diff mbox series

[v1,1/1] ASoC: SOF: Intel: bdw: remove duplicating driver data retrieval

Message ID 20220703143533.62173-1-andriy.shevchenko@linux.intel.com (mailing list archive)
State Superseded
Headers show
Series [v1,1/1] ASoC: SOF: Intel: bdw: remove duplicating driver data retrieval | expand

Commit Message

Andy Shevchenko July 3, 2022, 2:35 p.m. UTC
device_get_match_data() in ACPI case calls similar to acpi_match_device().
Hence there is no need to duplicate the call. Just assign what is in
the id->driver_data.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 sound/soc/sof/intel/bdw.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

Comments

Peter Ujfalusi July 4, 2022, 7:31 a.m. UTC | #1
On 03/07/2022 17:35, Andy Shevchenko wrote:
> device_get_match_data() in ACPI case calls similar to acpi_match_device().
> Hence there is no need to duplicate the call. Just assign what is in
> the id->driver_data.
> 
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
> ---
>  sound/soc/sof/intel/bdw.c | 7 ++-----
>  1 file changed, 2 insertions(+), 5 deletions(-)
> 
> diff --git a/sound/soc/sof/intel/bdw.c b/sound/soc/sof/intel/bdw.c
> index 26df780c702e..346813aa3768 100644
> --- a/sound/soc/sof/intel/bdw.c
> +++ b/sound/soc/sof/intel/bdw.c
> @@ -674,6 +674,7 @@ static int sof_broadwell_probe(struct platform_device *pdev)
>  	id = acpi_match_device(dev->driver->acpi_match_table, dev);
>  	if (!id)
>  		return -ENODEV;
> +	desc = (const struct sof_dev_desc *)id->driver_data;
>  
>  	ret = snd_intel_acpi_dsp_driver_probe(dev, id->id);
>  	if (ret != SND_INTEL_DSP_DRIVER_ANY && ret != SND_INTEL_DSP_DRIVER_SOF) {
> @@ -681,11 +682,7 @@ static int sof_broadwell_probe(struct platform_device *pdev)
>  		return -ENODEV;
>  	}
>  
> -	desc = device_get_match_data(dev);
> -	if (!desc)
> -		return -ENODEV;
> -
> -	return sof_acpi_probe(pdev, device_get_match_data(dev));
> +	return sof_acpi_probe(pdev, desc);
>  }

Would it not be simpler to just:

diff --git a/sound/soc/sof/intel/bdw.c b/sound/soc/sof/intel/bdw.c
index 26df780c702e..2ee22c2772ba 100644
--- a/sound/soc/sof/intel/bdw.c
+++ b/sound/soc/sof/intel/bdw.c
@@ -668,7 +668,6 @@ static int sof_broadwell_probe(struct platform_device *pdev)
 {
 	struct device *dev = &pdev->dev;
 	const struct acpi_device_id *id;
-	const struct sof_dev_desc *desc;
 	int ret;
 
 	id = acpi_match_device(dev->driver->acpi_match_table, dev);
@@ -681,11 +680,7 @@ static int sof_broadwell_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 
-	desc = device_get_match_data(dev);
-	if (!desc)
-		return -ENODEV;
-
-	return sof_acpi_probe(pdev, device_get_match_data(dev));
+	return sof_acpi_probe(pdev, (const struct sof_dev_desc *)id->driver_data);
 }
 
 /* acpi_driver definition */
---
We don't use the desc in sof_broadwell_probe(), so we can just pass the id->driver_data
Andy Shevchenko July 4, 2022, 4:19 p.m. UTC | #2
On Mon, Jul 04, 2022 at 10:31:48AM +0300, Péter Ujfalusi wrote:
> On 03/07/2022 17:35, Andy Shevchenko wrote:

...

> +	return sof_acpi_probe(pdev, (const struct sof_dev_desc *)id->driver_data);
>  }

> We don't use the desc in sof_broadwell_probe(), so we can just pass the id->driver_data

I prefer not to cast inside the function parameters when it's not about POD.
Peter Ujfalusi July 5, 2022, 12:09 p.m. UTC | #3
On 04/07/2022 19:19, Andy Shevchenko wrote:
> On Mon, Jul 04, 2022 at 10:31:48AM +0300, Péter Ujfalusi wrote:
>> On 03/07/2022 17:35, Andy Shevchenko wrote:
> 
> ...
> 
>> +	return sof_acpi_probe(pdev, (const struct sof_dev_desc *)id->driver_data);
>>  }
> 
>> We don't use the desc in sof_broadwell_probe(), so we can just pass the id->driver_data
> 
> I prefer not to cast inside the function parameters when it's not about POD.

Fair enough.

