[vlc-devel] [PATCH v3 2/3] video_output/mmal: Add mmal based video output plugin
Jean-Baptiste Kempf
jb at videolan.org
Sun Apr 13 20:04:38 CEST 2014
On 11 Apr, Julian Scheel wrote :
> Am 10.04.2014 12:28, schrieb Jean-Baptiste Kempf:
> >On 10 Apr, Julian Scheel wrote :
> >>+struct vout_display_sys_t {
> >>+ int layer;
> >>+ bool opaque;
> >>+ MMAL_COMPONENT_T *component;
> >>+ MMAL_PORT_T *input;
> >>+ MMAL_POOL_T *pool;
> >>+ unsigned num_buffers;
> >>+ uint32_t buffer_size;
> >>+ picture_t **pictures;
> >>+ picture_pool_t *picture_pool;
> >>+ plane_t planes[3];
> >>+ vlc_mutex_t buffer_mutex;
> >>+ vlc_cond_t buffer_cond;
> >>+ int buffers_in_transit;
> >>+ unsigned display_width;
> >>+ unsigned display_height;
> >>+ vlc_mutex_t manage_mutex;
> >>+ bool need_configure_display;
> >>+ DISPMANX_DISPLAY_HANDLE_T dmx_handle;
> >>+ struct dmx_region_t *dmx_region;
> >>+ DISPMANX_ELEMENT_HANDLE_T bkg_element;
> >>+ DISPMANX_RESOURCE_HANDLE_T bkg_resource;
> >>+};
> >
> >You should repack this structure
>
> What do you mean by repack this structure?
http://docs.oracle.com/cd/E19683-01/806-6543/auto20/index.html
> >You can reorder those to remove some declarations, if you want...
>
> I think it was merely a design decision by Dennis to just forward declare
> all internal functions in the order of later appearance.
> Dennis, could you add a short comment?
> If this is not desired by VLC core devs we can change it to only forward
> declare what's really needed.
I do _not_ care, but some devs prefer it.
And often, there are no choices...
> >This is valid for I420, right? Can't you get that from the fmt
> >description?
>
> That's I420 yes - actually the only raw format the vout supports as of now.
> As the number of planes is hardcoded for I420 in the whole vout I don't see
> an advantage of extracting it from the format.
OK. I just thought you would extend it, afterwards.
--
Jean-Baptiste Kempf
http://www.jbkempf.com/ - +33 672 704 734
Sent from my Electronic Device
More information about the vlc-devel
mailing list