I thought you were not making sense for a bit. Then I walked away,
came back, and I think you're making sense now. So:
* Crop always refers to the source image
* Compose always refers to the destination image
On Sat, Oct 12, 2013 at 1:00 AM, Hans Verkuil <hverkuil@xxxxxxxxx> wrote:
> On 10/12/2013 01:48 AM, John Sheu wrote:
>> On Wed, Oct 9, 2013 at 11:49 PM, Hans Verkuil <hverkuil@xxxxxxxxx> wrote:
>> In all cases, the crop boundary refers to the area in the source
>> image; for a CAPTURE device, this is the (presumably analog) sensor,
>
> Correct.
>
>> and for an OUTPUT device, this is the memory buffer.
>
> Correct.
Here you are referring to the crop boundary, which is _not_ always
what {G,S}_CROP refers to. (Confusing part). {G,S}_CROP refers to
the crop boundary only for a CAPTURE queue.
>> My particular
>> case is a memory-to-memory device, with both CAPTURE and OUTPUT
>> queues. In this case, {G,S}_CROP on either the CAPTURE or OUTPUT
>> queues should effect exactly the same operation: cropping on the
>> source image, i.e. whatever image buffer I'm providing to the OUTPUT
>> queue.
>
> Incorrect.
>
> S_CROP on an OUTPUT queue does the inverse: it refers to the area in
> the sink image.
This confused me for a bit (seeming contradiction with the above),
until I realized that you're referring to the S_CROP ioctl here, which
is _not_ the "crop boundary"; on an OUTPUT queue it refers to the
compose boundary.
> No, it adds the compose operation for capture and the crop operation for
> output, and it uses the terms 'cropping' and 'composing' correctly
> without the inversion that S_CROP introduced on the output side.
>
> Bottom line: S_CROP for capture is equivalent to S_SELECTION(V4L2_SEL_TGT_CROP).
> S_CROP for output is equivalent to S_SELECTION(V4L2_SEL_TGT_COMPOSE).
So yes. By adding the {G,S}_SELECTION ioctls we can now refer to the
compose boundary for CAPTURE, and crop boundary for OUTPUT.
Now, here's a question. It seems that for a mem2mem device, since
{G,S}_CROP on the CAPTURE queue covers the crop boundary, and
{G,S}_CROP on the OUTPUT queue capture the compose boundary, is there
any missing functionality that {G,S}_SELECTION is covering here. In
other words: for a mem2mem device, the crop and compose boundaries
should be identical for the CAPTURE and OUTPUT queues?
-John Sheu
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
