Re: [PATCH]:Replacing current->state with set_current_state in kernel/signal.c

[Eric Dumazet <dada1@cosmosbay.com>]

On Wed, 25 Apr 2007 12:08:58 +0530
Shani Moideen <shani.moideen@wipro.com> wrote:

> Hi,
> 
> Replacing current->state with set_current_state in kernel/signal.c
> 	

> @@ -2596,7 +2596,7 @@ sys_signal(int sig, __sighandler_t handler)
>  asmlinkage long
>  sys_pause(void)
>  {
> -       current->state = TASK_INTERRUPTIBLE;
> +       set_current_state(TASK_INTERRUPTIBLE);
>         schedule();

Hi Shani

Either you think you corrected a BUG, so please state it clearly in Changelog so that Linus immediatly apply your patch for 2.6.21 :)

Either you dont know the exact semantic of set_current_state() and think it's a cleaner way to set current->state.
It might looks better for you but it's not the *same* thing.

I suggest you carefully study the difference between set_current_state() and __set_current_state(), and submit a new patch, once you feel comfortable with it.

Here is the relevant extract from include/linux/sched.h

/*
 * set_current_state() includes a barrier so that the write of current->state
 * is correctly serialised wrt the caller's subsequent test of whether to
 * actually sleep:
 *
 *      set_current_state(TASK_UNINTERRUPTIBLE);
 *      if (do_i_need_to_sleep())
 *              schedule();
 *
 * If the caller does not need such serialisation then use __set_current_state()
 */
#define __set_current_state(state_value)                        \
        do { current->state = (state_value); } while (0)
#define set_current_state(state_value)          \
        set_mb(current->state, (state_value))


Thank you