diff mbox

[4/4,v2] ASoC: rsnd: spin lock for interrupt handler

Message ID 87382qtwto.wl%kuninori.morimoto.gx@renesas.com (mailing list archive)
State Not Applicable
Headers show

Commit Message

Kuninori Morimoto May 21, 2015, 3:50 a.m. UTC
From: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>

Renesas R-Car driver interrupt handler was not locked before.
But now, SSI/SRC interrupt handler calls restart function
which should be called under spin lock.
Below error might happen witout this patch.

Unable to handle kernel NULL pointer dereference at virtual address 00000048
pgd = edfac000
[00000048] *pgd=6e0f0831, *pte=00000000, *ppte=00000000
Internal error: Oops: 17 [#1] SMP ARM
CPU: 0 PID: 2009 Comm: aplay Not tainted 4.1.0-rc2-dirty #4
Hardware name: Generic R8A7790 (Flattened Device Tree)
task: eeac9040 ti: eebe8000 task.ti: eebe8000
PC is at rsnd_get_adinr+0x28/0x60
LR is at rsnd_src_ssiu_start+0xdc/0x19c
pc : [<c0409790>]    lr : [<c040c068>]    psr: a0000193
sp : eebe9e58  ip : eebe9e68  fp : eebe9e64
r10: c06ed9d0  r9 : ee919d10  r8 : 00000001
r7 : 00000001  r6 : ee1cb090  r5 : 00000000  r4 : edcaa418
r3 : 00000000  r2 : eea8ce00  r1 : 80000193  r0 : edcaa418
...

Reported-by: Cao Minh Hiep <cm-hiep@jinso.co.jp>
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Keita Kobayashi <keita.kobayashi.ym@renesas.com>
Tested by: Cao Minh Hiep <cm-hiep@jinso.co.jp>
---
v1 -> v2

 - no change, but added Hiep's Test-by

 sound/soc/sh/rcar/core.c | 12 ++++++++++--
 sound/soc/sh/rcar/rsnd.h |  3 +--
 sound/soc/sh/rcar/src.c  | 11 ++++++++---
 sound/soc/sh/rcar/ssi.c  | 14 +++++++++++---
 4 files changed, 30 insertions(+), 10 deletions(-)

Comments

Mark Brown May 21, 2015, 10:59 a.m. UTC | #1
On Thu, May 21, 2015 at 03:50:23AM +0000, Kuninori Morimoto wrote:
> From: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> 
> Renesas R-Car driver interrupt handler was not locked before.
> But now, SSI/SRC interrupt handler calls restart function
> which should be called under spin lock.
> Below error might happen witout this patch.

This looks fine but I guess it depends on patch 1 otherwise more errors
will be introduced?
Vinod Koul May 21, 2015, 4:15 p.m. UTC | #2
On Thu, May 21, 2015 at 11:59:40AM +0100, Mark Brown wrote:
> On Thu, May 21, 2015 at 03:50:23AM +0000, Kuninori Morimoto wrote:
> > From: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> > 
> > Renesas R-Car driver interrupt handler was not locked before.
> > But now, SSI/SRC interrupt handler calls restart function
> > which should be called under spin lock.
> > Below error might happen witout this patch.
> 
> This looks fine but I guess it depends on patch 1 otherwise more errors
> will be introduced?
The 1st one looks good to me and was about to apply, but saw this and to
sync would you the 1st patch thru ASoC tree or fine with me applying that
one
Mark Brown May 21, 2015, 4:46 p.m. UTC | #3
On Thu, May 21, 2015 at 09:45:52PM +0530, Vinod Koul wrote:
> On Thu, May 21, 2015 at 11:59:40AM +0100, Mark Brown wrote:

> > This looks fine but I guess it depends on patch 1 otherwise more errors
> > will be introduced?

> The 1st one looks good to me and was about to apply, but saw this and to
> sync would you the 1st patch thru ASoC tree or fine with me applying that
> one

Morimoto-san, do the two patches need to go through the same tree?  I
guess it's better to have them both in the ASoC tree given how much work
you're doing on these drivers.  Vinod, do you want to apply and send a
pull request or do you just want to ack?
Kuninori Morimoto May 21, 2015, 11:59 p.m. UTC | #4
Hi Mark

> > The 1st one looks good to me and was about to apply, but saw this and to
> > sync would you the 1st patch thru ASoC tree or fine with me applying that
> > one
> 
> Morimoto-san, do the two patches need to go through the same tree?  I
> guess it's better to have them both in the ASoC tree given how much work
> you're doing on these drivers.  Vinod, do you want to apply and send a
> pull request or do you just want to ack?

Thank you. Yes in same tree is understandable.

Best regards
---
Kuninori Morimoto
--
To unsubscribe from this list: send the line "unsubscribe dmaengine" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Mark Brown May 22, 2015, 1:15 p.m. UTC | #5
On Thu, May 21, 2015 at 03:50:23AM +0000, Kuninori Morimoto wrote:
> From: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
> 
> Renesas R-Car driver interrupt handler was not locked before.
> But now, SSI/SRC interrupt handler calls restart function
> which should be called under spin lock.
> Below error might happen witout this patch.

