[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <ZQGbn76+vd9mFmUa@kekkonen.localdomain>
Date: Wed, 13 Sep 2023 11:23:11 +0000
From: Sakari Ailus <sakari.ailus@...ux.intel.com>
To: Laurent Pinchart <laurent.pinchart@...asonboard.com>
Cc: Arnd Bergmann <arnd@...nel.org>,
Mauro Carvalho Chehab <mchehab@...nel.org>,
Hans Verkuil <hverkuil-cisco@...all.nl>,
Hans de Goede <hdegoede@...hat.com>,
Arnd Bergmann <arnd@...db.de>, linux-media@...r.kernel.org,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH] media: fix VIDEO_CAMERA_SENSOR dependencies
Hi Laurent,
On Tue, Sep 12, 2023 at 06:59:56PM +0300, Laurent Pinchart wrote:
> On Tue, Sep 12, 2023 at 12:16:03PM +0000, Sakari Ailus wrote:
> > Hi Arnd,
> >
> > On Tue, Sep 12, 2023 at 02:01:42PM +0200, Arnd Bergmann wrote:
> > > From: Arnd Bergmann <arnd@...db.de>
> > >
> > > The new Kconfig submenu now guards the visibility of camera sensor drivers,
> > > and any driver that selects one of them runs into a missing dependency warning
> > > if it's disabled:
> > >
> > > WARNING: unmet direct dependencies detected for VIDEO_OV2640
> > > Depends on [n]: MEDIA_SUPPORT [=y] && VIDEO_DEV [=y] && VIDEO_CAMERA_SENSOR [=n]
> > > Selected by [y]:
> > > - VIDEO_EM28XX_V4L2 [=y] && USB [=y] && MEDIA_SUPPORT [=y] && MEDIA_USB_SUPPORT [=y] && (MEDIA_CAMERA_SUPPORT [=y] || MEDIA_ANALOG_TV_SUPPORT [=y] || MEDIA_DIGITAL_TV_SUPPORT [=y]) && VIDEO_EM28XX [=y] && MEDIA_SUBDRV_AUTOSELECT [=y] && MEDIA_CAMERA_SUPPORT [=y]
> > > - VIDEO_GO7007 [=y] && MEDIA_SUPPORT [=y] && MEDIA_USB_SUPPORT [=y] && MEDIA_ANALOG_TV_SUPPORT [=y] && VIDEO_DEV [=y] && I2C [=y] && SND [=y] && USB [=y] && MEDIA_SUBDRV_AUTOSELECT [=y] && MEDIA_CAMERA_SUPPORT [=y]
> > >
> > > WARNING: unmet direct dependencies detected for VIDEO_MT9V011
> > > Depends on [n]: MEDIA_SUPPORT [=y] && VIDEO_DEV [=y] && VIDEO_CAMERA_SENSOR [=n]
> > > Selected by [y]:
> > > - VIDEO_EM28XX_V4L2 [=y] && USB [=y] && MEDIA_SUPPORT [=y] && MEDIA_USB_SUPPORT [=y] && (MEDIA_CAMERA_SUPPORT [=y] || MEDIA_ANALOG_TV_SUPPORT [=y] || MEDIA_DIGITAL_TV_SUPPORT [=y]) && VIDEO_EM28XX [=y] && MEDIA_SUBDRV_AUTOSELECT [=y] && MEDIA_CAMERA_SUPPORT [=y]
> > >
> > > WARNING: unmet direct dependencies detected for VIDEO_OV7670
> > > Depends on [n]: MEDIA_SUPPORT [=y] && VIDEO_DEV [=y] && VIDEO_CAMERA_SENSOR [=n]
> > > Selected by [y]:
> > > - VIDEO_CAFE_CCIC [=y] && MEDIA_SUPPORT [=y] && MEDIA_PLATFORM_SUPPORT [=y] && MEDIA_PLATFORM_DRIVERS [=y] && V4L_PLATFORM_DRIVERS [=y] && PCI [=y] && I2C [=y] && VIDEO_DEV [=y] && COMMON_CLK [=y]
> > > - VIDEO_MMP_CAMERA [=y] && MEDIA_SUPPORT [=y] && MEDIA_PLATFORM_SUPPORT [=y] && MEDIA_PLATFORM_DRIVERS [=y] && V4L_PLATFORM_DRIVERS [=y] && I2C [=y] && VIDEO_DEV [=y] && (ARCH_MMP [=y] || COMPILE_TEST [=y]) && COMMON_CLK [=y]
> > > - VIDEO_VIA_CAMERA [=y] && MEDIA_SUPPORT [=y] && MEDIA_PLATFORM_SUPPORT [=y] && MEDIA_PLATFORM_DRIVERS [=y] && V4L_PLATFORM_DRIVERS [=y] && FB_VIA [=y] && VIDEO_DEV [=y]
> > >
> > > Some of these are guarded by 'if MEDIA_SUBDRV_AUTOSELECT &&
> > > MEDIA_CAMERA_SUPPORT', which seems to be the right approach, so update
> > > those to use the new VIDEO_CAMERA_SENSOR symbol instead of
> > > MEDIA_CAMERA_SUPPORT and add the same condition to the ones that
> > > don't already have one.
> > >
> > > Fixes: 7d3c7d2a2914e ("media: i2c: Add a camera sensor top level menu")
> > > Signed-off-by: Arnd Bergmann <arnd@...db.de>
> >
> > I believe these matters have been fixed, but the fixes are only in the
> > media stage tree so far (and not in the fixes branch). They should go to
> > v6.6, not v6.7.
>
> That would be commits f59933e8b1d1 ("media: via: Use correct dependency
> for camera sensor drivers") and 8ba283f6c929 ("media: v4l: Use correct
> dependency for camera sensor drivers"). The latter is identical to the
> corresponding changes in this patch, but the former lacks the
> MEDIA_SUBDRV_AUTOSELECT check. Is this intentional ?
This is a bit grey area.
The via parallel receiver driver has been apparently used with a single
camera, the ov7670. It's a very old driver so I don't necessarily expect
this to change.
Changing this either way would have fixed the core issue.
Ideally there should be no connection between the via parallel receiver
driver and any sensor driver.
--
Regards,
Sakari Ailus
Powered by blists - more mailing lists