public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* [PATCH 1/3] wait_task_stopped: don't use task_pid_nr_ns() lockless
@ 2007-11-16 17:24 Oleg Nesterov
  2007-11-16 19:54 ` Roland McGrath
  2007-11-19  9:28 ` Pavel Emelyanov
  0 siblings, 2 replies; 3+ messages in thread
From: Oleg Nesterov @ 2007-11-16 17:24 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Alexey Dobriyan, Kees Cook, Linus Torvalds, Pavel Emelyanov,
	Roland McGrath, Scott James Remnant, linux-kernel

wait_task_stopped(WNOWAIT) does task_pid_nr_ns() without tasklist/rcu lock,
we can read an already freed memory. Use the cached pid_t value.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>

--- 24/kernel/exit.c~1_PID	2007-11-16 18:12:44.000000000 +0300
+++ 24/kernel/exit.c	2007-11-16 18:13:54.000000000 +0300
@@ -1357,7 +1357,7 @@ static int wait_task_stopped(struct task
 			     int __user *stat_addr, struct rusage __user *ru)
 {
 	int retval, exit_code;
-	struct pid_namespace *ns;
+	pid_t pid;
 
 	if (!p->exit_code)
 		return 0;
@@ -1376,12 +1376,11 @@ static int wait_task_stopped(struct task
 	 * keep holding onto the tasklist_lock while we call getrusage and
 	 * possibly take page faults for user memory.
 	 */
-	ns = current->nsproxy->pid_ns;
+	pid = task_pid_nr_ns(p, current->nsproxy->pid_ns);
 	get_task_struct(p);
 	read_unlock(&tasklist_lock);
 
 	if (unlikely(noreap)) {
-		pid_t pid = task_pid_nr_ns(p, ns);
 		uid_t uid = p->uid;
 		int why = (p->ptrace & PT_PTRACED) ? CLD_TRAPPED : CLD_STOPPED;
 
@@ -1451,11 +1450,11 @@ bail_ref:
 	if (!retval && infop)
 		retval = put_user(exit_code, &infop->si_status);
 	if (!retval && infop)
-		retval = put_user(task_pid_nr_ns(p, ns), &infop->si_pid);
+		retval = put_user(pid, &infop->si_pid);
 	if (!retval && infop)
 		retval = put_user(p->uid, &infop->si_uid);
 	if (!retval)
-		retval = task_pid_nr_ns(p, ns);
+		retval = pid;
 	put_task_struct(p);
 
 	BUG_ON(!retval);


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH 1/3] wait_task_stopped: don't use task_pid_nr_ns() lockless
  2007-11-16 17:24 [PATCH 1/3] wait_task_stopped: don't use task_pid_nr_ns() lockless Oleg Nesterov
@ 2007-11-16 19:54 ` Roland McGrath
  2007-11-19  9:28 ` Pavel Emelyanov
  1 sibling, 0 replies; 3+ messages in thread
From: Roland McGrath @ 2007-11-16 19:54 UTC (permalink / raw)
  To: Oleg Nesterov
  Cc: Andrew Morton, Alexey Dobriyan, Kees Cook, Linus Torvalds,
	Pavel Emelyanov, Scott James Remnant, linux-kernel

Looks good to me.


Thanks,
Roland

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH 1/3] wait_task_stopped: don't use task_pid_nr_ns() lockless
  2007-11-16 17:24 [PATCH 1/3] wait_task_stopped: don't use task_pid_nr_ns() lockless Oleg Nesterov
  2007-11-16 19:54 ` Roland McGrath
@ 2007-11-19  9:28 ` Pavel Emelyanov
  1 sibling, 0 replies; 3+ messages in thread
From: Pavel Emelyanov @ 2007-11-19  9:28 UTC (permalink / raw)
  To: Oleg Nesterov, Andrew Morton
  Cc: Alexey Dobriyan, Kees Cook, Linus Torvalds, Pavel Emelyanov,
	Roland McGrath, Scott James Remnant, linux-kernel

Oleg Nesterov wrote:
> wait_task_stopped(WNOWAIT) does task_pid_nr_ns() without tasklist/rcu lock,
> we can read an already freed memory. Use the cached pid_t value.

Indeed. Thanks!

> Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>

Acked-by: Pavel Emelyanov <xemul@openvz.org>

> --- 24/kernel/exit.c~1_PID	2007-11-16 18:12:44.000000000 +0300
> +++ 24/kernel/exit.c	2007-11-16 18:13:54.000000000 +0300
> @@ -1357,7 +1357,7 @@ static int wait_task_stopped(struct task
>  			     int __user *stat_addr, struct rusage __user *ru)
>  {
>  	int retval, exit_code;
> -	struct pid_namespace *ns;
> +	pid_t pid;
>  
>  	if (!p->exit_code)
>  		return 0;
> @@ -1376,12 +1376,11 @@ static int wait_task_stopped(struct task
>  	 * keep holding onto the tasklist_lock while we call getrusage and
>  	 * possibly take page faults for user memory.
>  	 */
> -	ns = current->nsproxy->pid_ns;
> +	pid = task_pid_nr_ns(p, current->nsproxy->pid_ns);
>  	get_task_struct(p);
>  	read_unlock(&tasklist_lock);
>  
>  	if (unlikely(noreap)) {
> -		pid_t pid = task_pid_nr_ns(p, ns);
>  		uid_t uid = p->uid;
>  		int why = (p->ptrace & PT_PTRACED) ? CLD_TRAPPED : CLD_STOPPED;
>  
> @@ -1451,11 +1450,11 @@ bail_ref:
>  	if (!retval && infop)
>  		retval = put_user(exit_code, &infop->si_status);
>  	if (!retval && infop)
> -		retval = put_user(task_pid_nr_ns(p, ns), &infop->si_pid);
> +		retval = put_user(pid, &infop->si_pid);
>  	if (!retval && infop)
>  		retval = put_user(p->uid, &infop->si_uid);
>  	if (!retval)
> -		retval = task_pid_nr_ns(p, ns);
> +		retval = pid;
>  	put_task_struct(p);
>  
>  	BUG_ON(!retval);
> 
> 


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2007-11-19  9:30 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2007-11-16 17:24 [PATCH 1/3] wait_task_stopped: don't use task_pid_nr_ns() lockless Oleg Nesterov
2007-11-16 19:54 ` Roland McGrath
2007-11-19  9:28 ` Pavel Emelyanov

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