ffmpeg.git
6 years agotiny_psnr: fix range calculation for sample size of 32 bits
Mans Rullgard [Sun, 14 Oct 2012 23:21:53 +0000 (00:21 +0100)]
tiny_psnr: fix range calculation for sample size of 32 bits

For a sample size of 32 bits, the shift would overflow producing
undefined results.  Incidentally, in the only test currently using
32-bit samples, the output matches the reference exactly on most
systems meaning the bad 'max' value is never used.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agotiny_psnr: check for specified sample size less than 1
Mans Rullgard [Sun, 14 Oct 2012 23:11:44 +0000 (00:11 +0100)]
tiny_psnr: check for specified sample size less than 1

A zero or negative sample size is impossible and should be
reported as an error.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agofate: improve md5sum utility selection
Mans Rullgard [Sun, 14 Oct 2012 14:49:25 +0000 (15:49 +0100)]
fate: improve md5sum utility selection

The 'md5sum' command is used with the -b flag so the presence test
must also use this flag.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agorangecoder-test: Drop timer output that clutters stderr
Diego Biurrun [Fri, 5 Oct 2012 16:53:12 +0000 (18:53 +0200)]
rangecoder-test: Drop timer output that clutters stderr

6 years agorangecoder-test: Return in case of an error
Diego Biurrun [Fri, 12 Oct 2012 19:14:56 +0000 (21:14 +0200)]
rangecoder-test: Return in case of an error

6 years agobuild: simplify enabling of compat objects
Mans Rullgard [Sun, 14 Oct 2012 16:56:00 +0000 (17:56 +0100)]
build: simplify enabling of compat objects

Add a configure function to pull in a compat object and set up
redirects in one operation.  This avoids duplicating conditions
across configure and makefiles.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoconfigure: simplify argument handling in check_ld
Mans Rullgard [Sun, 14 Oct 2012 00:35:56 +0000 (01:35 +0100)]
configure: simplify argument handling in check_ld

Use the existing filter functions instead of open-coding the
separation of general flags and libraries.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoconfigure: simplify get_version() function
Mans Rullgard [Sat, 13 Oct 2012 21:56:25 +0000 (22:56 +0100)]
configure: simplify get_version() function

awk alone can do this, no need for grep.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agobuild: support asan and tsan toolchain shortcuts
Luca Barbato [Sun, 14 Oct 2012 05:54:22 +0000 (07:54 +0200)]
build: support asan and tsan toolchain shortcuts

6 years agormdec: Move SIPR code shared with Matroska demuxer to a separate file
Diego Biurrun [Sun, 14 Oct 2012 21:41:12 +0000 (23:41 +0200)]
rmdec: Move SIPR code shared with Matroska demuxer to a separate file

6 years agoswscale: try to use mmap only if available
Mans Rullgard [Sat, 13 Oct 2012 16:00:16 +0000 (16:00 +0000)]
swscale: try to use mmap only if available

Some systems, e.g. Minix, have sys/mman.h defining MAP_ANONYMOUS without
providing (working) mmap and friends.  The mmx filter generation code
checks only for MAP_ANONYMOUS, not for availability of mmap itself which
leads to build errors on aforementioned systems.

This changes the conditional compilation to use mmap only if all the
required functions are available.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoconfigure: check for mprotect
Mans Rullgard [Sat, 13 Oct 2012 02:40:48 +0000 (02:40 +0000)]
configure: check for mprotect

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agowmapro: use planar sample format
Justin Ruggles [Wed, 29 Aug 2012 02:49:17 +0000 (22:49 -0400)]
wmapro: use planar sample format

6 years agowmalossless: output in planar sample format
Justin Ruggles [Sat, 25 Aug 2012 23:09:40 +0000 (19:09 -0400)]
wmalossless: output in planar sample format

6 years agowmadec: use float planar sample format output
Justin Ruggles [Wed, 29 Aug 2012 03:12:49 +0000 (23:12 -0400)]
wmadec: use float planar sample format output

6 years agoshorten: use planar sample format
Justin Ruggles [Tue, 28 Aug 2012 18:02:41 +0000 (14:02 -0400)]
shorten: use planar sample format

6 years agolavc: update documentation for AVFrame.extended_data
Justin Ruggles [Sun, 14 Oct 2012 18:31:47 +0000 (14:31 -0400)]
lavc: update documentation for AVFrame.extended_data

6 years agoavutil: Do not make ff_ symbols globally visible.
Diego Biurrun [Thu, 4 Oct 2012 17:20:03 +0000 (19:20 +0200)]
avutil: Do not make ff_ symbols globally visible.

6 years agoavutil: Rename ff_set_systematic_pal2() ---> avpriv_set_systematic_pal2()
Diego Biurrun [Thu, 4 Oct 2012 19:15:51 +0000 (21:15 +0200)]
avutil: Rename ff_set_systematic_pal2() ---> avpriv_set_systematic_pal2()

Functions used across libraries should have avpriv_ and not ff_ prefixes.

6 years agobuild: tms470: work around glibc math.h problems
Mans Rullgard [Sat, 13 Oct 2012 02:16:13 +0000 (03:16 +0100)]
build: tms470: work around glibc math.h problems

The glibc definitions of INFINITY and NAN do not work with the
tms470 compiler, nor do our usual fallbacks.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoconfigure: improve tms470 compiler usage with glibc
Mans Rullgard [Sat, 13 Oct 2012 01:28:41 +0000 (02:28 +0100)]
configure: improve tms470 compiler usage with glibc

Apply flags to work around glibc quirks only if glibc is detected,
and add a few more such flags.

Do not mess with as/ld settings in probe_cc.  This is not the
proper place.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoconfigure: tms470: add mapping for -mfpu=vfpv3-d16 flag
Mans Rullgard [Sat, 13 Oct 2012 00:34:42 +0000 (01:34 +0100)]
configure: tms470: add mapping for -mfpu=vfpv3-d16 flag

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoconfigure: recognise Minix as OS
Mans Rullgard [Fri, 12 Oct 2012 21:42:27 +0000 (21:42 +0000)]
configure: recognise Minix as OS

No special setup is required for Minix.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoconfigure: work around bug in ash shell
Mans Rullgard [Fri, 12 Oct 2012 21:40:07 +0000 (21:40 +0000)]
configure: work around bug in ash shell

The ash 'test' builtin misbehaves if the first operand of a binary
operator looks like a unary operator.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoeval-test: make table static const
Mans Rullgard [Fri, 12 Oct 2012 18:35:11 +0000 (19:35 +0100)]
eval-test: make table static const

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agolavr: handle clipping in the float to s32 conversion
Justin Ruggles [Fri, 12 Oct 2012 19:46:53 +0000 (15:46 -0400)]
lavr: handle clipping in the float to s32 conversion

We cannot clip to INT_MAX because that value cannot be exactly
represented by a float value and ends up overflowing during conversion
anyway. We need to use a slightly smaller float value, which ends up
with slightly inaccurate results for samples which clip or nearly clip,
but it is close enough. Using doubles as intermediates in the conversion
would be more accurate, but it takes about twice as much time.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>
6 years agonut: support pcm codecs not mapped in avi
Luca Barbato [Fri, 12 Oct 2012 15:29:43 +0000 (17:29 +0200)]
nut: support pcm codecs not mapped in avi

The native tags will be used when available.

6 years agonut: prioritize native tags
Luca Barbato [Fri, 12 Oct 2012 13:49:36 +0000 (15:49 +0200)]
nut: prioritize native tags

Use native tags instead of avi ones, simplifies a lot raw video codecs
handling.

6 years agovc1: Use codec ID from AVCodecContext while parsing frame header
Mashiat Sarker Shakkhar [Thu, 11 Oct 2012 17:13:56 +0000 (13:13 -0400)]
vc1: Use codec ID from AVCodecContext while parsing frame header

This fixes a segfault with samples that I have (both of them MPEG-TS). Looks like
avctx->codec is not being set during parsing.

Signed-off-by: Kostya Shishkov <kostya.shishkov@gmail.com>
6 years agoavplay: support mid-stream sample rate changes
Justin Ruggles [Sun, 30 Sep 2012 22:18:00 +0000 (18:18 -0400)]
avplay: support mid-stream sample rate changes

Resample to the rate that was configured in SDL.

6 years agoavplay: use audio parameters from the decoded frame instead of AVCodecContext
Justin Ruggles [Mon, 8 Oct 2012 03:55:32 +0000 (23:55 -0400)]
avplay: use audio parameters from the decoded frame instead of AVCodecContext

6 years agodca: allocate a secondary buffer for extra channels when downmixing
Justin Ruggles [Thu, 11 Oct 2012 18:21:56 +0000 (14:21 -0400)]
dca: allocate a secondary buffer for extra channels when downmixing

The output AVFrame buffer only has data for the downmix channels.
Fixes a segfault when decoding dca with request_channels == 2.

6 years agoconfigure: use utilities from /usr/xpg4/bin if it exists
Mans Rullgard [Fri, 7 Sep 2012 11:50:43 +0000 (12:50 +0100)]
configure: use utilities from /usr/xpg4/bin if it exists

Solaris defaults to non-standard utilities (grep, sed, ...) with
proper ones being in /usr/xpg4/bin.  Prefixing PATH with this
directory when it exists ensures we get correct variants.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoavstring-test: fix memory leaks
Mans Rullgard [Fri, 12 Oct 2012 21:20:09 +0000 (22:20 +0100)]
avstring-test: fix memory leaks

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agortpdec_xiph: fix function return type
Mans Rullgard [Thu, 11 Oct 2012 20:59:12 +0000 (21:59 +0100)]
rtpdec_xiph: fix function return type

parse_packed_headers() returns either zero or a negative error code
so its return type must be signed.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agosmjpeg: fix type of 'ret' variable in smjpeg_read_packet()
Mans Rullgard [Thu, 11 Oct 2012 20:53:22 +0000 (21:53 +0100)]
smjpeg: fix type of 'ret' variable in smjpeg_read_packet()

The 'ret' variable is used for negative error codes so it should
be a signed type.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agompegvideo: remove write-only variable
Mans Rullgard [Thu, 11 Oct 2012 20:51:41 +0000 (21:51 +0100)]
mpegvideo: remove write-only variable

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoUse proper return values in case of missing features
Diego Biurrun [Mon, 8 Oct 2012 22:41:34 +0000 (00:41 +0200)]
Use proper return values in case of missing features

6 years agofate: add avstring test
Diego Biurrun [Sun, 30 Oct 2011 21:02:10 +0000 (22:02 +0100)]
fate: add avstring test

6 years agorangecoder-test: Set error message log level to error, instead of debug
Diego Biurrun [Fri, 5 Oct 2012 17:05:48 +0000 (19:05 +0200)]
rangecoder-test: Set error message log level to error, instead of debug

6 years agorangecoder: K&R formatting cosmetics
Diego Biurrun [Fri, 5 Oct 2012 17:05:00 +0000 (19:05 +0200)]
rangecoder: K&R formatting cosmetics

6 years agoavcodec: Remove broken MMI optimizations
Diego Biurrun [Thu, 11 Oct 2012 14:48:27 +0000 (16:48 +0200)]
avcodec: Remove broken MMI optimizations

The code fails to compile and is broken beyond repair.

6 years agoRemove libmpeg2 #define remnants
Diego Biurrun [Thu, 11 Oct 2012 14:35:48 +0000 (16:35 +0200)]
Remove libmpeg2 #define remnants

6 years agoDe-doxygenize some top-level files
Diego Biurrun [Fri, 24 Aug 2012 11:31:50 +0000 (13:31 +0200)]
De-doxygenize some top-level files

6 years agoMove av_reverse table to libavcodec
Diego Biurrun [Thu, 11 Oct 2012 16:50:30 +0000 (18:50 +0200)]
Move av_reverse table to libavcodec

It is only used in that library.

6 years agoavutil: Duplicate ff_log2_tab instead of sharing it across libs
Diego Biurrun [Thu, 4 Oct 2012 19:12:32 +0000 (21:12 +0200)]
avutil: Duplicate ff_log2_tab instead of sharing it across libs

The table is so small that the space gain is not worth the
performance overhead of cross-library access.

6 years agoavfilter: call x86 init functions under if (ARCH_X86), not if (HAVE_MMX)
Diego Biurrun [Wed, 29 Aug 2012 17:37:14 +0000 (19:37 +0200)]
avfilter: call x86 init functions under if (ARCH_X86), not if (HAVE_MMX)

6 years agortspdec: Set the default port for listen mode, if none is specified
Martin Storsjö [Fri, 12 Oct 2012 08:10:09 +0000 (11:10 +0300)]
rtspdec: Set the default port for listen mode, if none is specified

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agotscc2: Fix an out of array access
Michael Niedermayer [Fri, 12 Oct 2012 00:26:50 +0000 (02:26 +0200)]
tscc2: Fix an out of array access

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agortmpproto: Fix an out of array write
Michael Niedermayer [Fri, 12 Oct 2012 00:06:51 +0000 (02:06 +0200)]
rtmpproto: Fix an out of array write

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agortspdec: Fix use of uninitialized byte
Michael Niedermayer [Thu, 11 Oct 2012 21:57:08 +0000 (23:57 +0200)]
rtspdec: Fix use of uninitialized byte

ffurl_read_complete can return 0 as well as negative error codes.

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agovp8: reset loopfilter delta values at keyframes.
Sami Pietila [Fri, 12 Oct 2012 14:12:49 +0000 (07:12 -0700)]
vp8: reset loopfilter delta values at keyframes.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>
6 years agoavutil: add yuva422p and yuva444p formats
Luca Barbato [Mon, 1 Oct 2012 02:18:02 +0000 (04:18 +0200)]
avutil: add yuva422p and yuva444p formats

6 years agolavu,lavd: do not use av_pix_fmt_descriptors directly.
Anton Khirnov [Sat, 6 Oct 2012 11:29:37 +0000 (13:29 +0200)]
lavu,lavd: do not use av_pix_fmt_descriptors directly.

6 years agolavfi: do not use av_pix_fmt_descriptors directly.
Anton Khirnov [Sat, 6 Oct 2012 11:29:37 +0000 (13:29 +0200)]
lavfi: do not use av_pix_fmt_descriptors directly.

6 years agolavc: do not use av_pix_fmt_descriptors directly.
Anton Khirnov [Sat, 6 Oct 2012 11:29:37 +0000 (13:29 +0200)]
lavc: do not use av_pix_fmt_descriptors directly.

6 years agompegvideo: fix indentation
Anton Khirnov [Sat, 6 Oct 2012 13:21:17 +0000 (15:21 +0200)]
mpegvideo: fix indentation

6 years agosws: do not use av_pix_fmt_descriptors directly.
Anton Khirnov [Sat, 6 Oct 2012 11:29:37 +0000 (13:29 +0200)]
sws: do not use av_pix_fmt_descriptors directly.

6 years agotools: do not use av_pix_fmt_descriptors directly.
Anton Khirnov [Sat, 6 Oct 2012 11:29:37 +0000 (13:29 +0200)]
tools: do not use av_pix_fmt_descriptors directly.

6 years agopixdesc: add functions for accessing pixel format descriptors.
Anton Khirnov [Sat, 6 Oct 2012 10:36:38 +0000 (12:36 +0200)]
pixdesc: add functions for accessing pixel format descriptors.

Make av_pix_fmt_descriptors table static on next major bump.

Making the table public is dangerous, since the caller has no way to
know how large it actually is. It also prevents adding new fields to
AVPixFmtDescriptor without a major bump.

6 years agobuild: add support for Tru64 (OSF/1)
Mans Rullgard [Thu, 11 Oct 2012 11:33:38 +0000 (12:33 +0100)]
build: add support for Tru64 (OSF/1)

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agomd5: Allocate a normal private context for the opaque md5 context pointer
Martin Storsjö [Thu, 11 Oct 2012 12:27:25 +0000 (15:27 +0300)]
md5: Allocate a normal private context for the opaque md5 context pointer

This avoids having to overestimate the md5 context size, which
isn't known beforehand, allowing us to use the new allocate functions
instead.

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agoUse the new aes/md5/sha/tree allocation functions
Martin Storsjö [Thu, 11 Oct 2012 12:10:45 +0000 (15:10 +0300)]
Use the new aes/md5/sha/tree allocation functions

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agoavutil: Add functions for allocating opaque contexts for algorithms
Martin Storsjö [Thu, 11 Oct 2012 12:08:04 +0000 (15:08 +0300)]
avutil: Add functions for allocating opaque contexts for algorithms

The current API where the plain size is exposed is not of much
use - in most cases it is allocated dynamically anyway.

If allocated e.g. on the stack via an uint8_t array, there's no
guarantee that the struct's members are aligned properly (unless
the array is overallocated and the opaque pointer within it
manually aligned to some unspecified alignment).

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agosvq3: fix pointer type warning
Mans Rullgard [Thu, 11 Oct 2012 16:39:47 +0000 (17:39 +0100)]
svq3: fix pointer type warning

Fixes:
libavcodec/svq3.c:661:9: warning: passing argument 2 of 'svq3_decode_block' from incompatible pointer type
libavcodec/svq3.c:208:19: note: expected 'DCTELEM *' but argument is of type 'DCTELEM (*)[32]'

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agosvq3: replace unsafe pointer casting with intreadwrite macros
Mans Rullgard [Thu, 11 Oct 2012 15:08:22 +0000 (16:08 +0100)]
svq3: replace unsafe pointer casting with intreadwrite macros

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoparseutils-test: various cleanups
Mans Rullgard [Thu, 11 Oct 2012 13:15:10 +0000 (14:15 +0100)]
parseutils-test: various cleanups

- make tables static const
- remove useless use of compound literal
- break long lines
- fix a comma/semicolon typo

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoparseutils: fix const removal warning
Mans Rullgard [Thu, 11 Oct 2012 13:12:34 +0000 (14:12 +0100)]
parseutils: fix const removal warning

The const qualifier is still removed although it happens inside
the strtol() function so no warning is generated.

Fixes:
libavutil/parseutils.c:110:11: warning: assignment discards qualifiers from pointer target type

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoprepare 9_beta1 release
Janne Grunau [Wed, 10 Oct 2012 21:40:04 +0000 (23:40 +0200)]
prepare 9_beta1 release

6 years agoMove avutil tables only used in libavcodec to libavcodec.
Diego Biurrun [Wed, 10 Oct 2012 23:16:05 +0000 (01:16 +0200)]
Move avutil tables only used in libavcodec to libavcodec.

6 years agooptions_table: refs option is not snow-only
Diego Biurrun [Mon, 8 Oct 2012 11:04:39 +0000 (13:04 +0200)]
options_table: refs option is not snow-only

6 years agorandom_seed: Support using CryptGenRandom on windows
Martin Storsjö [Thu, 11 Oct 2012 11:30:23 +0000 (14:30 +0300)]
random_seed: Support using CryptGenRandom on windows

Signed-off-by: Martin Storsjö <martin@martin.st>
6 years agodoc: update the faq entry about custom I/O
Luca Barbato [Sun, 7 Oct 2012 21:19:29 +0000 (21:19 +0000)]
doc: update the faq entry about custom I/O

URLProtocols are private and custom AVIOContexts are a simple and
easier solution for most of the situations.

6 years agomlpdsp: adding missing file
Luca Barbato [Thu, 11 Oct 2012 10:34:36 +0000 (12:34 +0200)]
mlpdsp: adding missing file

6 years agodsputil: split out mlp dsp function
Luca Barbato [Thu, 11 Oct 2012 00:01:41 +0000 (02:01 +0200)]
dsputil: split out mlp dsp function

6 years agosh4: add required #include, fix build
Mans Rullgard [Thu, 11 Oct 2012 01:08:24 +0000 (02:08 +0100)]
sh4: add required #include, fix build

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agoaverror: make error values proper negative values
Janne Grunau [Wed, 10 Oct 2012 20:25:51 +0000 (22:25 +0200)]
averror: make error values proper negative values

Fixes stupid mistake in 25227c3a78fefd74bc6ea34032cd49ec8c8d9d85.

6 years agobuild: do not use LIB as variable name
Mans Rullgard [Wed, 10 Oct 2012 22:02:57 +0000 (23:02 +0100)]
build: do not use LIB as variable name

The Microsoft linker uses the LIB environment variable which
clashes with a make variable of the same name.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agobuild: whitespace cosmetics
Mans Rullgard [Wed, 10 Oct 2012 22:00:00 +0000 (23:00 +0100)]
build: whitespace cosmetics

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agobuild: remove single-use variable THIS_LIB
Mans Rullgard [Wed, 10 Oct 2012 21:59:37 +0000 (22:59 +0100)]
build: remove single-use variable THIS_LIB

Replace the single use of THIS_LIB with its value.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agobuild: remove references to unused EXTRAOBJS variable
Mans Rullgard [Wed, 10 Oct 2012 21:36:54 +0000 (22:36 +0100)]
build: remove references to unused EXTRAOBJS variable

This was part of a ghastly hack that is long since gone.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agolavfi: convert input/ouput list compound literals to named objects
Mans Rullgard [Tue, 24 Jul 2012 13:14:01 +0000 (14:14 +0100)]
lavfi: convert input/ouput list compound literals to named objects

A number of compilers, for example those from TI and IBM, choke on
these initialisers.  The current style is also quite ugly.

Signed-off-by: Mans Rullgard <mans@mansr.com>
6 years agofate: add h263 obmc vsynth tests
Janne Grunau [Wed, 10 Oct 2012 12:22:57 +0000 (14:22 +0200)]
fate: add h263 obmc vsynth tests

6 years agoavconv: remove bogus warning when using avconv -h without parameter
Janne Grunau [Wed, 10 Oct 2012 18:35:26 +0000 (20:35 +0200)]
avconv: remove bogus warning when using avconv -h without parameter

6 years agoaverror: explicitly define AVERROR_* values
Janne Grunau [Wed, 10 Oct 2012 18:11:56 +0000 (20:11 +0200)]
averror: explicitly define AVERROR_* values

6 years agoflashsv: propagate inflateReset() errors
Janne Grunau [Wed, 10 Oct 2012 17:47:05 +0000 (19:47 +0200)]
flashsv: propagate inflateReset() errors

Fixes CID717493.

6 years agoindeo4/5: remove constant parameter num_bands from wavelet recomposition
Janne Grunau [Wed, 10 Oct 2012 17:16:09 +0000 (19:16 +0200)]
indeo4/5: remove constant parameter num_bands from wavelet recomposition

Fixes bogus uninitialized value compiler and coverity warnings.

6 years agomxfdec: return error if no segments are available in mxf_get_sorted_table_segments
Janne Grunau [Tue, 9 Oct 2012 19:45:23 +0000 (21:45 +0200)]
mxfdec: return error if no segments are available in mxf_get_sorted_table_segments

Fixes CID732275.

