[PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[Andy Shevchenko]

There is at least one user that will gain from the
i2c_acpi_find_client_by_adev() being exported.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 drivers/i2c/i2c-core-acpi.c | 3 ++-
 include/linux/i2c.h         | 6 ++++++
 2 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/drivers/i2c/i2c-core-acpi.c b/drivers/i2c/i2c-core-acpi.c
index 8ceaa88dd78f..5be37a5efcb4 100644
--- a/drivers/i2c/i2c-core-acpi.c
+++ b/drivers/i2c/i2c-core-acpi.c
@@ -387,7 +387,7 @@ struct i2c_adapter *i2c_acpi_find_adapter_by_handle(acpi_handle handle)
 }
 EXPORT_SYMBOL_GPL(i2c_acpi_find_adapter_by_handle);
 
-static struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev)
+struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev)
 {
 	struct device *dev;
 	struct i2c_client *client;
@@ -402,6 +402,7 @@ static struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev)
 
 	return client;
 }
+EXPORT_SYMBOL_GPL(i2c_acpi_find_client_by_adev);
 
 static int i2c_acpi_notify(struct notifier_block *nb, unsigned long value,
 			   void *arg)
diff --git a/include/linux/i2c.h b/include/linux/i2c.h
index e8f2ac8c9c3d..335dc4f5abbb 100644
--- a/include/linux/i2c.h
+++ b/include/linux/i2c.h
@@ -995,6 +995,7 @@ static inline int of_i2c_get_board_info(struct device *dev,
 
 #endif /* CONFIG_OF */
 
+struct acpi_device;
 struct acpi_resource;
 struct acpi_resource_i2c_serialbus;
 
@@ -1005,6 +1006,7 @@ u32 i2c_acpi_find_bus_speed(struct device *dev);
 struct i2c_client *i2c_acpi_new_device(struct device *dev, int index,
 				       struct i2c_board_info *info);
 struct i2c_adapter *i2c_acpi_find_adapter_by_handle(acpi_handle handle);
+struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev);
 #else
 static inline bool i2c_acpi_get_i2c_resource(struct acpi_resource *ares,
 					     struct acpi_resource_i2c_serialbus **i2c)
@@ -1024,6 +1026,10 @@ static inline struct i2c_adapter *i2c_acpi_find_adapter_by_handle(acpi_handle ha
 {
 	return NULL;
 }
+static inline struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev)
+{
+	return NULL;
+}
 #endif /* CONFIG_ACPI */
 
 #endif /* _LINUX_I2C_H */
-- 
2.30.2

[PATCH v1 2/6] staging: atomisp: Replace open-coded i2c_acpi_find_client_by_adev()

[Andy Shevchenko]

gmin_i2c_dev_exists() is using open-coded variant of
i2c_acpi_find_client_by_adev(). Replace it with a corresponding call.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 .../staging/media/atomisp/pci/atomisp_gmin_platform.c    | 9 ++-------
 1 file changed, 2 insertions(+), 7 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
index 135994d44802..a1064d1a3d6b 100644
--- a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
+++ b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
@@ -378,19 +378,14 @@ static struct i2c_client *gmin_i2c_dev_exists(struct device *dev, char *name,
 					      struct i2c_client **client)
 {
 	struct acpi_device *adev;
-	struct device *d;
 
 	adev = acpi_dev_get_first_match_dev(name, NULL, -1);
 	if (!adev)
 		return NULL;
 
-	d = bus_find_device_by_acpi_dev(&i2c_bus_type, adev);
-	acpi_dev_put(adev);
-	if (!d)
-		return NULL;
+	*client = i2c_acpi_find_client_by_adev(adev);
 
-	*client = i2c_verify_client(d);
-	put_device(d);
+	acpi_dev_put(adev);
 
 	dev_dbg(dev, "found '%s' at address 0x%02x, adapter %d\n",
 		(*client)->name, (*client)->addr, (*client)->adapter->nr);
-- 
2.30.2

Re: [PATCH v1 2/6] staging: atomisp: Replace open-coded i2c_acpi_find_client_by_adev()

[Mauro Carvalho Chehab]

Hi Andy,

Em Wed, 26 May 2021 15:43:18 +0300
Andy Shevchenko <andriy.shevchenko@linux.intel.com> escreveu:

> gmin_i2c_dev_exists() is using open-coded variant of
> i2c_acpi_find_client_by_adev(). Replace it with a corresponding call.
> 
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>

