[PATCH 0/2] fixed dump_backtrace() when task running on another cpu

[PATCH 0/2] fixed dump_backtrace() when task running on another cpu

[Wang Qing]

We cannot get FP and PC when the task is running on another CPU,
task->thread.cpu_context is the last time the task was switched out,
it's better to give a reminder than to provide wrong information.

We have no interface whether the task is running,
so we need to add an interface and distinguish CONFIG_SMP.

Wang Qing (2):
  [V2 1/2]sched:add task_running_oncpu
  [V2 2/2]ARM64:fixed dump_backtrace() when task running on another cpu

 arch/arm64/kernel/traps.c |  7 +++++++
 include/linux/sched.h     | 10 ++++++++++
 2 files changed, 17 insertions(+)

-- 
2.7.4


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 1/2] [V2 1/2]sched:add task_running_oncpu

[Wang Qing]

We have no interface whether the task is running,
so we need to add an interface and distinguish CONFIG_SMP.

Signed-off-by: Wang Qing <wangqing@vivo.com>
---
 include/linux/sched.h | 10 ++++++++++
 1 file changed, 10 insertions(+)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 4418f5c..13cc8f5 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1843,6 +1843,11 @@ static inline unsigned int task_cpu(const struct task_struct *p)
 
 extern void set_task_cpu(struct task_struct *p, unsigned int cpu);
 
+static inline int task_running_oncpu(const struct task_struct *p)
+{
+	return p->on_cpu;
+}
+
 #else
 
 static inline unsigned int task_cpu(const struct task_struct *p)
@@ -1854,6 +1859,11 @@ static inline void set_task_cpu(struct task_struct *p, unsigned int cpu)
 {
 }
 
+static inline int task_running_oncpu(const struct task_struct *p)
+{
+	return p == current;
+}
+
 #endif /* CONFIG_SMP */
 
 /*
-- 
2.7.4


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 2/2] [V2 2/2]ARM64:fixed dump_backtrace() when task running on another cpu

[Wang Qing]

We cannot get FP and PC when the task is running on another CPU,
task->thread.cpu_context is the last time the task was switched out,
it's better to give a reminder than to provide wrong information.

Signed-off-by: Wang Qing <wangqing@vivo.com>
---
 arch/arm64/kernel/traps.c | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c
index cf402be..831b8fd 100644
--- a/arch/arm64/kernel/traps.c
+++ b/arch/arm64/kernel/traps.c
@@ -106,6 +106,13 @@ void dump_backtrace(struct pt_regs *regs, struct task_struct *tsk)
 		start_backtrace(&frame,
 				(unsigned long)__builtin_frame_address(0),
 				(unsigned long)dump_backtrace);
+	} else if (task_running_oncpu(tsk)) {
+		/*
+		 * The task is running in another cpu, we cannot get it.
+		 */
+		pr_warn("tsk: %s is running in CPU%d, Don't call trace!\n",
+			tsk->comm, task_cpu(tsk));
+		return;
 	} else {
 		/*
 		 * task blocked in __switch_to
-- 
2.7.4


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 1/2] [V2 1/2]sched:add task_running_oncpu

[Vincent Guittot]

On Mon, 13 Apr 2020 at 14:04, Wang Qing <wangqing@vivo.com> wrote:
>
> We have no interface whether the task is running,
> so we need to add an interface and distinguish CONFIG_SMP.
>
> Signed-off-by: Wang Qing <wangqing@vivo.com>
> ---
>  include/linux/sched.h | 10 ++++++++++
>  1 file changed, 10 insertions(+)
>
> diff --git a/include/linux/sched.h b/include/linux/sched.h
> index 4418f5c..13cc8f5 100644
> --- a/include/linux/sched.h
> +++ b/include/linux/sched.h
> @@ -1843,6 +1843,11 @@ static inline unsigned int task_cpu(const struct task_struct *p)
>
>  extern void set_task_cpu(struct task_struct *p, unsigned int cpu);
>
> +static inline int task_running_oncpu(const struct task_struct *p)

This function name is too close from task_running_on_cpu() and can be
misleading as the difference is only "_"
Also, how task_running_oncpu() is different from task_running() ?

> +{
> +       return p->on_cpu;
> +}
> +
>  #else
>
>  static inline unsigned int task_cpu(const struct task_struct *p)
> @@ -1854,6 +1859,11 @@ static inline void set_task_cpu(struct task_struct *p, unsigned int cpu)
>  {
>  }
>
> +static inline int task_running_oncpu(const struct task_struct *p)
> +{
> +       return p == current;
> +}
> +
>  #endif /* CONFIG_SMP */
>
>  /*
> --
> 2.7.4
>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 1/2] [V2 1/2]sched:add task_running_oncpu

[Peter Zijlstra]

On Tue, Apr 14, 2020 at 09:20:57AM +0200, Vincent Guittot wrote:
> On Mon, 13 Apr 2020 at 14:04, Wang Qing <wangqing@vivo.com> wrote:
> >
> > We have no interface whether the task is running,
> > so we need to add an interface and distinguish CONFIG_SMP.
> >
> > Signed-off-by: Wang Qing <wangqing@vivo.com>
> > ---
> >  include/linux/sched.h | 10 ++++++++++
> >  1 file changed, 10 insertions(+)
> >
> > diff --git a/include/linux/sched.h b/include/linux/sched.h
> > index 4418f5c..13cc8f5 100644
> > --- a/include/linux/sched.h
> > +++ b/include/linux/sched.h
> > @@ -1843,6 +1843,11 @@ static inline unsigned int task_cpu(const struct task_struct *p)
> >
> >  extern void set_task_cpu(struct task_struct *p, unsigned int cpu);
> >
> > +static inline int task_running_oncpu(const struct task_struct *p)
> 
> This function name is too close from task_running_on_cpu() and can be
> misleading as the difference is only "_"
> Also, how task_running_oncpu() is different from task_running() ?

It doesn't have the (arguably superfluous) rq argument. But yes, agreed,
if anything lift that thing (without the argument).

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 1/2] [V2 1/2]sched:add task_running_oncpu

[Wang Qing]

>On Tue, Apr 14, 2020 at 09:20:57AM +0200, Vincent Guittot wrote:
>> On Mon, 13 Apr 2020 at 14:04, Wang Qing <wangqing@vivo.com> wrote:
>> >
>> > We have no interface whether the task is running,
>> > so we need to add an interface and distinguish CONFIG_SMP.
>> >
>> > Signed-off-by: Wang Qing <wangqing@vivo.com>
>> > ---
>> >  include/linux/sched.h | 10 ++++++++++
>> >  1 file changed, 10 insertions(+)
>> >
>> > diff --git a/include/linux/sched.h b/include/linux/sched.h
>> > index 4418f5c..13cc8f5 100644
>> > --- a/include/linux/sched.h
>> > +++ b/include/linux/sched.h
>> > @@ -1843,6 +1843,11 @@ static inline unsigned int task_cpu(const struct task_struct *p)
>> >
>> >  extern void set_task_cpu(struct task_struct *p, unsigned int cpu);
>> >
>> > +static inline int task_running_oncpu(const struct task_struct *p)
>> 
>> This function name is too close from task_running_on_cpu() and can be
>> misleading as the difference is only "_"
>> Also, how task_running_oncpu() is different from task_running() ?
>
>It doesn't have the (arguably superfluous) rq argument. But yes, agreed,
>if anything lift that thing (without the argument).

I think task_running() should be renamed to task_running_on_rq() like
the naming of task_running_on_cpu(), this is what it originally mean,
and add task_running() (with the task argument only).

I updated the patch for that.

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel