This ensures that the synchro object is allocated regardless of the
video output state.

Note that they are probably other ways to trigger similar NULL
dereference. It is highly questionable that this plugin should remain
in the tree with all its bugs and no upstream (libmpeg2) maintainance.

This set of "helpers" is only used by the libmpeg2 plugin, and was not
exposed to out-of-tree-plugins. There is apparently no point keeping it
inside the core.

This option had been removed from the filter module in 2013, see
4b4f62a0.

Fixes #15418

Filip Roséen authored 7 years ago and Jean-Baptiste Kempf committed 7 years ago

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Filip Roséen authored 7 years ago and Jean-Baptiste Kempf committed 7 years ago

As extensions, such as .001, are heavily coupled with access/concat,
this allows a certain pattern to force a specific access module to
handle the incoming data.

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Filip Roséen authored 7 years ago and Jean-Baptiste Kempf committed 7 years ago

This allows for easier implementation for modules using the
concat-list, as it can assume that all entries can be passed directly
to functions such as vlc_stream_NewURL (as they are now valid URIs).

These changes also makes it so that the contents of concat-list
matches that of the documentation, as it uses the notion "input URLs".

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

Filip Roséen authored 7 years ago and Jean-Baptiste Kempf committed 7 years ago

These changes allow for an implementation that is a little bit easier
to read, and maintain.

Signed-off-by: Jean-Baptiste Kempf <jb@videolan.org>

End of preroll time was not updated properly.

Initial discontinuities were notified
through es_out/input_DecoderDiscontinuity
79cb45a4
using a forged empty block with discontinuity
flag.

Current decoders/demuxers discontinuity flag
is set on first block after discontinuity, and
is no longer empty data. We need to test for
this.

refs #18245
First spu dropped on preroll (preroll > INT_MIN)
(due to another special case/preroll incompatibility)

The conversion from UTF-16 to UTF-8 could (until a few days ago) fail
before the end. In that case, the appended nul terminator would not be
converted. A nul terminator is anyway always appended after conversion
and trimming.

If conversion failed on the first character, that nul terminator would
be written at a negative offset, leading to heap buffer "underflow" and
memory corruption. This was fixed but lead to a mismatch in the value of
i_lines depending on the character width.

This change removes the useless pre-conversion nul terminator, and thus
makes trimming work again with single byte character width work. This
fixes reading text files formatted with MS-DOS line endings.

Since deinterlace filters are loaded manually.

Must not depend of ttml style presence

Removal of CR or LF final characters relied on the off-by-one bug fixed
in the previous changest, and no longer works. This adjusts it
accordingly.

Do not erase the last converted byte. This bug has apparently existed
ever since UTF-16 support was added.

If the conversion fails, this bug resulted in a heap underflow (writing
zero right before the beginning of the buffer).

xib compilation is changed to be not triggered if macosx module
is not built.

This reverts commit c2e45c4e.

Postproc is almost always useless nowadays and does not work in most
cases.

no comment

Filip Roséen authored 7 years ago and Rémi Denis-Courmont committed 7 years ago

Signed-off-by: Rémi Denis-Courmont <remi@remlab.net>