Re: [PATCH 2/9 v8] V4L: add two new ioctl()s for multi-size videobuffer management

[Guennadi Liakhovetski <g.liakhovetski@gmx.de>]

On Wed, 28 Sep 2011, Guennadi Liakhovetski wrote:

> Hi Hans
> 
> On Wed, 28 Sep 2011, Hans Verkuil wrote:
> 
> > On Tuesday, September 27, 2011 18:54:53 Guennadi Liakhovetski wrote:
> > > A possibility to preallocate and initialise buffers of different sizes
> > > in V4L2 is required for an efficient implementation of a snapshot
> > > mode. This patch adds two new ioctl()s: VIDIOC_CREATE_BUFS and
> > > VIDIOC_PREPARE_BUF and defines respective data structures.
> > > 
> > > Signed-off-by: Guennadi Liakhovetski <g.liakhovetski@gmx.de>
> > > ---
> > > 
> > > v8: addressed comments from Hans - thanks:
> > > 
> > >     1. added checks in ioctl() preprocessing
> > >     2. changed VIDIOC_PREPARE_BUF to _IOWR
> > > 
> > >  drivers/media/video/v4l2-compat-ioctl32.c |   67 +++++++++++++++++++++++++---
> > >  drivers/media/video/v4l2-ioctl.c          |   36 +++++++++++++++
> > >  include/linux/videodev2.h                 |   17 +++++++
> > >  include/media/v4l2-ioctl.h                |    2 +
> > >  4 files changed, 114 insertions(+), 8 deletions(-)
> > > 
> > 
> > Almost:
> > 
> > > diff --git a/include/linux/videodev2.h b/include/linux/videodev2.h
> > > index 9d14523..7d75dd1 100644
> > > --- a/include/linux/videodev2.h
> > > +++ b/include/linux/videodev2.h
> > > @@ -653,6 +653,9 @@ struct v4l2_buffer {
> > >  #define V4L2_BUF_FLAG_ERROR	0x0040
> > >  #define V4L2_BUF_FLAG_TIMECODE	0x0100	/* timecode field is valid */
> > >  #define V4L2_BUF_FLAG_INPUT     0x0200  /* input field is valid */
> > > +/* Cache handling flags */
> > > +#define V4L2_BUF_FLAG_NO_CACHE_INVALIDATE	0x0400
> > > +#define V4L2_BUF_FLAG_NO_CACHE_CLEAN		0x0800
> > >  
> > >  /*
> > >   *	O V E R L A Y   P R E V I E W
> > > @@ -2099,6 +2102,15 @@ struct v4l2_dbg_chip_ident {
> > >  	__u32 revision;    /* chip revision, chip specific */
> > >  } __attribute__ ((packed));
> > >  
> > > +/* VIDIOC_CREATE_BUFS */
> > > +struct v4l2_create_buffers {
> > > +	__u32			index;		/* output: buffers index...index + count - 1 have been created */
> > > +	__u32			count;
> > > +	enum v4l2_memory        memory;
> > > +	struct v4l2_format	format;		/* "type" is used always, the rest if sizeimage == 0 */
> > > +	__u32			reserved[8];
> > > +};
> > > +
> > >  /*
> > >   *	I O C T L   C O D E S   F O R   V I D E O   D E V I C E S
> > >   *
> > > @@ -2189,6 +2201,11 @@ struct v4l2_dbg_chip_ident {
> > >  #define	VIDIOC_SUBSCRIBE_EVENT	 _IOW('V', 90, struct v4l2_event_subscription)
> > >  #define	VIDIOC_UNSUBSCRIBE_EVENT _IOW('V', 91, struct v4l2_event_subscription)
> > >  
> > > +/* Experimental, the below two ioctls may change over the next couple of kernel
> > > +   versions */
> > > +#define VIDIOC_CREATE_BUFS	_IOWR('V', 92, struct v4l2_create_buffers)
> > > +#define VIDIOC_PREPARE_BUF	_IOWR('V', 93, struct v4l2_buffer)
> > > +
> > >  /* Reminder: when adding new ioctls please add support for them to
> > >     drivers/media/video/v4l2-compat-ioctl32.c as well! */
> > >  
> > > diff --git a/include/media/v4l2-ioctl.h b/include/media/v4l2-ioctl.h
> > > index dd9f1e7..55cf8ae 100644
> > > --- a/include/media/v4l2-ioctl.h
> > > +++ b/include/media/v4l2-ioctl.h
> > > @@ -122,6 +122,8 @@ struct v4l2_ioctl_ops {
> > >  	int (*vidioc_qbuf)    (struct file *file, void *fh, struct v4l2_buffer *b);
> > >  	int (*vidioc_dqbuf)   (struct file *file, void *fh, struct v4l2_buffer *b);
> > >  
> > > +	int (*vidioc_create_bufs)(struct file *file, void *fh, struct v4l2_create_buffers *b);
> > > +	int (*vidioc_prepare_buf)(struct file *file, void *fh, const struct v4l2_buffer *b);
> > 
> > If this is IOWR, then there shouldn't be a const here.
> 
> hrm... Sure.
> 
> > I have been thinking about this a bit more. Currently we only have a V4L2_BUF_FLAG_QUEUED
> > flag and no V4L2_BUF_FLAG_PREPARED flag. I do think we need this after all. The QUEUED flag
> > can't be used here as the buffer isn't queued yet, it's only prepared.
> 
> Ok, I can add it to this patch together with the other two cache-handling 
> flags. I presume, I shall also add it to
> 
> V4L2_BUFFER_STATE_FLAGS
> 
> in videobuf2-core.c, otherwise just let drivers and apps go figure?... We 
> could set that flag centrally in __buf_prepare(), but since 
> V4L2_BUF_FLAG_QUEUED is managed by individual drivers, we probably want to 
> leave V4L2_BUF_FLAG_PREPARED like that too?

Good, time to wake up. V4L2_BUF_FLAG_QUEUED is indeed set in 
videobuf2-core.c, so, I guess, the same should be done with 
V4L2_BUF_FLAG_PREPARED in __fill_v4l2_buffer().

Thanks
Guennadi

> > Regards,
> > 
> > 	Hans
> > 
> > >  
> > >  	int (*vidioc_overlay) (struct file *file, void *fh, unsigned int i);
> > >  	int (*vidioc_g_fbuf)   (struct file *file, void *fh,
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/