ALSA: seq: Avoid concurrent access to queue flags

The queue flags are represented in bit fields and the concurrent
access may result in unexpected results.  Although the current code
should be mostly OK as it's only reading a field while writing other
fields as KCSAN reported, it's safer to cover both with a proper
spinlock protection.

This patch fixes the possible concurrent read by protecting with
q->owner_lock.  Also the queue owner field is protected as well since
it's the field to be protected by the lock itself.

Reported-by: syzbot+65c6c92d04304d0a8efc@syzkaller.appspotmail.com
Reported-by: syzbot+e60ddfa48717579799dd@syzkaller.appspotmail.com
Link: https://lore.kernel.org/r/20200214111316.26939-2-tiwai@suse.de
Signed-off-by: Takashi Iwai <tiwai@suse.de>

diff --git a/sound/core/seq/seq_queue.c b/sound/core/seq/seq_queue.c
index caf68bf42f134be3e802c976b6b824f38d64c82d..20c552cf83987c8a0eeaa879e99403b5d28efc74 100644 (file)
--- a/sound/core/seq/seq_queue.c
+++ b/sound/core/seq/seq_queue.c
@@ -392,6 +392,7 @@ int snd_seq_queue_check_access(int queueid, int client)
 int snd_seq_queue_set_owner(int queueid, int client, int locked)
 {
        struct snd_seq_queue *q = queueptr(queueid);
+       unsigned long flags;
 
        if (q == NULL)
                return -EINVAL;
@@ -401,8 +402,10 @@ int snd_seq_queue_set_owner(int queueid, int client, int locked)
                return -EPERM;
        }
 
+       spin_lock_irqsave(&q->owner_lock, flags);
        q->locked = locked ? 1 : 0;
        q->owner = client;
+       spin_unlock_irqrestore(&q->owner_lock, flags);
        queue_access_unlock(q);
        queuefree(q);
 
@@ -539,15 +542,17 @@ void snd_seq_queue_client_termination(int client)
        unsigned long flags;
        int i;
        struct snd_seq_queue *q;
+       bool matched;
 
        for (i = 0; i < SNDRV_SEQ_MAX_QUEUES; i++) {
                if ((q = queueptr(i)) == NULL)
                        continue;
                spin_lock_irqsave(&q->owner_lock, flags);
-               if (q->owner == client)
+               matched = (q->owner == client);
+               if (matched)
                        q->klocked = 1;
                spin_unlock_irqrestore(&q->owner_lock, flags);
-               if (q->owner == client) {
+               if (matched) {
                        if (q->timer->running)
                                snd_seq_timer_stop(q->timer);
                        snd_seq_timer_reset(q->timer);
@@ -739,6 +744,8 @@ void snd_seq_info_queues_read(struct snd_info_entry *entry,
        int i, bpm;
        struct snd_seq_queue *q;
        struct snd_seq_timer *tmr;
+       bool locked;
+       int owner;
 
        for (i = 0; i < SNDRV_SEQ_MAX_QUEUES; i++) {
                if ((q = queueptr(i)) == NULL)
@@ -750,9 +757,14 @@ void snd_seq_info_queues_read(struct snd_info_entry *entry,
                else
                        bpm = 0;
 
+               spin_lock_irq(&q->owner_lock);
+               locked = q->locked;
+               owner = q->owner;
+               spin_unlock_irq(&q->owner_lock);
+
                snd_iprintf(buffer, "queue %d: [%s]\n", q->queue, q->name);
-               snd_iprintf(buffer, "owned by client    : %d\n", q->owner);
-               snd_iprintf(buffer, "lock status        : %s\n", q->locked ? "Locked" : "Free");
+               snd_iprintf(buffer, "owned by client    : %d\n", owner);
+               snd_iprintf(buffer, "lock status        : %s\n", locked ? "Locked" : "Free");
                snd_iprintf(buffer, "queued time events : %d\n", snd_seq_prioq_avail(q->timeq));
                snd_iprintf(buffer, "queued tick events : %d\n", snd_seq_prioq_avail(q->tickq));
                snd_iprintf(buffer, "timer state        : %s\n", tmr->running ? "Running" : "Stopped");