Ronald S. Bultje [Thu, 16 Jun 2011 16:04:25 +0000 (12:04 -0400)]
swscale: disentangle yuv2rgbX_c_full() into small functions.
This is part of the Great Evil Plan to simplify swscale.
Ronald Bultje [Tue, 28 Jun 2011 01:10:03 +0000 (18:10 -0700)]
swscale: split yuv2packed[12X]_c() remainders into small functions.
This is part of the Great Evil Plan to simplify swscale.
Ronald S. Bultje [Thu, 16 Jun 2011 16:04:24 +0000 (12:04 -0400)]
swscale: split yuv2packedX_altivec in smaller functions.
This will likely lead to a considerable performance boost,
since it removes a branch from the inner loop. Part of the
Great Evil Plan to simplify swscale.
Ronald S. Bultje [Sun, 26 Jun 2011 22:52:00 +0000 (15:52 -0700)]
swscale: don't use planar output functions to write to NV12/21.
This prevents a crash when converting to NV12/21 without the bitexact
flags enabled.
Justin Ruggles [Mon, 27 Jun 2011 03:58:19 +0000 (23:58 -0400)]
ac3enc: move ff_ac3_encode_frame() to ac3enc_template.c
This avoids using function pointers for quite a few small functions, most of
which just call DSP functions.
Justin Ruggles [Mon, 27 Jun 2011 11:46:27 +0000 (07:46 -0400)]
ac3enc: merge log2_tab() into normalize_samples()
Justin Ruggles [Tue, 21 Jun 2011 16:59:11 +0000 (12:59 -0400)]
ac3enc: Remove bit allocation fallbacks.
They introduce too much complexity for very little benefit. The result is that
encoding will be more likely to fail at extremely low bitrates where quality
would be awful anyway.
Mans Rullgard [Sat, 25 Jun 2011 10:51:09 +0000 (11:51 +0100)]
build: improve rules for test programs
This generates dependencies for the *-test.o files ensuring
rebuilds when necessary.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Sat, 25 Jun 2011 10:57:32 +0000 (11:57 +0100)]
build: factor out the .c and .S compile commands as a macro
These commands have the same form, and using a common macro allows
it to be used elsewhere without further duplication.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Ronald S. Bultje [Sun, 26 Jun 2011 21:41:44 +0000 (14:41 -0700)]
swscale: remove unused xInc/srcW arguments from hScale().
Jason Garrett-Glaser [Sun, 26 Jun 2011 07:21:54 +0000 (00:21 -0700)]
H.264: disable 2tap qpel with CODEC_FLAG2_FAST and >8-bit
2tap qpel isn't implemented yet for high bit depth, so it just breaks decoding.
Jason Garrett-Glaser [Sun, 26 Jun 2011 02:40:05 +0000 (19:40 -0700)]
H.264: make filter_mb_fast support 4:4:4
Diego Biurrun [Sun, 26 Jun 2011 14:37:07 +0000 (16:37 +0200)]
mpeg4videoenc: Remove disabled variant of mpeg4_encode_block().
As a nice sideeffect this fixes the following warning:
libavcodec/mpeg4videoenc.c:302:20: warning: variable ‘rl’ set but not used [-Wunused-but-set-variable]
Alexis Ballier [Wed, 22 Jun 2011 20:25:14 +0000 (22:25 +0200)]
configure: allow post-fixed cpu strings for athlon64, k8, and opteron
when setting the -march flag.
This is to match gcc's {athlon64,k8,opteron}-sse3 -march flags.
Signed-off-by: Justin Ruggles <justin.ruggles@gmail.com>
Diego Biurrun [Tue, 14 Jun 2011 22:30:03 +0000 (00:30 +0200)]
Move some variable declarations below the proper #ifdefs.
This avoids warnings about set-but-unused variables.
Mans Rullgard [Sat, 25 Jun 2011 17:51:00 +0000 (18:51 +0100)]
ARM: silence some annoying armcc warnings
This silences warnings about pointer target sign mismatches as
already done for gcc with -Wno-pointer-sign.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Diego Biurrun [Sat, 25 Jun 2011 12:09:40 +0000 (14:09 +0200)]
ffplay: Remove unused-but-set channels variable from update_sample_display().
Diego Biurrun [Sat, 25 Jun 2011 00:01:12 +0000 (02:01 +0200)]
build: Add DEP_LIBS dependency directly to the shared library build rule.
The dependency was added conditional to a variable that is always defined,
so it is safe to add it directly.
Diego Biurrun [Fri, 24 Jun 2011 22:32:12 +0000 (00:32 +0200)]
build: Remove multiple inclusion guards from config.mak.
config.mak is no longer included multiple times; the guards are pointless.
Diego Biurrun [Fri, 24 Jun 2011 22:19:39 +0000 (00:19 +0200)]
build: Remove redundant config.mak includes from subdirectory Makefiles.
Calling Make from subdirectories is not supported and config.mak has
multiple inclusion guards anyway, so the top-level include is enough.
Diego Biurrun [Fri, 24 Jun 2011 08:49:59 +0000 (10:49 +0200)]
aacenc: Mark psy_3gpp_window() as av_unused.
It is intentionally left in to allow adding 3GPP-style windowing in the future.
Marking it av_unused silences an annoying unused function warning.
Diego Biurrun [Fri, 24 Jun 2011 19:16:10 +0000 (21:16 +0200)]
configure: Add -Wno-format-zero-length to CFLAGS.
The C standard specifies that zero-length format strings are allowed.
Can Wu [Fri, 17 Jun 2011 05:58:01 +0000 (13:58 +0800)]
udp: Receive on the remote port number, if no local port is set
For a unicast udp stream to localhost:1234, currently
ffplay udp://:1234?localport=1234
works, but
ffplay udp://:1234
doesn't work.
Signed-off-by: Martin Storsjö <martin@martin.st>
Diego Biurrun [Thu, 23 Jun 2011 20:41:54 +0000 (22:41 +0200)]
doxygen: Consistently use '@' instead of '\' for Doxygen markup.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Martin Storsjö [Fri, 17 Jun 2011 07:31:11 +0000 (10:31 +0300)]
Use av_printf_format to check the usage of printf style functions
This helps catching cases where the format string doesn't
match what is passed in, or injection bugs where user data
is passed in as format string.
Signed-off-by: Martin Storsjö <martin@martin.st>
Martin Storsjö [Fri, 17 Jun 2011 07:30:15 +0000 (10:30 +0300)]
Add av_printf_format, for marking printf style format strings and their parameters
Signed-off-by: Martin Storsjö <martin@martin.st>
Mans Rullgard [Thu, 23 Jun 2011 14:42:19 +0000 (15:42 +0100)]
ARM: enable thumb for Cortex-M* CPUs
These CPUs are thumb-only and thus require this option.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Diego Biurrun [Thu, 23 Jun 2011 11:27:21 +0000 (13:27 +0200)]
nsvdec: Propagate error values instead of returning 0 in nsv_read_header().
This eliminates a warning about a set-but-unused variable.
Mans Rullgard [Mon, 6 Jun 2011 21:38:38 +0000 (22:38 +0100)]
build: remove SRC_PATH_BARE variable
Setting SRC_PATH to "." when building in-tree removes the need
for a quoted version of the source path since out-of-tree builds
are not possible if the pathname contains spaces.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Sun, 5 Jun 2011 21:29:37 +0000 (22:29 +0100)]
build: move basic rules and variables to main Makefile
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 6 Jun 2011 14:21:12 +0000 (15:21 +0100)]
build: move special targets to end of main Makefile
Signed-off-by: Mans Rullgard <mans@mansr.com>
Stefano Sabatini [Fri, 3 Jun 2011 08:50:28 +0000 (10:50 +0200)]
lavdev: improve feedback in case of invalid frame rate/size
Show the invalid string in the error message.
While at it also prefer "Could not" over "Couldn't", plain forms are
preferred over contractions (simplify readability, especially for non
English-savvy people).
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Fri, 3 Jun 2011 08:41:47 +0000 (10:41 +0200)]
vfwcap: prefer "framerate_q" over "fps" in vfw_read_header()
The variable is used for containing the parsed value of framerate,
using a lexically consistent name eases readability/understanding.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Fri, 3 Jun 2011 08:39:54 +0000 (10:39 +0200)]
v4l2: prefer "framerate_q" over "fps" in v4l2_set_parameters()
The variable is used for containing the parsed value of framerate,
using a lexically consistent name eases readability/understanding.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Fri, 3 Jun 2011 08:31:42 +0000 (10:31 +0200)]
fbdev: prefer "framerate_q" over "fps" in device context
The variable is used for containing the parsed value of framerate,
using a lexically consistent name eases readability/understanding.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Fri, 3 Jun 2011 08:54:10 +0000 (10:54 +0200)]
bktr: prefer "framerate" over "fps" for grab_read_header()
The variable is used for containing the parsed value of s1->framerate,
using a lexically consistent name ease readability/understanding.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Nicolas George [Thu, 10 Mar 2011 11:10:34 +0000 (12:10 +0100)]
ALSA: implement channel layout for playback.
Currently quad, 5.0, 5.1 and 7.1 are implemented.
Implementing support for other formats/layouts and capture should be
straightforward.
5.0 and 7.1 support by Carl Eugen Hoyos.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Carl Eugen Hoyos [Wed, 13 Apr 2011 23:04:18 +0000 (01:04 +0200)]
alsa: support unsigned variants of already supported signed formats.
Carl Eugen Hoyos [Mon, 13 Jun 2011 19:16:30 +0000 (21:16 +0200)]
alsa: add support for more formats.
Specifically, f32, f64, s32, s24, a-law and mu-law.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Mans Rullgard [Tue, 14 Jun 2011 10:29:48 +0000 (11:29 +0100)]
ARM: allow building in Thumb2 mode
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Sun, 19 Jun 2011 17:47:06 +0000 (18:47 +0100)]
configure: add --optflags option
This allows overriding the default optimisation flags selected by
configure.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 6 Jun 2011 17:03:22 +0000 (18:03 +0100)]
build: move documentation rules to doc/Makefile
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 6 Jun 2011 18:44:29 +0000 (19:44 +0100)]
build: move test rules to tests/Makefile
Signed-off-by: Mans Rullgard <mans@mansr.com>
Justin Ruggles [Tue, 21 Jun 2011 20:49:16 +0000 (16:49 -0400)]
ac3enc: remove unneeded local variable in asym_quant()
Justin Ruggles [Tue, 21 Jun 2011 20:14:19 +0000 (16:14 -0400)]
ac3enc: remove a branch in asym_quant() by doing 2 shifts
Justin Ruggles [Tue, 21 Jun 2011 17:39:02 +0000 (13:39 -0400)]
ac3enc: avoid masking output in asym_quant() by using signed values for
quantized mantissas.
Jason Garrett-Glaser [Wed, 22 Jun 2011 09:34:02 +0000 (02:34 -0700)]
H.264: fix 4:4:4 + deblocking + 8x8dct + cavlc + MBAFF
Jason Garrett-Glaser [Wed, 22 Jun 2011 09:05:14 +0000 (02:05 -0700)]
H.264: fix 4:4:4 + deblocking + MBAFF
Jason Garrett-Glaser [Tue, 21 Jun 2011 11:16:33 +0000 (04:16 -0700)]
H.264: fix 4:4:4 cropping warning
Jason Garrett-Glaser [Tue, 21 Jun 2011 08:10:37 +0000 (01:10 -0700)]
H.264: reference the correct SPS in decode_scaling_matrices
Jason Garrett-Glaser [Mon, 20 Jun 2011 22:56:35 +0000 (15:56 -0700)]
H.264: fix bug in lossless 4:4:4 decoding
Coefficient test for i16x16 add_pixels4 assumed luma plane.
ami_stuff [Tue, 24 May 2011 21:38:01 +0000 (23:38 +0200)]
rawdec: Fix decoding of QT WRAW files.
