From: Martijn Coenen <maco@android.com>
Date: Fri, 16 Feb 2018 08:47:15 +0000 (+0100)
Subject: UPSTREAM: ANDROID: binder: synchronize_rcu() when using POLLFREE.
X-Git-Url: https://git.stricted.de/?a=commitdiff_plain;h=c0bf025b856adc38e5cabfcd37736e126f7bb593;p=GitHub%2FLineageOS%2Fandroid_kernel_samsung_universal7580.git

UPSTREAM: ANDROID: binder: synchronize_rcu() when using POLLFREE.

To prevent races with ep_remove_waitqueue() removing the
waitqueue at the same time.

Reported-by: syzbot+a2a3c4909716e271487e@syzkaller.appspotmail.com
Signed-off-by: Martijn Coenen <maco@android.com>
Cc: stable <stable@vger.kernel.org> # 4.14+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
(cherry picked from commit 5eeb2ca02a2f6084fc57ae5c244a38baab07033a)

Change-Id: Ia0089448079c78d0ab0b57303faf838e9e5ee797
---

diff --git a/drivers/android/binder.c b/drivers/android/binder.c
index f4aaae464c4..f3f5d7a08f3 100644
--- a/drivers/android/binder.c
+++ b/drivers/android/binder.c
@@ -4543,6 +4543,15 @@ static int binder_thread_release(struct binder_proc *proc,
 
 	binder_inner_proc_unlock(thread->proc);
 
+	/*
+	 * This is needed to avoid races between wake_up_poll() above and
+	 * and ep_remove_waitqueue() called for other reasons (eg the epoll file
+	 * descriptor being closed); ep_remove_waitqueue() holds an RCU read
+	 * lock, so we can be sure it's done after calling synchronize_rcu().
+	 */
+	if (thread->looper & BINDER_LOOPER_STATE_POLL)
+		synchronize_rcu();
+
 	if (send_reply)
 		binder_send_failed_reply(send_reply, BR_DEAD_REPLY);
 	binder_release_work(proc, &thread->todo);