Re: [PATCH] sched_rt: fix overload bug on rt group scheduling -v2

[Gregory Haskins <ghaskins@novell.com>]

[-- Attachment #1: Type: text/plain, Size: 3231 bytes --]

Peter Zijlstra wrote:
> Greg, how's this?
>   

Hi Peter,
  Sorry for the delay getting back to you.

I can't vouch specifically for the rt-group specific part, but as far as
the general balancer modifications to the migration code, it looks good. 

> ---
> Subject: sched_rt: fix overload bug on rt group scheduling
> From: Peter Zijlstra <peterz@infradead.org>
> Date: Wed, 01 Apr 2009 18:40:15 +0200
>
> Fixes an easily triggerable BUG() when setting process affinities.
>
> Make sure to count the number of migratable tasks in the same place:
> the root rt_rq. Otherwise the number doesn't make sense and we'll hit
> the BUG in set_cpus_allowed_rt().
>
> Also, make sure we only count tasks, not groups (this is probably
> already taken care of by the fact that rt_se->nr_cpus_allowed will be 0
> for groups, but be more explicit)
>
> Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
>   

Acked-by: Gregory Haskins <ghaskins@novell.com>

> ---
>  kernel/sched_rt.c |   18 +++++++++++++++++-
>  1 file changed, 17 insertions(+), 1 deletion(-)
>
> Index: linux-2.6/kernel/sched_rt.c
> ===================================================================
> --- linux-2.6.orig/kernel/sched_rt.c
> +++ linux-2.6/kernel/sched_rt.c
> @@ -10,6 +10,8 @@ static inline struct task_struct *rt_tas
>  
>  #ifdef CONFIG_RT_GROUP_SCHED
>  
> +#define rt_entity_is_task(rt_se) (!(rt_se)->my_q)
> +
>  static inline struct rq *rq_of_rt_rq(struct rt_rq *rt_rq)
>  {
>  	return rt_rq->rq;
> @@ -22,6 +24,8 @@ static inline struct rt_rq *rt_rq_of_se(
>  
>  #else /* CONFIG_RT_GROUP_SCHED */
>  
> +#define rt_entity_is_task(rt_se) (1)
> +
>  static inline struct rq *rq_of_rt_rq(struct rt_rq *rt_rq)
>  {
>  	return container_of(rt_rq, struct rq, rt);
> @@ -73,7 +77,7 @@ static inline void rt_clear_overload(str
>  
>  static void update_rt_migration(struct rt_rq *rt_rq)
>  {
> -	if (rt_rq->rt_nr_migratory && (rt_rq->rt_nr_running > 1)) {
> +	if (rt_rq->rt_nr_migratory && rt_rq->rt_nr_total > 1) {
>  		if (!rt_rq->overloaded) {
>  			rt_set_overload(rq_of_rt_rq(rt_rq));
>  			rt_rq->overloaded = 1;
> @@ -86,6 +90,12 @@ static void update_rt_migration(struct r
>  
>  static void inc_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
>  {
> +	if (!rt_entity_is_task(rt_se))
> +		return;
> +
> +	rt_rq = &rq_of_rt_rq(rt_rq)->rt;
> +
> +	rt_rq->rt_nr_total++;
>  	if (rt_se->nr_cpus_allowed > 1)
>  		rt_rq->rt_nr_migratory++;
>  
> @@ -94,6 +104,12 @@ static void inc_rt_migration(struct sche
>  
>  static void dec_rt_migration(struct sched_rt_entity *rt_se, struct rt_rq *rt_rq)
>  {
> +	if (!rt_entity_is_task(rt_se))
> +		return;
> +
> +	rt_rq = &rq_of_rt_rq(rt_rq)->rt;
> +
> +	rt_rq->rt_nr_total--;
>  	if (rt_se->nr_cpus_allowed > 1)
>  		rt_rq->rt_nr_migratory--;
>  
> diff --git a/kernel/sched.c b/kernel/sched.c
> index fd3ac58..a07d520 100644
> --- a/kernel/sched.c
> +++ b/kernel/sched.c
> @@ -493,6 +493,7 @@ struct rt_rq {
>  #endif
>  #ifdef CONFIG_SMP
>  	unsigned long rt_nr_migratory;
> +	unsigned long rt_nr_total;
>  	int overloaded;
>  	struct plist_head pushable_tasks;
>  #endif
>
>   



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 266 bytes --]