Re: Query about the realtime processes scheduling.

2011-08-18 Thread Mulyadi Santosa
On Thu, Aug 18, 2011 at 14:21, nilesh  wrote:
> Please ignore the post.
> I should have looked into the code before posting the query. It looks
> like the bonus/penalty is only for non-realtime tasks.

Correct. Most (if not all) realtime priority scheduling is done in
FIFO style and their priority is static.

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies


Re: Query about the realtime processes scheduling.

2011-08-18 Thread nilesh
On Thursday 18 August 2011 12:18 PM, nilesh wrote:
> Hi,
>
> I am not too familiar with Linux scheduler. What I could understand
> theoretically is - the interactive processes get advantage (-5) in priority.
> And hence the effective priority is higher.
>
> Query:
> Is this true for realtime processes ( with SCHED_FIFO/SCHED_RR policy)
> as well? Could the realtime processes be I/O bound or interactive?
>
Please ignore the post.
I should have looked into the code before posting the query. It looks 
like the bonus/penalty is only for non-realtime tasks.

2101static int effective_prio(struct task_struct *p)
2102{
2103p->normal_prio = normal_prio(p);
2104/*
2105 * If we are RT tasks or we were boosted to RT priority,
2106 * keep the priority unchanged. Otherwise, update priority
2107 * to the normal priority:
2108 */
2109if (!rt_prio(p->prio))
2110return p->normal_prio;
2111return p->prio;
2112}
2113

-- 
Thanks,
Nilesh

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies


Query about the realtime processes scheduling.

2011-08-17 Thread nilesh
Hi,

I am not too familiar with Linux scheduler. What I could understand 
theoretically is - the interactive processes get advantage (-5) in priority.
And hence the effective priority is higher.

Query:
Is this true for realtime processes ( with SCHED_FIFO/SCHED_RR policy) 
as well? Could the realtime processes be I/O bound or interactive?

-- 
Thanks,
Nilesh

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies