GitLab

See gstreamer/gstreamer!590

Part-of: <gstreamer/gst-plugins-bad!1498>

Part-of: <gstreamer/gst-plugins-bad!1462>

Part-of: <gstreamer/gst-plugins-bad!1491>

MJPEG Tools may reencode pictures in a second pass to stick
closer to the target bitrate. This can result in slower than
real-time encoding for full HD content in certain situations,
as entire GOPs need reencoding when the reference picture is
reencoded.

See https://sourceforge.net/p/mjpeg/bugs/141/ for background

Part-of: <!1491>

Part-of: <gstreamer/gst-plugins-bad!1491>

Part-of: <gstreamer/gst-plugins-bad!1491>

Instead of storing the pad and then only using it to get the
element.

Part-of: <gstreamer/gst-plugins-bad!1491>

Part-of: <gstreamer/gst-plugins-bad!1491>

Part-of: <gstreamer/gst-plugins-bad!1491>

Gets rid of lots of code paths that no one has built,
used or tested for ages, and makes code more maintainable.

Part-of: <gstreamer/gst-plugins-bad!1491>

https://bugzilla.gnome.org/show_bug.cgi?id=685414

Part-of: <gstreamer/gst-plugins-bad!1491>

The PAL/NTSC widescreen modes were added after 1.16 but inserted before
the HD modes, which changed the integer values of the enums.

Move them to the very end instead to keep backwards compatibility.

Fixes gstreamer/gst-plugins-bad#1048

Part-of: <gstreamer/gst-plugins-bad!1492>

If an address can't be parsed as IP address, try resolving it via
GResolver instead. SRT URIs more often than not contain hostnames and
without trying to resolve them we won't be able to handle such URIs.

Part-of: <gstreamer/gst-plugins-bad!1493>

Part-of: <gstreamer/gst-plugins-bad!1497>

opencv plugin is pulling a header which makses clang++ 10
complain a lot and blocks -werror.

```
/usr/include/opencv4/opencv2/flann/logger.h:83:36: error: format string is not a string literal [-Werror,-Wformat-nonliteral]
        int ret = vfprintf(stream, fmt, arglist);
                                   ^~~
```

Part-of: <gstreamer/gst-plugins-bad!1494>

Part-of: <!1495>

Part-of: <!1487>

clang 10 is complaining about incompatible types due to the
glib typesystem.

```
gst-plugins-bad/gst/autoconvert/b5c3019@@gstautoconvert@sha/gstautoconvert.c.o' -c ../subprojects/gst-plugins-bad/gst/autoconvert/gstautoconvert.c
../subprojects/gst-plugins-bad/gst/autoconvert/gstautoconvert.c:898:8: error: incompatible pointer types passing 'typeof ((((void *)0))) *' (aka 'void **') to parameter of type 'GList **' (aka 'struct _GList **') [-Werror,-Wincompatible-pointer-types]
  if (!g_atomic_pointer_compare_and_exchange (&autoconvert->factories, NULL,
       ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/usr/include/glib-2.0/glib/gatomic.h:192:44: note: expanded from macro 'g_atomic_pointer_compare_and_exchange'
    __atomic_compare_exchange_n ((atomic), &gapcae_oldval, (newval), FALSE, __ATOMIC_SEQ_CST, __ATOMIC_SEQ_CST) ? TRUE : FALSE; \
                                           ^~~~~~~~~~~~~~
1 error generated.
```

Part-of: <!1487>

Part-of: <!1487>

Fixes cerbero#283

Part-of: <!1470>

On Windows, MinGW-GCC Objective-C/C++ compilers can be in PATH and
mess up the build since they may not match the CPU family of the C/C++
compilers we are using.

Also require them on macOS/iOS, because they should always be present.

Fixes gstreamer/gst-build#88

Part-of: <gstreamer/gst-plugins-bad!1488>

Part-of: <gstreamer/gst-plugins-bad!1489>

display paramater can be NULL, but if it's not, it has to be a
GstVaDisplay.

Part-of: <gstreamer/gst-plugins-bad!1489>

Part-of: <!1489>

Part-of: <gstreamer/gst-plugins-bad!1477>

This makes `gst_srt_object_validate_parameters` work properly since
`localaddress` and `localport` will be missing if the URL did not
provide them.

Part-of: <gstreamer/gst-plugins-bad!1477>

This fixes `gst_srt_object_set_string_value` in particular because the
value might not be a static string.

Part-of: <gstreamer/gst-plugins-bad!1477>

Part-of: <gstreamer/gst-plugins-bad!1477>

If the user/pass aren't escaped, the userinfo will be ambiguous and we
won't know where to split. We will accidentally get it right if the :
belongs in the password.

Part-of: <gstreamer/gst-plugins-bad!1481>

While parsing the string, `gst_uri_from_string()` also unescapes the
userinfo. This is bad if your username contains a `:` character, since
we will then split the userinfo at the wrong location when parsing it.

To fix this, we can use the new `gst_uri_from_string_escaped()` API
that was added in
gstreamer/gstreamer!583

Fixes gstreamer/gst-plugins-bad#831

Part-of: <gstreamer/gst-plugins-bad!1481>

This a GTK+ example will share, through GstContext, a custom X11
VADisplay to a pipeline using vah264dec and appsink.

When the frames are processed for rendering, the VASurfaceID is
fetched from the buffer and it is rendered using vaPutSurface in a X11
widget.

This flag will return the VASurface value at mapping

Add all the machinery to instanciate VA display through GstContext,
thus all va elements can share the same display and the user can set
a custom one.

Otherwise the VASurfaceID is lost.

It it's not copied both cdata and klass, the string is lost. Thus
also it's freed from cdata when freeing it.

And that changes function's namespace

This is a bug in Gallium RadeonSI driver for Polaris10, which doesn't
report sufrace formats for reported chroma.

If one chroma doesn't report surface formats, skip the generated caps.

Part-of: <gstreamer/gst-plugins-bad!1483>

For the allocator to create surfaces with the correct chroma an
fourcc, it should use a surface format, not necessarily the negotiated
format.

Instead of the previous arbitrary extra formats list, the decoder
extracts the valid pixel formats from the given VA config, and pass
that list to the allocator which stores it (full transfer).

Then, when the allocator allocates a new surface, it looks for a
surface color format chroma-compatible with the negotiated image color
format.

Part-of: <gstreamer/gst-plugins-bad!1483>