At least on the top of v5.14-rc1, this patch causes a compilation
issue:

	drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c: In function ‘gmin_i2c_dev_exists’:
	drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c:386:19: error: implicit declaration of function ‘i2c_acpi_find_client_by_adev’; did you mean ‘i2c_acpi_find_adapter_by_handle’? [-Werror=implicit-function-declaration]
	  386 |         *client = i2c_acpi_find_client_by_adev(adev);
	      |                   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
	      |                   i2c_acpi_find_adapter_by_handle
	drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c:386:17: warning: assignment to ‘struct i2c_client *’ from ‘int’ makes pointer from integer without a cast [-Wint-conversion]
	  386 |         *client = i2c_acpi_find_client_by_adev(adev);
	      |                 ^

The reason is because such function is static:

	$ git grep i2c_acpi_find_client_by_adev
	drivers/i2c/i2c-core-acpi.c:static struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev)

IMO, a patch like that should be applied at the same tree as a patch
dropping "static" from drivers/i2c/i2c-core-acpi.c. If you want to do
so, feel free to add:

Reviewed-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>

> ---
>  .../staging/media/atomisp/pci/atomisp_gmin_platform.c    | 9 ++-------
>  1 file changed, 2 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> index 135994d44802..a1064d1a3d6b 100644
> --- a/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> +++ b/drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c
> @@ -378,19 +378,14 @@ static struct i2c_client *gmin_i2c_dev_exists(struct device *dev, char *name,
>  					      struct i2c_client **client)
>  {
>  	struct acpi_device *adev;
> -	struct device *d;
>  
>  	adev = acpi_dev_get_first_match_dev(name, NULL, -1);
>  	if (!adev)
>  		return NULL;
>  
> -	d = bus_find_device_by_acpi_dev(&i2c_bus_type, adev);
> -	acpi_dev_put(adev);
> -	if (!d)
> -		return NULL;
> +	*client = i2c_acpi_find_client_by_adev(adev);
>  
> -	*client = i2c_verify_client(d);
> -	put_device(d);
> +	acpi_dev_put(adev);
>  
>  	dev_dbg(dev, "found '%s' at address 0x%02x, adapter %d\n",
>  		(*client)->name, (*client)->addr, (*client)->adapter->nr);



Thanks,
Mauro

Re: [PATCH v1 2/6] staging: atomisp: Replace open-coded i2c_acpi_find_client_by_adev()

[Andy Shevchenko]

On Thu, Jul 22, 2021 at 10:57:44AM +0200, Mauro Carvalho Chehab wrote:
> Em Wed, 26 May 2021 15:43:18 +0300
> Andy Shevchenko <andriy.shevchenko@linux.intel.com> escreveu:
> 
> > gmin_i2c_dev_exists() is using open-coded variant of
> > i2c_acpi_find_client_by_adev(). Replace it with a corresponding call.
> > 
> > Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
> 
> At least on the top of v5.14-rc1, this patch causes a compilation
> issue:
> 
> 	drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c: In function ‘gmin_i2c_dev_exists’:
> 	drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c:386:19: error: implicit declaration of function ‘i2c_acpi_find_client_by_adev’; did you mean ‘i2c_acpi_find_adapter_by_handle’? [-Werror=implicit-function-declaration]
> 	  386 |         *client = i2c_acpi_find_client_by_adev(adev);
> 	      |                   ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 	      |                   i2c_acpi_find_adapter_by_handle
> 	drivers/staging/media/atomisp/pci/atomisp_gmin_platform.c:386:17: warning: assignment to ‘struct i2c_client *’ from ‘int’ makes pointer from integer without a cast [-Wint-conversion]
> 	  386 |         *client = i2c_acpi_find_client_by_adev(adev);
> 	      |                 ^
> 
> The reason is because such function is static:
> 
> 	$ git grep i2c_acpi_find_client_by_adev
> 	drivers/i2c/i2c-core-acpi.c:static struct i2c_client *i2c_acpi_find_client_by_adev(struct acpi_device *adev)
> 
> IMO, a patch like that should be applied at the same tree as a patch
> dropping "static" from drivers/i2c/i2c-core-acpi.c. If you want to do
> so, feel free to add:
> 
> Reviewed-by: Mauro Carvalho Chehab <mchehab+huawei@kernel.org>

Thanks!

There is a v2 of this where the patch is dropped from.


-- 
With Best Regards,
Andy Shevchenko

[PATCH v1 3/6] staging: atomisp: Remove unused declarations

[Andy Shevchenko]

There is a few static declarations that are not used anywhere, remove them.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 .../atomisp/pci/isp2400_input_system_local.h    | 17 -----------------
 1 file changed, 17 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h b/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
index 072a92199e05..27f3c8a45730 100644
--- a/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
+++ b/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
@@ -362,23 +362,6 @@ static const hrt_address MIPI_PORT_OFFSET[N_MIPI_PORT_ID] = {
 	0x00000200UL
 };
 
-static const mipi_lane_cfg_t MIPI_PORT_MAXLANES[N_MIPI_PORT_ID] = {
-	MIPI_4LANE_CFG,
-	MIPI_1LANE_CFG,
-	MIPI_2LANE_CFG
-};
-
-static const bool MIPI_PORT_ACTIVE[N_RX_MODE][N_MIPI_PORT_ID] = {
-	{true, true, false},
-	{true, true, false},
-	{true, true, false},
-	{true, true, false},
-	{true, true, true},
-	{true, true, true},
-	{true, true, true},
-	{true, true, true}
-};
-
 static const mipi_lane_cfg_t MIPI_PORT_LANES[N_RX_MODE][N_MIPI_PORT_ID] = {
 	{MIPI_4LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
 	{MIPI_3LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
-- 
2.30.2

[PATCH v1 4/6] staging: atomisp: Annotate a couple of definitions with __maybe_unused

[Andy Shevchenko]

There are definitions in the header that are not used by all modules inside
the driver. Annotate them with __maybe_unused to avoid compiler warnings.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 .../staging/media/atomisp/pci/isp2400_input_system_local.h    | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h b/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
index 27f3c8a45730..21f11bdfc85f 100644
--- a/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
+++ b/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
@@ -356,7 +356,7 @@ struct rx_cfg_s {
 };
 
 /* NOTE: The base has already an offset of 0x0100 */
-static const hrt_address MIPI_PORT_OFFSET[N_MIPI_PORT_ID] = {
+static const hrt_address __maybe_unused MIPI_PORT_OFFSET[N_MIPI_PORT_ID] = {
 	0x00000000UL,
 	0x00000100UL,
 	0x00000200UL
@@ -373,7 +373,7 @@ static const mipi_lane_cfg_t MIPI_PORT_LANES[N_RX_MODE][N_MIPI_PORT_ID] = {
 	{MIPI_1LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG}
 };
 
-static const hrt_address SUB_SYSTEM_OFFSET[N_SUB_SYSTEM_ID] = {
+static const hrt_address __maybe_unused SUB_SYSTEM_OFFSET[N_SUB_SYSTEM_ID] = {
 	0x00001000UL,
 	0x00002000UL,
 	0x00003000UL,
-- 
2.30.2

[PATCH v1 5/6] staging: atomisp: Move MIPI_PORT_LANES to the only user

[Andy Shevchenko]

Move MIPI_PORT_LANES to the only user of it, i.e. pci/runtime/isys/src/rx.c.

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 .../media/atomisp/pci/isp2400_input_system_local.h    | 11 -----------
 .../staging/media/atomisp/pci/runtime/isys/src/rx.c   | 11 +++++++++++
 2 files changed, 11 insertions(+), 11 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h b/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
index 21f11bdfc85f..2614b89b8e34 100644
--- a/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
+++ b/drivers/staging/media/atomisp/pci/isp2400_input_system_local.h
@@ -362,17 +362,6 @@ static const hrt_address __maybe_unused MIPI_PORT_OFFSET[N_MIPI_PORT_ID] = {
 	0x00000200UL
 };
 
-static const mipi_lane_cfg_t MIPI_PORT_LANES[N_RX_MODE][N_MIPI_PORT_ID] = {
-	{MIPI_4LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
-	{MIPI_3LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
-	{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
-	{MIPI_1LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
-	{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_2LANE_CFG},
-	{MIPI_3LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG},
-	{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG},
-	{MIPI_1LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG}
-};
-
 static const hrt_address __maybe_unused SUB_SYSTEM_OFFSET[N_SUB_SYSTEM_ID] = {
 	0x00001000UL,
 	0x00002000UL,
diff --git a/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c b/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c
index 4a18da6bf0c1..29486dcc9a04 100644
--- a/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c
+++ b/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c
@@ -20,6 +20,17 @@
 #include "ia_css_irq.h"
 #include "sh_css_internal.h"
 
+static const mipi_lane_cfg_t MIPI_PORT_LANES[N_RX_MODE][N_MIPI_PORT_ID] = {
+	{MIPI_4LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
+	{MIPI_3LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
+	{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
+	{MIPI_1LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
+	{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_2LANE_CFG},
+	{MIPI_3LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG},
+	{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG},
+	{MIPI_1LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG}
+};
+
 #if !defined(ISP2401)
 void ia_css_isys_rx_enable_all_interrupts(enum mipi_port_id port)
 {
-- 
2.30.2

Re: [PATCH v1 5/6] staging: atomisp: Move MIPI_PORT_LANES to the only user

[kernel test robot]

[-- Attachment #1: Type: text/plain, Size: 2641 bytes --]

Hi Andy,

I love your patch! Yet something to improve:

[auto build test ERROR on linuxtv-media/master]
[also build test ERROR on wsa/i2c/for-next v5.13-rc3 next-20210526]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Andy-Shevchenko/i2c-acpi-Export-i2c_acpi_find_client_by_adev-for-users/20210526-204453
base:   git://linuxtv.org/media_tree.git master
config: i386-allyesconfig (attached as .config)
compiler: gcc-9 (Debian 9.3.0-22) 9.3.0
reproduce (this is a W=1 build):
        # https://github.com/0day-ci/linux/commit/8f95e744d93a29bc02a3e8487a9fbb49e68d38dc
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Andy-Shevchenko/i2c-acpi-Export-i2c_acpi_find_client_by_adev-for-users/20210526-204453
        git checkout 8f95e744d93a29bc02a3e8487a9fbb49e68d38dc
        # save the attached .config to linux build tree
        make W=1 ARCH=i386 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

>> drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c:23:46: error: 'N_RX_MODE' undeclared here (not in a function); did you mean 'N_RX_ID'?
      23 | static const mipi_lane_cfg_t MIPI_PORT_LANES[N_RX_MODE][N_MIPI_PORT_ID] = {
         |                                              ^~~~~~~~~
         |                                              N_RX_ID
   drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c:23:30: warning: 'MIPI_PORT_LANES' defined but not used [-Wunused-variable]
      23 | static const mipi_lane_cfg_t MIPI_PORT_LANES[N_RX_MODE][N_MIPI_PORT_ID] = {
         |                              ^~~~~~~~~~~~~~~


vim +23 drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c

    22	
  > 23	static const mipi_lane_cfg_t MIPI_PORT_LANES[N_RX_MODE][N_MIPI_PORT_ID] = {
    24		{MIPI_4LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
    25		{MIPI_3LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
    26		{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
    27		{MIPI_1LANE_CFG, MIPI_1LANE_CFG, MIPI_0LANE_CFG},
    28		{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_2LANE_CFG},
    29		{MIPI_3LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG},
    30		{MIPI_2LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG},
    31		{MIPI_1LANE_CFG, MIPI_1LANE_CFG, MIPI_1LANE_CFG}
    32	};
    33	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 65239 bytes --]

[PATCH v1 6/6] staging: atomisp: Remove unused port_enabled variable

[Andy Shevchenko]

Remove unused port_enabled variable in ia_css_isys_rx_configure().

Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
---
 drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c b/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c
index 29486dcc9a04..79e82fc08aa9 100644
--- a/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c
+++ b/drivers/staging/media/atomisp/pci/runtime/isys/src/rx.c
@@ -490,7 +490,6 @@ unsigned int ia_css_csi2_calculate_input_system_alignment(
 void ia_css_isys_rx_configure(const rx_cfg_t *config,
 			      const enum ia_css_input_mode input_mode)
 {
-	bool port_enabled[N_MIPI_PORT_ID];
 	bool any_port_enabled = false;
 	enum mipi_port_id port;
 
@@ -527,8 +526,6 @@ void ia_css_isys_rx_configure(const rx_cfg_t *config,
 					_HRT_CSS_RECEIVER_2400_RX_COUNT_REG_IDX,
 					config->rxcount);
 
-		port_enabled[port] = true;
-
 		if (input_mode != IA_CSS_INPUT_MODE_BUFFERED_SENSOR) {
 			/* MW: A bit of a hack, straight wiring of the capture
 			 * units,assuming they are linearly enumerated. */
-- 
2.30.2

Re: [PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 332 bytes --]

On Wed, May 26, 2021 at 03:43:17PM +0300, Andy Shevchenko wrote:
> There is at least one user that will gain from the
> i2c_acpi_find_client_by_adev() being exported.

