trace:
May 12 12:05:01 euleros-pxe kernel: [ 4171.952275] <IRQ>
May 12 12:05:01 euleros-pxe kernel: [ 4171.952282] ttwu_do_activate+0x5c/0xb0
May 12 12:05:01 euleros-pxe kernel: [ 4171.952289] try_to_wake_up+0x3fd/0xa30
May 12 12:05:01 euleros-pxe kernel: [ 4171.952296] ? migrate_swap_stop+0x360/0x360
May 12 12:05:01 euleros-pxe kernel: [ 4171.952305] ? timerqueue_del+0x58/0x90
May 12 12:05:01 euleros-pxe kernel: [ 4171.952314] hrtimer_wakeup+0x36/0x40
May 12 12:05:01 euleros-pxe kernel: [ 4171.952321] __hrtimer_run_queues+0x253/0x540
May 12 12:05:01 euleros-pxe kernel: [ 4171.952328] ? __hrtimer_init+0x180/0x180
May 12 12:05:01 euleros-pxe kernel: [ 4171.952335] ? hrtimer_cancel+0x40/0x40
May 12 12:05:01 euleros-pxe kernel: [ 4171.952344] ? sched_clock+0x5/0x10
May 12 12:05:01 euleros-pxe kernel: [ 4171.952351] ? sched_clock_cpu+0x18/0x180
May 12 12:05:01 euleros-pxe kernel: [ 4171.952357] ? recalibrate_cpu_khz+0x10/0x10
May 12 12:05:01 euleros-pxe kernel: [ 4171.952365] ? ktime_get_update_offsets_now+0x163/0x200
May 12 12:05:01 euleros-pxe kernel: [ 4171.952373] hrtimer_interrupt+0x16f/0x440
May 12 12:05:01 euleros-pxe kernel: [ 4171.952384] smp_apic_timer_interrupt+0x88/0x1c0
May 12 12:05:01 euleros-pxe kernel: [ 4171.952398] apic_timer_interrupt+0xf/0x20

GVP openEuler/kernel

内容风险标识

评论 (1)

GVPopenEuler/kernel

内容风险标识