From: Hans Verkuil <hans.verkuil@xxxxxxxxx> There is a lot of confusion about the correct buffer type to use when calling the new selection and old crop APIs. Specifically whether the _MPLANE variant of a buf type should be used or not if the device is multi-planar. The spec said na, but that was unexpected to applications and drivers actually did different things as well. This patch series allows both to be used and updates the documentation accordingly. In the end, these APIs don't care whether it is a single or multiplanar device, that information is irrelevant to these ioctls. So allowing both is not unreasonable, especially given the mess we created. The first patch is unchanged from the original RFC here: https://patchwork.linuxtv.org/patch/41210/ The second patch was updated from this original RFC: - the note was moved after the struct containing the 'type' field. - kernel 4.12 was replaced with 4.14 (I'm assuming this will be too late for 4.13). - The phrase 'The Samsung Exynos drivers' was replaced by 'Some drivers'. Regards, Hans Hans Verkuil (2): v4l2-ioctl/exynos: fix G/S_SELECTION's type handling media/uapi/v4l: clarify cropcap/crop/selection behavior Documentation/media/uapi/v4l/vidioc-cropcap.rst | 23 ++++++---- Documentation/media/uapi/v4l/vidioc-g-crop.rst | 22 +++++---- .../media/uapi/v4l/vidioc-g-selection.rst | 22 +++++---- drivers/media/platform/exynos-gsc/gsc-core.c | 4 +- drivers/media/platform/exynos-gsc/gsc-m2m.c | 8 ++-- drivers/media/platform/exynos4-is/fimc-capture.c | 4 +- drivers/media/platform/exynos4-is/fimc-lite.c | 4 +- drivers/media/v4l2-core/v4l2-ioctl.c | 53 +++++++++++++++++++--- 8 files changed, 95 insertions(+), 45 deletions(-) -- 2.11.0
