The function kirkwood_i2s_dev_remove() may be used when probe fails.
Signed-off-by: Jean-Francois Moine <[email protected]>
---
sound/soc/kirkwood/kirkwood-i2s.c | 40 +++++++++++-----------
1 file changed, 20 insertions(+), 20 deletions(-)
diff --git a/sound/soc/kirkwood/kirkwood-i2s.c b/sound/soc/kirkwood/kirkwood-i2s.c
index 0109b1e..319086f 100644
--- a/sound/soc/kirkwood/kirkwood-i2s.c
+++ b/sound/soc/kirkwood/kirkwood-i2s.c
@@ -449,6 +449,19 @@ static const struct snd_soc_component_driver kirkwood_i2s_component = {
.name = DRV_NAME,
};
+static int kirkwood_i2s_dev_remove(struct platform_device *pdev)
+{
+ struct kirkwood_dma_data *priv = dev_get_drvdata(&pdev->dev);
+
+ snd_soc_unregister_component(&pdev->dev);
+
+ if (!IS_ERR(priv->extclk))
+ clk_disable_unprepare(priv->extclk);
+ clk_disable_unprepare(priv->clk);
+
+ return 0;
+}
+
static int kirkwood_i2s_dev_probe(struct platform_device *pdev)
{
struct kirkwood_asoc_platform_data *data = pdev->dev.platform_data;
@@ -519,30 +532,17 @@ static int kirkwood_i2s_dev_probe(struct platform_device *pdev)
err = snd_soc_register_component(&pdev->dev, &kirkwood_i2s_component,
soc_dai, 1);
- if (!err)
- return 0;
- dev_err(&pdev->dev, "snd_soc_register_component failed\n");
-
- if (!IS_ERR(priv->extclk))
- clk_disable_unprepare(priv->extclk);
- clk_disable_unprepare(priv->clk);
+ if (err) {
+ dev_err(&pdev->dev, "snd_soc_register_component failed\n");
+ goto fail;
+ }
+ return 0;
+fail:
+ kirkwood_i2s_dev_remove(pdev);
return err;
}
-static int kirkwood_i2s_dev_remove(struct platform_device *pdev)
-{
- struct kirkwood_dma_data *priv = dev_get_drvdata(&pdev->dev);
-
- snd_soc_unregister_component(&pdev->dev);
-
- if (!IS_ERR(priv->extclk))
- clk_disable_unprepare(priv->extclk);
- clk_disable_unprepare(priv->clk);
-
- return 0;
-}
-
static struct platform_driver kirkwood_i2s_driver = {
.probe = kirkwood_i2s_dev_probe,
.remove = kirkwood_i2s_dev_remove,
--
Ken ar c'hentañ | ** Breizh ha Linux atav! **
Jef | http://moinejf.free.fr/
On Wed, Jul 31, 2013 at 08:17:39AM +0200, Jean-Francois Moine wrote:
> @@ -519,30 +532,17 @@ static int kirkwood_i2s_dev_probe(struct platform_device *pdev)
>
> err = snd_soc_register_component(&pdev->dev, &kirkwood_i2s_component,
> soc_dai, 1);
> - if (!err)
> - return 0;
> - dev_err(&pdev->dev, "snd_soc_register_component failed\n");
> -
> - if (!IS_ERR(priv->extclk))
> - clk_disable_unprepare(priv->extclk);
> - clk_disable_unprepare(priv->clk);
> + if (err) {
> + dev_err(&pdev->dev, "snd_soc_register_component failed\n");
> + goto fail;
> + }
> + return 0;
This doesn't apply to mainline as there's a clk_put() there.
On Wed, Jul 31, 2013 at 08:17:39AM +0200, Jean-Francois Moine wrote:
> The function kirkwood_i2s_dev_remove() may be used when probe fails.
Looking at this deeper, I'm not happy with this.
