Re: [PATCH v2 0/5] adds ovm6211 driver to staging

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Quoting Petko Manolov (2022-01-03 21:19:21)
> On 22-01-03 13:19:22, Kieran Bingham wrote:
> > Hi Petko,
> > 
> > Quoting Petko Manolov (2022-01-03 11:09:17)
> > > v2: Removes an unused function (ovm6211_set_pix_clk) and this patch series is
> > > now based on media/master; Didn't receive any comments about the RFC version,
> > > thus i assume everything is perfect... :P
> > 
> > Did you see
> > https://lore.kernel.org/linux-media/Ya9XHiz%2FPm4CjQ13@xxxxxxxxxxxxxxxxxxxxxx/?
> > 
> > Sakari provided quite a few review comments to consider.
> 
> Nope, somehow his message has slipped from my attention.  I'd like to thank
> Sakari for the thorough review.  This is my first v4l2 driver and i have most
> likely made a lot of mistakes.  I'll address all his comments in v3 of the
> series along with some elaboration on my part.
> 
> > I don't think we need to add new sensor drivers to the staging directory which
> > would simplify your series quite a bit, and Sakari also stated the ovm6211
> > KConfig and Makefile entry should be in the patch along with the new driver
> > code (not in staging).
> 
> This is the exact opposite to what i've done for the netdev tree, where each
> change should be in a separate patch.  Anyway, i'll follow the media tree rules.
> 
> > So you would need to refactor this series to a single patch adding the driver
> > do drivers/media/i2c/, and a second patch which adds the DT-bindings
> > accordingly.
> 
> I am not sure about how practical the DT will be in this case.  The sensor was
> used on a custom board and a rather specific reset pin wiring.  I've tried to
> remove this logic from the driver, but it is still reflected in the DT that
> we've been using so far.  I've got to think about this one some more...

That's why the bindings are different from the DT itself. The specifics
of how you connect it up belong to your device tree, but the bindings
are there to state how you describe the hardware.

Think of the bindings as the documentation part for how someone would
write the device-tree nodes for your sensor driver to use it on their
custom platform.

It should be something like: 
  Documentation/devicetree/bindings/media/i2c/ovti,ov772x.yaml
for example.

As I understand it, any driver which adds a .compatible string must
provide bindings.

--
Kieran




[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]

  Powered by Linux