Quoting Niklas Söderlund (2021-11-01 16:55:02)

Hi Kieran,

On 2021-11-01 16:30:09 +0000, Kieran Bingham wrote:

quoted

Hi Sakari,

Quoting Sakari Ailus (2021-11-01 13:25:02)

quoted

Select VIDEO_V4L2 for the driver actually depends on it, failing to
compile otherwise.

Fixes: 51758f8b32134bacbf30bd217f7c2074e9b4b51e ("media: staging: max96712: Add basic support for MAX96712 GMSL2 deserializer")

I guess this is a local commit in your tree so far, I can't identify it.

quoted

Reported-by: kernel test robot <redacted>
Signed-off-by: Sakari Ailus <sakari.ailus@linux.intel.com>
---
 drivers/staging/media/max96712/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/staging/media/max96712/Kconfig b/drivers/staging/media/max96712/Kconfig
index 258d47644cbd..492a7ff70dd8 100644
--- a/drivers/staging/media/max96712/Kconfig
+++ b/drivers/staging/media/max96712/Kconfig

@@ -6,6 +6,7 @@ config VIDEO_MAX96712
        select V4L2_FWNODE
        select VIDEO_V4L2_SUBDEV_API
        select MEDIA_CONTROLLER
+       select VIDEO_V4L2

I don't see any other driver 'select' VIDEO_V4L2...
Most 'depend' on VIDEO_V4L2 instead.

And curiously, I already see
 - Add 'depends on VIDEO_V4L2'

in Niklas' v4.

I noticed too late that v3 was already in the media staging tree so I 
submitted a v4 with this fix. I should probably have submitted a Fixes 
patch like this one instead of a v4 if I knew it was already picked-up.

quoted

Is this one a special case that 'needs' to select it?

Both 'select' and 'depends on' solves the problem reported by the bot, 
which one is more correct I do not know.

Aha, I see. Well - consistency would tell me that 'depends on' would
match almost every other driver.

I notice that the MAX9286 suffers the same problem. I'll send a patch.
--
Kieran

quoted

--
Kieran

quoted

        help
          This driver supports the Maxim MAX96712 Quad GMSL2 Deserializer.
 
-- 
2.30.2

-- 
Regards,
Niklas Söderlund