Based on code by Janne Grunau
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

prevents regressions from h264-mt frame size changing code
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

avoids regression from the h264-mt frame size change code.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Based on a patch by Janne Grunau
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Based on patch by Janne Grunau
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Signed-off-by: Paul B Mahol <onemda@gmail.com>

* commit 'f1d8763a':
  mpegvideo: allocate scratch buffers after linesize is known

Conflicts:
	libavcodec/mpegvideo.c
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit 'ed2d7d58':
  ff_h264_direct_ref_list_init: fix B slice check.
  h264: increase dist_scale_factor for up to 32 references
Merged-by: Michael Niedermayer <michaelni@gmx.at>

* commit '523c7bd2':
  misc typo, style and wording fixes

Conflicts:
	libavcodec/options_table.h
	libavutil/pixfmt.h
Merged-by: Michael Niedermayer <michaelni@gmx.at>

It's confusing and not necessary, especially in the introduction.

In particular, make "See Also" chapter rendered in HTML output.

In particular, make "See Also" chapter rendered in HTML output.

Add more references, and generate alternative variants for HTML and POD
output. In particular, this allows to render a "See Also" chapter in the
HTML output.

Fix syntax, remove pointless introductory sentence, improve overall
consistency.

Factorize.

Ignore @c man begin ... @c man end comments, rely on @chapter for marking
the beginning of the various manual top level sections.

This allows us to write markup which is not dependent on the specific
texi2pod.pl implementation.

This change causes a few rendering issues, which will be fixed in further
patches.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Note that examples are changed.

This breaks usage for out-of-range values.

This will prevent an overflow in the SSSE3 and MMX filter_line code:
delta is expressed as an u16 being shifted by 2 to the left. If it
overflows, having a strength not above 64 will make sure that m is set
to 0 (making the m*m*delta >> 14 expression void).

A value above 64 should not make any sense unless gradfun is used as a
blur filter.

There is no noticable benefit for such precision.

Current dithering only use the first 4w instead of the whole 8 random values.

Current code divide before increasing precision.

First DC is only used once otherwise. This also makes the code
consistent with ASM versions.

Signed-off-by: Paul B Mahol <onemda@gmail.com>

Encoders that buffer a significant number of frames where a "few"
frames off with -shortest
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Fixes ticket #1557.

Since we can't know which stride a custom get_buffer() implementation is
going to use we have to allocate this scratch buffers after the linesize
is known. It was pretty safe for 8 bit per pixel pixel formats since we
always allocated memory for up to 16 bits per pixel. It broke hoever
with cmdutis.c's alloc_buffer() and high pixel bit depth since it
allocated larger edges than mpegvideo expected.

Fixes fuzzed sample nasa-8s2.ts_s244342.

Fixes null pointer dereference.

Found-by: Mateusz "j00ru" Jurczyk and Gynvael Coldwind

Compute dist_scale_factor_field only for MBAFF since that is the only
case in which it is used.