Hi Amit,
repeating my private reply.
Le vendredi 19 février 2021 à 03:04 -0600, Amit Pandya a écrit :
> Hi Nicolas,
>
> 1) Do reverse playback seek is supported with gst-v4l2 decoders and ever
> verified on any of the supported hw platforms.
>
> If not supported with gst-v4l2 decoders currently then let us know if you
> have any plans to support it with gst-v4l2 decoders?
As far as I remember it didn't work. Some work is needed, though your may ran
into memory size limitation as you need to allocate a full GOP worth of frames
(which requires dynamic allocation/vidioc_create_bufs).
>
> NOTE: Tried to check reverse playback with gst-v4l2 decoder on Exynos but
> found not working.
>
> 2) Also, with earlier communication on gst-devel thread titled "Gstreamer
> V4l2 Codec elements for video encoders/decoders" you informed that currently
> the most tested platforms (non scientific) for gst-v4l2 decode/encode
> implementation would be:
> - IMX.6/8 CODA driver
> - Qualcomn Venus driver (dynamic resolution is broken)
This is being actively worked on and being fix. It would look like a major
product using the upstream driver is coming, though with Chromium stack, but has
the same requirement as GStreamer.
https://www.qualcomm.com/products/mobile-computing/chromebooks> - Exynos (some stability issue, which seems to be in IOMMU driver)
>
> Do gst-v4l2 decode/encode implementation now been verified on any other hw
> platforms(specifically for 64-bit) other than the listed above?
Decode is used on many SoC with quite some success, though as of today, to
automated / regressions testing is being done. Perhaps the Khodi folks runs the
FFMPEG Fate suite from time to time (or at least against their patched kernel in
libreelecTV project).
_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel
Locked
