public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* Problem with set_current_state() and schedule()
@ 2008-10-15 14:50 Peter Teoh
  2008-10-15 15:16 ` Gilad Ben-Yossef
  0 siblings, 1 reply; 2+ messages in thread
From: Peter Teoh @ 2008-10-15 14:50 UTC (permalink / raw)
  To: LKML

I am puzzled over this:

The module just consist of three function:

static int myfunction(void)
{
       dbg("%s - enter....", __FUNCTION__);

       while (!kthread_should_stop()) {
               dbg("%s - executing....", __FUNCTION__);
               set_current_state(TASK_INTERRUPTIBLE);
               printk("sssssssssssssssssssssssssssssssss\n");


               schedule(); ======>       ////// removed for part 2
               printk("sssssssssssssssssssssssssssssssss\n");======>//////
removed for part 2


               msleep(1000);
               schedule();
       }
       return 0;
       dbg("%s exit", __FUNCTION__);
}

int init_module(void)
{

       g_task = kthread_run(myfunction, NULL, "mykthread");

       return 0;
}

void cleanup_module(void)
{
       printk(KERN_INFO "Goodbye mykthread.\n");
       kthread_stop(g_task);
}

Below is the dmesg out .... noticed two parts (based on timing output):

Part 1:
[ 1160.348000] mykthread: myfunction - enter....
[ 1160.348000] mykthread: myfunction - executing....
[ 1160.348000] sssssssssssssssssssssssssssssssss
[ 1208.632000] Goodbye mykthread.=======> this is because I "rmmod",
otherwise it got stuck in the previous statement.

Part 2:
[ 1208.632000] sssssssssssssssssssssssssssssssss
[ 1225.612000] mykthread: myfunction - enter....
[ 1225.612000] mykthread: myfunction - executing....
[ 1225.612000] sssssssssssssssssssssssssssssssss
[ 1226.616000] mykthread: myfunction - executing....
[ 1226.616000] sssssssssssssssssssssssssssssssss
[ 1227.620000] mykthread: myfunction - executing....
[ 1227.620000] sssssssssssssssssssssssssssssssss
[ 1228.624000] mykthread: myfunction - executing....
[ 1228.624000] sssssssssssssssssssssssssssssssss
[ 1229.628000] mykthread: myfunction - executing....
[ 1229.628000] sssssssssssssssssssssssssssssssss
[ 1230.632000] mykthread: myfunction - executing....
[ 1230.632000] sssssssssssssssssssssssssssssssss
[ 1231.636000] mykthread: myfunction - executing....
[ 1231.636000] sssssssssssssssssssssssssssssssss
[ 1232.640000] mykthread: myfunction - executing....
[ 1232.640000] sssssssssssssssssssssssssssssssss
[ 1233.644000] mykthread: myfunction - executing....
[ 1233.644000] sssssssssssssssssssssssssssssssss
[ 1234.648000] mykthread: myfunction - executing....
[ 1234.648000] sssssssssssssssssssssssssssssssss
[ 1235.652000] mykthread: myfunction - executing....
[ 1235.652000] sssssssssssssssssssssssssssssssss
[ 1236.656000] mykthread: myfunction - executing....
[ 1236.656000] sssssssssssssssssssssssssssssssss
[ 1237.660000] mykthread: myfunction - executing....
[ 1237.660000] sssssssssssssssssssssssssssssssss
[ 1238.664000] mykthread: myfunction - executing....
[ 1238.664000] sssssssssssssssssssssssssssssssss
[ 1239.668000] mykthread: myfunction - executing....
[ 1239.668000] sssssssssssssssssssssssssssssssss
[ 1240.672000] mykthread: myfunction - executing....
[ 1240.672000] sssssssssssssssssssssssssssssssss
[ 1241.676000] mykthread: myfunction - executing....
[ 1241.676000] sssssssssssssssssssssssssssssssss
[ 1242.680000] mykthread: myfunction - executing....
[ 1242.680000] sssssssssssssssssssssssssssssssss

For part 1:

After printing one line, the program never seemed to continue anymore.
  why?   It got stuck in the first schedule().

For part 2:

Just remove the "remove for part 2".   And u can immediately see that
it starts to loop forever.   So effectively there is only one
schedule() in the while loop.

Why?

I am quite puzzled......thanks in advance for the help.   :-).

-- 
Regards,
Peter Teoh

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: Problem with set_current_state() and schedule()
  2008-10-15 14:50 Problem with set_current_state() and schedule() Peter Teoh
@ 2008-10-15 15:16 ` Gilad Ben-Yossef
  0 siblings, 0 replies; 2+ messages in thread
From: Gilad Ben-Yossef @ 2008-10-15 15:16 UTC (permalink / raw)
  To: Peter Teoh; +Cc: LKML

Peter Teoh wrote:

> I am puzzled over this:
>
> The module just consist of three function:
>
> static int myfunction(void)
> {
>        dbg("%s - enter....", __FUNCTION__);
>
>        while (!kthread_should_stop()) {
>                dbg("%s - executing....", __FUNCTION__);
>                set_current_state(TASK_INTERRUPTIBLE);
>                printk("sssssssssssssssssssssssssssssssss\n");
>
>
>                schedule(); ======>       ////// removed for part 2
>                printk("sssssssssssssssssssssssssssssssss\n");======>//////
> removed for part 2
>
>
>                msleep(1000);
>                schedule();
>        }
>        return 0;
>        dbg("%s exit", __FUNCTION__);
> ...
>   
> For part 1:
>
> After printing one line, the program never seemed to continue anymore.
>   why?   It got stuck in the first schedule().
>   
You set the state of your task to TASK_INTERRUPTIBLE, that is, not valid 
to run and then call scheduler to give control to some other task. You 
don't include any code that actually sets the task state back to 
TASK_RUNNING, that is legible for scheduling so the schedule will never 
enter your task again. It effectively sleeps for ever.

You either need to use schedule_timeout with some delay or have some 
other means to set your task state back to TASK_RUNNING or better yet, 
use wait_event and friends and don't call schedule() directly.

> For part 2:
>
> Just remove the "remove for part 2".   And u can immediately see that
> it starts to loop forever.   So effectively there is only one
> schedule() in the while loop.
>   

> Why?
>   
Thats' because msleep() changes your task state to TASK_RUNNING when it 
finishes sleeping, so you're calling the scheduler with a TASK_RUNNING 
state and it will schedule your task back sometime (perhaps even 
immediately).
> I am quite puzzled......thanks in advance for the help.   :-).
>
>   
Go read chapter 7 of "Linux Device Drivers, Third Edition" :-)

Gilad


-- 
Gilad Ben-Yossef 
Chief Coffee Drinker

Codefidence Ltd.
The code is free, your time isn't.(TM)

Web:    http://codefidence.com
Email:  gilad@codefidence.com
Office: +972-8-9316883 ext. 201
Fax:    +972-8-9316885
Mobile: +972-52-8260388

	The Doctor: Don't worry, Reinette, just a nightmare. 
	Everyone has nightmares. Even monsters from under the 
	bed have nightmares, don't you, monster?
	Reinette: What do monsters have nightmares about?
	The Doctor: Me! 


^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2008-10-15 15:16 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-10-15 14:50 Problem with set_current_state() and schedule() Peter Teoh
2008-10-15 15:16 ` Gilad Ben-Yossef

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox