Re: [PATCH v3 0/9] Media Controller capture driver for DM365

From: Hans Verkuil
Date: Thu Nov 29 2012 - 02:45:02 EST


On Wed November 28 2012 20:30:21 Greg Kroah-Hartman wrote:
> On Wed, Nov 28, 2012 at 08:18:20PM +0100, Hans Verkuil wrote:
> > On Wed November 28 2012 18:22:48 Greg Kroah-Hartman wrote:
> > > On Wed, Nov 28, 2012 at 10:18:02AM -0200, Mauro Carvalho Chehab wrote:
> > > > Em Wed, 28 Nov 2012 12:56:10 +0100
> > > > Hans Verkuil <hansverk@xxxxxxxxx> escreveu:
> > > >
> > > > > On Wed 28 November 2012 12:45:37 Dan Carpenter wrote:
> > > > > > I wish people wouldn't submit big patches right before the merge
> > > > > > window opens... :/ It's better to let it sit in linux-next for a
> > > > > > couple weeks so people can mess with it a bit.
> > > > >
> > > > > It's been under review for quite some time now, and the main change since
> > > > > the last posted version is that this is now moved to staging/media.
> > > > >
> > > > > So it is not yet ready for prime time, but we do want it in to simplify
> > > > > the last remaining improvements needed to move it to drivers/media.
> > > >
> > > > "last remaining improvements"? I didn't review the patchset, but
> > > > the TODO list seems to have several pending stuff there:
> > > >
> > > > +- User space interface refinement
> > > > + - Controls should be used when possible rather than private ioctl
> > > > + - No enums should be used
> > > > + - Use of MC and V4L2 subdev APIs when applicable
> > > > + - Single interface header might suffice
> > > > + - Current interface forces to configure everything at once
> > > > +- Get rid of the dm365_ipipe_hw.[ch] layer
> > > > +- Active external sub-devices defined by link configuration; no strcmp
> > > > + needed
> > > > +- More generic platform data (i2c adapters)
> > > > +- The driver should have no knowledge of possible external subdevs; see
> > > > + struct vpfe_subdev_id
> > > > +- Some of the hardware control should be refactorede
> > > > +- Check proper serialisation (through mutexes and spinlocks)
> > > > +- Names that are visible in kernel global namespace should have a common
> > > > + prefix (or a few)
> > > >
> > > > From the above comments, both Kernelspace and Userspace APIs require
> > > > lots of work.
> >
> > And that's why it is in staging. Should a long TODO list now suddenly
> > prevent staging from being used? In Barcelona we discussed this and the
> > only requirement we came up was was that it should compile.
>
> Yes, that's all I care about in staging, but as I stated, I don't
> maintain drivers/staging/media/ that area is under Mauro's control
> (MAINTAINERS even says this), and I'm a bit leery of going against the
> wishes of an existing subsystem maintainer for adding staging drivers
> that tie into their subsystem.
>
> So if you get Mauro's approval, I'll be glad to queue it up.

Just for the record, my comments were addressed to Mauro, not to you.
I should have stated that in my mail explicitly. As you said, it's
Mauro's decision.

Regards,

Hans
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/