ALSA: seq: Don't handle loop timeout at snd_seq_pool_done()
authorTakashi Iwai <tiwai@suse.de>
Mon, 6 Feb 2017 14:09:48 +0000 (15:09 +0100)
committerTakashi Iwai <tiwai@suse.de>
Mon, 6 Feb 2017 14:09:48 +0000 (15:09 +0100)
snd_seq_pool_done() syncs with closing of all opened threads, but it
aborts the wait loop with a timeout, and proceeds to the release
resource even if not all threads have been closed.  The timeout was 5
seconds, and if you run a crazy stuff, it can exceed easily, and may
result in the access of the invalid memory address -- this is what
syzkaller detected in a bug report.

As a fix, let the code graduate from naiveness, simply remove the loop
timeout.

BugLink: http://lkml.kernel.org/r/CACT4Y+YdhDV2H5LLzDTJDVF-qiYHUHhtRaW4rbb4gUhTCQB81w@mail.gmail.com
Reported-by: Dmitry Vyukov <dvyukov@google.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
sound/core/seq/seq_memory.c

index c850345c43b53dd5616b155f34f741d0ca30701c..dfa5156f35856324d86f05315edfb2ef9cf4e74e 100644 (file)
@@ -419,7 +419,6 @@ int snd_seq_pool_done(struct snd_seq_pool *pool)
 {
        unsigned long flags;
        struct snd_seq_event_cell *ptr;
-       int max_count = 5 * HZ;
 
        if (snd_BUG_ON(!pool))
                return -EINVAL;
@@ -432,14 +431,8 @@ int snd_seq_pool_done(struct snd_seq_pool *pool)
        if (waitqueue_active(&pool->output_sleep))
                wake_up(&pool->output_sleep);
 
-       while (atomic_read(&pool->counter) > 0) {
-               if (max_count == 0) {
-                       pr_warn("ALSA: snd_seq_pool_done timeout: %d cells remain\n", atomic_read(&pool->counter));
-                       break;
-               }
+       while (atomic_read(&pool->counter) > 0)
                schedule_timeout_uninterruptible(1);
-               max_count--;
-       }
        
        /* release all resources */
        spin_lock_irqsave(&pool->lock, flags);