Applied, thanks.
diff mbox

Patch

diff --git a/sound/soc/sh/rcar/core.c b/sound/soc/sh/rcar/core.c
index 2b7323c..d460d2a 100644
--- a/sound/soc/sh/rcar/core.c
+++ b/sound/soc/sh/rcar/core.c
@@ -170,6 +170,14 @@  void rsnd_mod_quit(struct rsnd_mod *mod)
 		clk_unprepare(mod->clk);
 }
 
+int rsnd_mod_is_working(struct rsnd_mod *mod)
+{
+	struct rsnd_dai_stream *io = rsnd_mod_to_io(mod);
+
+	/* see rsnd_dai_stream_init/quit() */
+	return !!io->substream;
+}
+
 /*
  *	settting function
  */
@@ -362,7 +370,7 @@  static int rsnd_soc_dai_trigger(struct snd_pcm_substream *substream, int cmd,
 	int ret;
 	unsigned long flags;
 
-	rsnd_lock(priv, flags);
+	spin_lock_irqsave(&priv->lock, flags);
 
 	switch (cmd) {
 	case SNDRV_PCM_TRIGGER_START:
@@ -400,7 +408,7 @@  static int rsnd_soc_dai_trigger(struct snd_pcm_substream *substream, int cmd,
 	}
 
 dai_trigger_end:
-	rsnd_unlock(priv, flags);
+	spin_unlock_irqrestore(&priv->lock, flags);
 
 	return ret;
 }
diff --git a/sound/soc/sh/rcar/rsnd.h b/sound/soc/sh/rcar/rsnd.h
index 4e6de68..03ff071 100644
--- a/sound/soc/sh/rcar/rsnd.h
+++ b/sound/soc/sh/rcar/rsnd.h
@@ -303,6 +303,7 @@  int rsnd_mod_init(struct rsnd_mod *mod,
 		   int id);
 void rsnd_mod_quit(struct rsnd_mod *mod);
 char *rsnd_mod_name(struct rsnd_mod *mod);
+int rsnd_mod_is_working(struct rsnd_mod *mod);
 struct dma_chan *rsnd_mod_dma_req(struct rsnd_mod *mod);
 
 /*
@@ -449,8 +450,6 @@  struct rsnd_priv {
 #define rsnd_priv_to_pdev(priv)	((priv)->pdev)
 #define rsnd_priv_to_dev(priv)	(&(rsnd_priv_to_pdev(priv)->dev))
 #define rsnd_priv_to_info(priv)	((priv)->info)
-#define rsnd_lock(priv, flags) spin_lock_irqsave(&priv->lock, flags)
-#define rsnd_unlock(priv, flags) spin_unlock_irqrestore(&priv->lock, flags)
 
 /*
  *	rsnd_kctrl
diff --git a/sound/soc/sh/rcar/src.c b/sound/soc/sh/rcar/src.c
index 3beb32e..fbe9166 100644
--- a/sound/soc/sh/rcar/src.c
+++ b/sound/soc/sh/rcar/src.c
@@ -673,10 +673,13 @@  static int _rsnd_src_stop_gen2(struct rsnd_mod *mod)
 static irqreturn_t rsnd_src_interrupt_gen2(int irq, void *data)
 {
 	struct rsnd_mod *mod = data;
-	struct rsnd_dai_stream *io = rsnd_mod_to_io(mod);
+	struct rsnd_priv *priv = rsnd_mod_to_priv(mod);
+
+	spin_lock(&priv->lock);
 
-	if (!io)
-		return IRQ_NONE;
+	/* ignore all cases if not working */
+	if (!rsnd_mod_is_working(mod))
+		goto rsnd_src_interrupt_gen2_out;
 
 	if (rsnd_src_error_record_gen2(mod)) {
 		struct rsnd_priv *priv = rsnd_mod_to_priv(mod);
@@ -692,6 +695,8 @@  static irqreturn_t rsnd_src_interrupt_gen2(int irq, void *data)
 		else
 			dev_warn(dev, "no more SRC restart\n");
 	}
+rsnd_src_interrupt_gen2_out:
+	spin_unlock(&priv->lock);
 
 	return IRQ_HANDLED;
 }
diff --git a/sound/soc/sh/rcar/ssi.c b/sound/soc/sh/rcar/ssi.c
index 927ac52..50fa392 100644
--- a/sound/soc/sh/rcar/ssi.c
+++ b/sound/soc/sh/rcar/ssi.c
@@ -423,10 +423,15 @@  static irqreturn_t rsnd_ssi_interrupt(int irq, void *data)
 	struct rsnd_priv *priv = rsnd_mod_to_priv(mod);
 	struct rsnd_dai_stream *io = rsnd_mod_to_io(mod);
 	int is_dma = rsnd_ssi_is_dma_mode(mod);
-	u32 status = rsnd_mod_read(mod, SSISR);
+	u32 status;
+
+	spin_lock(&priv->lock);
 
-	if (!io)
-		return IRQ_NONE;
+	/* ignore all cases if not working */
+	if (!rsnd_mod_is_working(mod))
+		goto rsnd_ssi_interrupt_out;
+
+	status = rsnd_mod_read(mod, SSISR);
 
 	/* PIO only */
 	if (!is_dma && (status & DIRQ)) {
@@ -466,6 +471,9 @@  static irqreturn_t rsnd_ssi_interrupt(int irq, void *data)
 
 	rsnd_ssi_record_error(ssi, status);
 
+rsnd_ssi_interrupt_out:
+	spin_unlock(&priv->lock);
+
 	return IRQ_HANDLED;
 }