[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20090208184727.GA27081@redhat.com>
Date: Sun, 8 Feb 2009 19:47:27 +0100
From: Oleg Nesterov <oleg@...hat.com>
To: Andrew Morton <akpm@...ux-foundation.org>
Cc: Jerome Marchand <jmarchan@...hat.com>,
Roland McGrath <roland@...hat.com>,
Denys Vlasenko <dvlasenk@...hat.com>,
linux-kernel@...r.kernel.org
Subject: [PATCH 2/3] ptrace_untrace: use wake_up_process() instead of bogus
signal_wake_up()
Both ptrace_stop() and do_signal_stop() pathes always take ->siglock and
do recalc_sigpending() after wakeup.
This means that if the tracer sees task_is_traced(child) == T (perhaps it
it was actually TASK_STOPPED before ptrace_check_attach) under ->siglock,
it can use the plain wake_up_process() instead of signal_wake_up().
(and note that ptrace_resume() does wake_up_process() too).
We also have sys_clone(CLONE_STOPPED), but a) it must set TIF_SIGPENDING
by hand anyway, and b) it is deprecated.
Signed-off-by: Oleg Nesterov <oleg@...hat.com>
--- 6.29-rc3/kernel/ptrace.c~2_KILL_SIGWAKE 2009-02-08 05:11:52.000000000 +0100
+++ 6.29-rc3/kernel/ptrace.c 2009-02-08 06:22:26.000000000 +0100
@@ -60,11 +60,10 @@ static void ptrace_untrace(struct task_s
{
spin_lock(&child->sighand->siglock);
if (task_is_traced(child)) {
- if (child->signal->flags & SIGNAL_STOP_STOPPED) {
+ if (child->signal->flags & SIGNAL_STOP_STOPPED)
__set_task_state(child, TASK_STOPPED);
- } else {
- signal_wake_up(child, 1);
- }
+ else
+ wake_up_process(child);
}
spin_unlock(&child->sighand->siglock);
}
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists