Re: [PATCH] hungtask: add filter kthread/check comm

From: Vlastimil Babka
Date: Wed May 19 2021 - 07:28:59 EST


On 5/18/21 1:16 PM, chenguanyou wrote:

The description is missing.

> Signed-off-by: chenguanyou <chenguanyou@xxxxxxxxxx>
> ---
> include/linux/sched/sysctl.h | 4 ++++
> kernel/hung_task.c | 17 +++++++++++++++++
> kernel/sysctl.c | 15 +++++++++++++++
> lib/Kconfig.debug | 23 +++++++++++++++++++++++
> 4 files changed, 59 insertions(+)
>
> diff --git a/include/linux/sched/sysctl.h b/include/linux/sched/sysctl.h
> index 3c31ba88aca5..e8a9a28215bf 100644
> --- a/include/linux/sched/sysctl.h
> +++ b/include/linux/sched/sysctl.h
> @@ -14,11 +14,15 @@ extern unsigned int sysctl_hung_task_all_cpu_backtrace;
> #define sysctl_hung_task_all_cpu_backtrace 0
> #endif /* CONFIG_SMP */
>
> +#define TASK_COMM_LEN 16
> +
> extern int sysctl_hung_task_check_count;
> extern unsigned int sysctl_hung_task_panic;
> extern unsigned long sysctl_hung_task_timeout_secs;
> extern unsigned long sysctl_hung_task_check_interval_secs;
> extern int sysctl_hung_task_warnings;
> +extern unsigned int sysctl_hung_task_filter_kthread;
> +extern char sysctl_hung_task_check_comm[TASK_COMM_LEN];
> int proc_dohung_task_timeout_secs(struct ctl_table *table, int write,
> void *buffer, size_t *lenp, loff_t *ppos);
> #else
> diff --git a/kernel/hung_task.c b/kernel/hung_task.c
> index 396ebaebea3f..e018563d4882 100644
> --- a/kernel/hung_task.c
> +++ b/kernel/hung_task.c
> @@ -23,6 +23,7 @@
> #include <linux/sched/sysctl.h>
>
> #include <trace/events/sched.h>
> +#include <linux/string.h>
>
> /*
> * The number of tasks checked:
> @@ -48,6 +49,16 @@ unsigned long __read_mostly sysctl_hung_task_timeout_secs = CONFIG_DEFAULT_HUNG_
> */
> unsigned long __read_mostly sysctl_hung_task_check_interval_secs;
>
> +/*
> + * Non zero means no checking kthread
> + */
> +unsigned int __read_mostly sysctl_hung_task_filter_kthread = CONFIG_DEFAULT_HUNG_TASK_FILTER_KTHREAD;
> +
> +/*
> + * Only one
> + */
> +char __read_mostly sysctl_hung_task_check_comm[TASK_COMM_LEN] = CONFIG_DEFAULT_HUNG_TASK_CHECK_COMM;
> +
> int __read_mostly sysctl_hung_task_warnings = 10;
>
> static int __read_mostly did_panic;
> @@ -88,6 +99,12 @@ static void check_hung_task(struct task_struct *t, unsigned long timeout)
> {
> unsigned long switch_count = t->nvcsw + t->nivcsw;
>
> + if (unlikely(strlen(sysctl_hung_task_check_comm) && !strstr(t->comm, sysctl_hung_task_check_comm)))
> + return;
> +
> + if (unlikely(sysctl_hung_task_filter_kthread && t->flags & PF_KTHREAD))
> + return;
> +
> /*
> * Ensure the task is not frozen.
> * Also, skip vfork and any other user process that freezer should skip.
> diff --git a/kernel/sysctl.c b/kernel/sysctl.c
> index 62fbd09b5dc1..157c47a8430a 100644
> --- a/kernel/sysctl.c
> +++ b/kernel/sysctl.c
> @@ -2519,6 +2519,21 @@ static struct ctl_table kern_table[] = {
> .proc_handler = proc_dointvec_minmax,
> .extra1 = &neg_one,
> },
> + {
> + .procname = "hung_task_filter_kthread",
> + .data = &sysctl_hung_task_filter_kthread,
> + .maxlen = sizeof(int),
> + .mode = 0644,
> + .extra1 = SYSCTL_ZERO,
> + .extra2 = SYSCTL_ONE,
> + },
> + {
> + .procname = "hung_task_check_comm",
> + .data = &sysctl_hung_task_check_comm,
> + .maxlen = TASK_COMM_LEN,
> + .mode = 0644,
> + .proc_handler = proc_dostring,
> + },
> #endif
> #ifdef CONFIG_RT_MUTEXES
> {
> diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
> index 2c7f46b366f1..59cfa9e230ee 100644
> --- a/lib/Kconfig.debug
> +++ b/lib/Kconfig.debug
> @@ -1101,6 +1101,29 @@ config DEFAULT_HUNG_TASK_TIMEOUT
> A timeout of 0 disables the check. The default is two minutes.
> Keeping the default should be fine in most cases.
>
> +config DEFAULT_HUNG_TASK_FILTER_KTHREAD
> + bool "Default filter kthread for hung task"
> + depends on DETECT_HUNG_TASK
> + default 0
> + help
> + This option controls filter kthread used to determine when
> + a kernel task has become "state=TASK_UNINTERRUPTIBLE" and should be skip.
> +
> + It can be adjusted at runtime via the kernel.hung_task_filter_kthread
> + sysctl or by writing a value to
> + /proc/sys/kernel/hung_task_filter_kthread.
> +
> + A filter of 1 disables the check
> +
> +config DEFAULT_HUNG_TASK_CHECK_COMM
> + string "Default check only one comm"
> + depends on DETECT_HUNG_TASK
> + default ""
> + help
> + It can be adjusted at runtime via the kernel.hung_task_check_comm
> + sysctl or by writing a value to
> + /proc/sys/kernel/hung_task_check_comm.
> +
> config BOOTPARAM_HUNG_TASK_PANIC
> bool "Panic (Reboot) On Hung Tasks"
> depends on DETECT_HUNG_TASK
>