signals: send_sigqueue: don't forget about handle_stop_signal()
authorOleg Nesterov <oleg@tv-sign.ru>
Wed, 30 Apr 2008 07:52:48 +0000 (00:52 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Wed, 30 Apr 2008 15:29:35 +0000 (08:29 -0700)
send_group_sigqueue() calls handle_stop_signal(), send_sigqueue() doesn't.
This is not consistent and in fact I'd say this is (minor) bug.

Move handle_stop_signal() from send_group_sigqueue() to do_send_sigqueue(),
the latter is called by send_sigqueue() too.

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/signal.c

index 0a8b0aece80d9f7b52a62d41b4804d2796753cd9..8259262eaa6047596efb285d843ef245269adf18 100644 (file)
@@ -1274,8 +1274,10 @@ void sigqueue_free(struct sigqueue *q)
 }
 
 static int do_send_sigqueue(int sig, struct sigqueue *q, struct task_struct *t,
-               struct sigpending *pending)
+                               struct sigpending *pending)
 {
+       handle_stop_signal(sig, t);
+
        if (unlikely(!list_empty(&q->list))) {
                /*
                 * If an SI_TIMER entry is already queue just increment
@@ -1335,7 +1337,6 @@ send_group_sigqueue(int sig, struct sigqueue *q, struct task_struct *p)
        read_lock(&tasklist_lock);
        /* Since it_lock is held, p->sighand cannot be NULL. */
        spin_lock_irqsave(&p->sighand->siglock, flags);
-       handle_stop_signal(sig, p);
 
        ret = do_send_sigqueue(sig, q, p, &p->signal->shared_pending);