ASoC: rsnd: don't call snd_pcm_period_elapsed() under spin lock

'a9e1ac1a9e4585b5("ASoC: rsnd: spin lock for interrupt handler")'
added spin lock under interrupt handler to solve HW restart issue.

OTOH, current rsnd driver calls snd_pcm_period_elapsed() from
rsnd_dai_pointer_update(). but, it will be called under spin lock
if SSI was PIO mode.

If it was called under spin lock, it will call
snd_pcm_update_state() -> snd_pcm_drain_done().
Then, it calls rsnd_soc_dai_trigger() and will be dead-lock.
This patch doesn't call rsnd_dai_pointer_update() under spin lock

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Keita Kobayashi <keita.kobayashi.ym@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>

1 files changed, 5 insertions, 1 deletions

diff --git a/sound/soc/sh/rcar/ssi.c b/sound/soc/sh/rcar/ssi.c
index 856917392b79..25483211a349 100644
--- a/sound/soc/sh/rcar/ssi.c
+++ b/sound/soc/sh/rcar/ssi.c
@@ -426,6 +426,7 @@ static irqreturn_t rsnd_ssi_interrupt(int irq, void *data)
 	struct rsnd_dai_stream *io = rsnd_mod_to_io(mod);
 	int is_dma = rsnd_ssi_is_dma_mode(mod);
 	u32 status;
+	bool elapsed = false;
 
 	spin_lock(&priv->lock);
 
@@ -451,7 +452,7 @@ static irqreturn_t rsnd_ssi_interrupt(int irq, void *data)
 		else
 			*buf = rsnd_mod_read(mod, SSIRDR);
 
-		rsnd_dai_pointer_update(io, sizeof(*buf));
+		elapsed = rsnd_dai_pointer_update(io, sizeof(*buf));
 	}
 
 	/* DMA only */
@@ -476,6 +477,9 @@ static irqreturn_t rsnd_ssi_interrupt(int irq, void *data)
 rsnd_ssi_interrupt_out:
 	spin_unlock(&priv->lock);
 
+	if (elapsed)
+		rsnd_dai_period_elapsed(io);
+
 	return IRQ_HANDLED;
 }