[PATCH] account_group_exec_runtime: fix the racy usage of ->signal

[Oleg Nesterov <oleg@redhat.com>]

Compile tested.

Unlike other similar routines, account_group_exec_runtime() could be
called "implicitly" after exit_notify(). This means we can race with
the parent doing release_task(), we can't just check ->signal != NULL.

Take ->siglock to make sure ->signal can't go away.

This is the minimal fix, with this patch we don't need need get/put cpu,
and I think we should uninline this function.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>

--- K-28/kernel/sched_stats.h~A_G_E_R_FIX	2008-11-07 17:32:02.000000000 +0100
+++ K-28/kernel/sched_stats.h	2008-11-07 17:44:39.000000000 +0100
@@ -351,10 +351,12 @@ static inline void account_group_exec_ru
 					      unsigned long long ns)
 {
 	struct signal_struct *sig;
+	unsigned long flags;
 
-	sig = tsk->signal;
-	if (unlikely(!sig))
+	if (unlikely(!lock_task_sighand(tsk, &flags)))
 		return;
+
+	sig = tsk->signal;
 	if (sig->cputime.totals) {
 		struct task_cputime *times;
 
@@ -362,4 +364,6 @@ static inline void account_group_exec_ru
 		times->sum_exec_runtime += ns;
 		put_cpu_no_resched();
 	}
+
+	unlock_task_sighand(tsk, &flags);
 }