Re: [PATCH] ptrace: add ability to get/set signal-blocked mask (v2)

[Oleg Nesterov <oleg@redhat.com>]

On 04/23, Andrey Vagin wrote:
>
> @@ -841,6 +841,48 @@ int ptrace_request(struct task_struct *child, long request,
>  			ret = ptrace_setsiginfo(child, &siginfo);
>  		break;
>  
> +	case PTRACE_GETSIGMASK:
> +		if (addr != sizeof(sigset_t)) {
> +			ret = -EINVAL;
> +			break;
> +		}
> +
> +		if (copy_to_user(datavp, &child->blocked, sizeof(sigset_t)))
> +			ret = -EFAULT;

Yes, we can do this lockless. Only the task itself (or the debugger with
this patch) can change ->blocked, and the tracee can't run.

> +	case PTRACE_SETSIGMASK:
> +	{
> +		sigset_t new_set;
> +
> +		if (addr != sizeof(sigset_t)) {
> +			ret = -EINVAL;
> +			break;
> +		}
> +
> +		if (copy_from_user(&new_set, datavp, sizeof(sigset_t))) {
> +			ret = -EFAULT;
> +			break;
> +		}
> +
> +		sigdelsetmask(&new_set, sigmask(SIGKILL)|sigmask(SIGSTOP));
> +
> +		/*
> +		 * Every thread does recalc_sigpending() after resume, so
> +		 * retarget_shared_pending() and recalc_sigpending() are not
> +		 * called here.
> +		 */
> +		spin_lock_irq(&child->sighand->siglock);
> +		child->blocked = new_set;
> +		spin_unlock_irq(&child->sighand->siglock);
> +
> +		ret = 0;
> +		break;
> +	}

And this looks fine too.

Personally I am not sure we need to check addr == sizeof(), debugger
should know what it does... But I won't argue.

Oleg.