No objections per se. But as the user is in staging, I want to ask if
the use there is really a solution we would also accept outside of
staging? Or is it a hack?


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[Andy Shevchenko]

On Thu, May 27, 2021 at 10:26:33PM +0200, Wolfram Sang wrote:
> On Wed, May 26, 2021 at 03:43:17PM +0300, Andy Shevchenko wrote:
> > There is at least one user that will gain from the
> > i2c_acpi_find_client_by_adev() being exported.
> 
> No objections per se. But as the user is in staging, I want to ask if
> the use there is really a solution we would also accept outside of
> staging? Or is it a hack?

The similar OF API is exported for users, although amount of users and their
locations are different. The AtomISP driver is not in the best shape, I agree,
but for now any possible steps to make it better would be good steps in my
opinion. Later we may see if we can do this piece of code differently (IIRC
current way is probably the best taking into account legacy platforms support).

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[Andy Shevchenko]

On Fri, May 28, 2021 at 12:23:31PM +0300, Andy Shevchenko wrote:
> On Thu, May 27, 2021 at 10:26:33PM +0200, Wolfram Sang wrote:
> > On Wed, May 26, 2021 at 03:43:17PM +0300, Andy Shevchenko wrote:
> > > There is at least one user that will gain from the
> > > i2c_acpi_find_client_by_adev() being exported.
> > 
> > No objections per se. But as the user is in staging, I want to ask if
> > the use there is really a solution we would also accept outside of
> > staging? Or is it a hack?
> 
> The similar OF API is exported for users, although amount of users and their
> locations are different. The AtomISP driver is not in the best shape, I agree,
> but for now any possible steps to make it better would be good steps in my
> opinion. Later we may see if we can do this piece of code differently (IIRC
> current way is probably the best taking into account legacy platforms support).

Btw, we may move all current exports from I2C ACPI to its own namespace, then
we won't really care if it'e exported or not, only explicit consumers will use
it.

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[Greg Kroah-Hartman]

On Thu, May 27, 2021 at 10:26:33PM +0200, Wolfram Sang wrote:
> On Wed, May 26, 2021 at 03:43:17PM +0300, Andy Shevchenko wrote:
> > There is at least one user that will gain from the
> > i2c_acpi_find_client_by_adev() being exported.
> 
> No objections per se. But as the user is in staging, I want to ask if
> the use there is really a solution we would also accept outside of
> staging? Or is it a hack?

staging drivers should be self-contained, do not accept code in the core
kernel that only is used by staging drivers.

So I would not recommend this be accepted at this point in time.

Andy, work to get the driver out of staging first before doing stuff
like this.

thanks,

greg k-h

Re: [PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[Andy Shevchenko]

On Fri, May 28, 2021 at 11:25:51AM +0200, Greg Kroah-Hartman wrote:
> On Thu, May 27, 2021 at 10:26:33PM +0200, Wolfram Sang wrote:
> > On Wed, May 26, 2021 at 03:43:17PM +0300, Andy Shevchenko wrote:
> > > There is at least one user that will gain from the
> > > i2c_acpi_find_client_by_adev() being exported.
> > 
> > No objections per se. But as the user is in staging, I want to ask if
> > the use there is really a solution we would also accept outside of
> > staging? Or is it a hack?
> 
> staging drivers should be self-contained, do not accept code in the core
> kernel that only is used by staging drivers.
> 
> So I would not recommend this be accepted at this point in time.

Fair enough.

> Andy, work to get the driver out of staging first before doing stuff
> like this.

Okay, I'll drop first one and patches related to it in the v2.
It should bring us closer to the mentioned point.

Thanks for clarification!

-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v1 1/6] i2c: acpi: Export i2c_acpi_find_client_by_adev() for users

[Wolfram Sang]

[-- Attachment #1: Type: text/plain, Size: 316 bytes --]


> > staging drivers should be self-contained, do not accept code in the core
> > kernel that only is used by staging drivers.
> > 
> > So I would not recommend this be accepted at this point in time.
> 
> Fair enough.

You could add a comment, though, so we won't forget to remove the open
coding then.


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]