Re: [PATCH 08/10] posix-cpu-timers: Migrate to use new tick dependency mask model

[Chris Metcalf <cmetcalf@ezchip.com>]

On 07/23/2015 12:42 PM, Frederic Weisbecker wrote:
> +static void cpu_timer_list_dequeue(struct cpu_timer_list *t)
> +{
> +	if (!list_empty(&t->entry))
> +		cpu_timer_dec_tick_dependency();
> +	list_del_init(&t->entry);
> +}

Is the list_empty() test necessary? It wasn't in the original 
posix-timers code, and it feels like a pretty serious bug if you're 
doing a list_del on an empty list.

At a higher level, is the posix-cpu-timers code here really providing 
the right semantics? It seems like before, the code was checking a 
struct task-specific state, and now you are setting a global state such 
that if ANY task anywhere in the system (even on housekeeping cores) has 
a pending posix cpu timer, then nothing can go into nohz_full mode.

Perhaps what is needed is a task_struct->tick_dependency to go along 
with the system-wide and per-cpu flag words?
-- 
Chris Metcalf, EZChip Semiconductor
http://www.ezchip.com