How about assigning desc before it is used?

diff --git a/sound/soc/sof/intel/bdw.c b/sound/soc/sof/intel/bdw.c
index 26df780c702e..a446154f2803 100644
--- a/sound/soc/sof/intel/bdw.c
+++ b/sound/soc/sof/intel/bdw.c
@@ -681,11 +681,8 @@ static int sof_broadwell_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 
-	desc = device_get_match_data(dev);
-	if (!desc)
-		return -ENODEV;
-
-	return sof_acpi_probe(pdev, device_get_match_data(dev));
+	desc = (const struct sof_dev_desc *)id->driver_data;
+	return sof_acpi_probe(pdev, desc);
 }
 
 /* acpi_driver definition */
Pierre-Louis Bossart July 5, 2022, 2:42 p.m. UTC | #4
On 7/3/22 09:35, Andy Shevchenko wrote:
> device_get_match_data() in ACPI case calls similar to acpi_match_device().
> Hence there is no need to duplicate the call. Just assign what is in
> the id->driver_data.
> 
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
> ---
>  sound/soc/sof/intel/bdw.c | 7 ++-----
>  1 file changed, 2 insertions(+), 5 deletions(-)
> 
> diff --git a/sound/soc/sof/intel/bdw.c b/sound/soc/sof/intel/bdw.c
> index 26df780c702e..346813aa3768 100644
> --- a/sound/soc/sof/intel/bdw.c
> +++ b/sound/soc/sof/intel/bdw.c
> @@ -674,6 +674,7 @@ static int sof_broadwell_probe(struct platform_device *pdev)
>  	id = acpi_match_device(dev->driver->acpi_match_table, dev);
>  	if (!id)
>  		return -ENODEV;
> +	desc = (const struct sof_dev_desc *)id->driver_data;
>  
>  	ret = snd_intel_acpi_dsp_driver_probe(dev, id->id);
>  	if (ret != SND_INTEL_DSP_DRIVER_ANY && ret != SND_INTEL_DSP_DRIVER_SOF) {
> @@ -681,11 +682,7 @@ static int sof_broadwell_probe(struct platform_device *pdev)
>  		return -ENODEV;
>  	}
>  
> -	desc = device_get_match_data(dev);
> -	if (!desc)
> -		return -ENODEV;
> -
> -	return sof_acpi_probe(pdev, device_get_match_data(dev));
> +	return sof_acpi_probe(pdev, desc);

Thanks Andy, we have the same code pattern in sound/soc/sof/intel/byt.c,
can we change it as part of the series to keep the two files aligned?

>  }
>  
>  /* acpi_driver definition */
Andy Shevchenko July 5, 2022, 3:25 p.m. UTC | #5
On Tue, Jul 5, 2022 at 3:19 PM Péter Ujfalusi
<peter.ujfalusi@linux.intel.com> wrote:
> On 04/07/2022 19:19, Andy Shevchenko wrote:
> > On Mon, Jul 04, 2022 at 10:31:48AM +0300, Péter Ujfalusi wrote:
> >> On 03/07/2022 17:35, Andy Shevchenko wrote:

...

> >> +    return sof_acpi_probe(pdev, (const struct sof_dev_desc *)id->driver_data);
> >>  }
> >
> >> We don't use the desc in sof_broadwell_probe(), so we can just pass the id->driver_data
> >
> > I prefer not to cast inside the function parameters when it's not about POD.
>
> Fair enough.
>
> How about assigning desc before it is used?

As answered in the other email, I'm fine to modify the change accordingly.
Andy Shevchenko July 5, 2022, 3:26 p.m. UTC | #6
On Tue, Jul 5, 2022 at 5:02 PM Pierre-Louis Bossart
<pierre-louis.bossart@linux.intel.com> wrote:
> On 7/3/22 09:35, Andy Shevchenko wrote:

...

> Thanks Andy, we have the same code pattern in sound/soc/sof/intel/byt.c,
> can we change it as part of the series to keep the two files aligned?

Sure. I was looking for the ID of WPT SPI (I2S) controller and that's
how I found these two.
diff mbox series

Patch

diff --git a/sound/soc/sof/intel/bdw.c b/sound/soc/sof/intel/bdw.c
index 26df780c702e..346813aa3768 100644
--- a/sound/soc/sof/intel/bdw.c
+++ b/sound/soc/sof/intel/bdw.c
@@ -674,6 +674,7 @@  static int sof_broadwell_probe(struct platform_device *pdev)
 	id = acpi_match_device(dev->driver->acpi_match_table, dev);
 	if (!id)
 		return -ENODEV;
+	desc = (const struct sof_dev_desc *)id->driver_data;
 
 	ret = snd_intel_acpi_dsp_driver_probe(dev, id->id);
 	if (ret != SND_INTEL_DSP_DRIVER_ANY && ret != SND_INTEL_DSP_DRIVER_SOF) {
@@ -681,11 +682,7 @@  static int sof_broadwell_probe(struct platform_device *pdev)
 		return -ENODEV;
 	}
 
-	desc = device_get_match_data(dev);
-	if (!desc)
-		return -ENODEV;
-
-	return sof_acpi_probe(pdev, device_get_match_data(dev));
+	return sof_acpi_probe(pdev, desc);
 }
 
 /* acpi_driver definition */