On (19/11/13 10:39), Steven Rostedt wrote:
[..]
> > void show_stack(struct task_struct *task, unsigned long *sp, int log_level)
> > {
> > printk_emergency_enter(log_level);
> > __show_stack(task, sp);
> > printk_emergency_exit();
> > }
> > // - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - //
> >
> > show_stack() never schedules, disabling preemption around it should
> > not change anything. Should it be interrupted, we will handle it via
> > preempt count.
>
> Please no! The whole point of the printk rewrite was to allow for
> printk to be preemptible and used in more contexts. The show_stack() can
> be all over the place and is not a fast function. Let's not disable
> preemption for it.
I never said that this code should be used all over the place. What
I did say several times was that this code is for quick debugging,
when one sits behind a slow serial console and wants to tweak loglevel
of all printk-s of a particular context/function only.
-ss
_______________________________________________
Kgdb-bugreport mailing list
Kgdb-bugreport@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/kgdb-bugreport
