Re: [PATCH v2 07/12] ptrace: Don't change __state

[Oleg Nesterov]

On 04/29, Eric W. Biederman wrote:
>
>  static void ptrace_unfreeze_traced(struct task_struct *task)
>  {
> -	if (READ_ONCE(task->__state) != __TASK_TRACED)
> -		return;
> -
> -	WARN_ON(!task->ptrace || task->parent != current);
> +	unsigned long flags;
>
>  	/*
> -	 * PTRACE_LISTEN can allow ptrace_trap_notify to wake us up remotely.
> -	 * Recheck state under the lock to close this race.
> +	 * The child may be awake and may have cleared
> +	 * JOBCTL_PTRACE_FROZEN (see ptrace_resume).  The child will
> +	 * not set JOBCTL_PTRACE_FROZEN or enter __TASK_TRACED anew.
>  	 */
> -	spin_lock_irq(&task->sighand->siglock);
> -	if (READ_ONCE(task->__state) == __TASK_TRACED) {
> +	if (lock_task_sighand(task, &flags)) {
> +		task->jobctl &= ~JOBCTL_PTRACE_FROZEN;

Well, I think that the fast-path

	if (!(task->jobctl & JOBCTL_PTRACE_FROZEN))
		return;

at the start makes sense, we can avoid lock_task_sighand() if the tracee
was resumed.

Oleg.