【OLK-6.6】arm版本hungtask_monitor触发系统自愈时不输出CPU调用栈

【特性描述】
arm版本hungtask_monitor触发系统自愈时输出CPU调用栈

【特性竞争力】
hungtask_monitor在系统自愈后能正常在dmesg中打印调用栈

【硬件架构】
ARM

【特性约束】
NA

【涉及仓库】
OLK-5.10 OLK-6.6

【交付个人/团队】
 @jeffrey-0liao

【验证信息】
特性依赖：
CONFIG_DETECT_HUNG_TASK=y
CONFIG_WATCHDOG=y

启动项：
gic-version=3
irqchip.gicv3_pseudo_nmi=1
kgdboc=ttyAMA0,115200

用例：
打补丁前：
fsfreeze --freeze /
echo a > /a

363.535748] INFO: task sh:207 blocked for more than 120 seconds.
363.536419]	Not tainted 6.6.0-01760-gf7998b87456a #5
363.5369931	"echo 0 > /proc/sys/kernel/hung task timeout secs" disables this message.
363.537529]	task:sh	state:D stack:0   pid:207  ppid:1	flags:0x00000000
363.538507]	Call trace:
363.5392201	switch to+0xd0/0x1a4
363.5399661	schedule+0x2a0/0xa9c
363.540616]	schedule+0x5c/0xc4
363.541157]	percpu rwsem wait+0xe0/0xfc
363.541729]	percpu down read+0x4c/0x98
363.542334]	mnt want write+0x17c/0x1f0
363.543085]	path openat+0xe88/0xe98
363.543734]	do filp_open+0x84/0x108
363.5444151	do_sys_openat2+0x210/0x248
363.545144]	do sys open+0x58/0xa4
363.545691]	arm64 sys openat+0x24/0x30
363.5464361	invoke syscall.constprop.0+0x4c/0xd8  
363.547487]	do el0 svc+0x40/0xc8
363.547974]	el0 svc+0x54/0x1b0
363.5485121	elOt 64 sync handler+0xc0/0xc4
363.549176]	elOt 64 sync+0x1a4/0x1a8
363.550226] Kernel panic - not syncing: hung_task: blocked tasks
363.551678] CPU: 5 PID: 58 Comm: khungtaskd Not tainted 6.6.0-01760-gf7998b87456a #5
363.553341] Hardware name: Linux,dummy-virt (DT)
363.5542721	Call trace:
363.5546881	dump backtrace+0x94/0xec
363.5552311	show stack+0x18/0x24
363.555829]	dump_stack lvl+0x74/0xcO
363.5563031	dump stack+0x18/0x24
363.556557]	panic+0x378/0x3f0
363.556800]	watchdog+0x2dc/0x514
363.5570731	kthread+0x11c/0x128
363.557325]	ret from fork+0x10/0x20
363.558337] SMP: stopping secondary CPUse
363.559480] Kernel Offset: 0x285dace00000 from Oxffff800080000000
363.559988] PHYS OFFSET: 0xffff93f280000000
363.560377]	CPU features: 0x1,10000003,3c020000,7000421b
363.561084] Memory Limit: none
363.561756]	__-[ end Kernel panic - not syncing: hung task: blocked tasks

kernel panic，无调用栈打印

打补丁后：
echo 1 > /proc/sys/kernel/hung_task_all_cpu_backtrace
fsfreeze --freeze /
echo a > /a

[ 216.067707] INFO: task sh:206 blocked for more than 30 seconds.
216.0682407	Not tainted 6.6.0-01771-ga2b9a1cb048f #2
216.068617]	"echo 0 > /proc/sys/kernel/hung task timeout secs" disables this message.
216.069108]	task:sh	state:D stack:0	pid:206  ppid:1	flags:0x00000000
216.069931]	Call trace:
216.0702011	switch to+0xd0/0x1a4
216.0706511	schedule+0x29c/0xa64
216.0708851	schedule+0x5c/0xc4
216.071069]	percpu_rwsem wait+0xe0/0xfc
216.0716911	percpu down read+0x4c/0x98
216.0719451	mnt want write+0x17c/0x1f0
216.072122]	path openat+0xe88/0xe98
216.0723221	do filp open+0x84/0x108
216.072534]	do_sys_openat2+0x210/0x248
216.072720]	do sys open+0x58/0xa4
216.0729431	arm64 sys openat+0x24/0x30
216.0732021	invoke syscall.constprop.0+0x4c/0xd8
216.0735501	do el0 svc+0x40/0xc8
216.073790]	el0 svc+0x4c/0x158
216.073997]	elOt 64 sync handler+0xc0/0xc4
216.0742311	elOt 64 sync+0x190/0x194
216.0749561	NMI backtrace for cpu 6
216.075576] CPU: 6 PID: 58 Comm: khungtaskd Not tainted 6.6.0-01771-ga2b9a1cb048f #2
216.076248]	Hardware name: Linux,dummy-virt (DT)
216.0765721	Call trace:
216.0768411	dump backtrace+0x94/0xec
216.077020]	show stack+0x18/0x24
216.0771281	dump stack lvl+0x48/0x60
216.0773381	dump stack+0x18/0x24
216.0775311	nmi cpu backtrace+0x110/0x178
216.077764]	nmi_trigger_cpumask backtrace+0x1f8/0x258
216.0779351	arch trigger cpumask backtrace+0x40/0x50
216.0781807	watchdog+0x4f0/0x50c
216.0783811	kthread+0x11c/0x128
216.0785651	ret from fork+0x10/0x20
216.079158] Sending NMI from CPU 6 to CPUs 0-5,7:
216.0800011	NMI backtrace for cpu 0 skipped: idling at default idle call+0x28/0x3c
216.080256] NMI backtrace for cpu 2 skipped: idling at default idle_call+0x28/0x3ce
216.0803431	NMI backtrace for cpu 7
216.080461]	CPU: 7 PID:110 Comm: syslogd Not tainted 6.6.0-01771-ga2b9a1cb048f #2
216.080488] Hardware name: Linux,dummy-virt (DT)
216.080639]	pstate: 20000000 (nzCv daif -PAN -UAO -TCO -DIT -SSBS BTYPE=--
216.080759]	pc : 0000ffffa700e410
216.0808721	lr : 0000aaaad083f79c
216.080982] sp : 0000ffffc16d6b20
216.081177] x29: 0000ffffc16d6b20 x28: 0000aaaad0906000 x27:0000000000000000

正常打印调用栈

GVP openEuler/kernel

内容风险标识

评论 (1)

GVPopenEuler/kernel

内容风险标识