Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-25 Thread Nicolas George
Anton Khirnov (12021-01-23): > But does anyone use anything else than ffmpeg.c for that? > > The question here is not whether this functionality should exist at all, > but whether libavdevice/libavformat is the right place for it. If something is useful in ffmpeg.c, then it is likely to be useful

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-23 Thread Anton Khirnov
Quoting Marton Balint (2021-01-20 00:40:30) > > > On Tue, 19 Jan 2021, Anton Khirnov wrote: > > > Quoting Mark Thompson (2021-01-19 00:37:09) > >> On 16/01/2021 22:12, Nicolas Caramelli wrote: > >> > This patch adds KMS/DRM output device for rendering a video stream > >> > using KMS/DRM dumb buf

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-20 Thread Nicolas George
Mark Thompson (12021-01-20): > Because the libavformat API is not suitable for (especially) these > use-cases, so including them in the library feels like a hack that > happened to be convenient for some subset of people even though it > doesn't really fit and wants something else. Again, I have u

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-20 Thread Mark Thompson
On 20/01/2021 11:53, Nicolas George wrote: Mark Thompson (12021-01-19): Can you explain in more detail your use of libavdevice here? My general sense is that while they might sometimes be convenient to use if you are already working in a libav* context, they are so limited (by both the API and

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-20 Thread Nicolas George
Mark Thompson (12021-01-19): > Can you explain in more detail your use of libavdevice here? > > My general sense is that while they might sometimes be convenient to > use if you are already working in a libav* context, they are so > limited (by both the API and the implementation) that most users

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-20 Thread Nicolas George
James Almer (12021-01-19): > On 1/19/2021 4:04 PM, Nicolas George wrote: > > Anton Khirnov (12021-01-19): > > > > (Note that I would argue against adding more general display output > > > > devices which are already present, like fb and xv, because they are > > > > of essentially no value to libavd

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-19 Thread Marton Balint
On Tue, 19 Jan 2021, Anton Khirnov wrote: Quoting Mark Thompson (2021-01-19 00:37:09) On 16/01/2021 22:12, Nicolas Caramelli wrote: > This patch adds KMS/DRM output device for rendering a video stream > using KMS/DRM dumb buffer. > The proposed implementation is very basic, only bgr0 pixel fo

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-19 Thread Mark Thompson
On 19/01/2021 19:04, Nicolas George wrote: Anton Khirnov (12021-01-19): (Note that I would argue against adding more general display output devices which are already present, like fb and xv, because they are of essentially no value to libavdevice users. Removing legacy code is harder, though.)

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-19 Thread James Almer
On 1/19/2021 4:04 PM, Nicolas George wrote: Anton Khirnov (12021-01-19): (Note that I would argue against adding more general display output devices which are already present, like fb and xv, because they are of essentially no value to libavdevice users. Removing legacy code is harder, though.)

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-19 Thread Nicolas George
Anton Khirnov (12021-01-19): > > (Note that I would argue against adding more general display output > > devices which are already present, like fb and xv, because they are > > of essentially no value to libavdevice users. Removing legacy code > > is harder, though.) > > +1 > > IMO libavdevice o

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-19 Thread Nicolas Caramelli
I appreciate to directly experiment ffmpeg without X11 or SDL dependencies, just using the ffmpeg binary. This is especially true when I work on embedded systems and I use it a lot. Even though the fbdev output device is available, the Linux Framebuffer is often considered obsolete in favor of KMS

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-19 Thread John Cox
On Mon, 18 Jan 2021 23:37:09 +, you wrote: >On 16/01/2021 22:12, Nicolas Caramelli wrote: >> This patch adds KMS/DRM output device for rendering a video stream >> using KMS/DRM dumb buffer. >> The proposed implementation is very basic, only bgr0 pixel format is >> currently supported (the most

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-19 Thread Anton Khirnov
Quoting Mark Thompson (2021-01-19 00:37:09) > On 16/01/2021 22:12, Nicolas Caramelli wrote: > > This patch adds KMS/DRM output device for rendering a video stream > > using KMS/DRM dumb buffer. > > The proposed implementation is very basic, only bgr0 pixel format is > > currently supported (the mos

Re: [FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-18 Thread Mark Thompson
On 16/01/2021 22:12, Nicolas Caramelli wrote: This patch adds KMS/DRM output device for rendering a video stream using KMS/DRM dumb buffer. The proposed implementation is very basic, only bgr0 pixel format is currently supported (the most common format with KMS/DRM). To enable this output device

[FFmpeg-devel] [PATCH] libavdevice: Add KMS/DRM output device

2021-01-16 Thread Nicolas Caramelli
This patch adds KMS/DRM output device for rendering a video stream using KMS/DRM dumb buffer. The proposed implementation is very basic, only bgr0 pixel format is currently supported (the most common format with KMS/DRM). To enable this output device you need to configure FFmpeg with --enable-libdr