ALSA: seq: Improve data consistency at polling

snd_seq_poll() calls snd_seq_write_pool_allocated() that reads out a
field in client->pool object, while it can be updated concurrently via
ioctls, as reported by syzbot.  The data race itself is harmless, as
it's merely a poll() call, and the state is volatile.  OTOH, the read
out of poll object info from the caller side is fragile, and we can
leave it better in snd_seq_pool_poll_wait() alone.

A similar pattern is seen in snd_seq_kernel_client_write_poll(), too,
which is called from the OSS sequencer.

This patch drops the pool checks from the caller side and add the
pool->lock in snd_seq_pool_poll_wait() for better data consistency.

Reported-by: [email protected]
Closes: https://lore.kernel.org/[email protected]
Link: https://patch.msgid.link/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>

Author: tiwai

sound/core/seq/seq_clientmgr.c

@@ -1130,8 +1130,7 @@ static __poll_t snd_seq_poll(struct file *file, poll_table * wait)
 	if (snd_seq_file_flags(file) & SNDRV_SEQ_LFLG_OUTPUT) {
 
 		/* check if data is available in the pool */
-		if (!snd_seq_write_pool_allocated(client) ||
-		    snd_seq_pool_poll_wait(client->pool, file, wait))
+		if (snd_seq_pool_poll_wait(client->pool, file, wait))
 			mask |= EPOLLOUT | EPOLLWRNORM;
 	}
 
@@ -2566,8 +2565,6 @@ int snd_seq_kernel_client_write_poll(int clientid, struct file *file, poll_table
 	if (client == NULL)
 		return -ENXIO;
 
-	if (! snd_seq_write_pool_allocated(client))
-		return 1;
 	if (snd_seq_pool_poll_wait(client->pool, file, wait))
 		return 1;
 	return 0;

sound/core/seq/seq_memory.c

@@ -427,6 +427,7 @@ int snd_seq_pool_poll_wait(struct snd_seq_pool *pool, struct file *file,
 			   poll_table *wait)
 {
 	poll_wait(file, &pool->output_sleep, wait);
+	guard(spinlock_irq)(&pool->lock);
 	return snd_seq_output_ok(pool);
 }