[PATCH v1 0/2] media: atomisp: Fix and refactor PMIC I²C discovery

[PATCH v1 0/2] media: atomisp: Fix and refactor PMIC I²C discovery

[Andy Shevchenko]

Drop the reference count when detecting PMIC.
Refactor PMIC detection using new API.

Andy Shevchenko (2):
  media: atomisp: Put PMIC device after getting its I²C address
  media: atomisp: Replace open-coded i2c_find_device_by_fwnode()

 .../media/atomisp/pci/atomisp_gmin_platform.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)

-- 
2.43.0.rc1.1.gbec44491f096

[PATCH v1 1/2] media: atomisp: Put PMIC device after getting its I²C address

[Andy Shevchenko]

We don't use the PMIC I²C client device after getting its address.
Drop the reference to it. We do not expect device to disappear
as it should be taken care by the OpRegion drivers.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 .../media/atomisp/pci/atomisp_gmin_platform.c       | 13 ++++++++-----
 1 file changed, 8 insertions(+), 5 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
index 139ad7ad1dcf..80aa2211cdc3 100644
--- a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
+++ b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
@@ -497,16 +497,19 @@ static u8 gmin_get_pmic_id_and_addr(struct device *dev)
 	if (pmic_id)
 		return pmic_i2c_addr;
 
-	if (gmin_i2c_dev_exists(dev, PMIC_ACPI_TI, &power))
+	if (gmin_i2c_dev_exists(dev, PMIC_ACPI_TI, &power)) {
 		pmic_id = PMIC_TI;
-	else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_AXP, &power))
+	} else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_AXP, &power)) {
 		pmic_id = PMIC_AXP;
-	else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_CRYSTALCOVE, &power))
+	} else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_CRYSTALCOVE, &power)) {
 		pmic_id = PMIC_CRYSTALCOVE;
-	else
+	} else {
 		pmic_id = PMIC_REGULATOR;
+		return 0;
+	}
 
-	pmic_i2c_addr = power ? power->addr : 0;
+	pmic_i2c_addr = power->addr;
+	put_device(&power->dev);
 	return pmic_i2c_addr;
 }
 
-- 
2.43.0.rc1.1.gbec44491f096

Re: [PATCH v1 1/2] media: atomisp: Put PMIC device after getting its I²C address

[Hans de Goede]

Hi,

On 3/26/24 9:27 PM, Andy Shevchenko wrote:
> We don't use the PMIC I²C client device after getting its address.
> Drop the reference to it. We do not expect device to disappear
> as it should be taken care by the OpRegion drivers.
> 
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

Thank you for your patch.

I have merged this in my media-atomisp branch:
https://git.kernel.org/pub/scm/linux/kernel/git/hansg/linux.git/log/?h=media-atomisp

And this patch will be included in my next
pull-request to Mauro (to media subsystem maintainer)

Regards,

Hans



> ---
>  .../media/atomisp/pci/atomisp_gmin_platform.c       | 13 ++++++++-----
>  1 file changed, 8 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> index 139ad7ad1dcf..80aa2211cdc3 100644
> --- a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> +++ b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> @@ -497,16 +497,19 @@ static u8 gmin_get_pmic_id_and_addr(struct device *dev)
>  	if (pmic_id)
>  		return pmic_i2c_addr;
>  
> -	if (gmin_i2c_dev_exists(dev, PMIC_ACPI_TI, &power))
> +	if (gmin_i2c_dev_exists(dev, PMIC_ACPI_TI, &power)) {
>  		pmic_id = PMIC_TI;
> -	else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_AXP, &power))
> +	} else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_AXP, &power)) {
>  		pmic_id = PMIC_AXP;
> -	else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_CRYSTALCOVE, &power))
> +	} else if (gmin_i2c_dev_exists(dev, PMIC_ACPI_CRYSTALCOVE, &power)) {
>  		pmic_id = PMIC_CRYSTALCOVE;
> -	else
> +	} else {
>  		pmic_id = PMIC_REGULATOR;
> +		return 0;
> +	}
>  
> -	pmic_i2c_addr = power ? power->addr : 0;
> +	pmic_i2c_addr = power->addr;
> +	put_device(&power->dev);
>  	return pmic_i2c_addr;
>  }
>  

[PATCH v1 2/2] media: atomisp: Replace open-coded i2c_find_device_by_fwnode()

[Andy Shevchenko]

gmin_i2c_dev_exists() is using open-coded variant of
i2c_find_device_by_fwnode(). Replace it with a corresponding call.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
index 80aa2211cdc3..b7c477280986 100644
--- a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
+++ b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
@@ -394,12 +394,10 @@ static struct i2c_client *gmin_i2c_dev_exists(struct device *dev, char *name,
 	if (!adev)
 		return NULL;
 
-	d = bus_find_device_by_acpi_dev(&i2c_bus_type, adev);
+	d = get_device(acpi_get_first_physical_node(adev));
 	acpi_dev_put(adev);
-	if (!d)
-		return NULL;
 
-	*client = i2c_verify_client(d);
+	*client = i2c_find_device_by_fwnode(dev_fwnode(d));
 	put_device(d);
 
 	dev_dbg(dev, "found '%s' at address 0x%02x, adapter %d\n",
-- 
2.43.0.rc1.1.gbec44491f096

Re: [PATCH v1 2/2] media: atomisp: Replace open-coded i2c_find_device_by_fwnode()

[Hans de Goede]

Hi Andy,

Thank you for the patch.

