include/linux/sched/debug.h | 1 + kernel/hung_task.c | 4 ++-- kernel/sched/core.c | 19 +++++++++++++------ 3 files changed, 16 insertions(+), 8 deletions(-)
Hi,
A request for comments, not a formal patch.
Is there any reason why hung-task watchdog prints error header
with KERN_ERR log level while task's backtrace is printed with
KERN_INFO? Will it make sense to unify log levels and introduce
sched_show_task() variant that accept log level param so that
everything that hung-task watchdog prints becomes KERN_ERR?
Something like this:
---
include/linux/sched/debug.h | 1 +
kernel/hung_task.c | 4 ++--
kernel/sched/core.c | 19 +++++++++++++------
3 files changed, 16 insertions(+), 8 deletions(-)
diff --git a/include/linux/sched/debug.h b/include/linux/sched/debug.h
index b5035afa2396..8e46854d16fc 100644
--- a/include/linux/sched/debug.h
+++ b/include/linux/sched/debug.h
@@ -34,6 +34,7 @@ extern void show_stack(struct task_struct *task, unsigned long *sp,
const char *loglvl);
extern void sched_show_task(struct task_struct *p);
+extern void sched_show_task_log_lvl(struct task_struct *p, const char *lvl);
#ifdef CONFIG_SCHED_DEBUG
struct seq_file;
diff --git a/kernel/hung_task.c b/kernel/hung_task.c
index ccd7217fcec1..23dec32a05f3 100644
--- a/kernel/hung_task.c
+++ b/kernel/hung_task.c
@@ -119,7 +119,7 @@ static void debug_show_blocker(struct task_struct *task)
if ((unsigned long)t == owner) {
pr_err("INFO: task %s:%d is blocked on a mutex likely owned by task %s:%d.\n",
task->comm, task->pid, t->comm, t->pid);
- sched_show_task(t);
+ sched_show_task_log_lvl(t, KERN_ERR);
return;
}
}
@@ -186,7 +186,7 @@ static void check_hung_task(struct task_struct *t, unsigned long timeout)
pr_err(" Blocked by coredump.\n");
pr_err("\"echo 0 > /proc/sys/kernel/hung_task_timeout_secs\""
" disables this message.\n");
- sched_show_task(t);
+ sched_show_task_log_lvl(t, KERN_ERR);
debug_show_blocker(t);
hung_task_show_lock = true;
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 00ea6f31973c..b6a5a41fbbfc 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -7687,7 +7687,7 @@ void __sched io_schedule(void)
}
EXPORT_SYMBOL(io_schedule);
-void sched_show_task(struct task_struct *p)
+void sched_show_task_log_lvl(struct task_struct *p, const char *lvl)
{
unsigned long free;
int ppid;
@@ -7695,7 +7695,8 @@ void sched_show_task(struct task_struct *p)
if (!try_get_task_stack(p))
return;
- pr_info("task:%-15.15s state:%c", p->comm, task_state_to_char(p));
+ printk("%stask:%-15.15s state:%c", lvl,
+ p->comm, task_state_to_char(p));
if (task_is_running(p))
pr_cont(" running task ");
@@ -7709,12 +7710,18 @@ void sched_show_task(struct task_struct *p)
free, task_pid_nr(p), task_tgid_nr(p),
ppid, p->flags, read_task_thread_flags(p));
- print_worker_info(KERN_INFO, p);
- print_stop_info(KERN_INFO, p);
- print_scx_info(KERN_INFO, p);
- show_stack(p, NULL, KERN_INFO);
+ print_worker_info(lvl, p);
+ print_stop_info(lvl, p);
+ print_scx_info(lvl, p);
+ show_stack(p, NULL, lvl);
put_task_stack(p);
}
+EXPORT_SYMBOL_GPL(sched_show_task_log_lvl);
+
+void sched_show_task(struct task_struct *p)
+{
+ sched_show_task_log_lvl(p, KERN_INFO);
+}
EXPORT_SYMBOL_GPL(sched_show_task);
static inline bool
--
2.49.0.rc0.332.g42c0ae87b1-googHi Sergey,
On Mon, Mar 10, 2025 at 5:25 PM Sergey Senozhatsky
<senozhatsky@chromium.org> wrote:
>
> Hi,
>
> A request for comments, not a formal patch.
>
> Is there any reason why hung-task watchdog prints error header
> with KERN_ERR log level while task's backtrace is printed with
> KERN_INFO? Will it make sense to unify log levels and introduce
> sched_show_task() variant that accept log level param so that
> everything that hung-task watchdog prints becomes KERN_ERR?
Thanks a lot for looking into this.
Let me just add that I've been looking into ways to automatically
analyze kernel crash reports and I noticed that when I filter out
KERN_INFO and higher log levels, I end up losing useful information
for hung-task watchdog-kind of problems. This change would greatly
help in filtering out unnecessary noise from the logs.
By the way, if having it as KERN_INFO by default would still be
desirable, I suppose we could add a KConfig option to set the desired
log level?
