Signed-off-by: Martin Storsjö <martin@martin.st>

Signed-off-by: Martin Storsjö <martin@martin.st>

Fixes MSVR-11-0088.
Credit:  Jeong Wook Oh of Microsoft and Microsoft Vulnerability Research (MSVR)

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Martin Storsjö <martin@martin.st>

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Signed-off-by: Anton Khirnov <anton@khirnov.net>

They will only ever be <=0 if the option is broken, in which case this
check is hiding a bug.

It's a hack which was created to allow for multiple options with
different defaults to refer to same field (e.g. 'b' vs 'ab'). There is
no need for it anymore.

Also change it to be more readable.

Specifically, trellis, nr, me_range, me_method, subq, b_strategy,
keyint_min, chroma_me and coder.
Change their defaults to -1 and apply them after
x264_param_default_preset() if the user explicitly set them.

Move setting it after x264_param_default_preset()

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Deprecate AVCodecContext.directpred

Deprecate AVCodecContext.partitions.

Deprecate AVCodecContext.complexityblur

Deprecate AVCodecContext.deblockalpha/deblockbeta

Deprecate AVCodecContext.bframebias.

Specifically: gop_size, max_b_frames, scenechange_threshold, qmin, qmax,
max_qdiff, qblur, qcompress and refs.

Change their default values to -1 and only use them if the user
explicitly set them. Otherwise x264 defaults are used.

Move setting those options after x264_param_default_preset(), so they
don't get overwritten by it.

priv_data is memset to 0.

Since we now always allocate it, it can simply be made part of the context
instead.

Update FATE references accordingly.

scale_coefficients() rather than in apply_channel_coupling()

The number of rematrixing bands still needs to be calculated.

This function should return number of samples decoded, not number of bytes
decoded.
Spotted by Uoti Urpala.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

After 039267f192f335144d9251e6f05a1513dd1e8ef1, metadata specifiers are
passed without the leading ':'. Remove a ++ that didn't take this into
account.

Signed-off-by: Anton Khirnov <anton@khirnov.net>

This fixes improper flushing in the cases when the same frame is decoded in
several iterations (for being too large to fit into output buffer) and flush is
called mid-decoding and it also resets context in case of decoding errors.

Signed-off-by: Anton Khirnov <anton@khirnov.net>

Change syntax -- -[vas]bsf are replaced by -bsf:[stream specifier], the
argument is now a comma-separated list of bitstream filters.

Make it work properly with multiple input files.

Add a -frames option that uses generic stream specifiers, change
-[vad]frames into aliases to it.

It is now possible to set metadata on streams and chapters.

It should be flags, not int.

Move stream maps to it.

Eventually all non-global options should be stored in it.

This allows options like -target, which are just shortcuts for other
options, to work without dummy function for all options they invoke.

This is the first step to removing the globals plague from avtools.

This will be useful in the following commit.

Allows cmdutils to call each tool's own cleanup function.

The initial request contains "Range: 0-", which servers normally
have responded with "HTTP/1.1 206 Partial Content" reply with
a Content-Range header, which was used as indicator for seekability.

Apache, since 2.2.20, responds with "HTTP/1.1 200 OK" for these
requests, which is more friendly to caches and proxies, but the
seekability still is indicated via the Accept-Ranges: bytes header.

Signed-off-by: Martin Storsjö <martin@martin.st>