__exit_signal: don't take rcu lock
authorOleg Nesterov <oleg@tv-sign.ru>
Fri, 25 Jul 2008 08:47:29 +0000 (01:47 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Fri, 25 Jul 2008 17:53:38 +0000 (10:53 -0700)
There is no reason for rcu_read_lock() in __exit_signal().  tsk->sighand
can only be changed if tsk does exec, obviously this is not possible.

Signed-off-by: Oleg Nesterov <oleg@tv-sign.ru>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
kernel/exit.c

index 93d2711b938123d34f1c0011e8f91a589b465615..a7799d8a640495756f63c579f48f048e568addae 100644 (file)
@@ -85,7 +85,6 @@ static void __exit_signal(struct task_struct *tsk)
        BUG_ON(!sig);
        BUG_ON(!atomic_read(&sig->count));
 
-       rcu_read_lock();
        sighand = rcu_dereference(tsk->sighand);
        spin_lock(&sighand->siglock);
 
@@ -136,7 +135,6 @@ static void __exit_signal(struct task_struct *tsk)
        tsk->signal = NULL;
        tsk->sighand = NULL;
        spin_unlock(&sighand->siglock);
-       rcu_read_unlock();
 
        __cleanup_sighand(sighand);
        clear_tsk_thread_flag(tsk,TIF_SIGPENDING);