Re: [PATCH] de_thread: Move notify_count write under lock

[Oleg Nesterov <oleg@redhat.com>]

I still think that the changelog and the comment look a bit confusing...
it could simply say that exit_notify() can see these STORE's out of order.
And we can set ->notify_count after ->exit_state check, but again this is
cosmetic, I won't insist. The main problem with this patch is that it was
ignored ;)

Kirill, could you resend? Feel free to add my ack.

On 02/05, Kirill Tkhai wrote:
> 
> The write operation may be reordered with the setting of group_exit_task.
> If so, this fires in exit_notify().
> 
> Looks like, it's not good to add smp barriers for this case, especially
> in exit_notify(), so let's put the notify_count write under write lock.
> 
> Signed-off-by: Kirill Tkhai <ktkhai@parallels.com>
> ---
>  fs/exec.c |    8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/exec.c b/fs/exec.c
> index ad8798e..42782d5 100644
> --- a/fs/exec.c
> +++ b/fs/exec.c
> @@ -920,10 +920,16 @@ static int de_thread(struct task_struct *tsk)
>  	if (!thread_group_leader(tsk)) {
>  		struct task_struct *leader = tsk->group_leader;
>  
> -		sig->notify_count = -1;	/* for exit_notify() */
>  		for (;;) {
>  			threadgroup_change_begin(tsk);
>  			write_lock_irq(&tasklist_lock);
> +			/*
> +			 * We could set it once outside the for() cycle, but
> +			 * this requires to use SMP barriers there and in
> +			 * exit_notify(), because the write operation may
> +			 * be reordered with the setting of group_exit_task.
> +			 */
> +			sig->notify_count = -1;	/* for exit_notify() */
>  			if (likely(leader->exit_state))
>  				break;
>  			__set_current_state(TASK_KILLABLE);
> 
> 
>