On 3/26/24 9:27 PM, Andy Shevchenko wrote:
> gmin_i2c_dev_exists() is using open-coded variant of
> i2c_find_device_by_fwnode(). Replace it with a corresponding call.
> 
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
> ---
>  drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c | 6 ++----
>  1 file changed, 2 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> index 80aa2211cdc3..b7c477280986 100644
> --- a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> +++ b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> @@ -394,12 +394,10 @@ static struct i2c_client *gmin_i2c_dev_exists(struct device *dev, char *name,
>  	if (!adev)
>  		return NULL;
>  
> -	d = bus_find_device_by_acpi_dev(&i2c_bus_type, adev);
> +	d = get_device(acpi_get_first_physical_node(adev));
>  	acpi_dev_put(adev);
> -	if (!d)
> -		return NULL;
>  
> -	*client = i2c_verify_client(d);
> +	*client = i2c_find_device_by_fwnode(dev_fwnode(d));
>  	put_device(d);
>  
>  	dev_dbg(dev, "found '%s' at address 0x%02x, adapter %d\n",

If we are ging to simplify this I think the change should be:

From ea66d15b9a72fcb8baf22a2ff059f2b842a91b67 Mon Sep 17 00:00:00 2001
From: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Date: Wed, 10 Apr 2024 12:32:11 +0200
Subject: [PATCH] media: atomisp: Replace open-coded
 i2c_find_device_by_fwnode()

gmin_i2c_dev_exists() is using open-coded variant of
i2c_find_device_by_fwnode(). Replace it with a corresponding call.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
---
Changes in v2:
- Directly use the ACPI fwnode through acpi_fwnode_handle(adev) instead
  of first calling acpi_get_first_physical_node(adev)
---
 drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
index 804ffff245f3..ed6214327ce5 100644
--- a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
+++ b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
@@ -394,14 +394,11 @@ static struct i2c_client *gmin_i2c_dev_exists(struct device *dev, char *name,
 	if (!adev)
 		return NULL;
 
-	d = bus_find_device_by_acpi_dev(&i2c_bus_type, adev);
+	*client = i2c_find_device_by_fwnode(acpi_fwnode_handle(adev));
 	acpi_dev_put(adev);
-	if (!d)
+	if (!*client)
 		return NULL;
 
-	*client = i2c_verify_client(d);
-	put_device(d);
-
 	dev_dbg(dev, "found '%s' at address 0x%02x, adapter %d\n",
 		(*client)->name, (*client)->addr, (*client)->adapter->nr);
 	return *client;
-- 
2.44.0

Note I added a:

	if (!*client)
		return NULL;

To fix a bug in the original code where the dev_dbg() arguments
are dereferencing a potential NULL *client ptr.

I'm going to merge this variant into my media-atomisp branch
instead of the orignal.

Regards,

Hans

Re: [PATCH v1 2/2] media: atomisp: Replace open-coded i2c_find_device_by_fwnode()

[Andy Shevchenko]

On Wed, Apr 10, 2024 at 12:35:29PM +0200, Hans de Goede wrote:
> On 3/26/24 9:27 PM, Andy Shevchenko wrote:

...

> I'm going to merge this variant into my media-atomisp branch
> instead of the orignal.

Thanks!

This explains why changelog made the commit message.
Was it done deliberately?

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v1 2/2] media: atomisp: Replace open-coded i2c_find_device_by_fwnode()

[Hans de Goede]

Hi Andy,

On 4/10/24 3:35 PM, Andy Shevchenko wrote:
> On Wed, Apr 10, 2024 at 12:35:29PM +0200, Hans de Goede wrote:
>> On 3/26/24 9:27 PM, Andy Shevchenko wrote:
> 
> ...
> 
>> I'm going to merge this variant into my media-atomisp branch
>> instead of the orignal.
> 
> Thanks!
> 
> This explains why changelog made the commit message.
> Was it done deliberately?

Oops, no that was by accident. I normally git send-email
+ git am my patches and then git am cuts it off...

I just send out a pull-request with this minor wart in it
(after testing), so unless there are other reasons to respin
I guess we'll have to live with it.

Regards,

Hans

Re: [PATCH v1 2/2] media: atomisp: Replace open-coded i2c_find_device_by_fwnode()

[Andy Shevchenko]

On Wed, Apr 10, 2024 at 03:41:15PM +0200, Hans de Goede wrote:
> On 4/10/24 3:35 PM, Andy Shevchenko wrote:
> > On Wed, Apr 10, 2024 at 12:35:29PM +0200, Hans de Goede wrote:
> >> On 3/26/24 9:27 PM, Andy Shevchenko wrote:

...

> >> I'm going to merge this variant into my media-atomisp branch
> >> instead of the orignal.
> > 
> > Thanks!
> > 
> > This explains why changelog made the commit message.
> > Was it done deliberately?
> 
> Oops, no that was by accident. I normally git send-email
> + git am my patches and then git am cuts it off...
> 
> I just send out a pull-request with this minor wart in it
> (after testing), so unless there are other reasons to respin
> I guess we'll have to live with it.

You might check the linux-next scripts or ask Stephen if this is an issue.
In any case we will know sooner or later :-)

-- 
With Best Regards,
Andy Shevchenko