6 years agoDouble motion vector range for HPEL interlaced picture in proper place
Mashiat Sarker Shakkhar [Tue, 9 Oct 2012 19:38:19 +0000 (19:38 +0000)]
Double motion vector range for HPEL interlaced picture in proper place

The existing code is not in the right place and it should cover both
interlaced frame and field pictures.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
6 years agovc1dec: Set opposite to the correct value for 1REF field pictures
Mashiat Sarker Shakkhar [Tue, 9 Oct 2012 19:38:18 +0000 (19:38 +0000)]
vc1dec: Set opposite to the correct value for 1REF field pictures

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
6 years agovc1dec: Invoke edge emulation regardless of MV precision for 1-MV chroma
Mashiat Sarker Shakkhar [Tue, 9 Oct 2012 19:38:16 +0000 (19:38 +0000)]
vc1dec: Invoke edge emulation regardless of MV precision for 1-MV chroma

This is required due to the way VC-1 handles chroma pull-back which may end
up causing negative chroma MV for zero luma MV. Edge emulation needs to be
invoked in such cases.

This only affects vertical component of chroma motion vector.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
6 years agovc1dec: Set chroma reference field from REFFIELD for 1REF field pictures
Mashiat Sarker Shakkhar [Tue, 9 Oct 2012 19:38:15 +0000 (19:38 +0000)]
vc1dec: Set chroma reference field from REFFIELD for 1REF field pictures

Interlaced field pictures can have one or two reference pictures, signaled
by NUMREF syntax element. For single reference pictures, reference picture
is determined by REFFIELD syntax element.

Signed-off-by: Derek Buitenhuis <derek.buitenhuis@gmail.com>
6 years agosegment: support applehttp style list
Luca Barbato [Tue, 9 Oct 2012 00:49:42 +0000 (02:49 +0200)]
segment: support applehttp style list

6 years agoavio: introduce avio_closep
Luca Barbato [Tue, 9 Oct 2012 23:39:06 +0000 (01:39 +0200)]
avio: introduce avio_closep

6 years agompegtsenc: set muxing type notification to verbose
Luca Barbato [Tue, 9 Oct 2012 20:00:22 +0000 (22:00 +0200)]
mpegtsenc: set muxing type notification to verbose

6 years agovc1dec: Use correct spelling of "opposite"
Mashiat Sarker Shakkhar [Tue, 9 Oct 2012 20:38:17 +0000 (16:38 -0400)]
vc1dec: Use correct spelling of "opposite"

Signed-off-by: Diego Biurrun <diego@biurrun.de>
6 years agoa64multienc: change mc_frame_counter to unsigned
Janne Grunau [Thu, 20 Sep 2012 09:34:35 +0000 (09:34 +0000)]
a64multienc: change mc_frame_counter to unsigned

GCC cannot proof that c->mc_frame_counter is always >- 0, changing the
type from int to unsigned fixes following warning:

libavcodec/a64multienc.c: In function ‘a64multi_encode_frame’:
libavcodec/a64multienc.c:342:17: warning: ‘buf’ may be used uninitialized in this function[-Wuninitialized]

Signed-off-by: Janne Grunau <janne-libav@jannau.net>
6 years agoarm: call arm-specific rv34dsp init functions under if (ARCH_ARM)
Jean-Baptiste Kempf [Fri, 5 Oct 2012 09:12:04 +0000 (09:12 +0000)]
arm: call arm-specific rv34dsp init functions under if (ARCH_ARM)

Assign NEON specific function pointers after runtime check via
av_get_cpu_flags().

Signed-off-by: Janne Grunau <janne-libav@jannau.net>
6 years agosvq1: Drop a bunch of useless parentheses
Diego Biurrun [Wed, 10 Oct 2012 12:08:21 +0000 (14:08 +0200)]
svq1: Drop a bunch of useless parentheses

6 years agoparseutils-test: do not print numerical error codes
Mans Rullgard [Wed, 10 Oct 2012 11:51:04 +0000 (12:51 +0100)]
parseutils-test: do not print numerical error codes

The error codes differ between systems so printing the value makes
the fate test fail on some systems.

Signed-off-by: Mans Rullgard <mans@mansr.com>