> +static int kirkwood_i2s_dev_remove(struct platform_device *pdev)
> +{
> + struct kirkwood_dma_data *priv = dev_get_drvdata(&pdev->dev);
> +
> + snd_soc_unregister_component(&pdev->dev);
...
> @@ -519,30 +532,17 @@ static int kirkwood_i2s_dev_probe(struct platform_device *pdev)
>
> err = snd_soc_register_component(&pdev->dev, &kirkwood_i2s_component,
> soc_dai, 1);
> + if (err) {
> + dev_err(&pdev->dev, "snd_soc_register_component failed\n");
> + goto fail;
> + }
> + return 0;
>
> +fail:
> + kirkwood_i2s_dev_remove(pdev);
What this means is that if snd_soc_register_component() fails, we end
up calling snd_soc_unregister_component(). This may be fine with the
way snd_soc_unregister_component() is currently implemented, but you're
making the assumption that it's fine to call snd_soc_unregister_component()
for a device which hasn't been registered. Technically, this is a
layering violation, which makes this change fragile if the behaviour
of snd_soc_unregister_component() changes in the future.
For the sake of two calls in the error path, I don't think the benefits
of this patch outweigh the risk.
On Sat, 3 Aug 2013 13:39:54 +0100
Russell King - ARM Linux <[email protected]> wrote:
> This doesn't apply to mainline as there's a clk_put() there.
Right, it applies to linux-next:
On Fri, 26 Jul 2013 11:32:50 +0100, Mark Brown <[email protected]> wrote:
> Always submit against where the code will be applied - -next in this
> case.
--
Ken ar c'hentañ | ** Breizh ha Linux atav! **
Jef | http://moinejf.free.fr/
On Sat, Aug 03, 2013 at 05:55:13PM +0200, Jean-Francois Moine wrote:
> On Sat, 3 Aug 2013 13:39:54 +0100
> Russell King - ARM Linux <[email protected]> wrote:
>
> > This doesn't apply to mainline as there's a clk_put() there.
>
> Right, it applies to linux-next:
>
> On Fri, 26 Jul 2013 11:32:50 +0100, Mark Brown <[email protected]> wrote:
>
> > Always submit against where the code will be applied - -next in this
> > case.
Sorry, I'm just not going to bother with your patches unless you start
posting them in such a way that interested parties can test them without
having to run around and pull random trees. In other words, you should
have Cc'd interested parties on that devm_* change patch - especially
myself as I was the one who introduced that code.
On Sat, Aug 03, 2013 at 05:10:29PM +0100, Russell King - ARM Linux wrote:
> having to run around and pull random trees. In other words, you should
> have Cc'd interested parties on that devm_* change patch - especially
> myself as I was the one who introduced that code.
That was me, sorry - Jean-Francios had nothing to do with the patch.
On Sat, 3 Aug 2013 13:46:52 +0100
Russell King - ARM Linux <[email protected]> wrote:
> On Wed, Jul 31, 2013 at 08:17:39AM +0200, Jean-Francois Moine wrote:
> > The function kirkwood_i2s_dev_remove() may be used when probe fails.
>
> Looking at this deeper, I'm not happy with this.
[snip]
> What this means is that if snd_soc_register_component() fails, we end
> up calling snd_soc_unregister_component(). This may be fine with the
> way snd_soc_unregister_component() is currently implemented, but you're
> making the assumption that it's fine to call snd_soc_unregister_component()
> for a device which hasn't been registered. Technically, this is a
> layering violation, which makes this change fragile if the behaviour
> of snd_soc_unregister_component() changes in the future.
>
> For the sake of two calls in the error path, I don't think the benefits
> of this patch outweigh the risk.
You are right, but if snd_soc_unregister_component() could be
officially used safely with no previous call to
snd_soc_register_component(), this would simplify error handling in
other drivers too...
--
Ken ar c'hentañ | ** Breizh ha Linux atav! **
Jef | http://moinejf.free.fr/