Best regards,
Tomasz
>
> Something like this:
>
> ---
> include/linux/sched/debug.h | 1 +
> kernel/hung_task.c | 4 ++--
> kernel/sched/core.c | 19 +++++++++++++------
> 3 files changed, 16 insertions(+), 8 deletions(-)
>
> diff --git a/include/linux/sched/debug.h b/include/linux/sched/debug.h
> index b5035afa2396..8e46854d16fc 100644
> --- a/include/linux/sched/debug.h
> +++ b/include/linux/sched/debug.h
> @@ -34,6 +34,7 @@ extern void show_stack(struct task_struct *task, unsigned long *sp,
> const char *loglvl);
>
> extern void sched_show_task(struct task_struct *p);
> +extern void sched_show_task_log_lvl(struct task_struct *p, const char *lvl);
>
> #ifdef CONFIG_SCHED_DEBUG
> struct seq_file;
> diff --git a/kernel/hung_task.c b/kernel/hung_task.c
> index ccd7217fcec1..23dec32a05f3 100644
> --- a/kernel/hung_task.c
> +++ b/kernel/hung_task.c
> @@ -119,7 +119,7 @@ static void debug_show_blocker(struct task_struct *task)
> if ((unsigned long)t == owner) {
> pr_err("INFO: task %s:%d is blocked on a mutex likely owned by task %s:%d.\n",
> task->comm, task->pid, t->comm, t->pid);
> - sched_show_task(t);
> + sched_show_task_log_lvl(t, KERN_ERR);
> return;
> }
> }
> @@ -186,7 +186,7 @@ static void check_hung_task(struct task_struct *t, unsigned long timeout)
> pr_err(" Blocked by coredump.\n");
> pr_err("\"echo 0 > /proc/sys/kernel/hung_task_timeout_secs\""
> " disables this message.\n");
> - sched_show_task(t);
> + sched_show_task_log_lvl(t, KERN_ERR);
> debug_show_blocker(t);
> hung_task_show_lock = true;
>
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index 00ea6f31973c..b6a5a41fbbfc 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -7687,7 +7687,7 @@ void __sched io_schedule(void)
> }
> EXPORT_SYMBOL(io_schedule);
>
> -void sched_show_task(struct task_struct *p)
> +void sched_show_task_log_lvl(struct task_struct *p, const char *lvl)
> {
> unsigned long free;
> int ppid;
> @@ -7695,7 +7695,8 @@ void sched_show_task(struct task_struct *p)
> if (!try_get_task_stack(p))
> return;
>
> - pr_info("task:%-15.15s state:%c", p->comm, task_state_to_char(p));
> + printk("%stask:%-15.15s state:%c", lvl,
> + p->comm, task_state_to_char(p));
>
> if (task_is_running(p))
> pr_cont(" running task ");
> @@ -7709,12 +7710,18 @@ void sched_show_task(struct task_struct *p)
> free, task_pid_nr(p), task_tgid_nr(p),
> ppid, p->flags, read_task_thread_flags(p));
>
> - print_worker_info(KERN_INFO, p);
> - print_stop_info(KERN_INFO, p);
> - print_scx_info(KERN_INFO, p);
> - show_stack(p, NULL, KERN_INFO);
> + print_worker_info(lvl, p);
> + print_stop_info(lvl, p);
> + print_scx_info(lvl, p);
> + show_stack(p, NULL, lvl);
> put_task_stack(p);
> }
> +EXPORT_SYMBOL_GPL(sched_show_task_log_lvl);
> +
> +void sched_show_task(struct task_struct *p)
> +{
> + sched_show_task_log_lvl(p, KERN_INFO);
> +}
> EXPORT_SYMBOL_GPL(sched_show_task);
>
> static inline bool
> --
> 2.49.0.rc0.332.g42c0ae87b1-goog
>
On Wed, Mar 12, 2025 at 11:51 PM Tomasz Figa <tfiga@chromium.org> wrote: > > Hi Sergey, > > On Mon, Mar 10, 2025 at 5:25 PM Sergey Senozhatsky > <senozhatsky@chromium.org> wrote: > > > > Hi, > > > > A request for comments, not a formal patch. > > > > Is there any reason why hung-task watchdog prints error header > > with KERN_ERR log level while task's backtrace is printed with > > KERN_INFO? Will it make sense to unify log levels and introduce > > sched_show_task() variant that accept log level param so that > > everything that hung-task watchdog prints becomes KERN_ERR? > > Thanks a lot for looking into this. > > Let me just add that I've been looking into ways to automatically > analyze kernel crash reports and I noticed that when I filter out > KERN_INFO and higher log levels, I end up losing useful information > for hung-task watchdog-kind of problems. This change would greatly > help in filtering out unnecessary noise from the logs. > > By the way, if having it as KERN_INFO by default would still be > desirable, I suppose we could add a KConfig option to set the desired > log level? Gentle ping on this patch. Best regards, Tomasz
On (25/04/24 13:34), Tomasz Figa wrote: > > > Is there any reason why hung-task watchdog prints error header > > > with KERN_ERR log level while task's backtrace is printed with > > > KERN_INFO? Will it make sense to unify log levels and introduce > > > sched_show_task() variant that accept log level param so that > > > everything that hung-task watchdog prints becomes KERN_ERR? > > > > Thanks a lot for looking into this. > > > > Let me just add that I've been looking into ways to automatically > > analyze kernel crash reports and I noticed that when I filter out > > KERN_INFO and higher log levels, I end up losing useful information > > for hung-task watchdog-kind of problems. This change would greatly > > help in filtering out unnecessary noise from the logs. > > > > By the way, if having it as KERN_INFO by default would still be > > desirable, I suppose we could add a KConfig option to set the desired > > log level? > > Gentle ping on this patch. Yeah, sorry. Gentle `git send-email $1` on this. $1 https://lore.kernel.org/linux-kernel/20250424070436.2380215-1-senozhatsky@chromium.org
© 2016 - 2026 Red Hat, Inc.