Blob Blame History Raw
From: Oliver Neukum <oneukum@suse.com>
Date: Fri, 15 Nov 2019 11:35:05 -0800
Subject: Input: ff-memless - kill timer in destroy()
Git-commit: fa3a5a1880c91bb92594ad42dfe9eedad7996b86
Patch-mainline: 5.4-rc8
References: CVE-2019-19524 bsc#1158413 bnc#1151927 5.3.12

No timer must be left running when the device goes away.

Signed-off-by: Oliver Neukum <oneukum@suse.com>
Reported-and-tested-by: syzbot+b6c55daa701fc389e286@syzkaller.appspotmail.com
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/1573726121.17351.3.camel@suse.com
Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
---
 drivers/input/ff-memless.c |    9 +++++++++
 1 file changed, 9 insertions(+)

--- a/drivers/input/ff-memless.c
+++ b/drivers/input/ff-memless.c
@@ -489,6 +489,15 @@ static void ml_ff_destroy(struct ff_devi
 {
 	struct ml_device *ml = ff->private;
 
+	/*
+	 * Even though we stop all playing effects when tearing down
+	 * an input device (via input_device_flush() that calls into
+	 * input_ff_flush() that stops and erases all effects), we
+	 * do not actually stop the timer, and therefore we should
+	 * do it here.
+	 */
+	del_timer_sync(&ml->timer);
+
 	kfree(ml->private);
 }