The recent rewrite of PCM link lock management introduced the refcount
in snd_pcm_group object, managed by the kernel refcount_t API. This
caused unexpected kernel warnings when the kernel is built with
CONFIG_REFCOUNT_FULL=y. As the warning line indicates, the problem is
obviously that we start with refcount=0 and do refcount_inc() for
adding each PCM link, while refcount_t API doesn't like refcount_inc()
performed on zero.
For adapting the proper refcount_t usage, this patch changes the logic
slightly:
- The initial refcount is 1, assuming the single list entry
- The refcount is incremented / decremented at each PCM link addition
and deletion
- ... which allows us concentrating only on the refcount as a release
condition
Fixes: f57f3df03a8e ("ALSA: pcm: More fine-grained PCM link locking")
BugLink: https://bugzilla.kernel.org/show_bug.cgi?id=204221
Reported-and-tested-by: Duncan Overbruck <kernel@xxxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx>
Signed-off-by: Takashi Iwai <tiwai@xxxxxxx>
---
sound/core/pcm_native.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c
index 860543a4c840..12dd9b318db1 100644
--- a/sound/core/pcm_native.c
+++ b/sound/core/pcm_native.c
@@ -77,7 +77,7 @@ void snd_pcm_group_init(struct snd_pcm_group *group)
spin_lock_init(&group->lock);
mutex_init(&group->mutex);
INIT_LIST_HEAD(&group->substreams);
- refcount_set(&group->refs, 0);
+ refcount_set(&group->refs, 1);
}
/* define group lock helpers */
@@ -1096,8 +1096,7 @@ static void snd_pcm_group_unref(struct snd_pcm_group *group,
if (!group)
return;
- do_free = refcount_dec_and_test(&group->refs) &&
- list_empty(&group->substreams);
+ do_free = refcount_dec_and_test(&group->refs);
snd_pcm_group_unlock(group, substream->pcm->nonatomic);
if (do_free)
kfree(group);
@@ -2020,6 +2019,7 @@ static int snd_pcm_link(struct snd_pcm_substream *substream, int fd)
snd_pcm_group_lock_irq(target_group, nonatomic);
snd_pcm_stream_lock(substream1);
snd_pcm_group_assign(substream1, target_group);
+ refcount_inc(&target_group->refs);
snd_pcm_stream_unlock(substream1);
snd_pcm_group_unlock_irq(target_group, nonatomic);
_end:
@@ -2056,13 +2056,14 @@ static int snd_pcm_unlink(struct snd_pcm_substream *substream)
snd_pcm_group_lock_irq(group, nonatomic);
relink_to_local(substream);
+ refcount_dec(&group->refs);
/* detach the last stream, too */
if (list_is_singular(&group->substreams)) {
relink_to_local(list_first_entry(&group->substreams,
struct snd_pcm_substream,
link_list));
- do_free = !refcount_read(&group->refs);
+ do_free = refcount_dec_and_test(&group->refs);
}
snd_pcm_group_unlock_irq(group, nonatomic);
--
2.16.4
_______________________________________________
Alsa-devel mailing list
Alsa-devel@xxxxxxxxxxxxxxxx
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