From some tests it results that:
1. All of the AVI/MOV WRAW files need to be flipped.
2. MOV WRAW files need to use AVI color modes.
3. Assigning PAL8 mode by default to WRAW codec is not correct.
Mans Rullgard [Sun, 19 Jun 2011 17:41:14 +0000 (18:41 +0100)]
configure: report optimization for size separately
This removes an unsightly override of the 'optimizations' setting
only to make the configure report print 'small' when --enable-small
is used.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Carl Eugen Hoyos [Thu, 12 May 2011 23:08:01 +0000 (01:08 +0200)]
mov: Support Digital Voodoo SD 8 Bit and DTS codec identifiers.
ami_stuff [Thu, 12 May 2011 22:49:29 +0000 (00:49 +0200)]
mov: Support R10g codec identifier.
Kamil Nowosad [Mon, 23 May 2011 21:13:34 +0000 (23:13 +0200)]
riff/img2: Add JPEG 2000 codec IDs.
ami_stuff [Wed, 11 May 2011 20:18:15 +0000 (22:18 +0200)]
riff: Add DAVC fourcc.
This fourcc is used by the "mpegable AVC" codec and files encoded with
this codec decode correctly with our H.264 decoder.
Carl Eugen Hoyos [Sun, 24 Apr 2011 22:27:03 +0000 (00:27 +0200)]
riff: Add M263, XVIX, MMJP, CDV5 fourccs.
ami_stuff [Sun, 24 Apr 2011 22:20:27 +0000 (00:20 +0200)]
rawvideo: Support auv2 fourcc.
Diego Biurrun [Wed, 15 Jun 2011 11:45:24 +0000 (13:45 +0200)]
swscale: Remove unused variable from ff_bfin_get_unscaled_swscale().
Diego Biurrun [Mon, 20 Jun 2011 23:11:28 +0000 (01:11 +0200)]
h264: Fix assert that failed to compile with -DDEBUG.
The assert referenced a variable that no longer exists since 4:4:4 support.
Daniel Kang [Tue, 21 Jun 2011 04:30:23 +0000 (00:30 -0400)]
h264: Add x86 assembly for 10-bit weight/biweight H.264 functions.
Mainly ported from 8-bit H.264 weight/biweight.
Signed-off-by: Diego Biurrun <diego@biurrun.de>
Mans Rullgard [Sun, 19 Jun 2011 15:19:06 +0000 (16:19 +0100)]
fate: remove output redirections from old regtest scripts
All tests are run through the fate-run.sh script which already
sets up redirections. Using the outputs set there simplifies
things somewhat.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Justin Ruggles [Tue, 7 Jun 2011 17:40:22 +0000 (13:40 -0400)]
Replace usages of av_get_bits_per_sample_fmt() with av_get_bytes_per_sample().
av_get_bits_per_sample_fmt() is deprecated.
Mans Rullgard [Mon, 20 Jun 2011 22:10:26 +0000 (23:10 +0100)]
x86: cabac: fix register constraints for 32-bit mode
Some operands need to be accessed in byte mode, which restricts the
available registers in 32-bit mode. Using the 'q' constraint selects
a suitable register.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 09:53:41 +0000 (10:53 +0100)]
cabac: move x86 asm to libavcodec/x86/cabac.h
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 09:24:26 +0000 (10:24 +0100)]
x86: h264: cast pointers to intptr_t rather than int
Only the low-order bits are used here so the type is not important,
but this avoids a compiler warning.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 09:20:15 +0000 (10:20 +0100)]
x86: h264: remove hardcoded edi in decode_significance_8x8_x86()
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 09:12:29 +0000 (10:12 +0100)]
x86: h264: remove hardcoded esi in decode_significance[_8x8]_x86()
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 09:06:07 +0000 (10:06 +0100)]
x86: h264: remove hardcoded edx in decode_significance[_8x8]_x86()
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 08:23:26 +0000 (09:23 +0100)]
x86: h264: remove hardcoded eax in decode_significance[_8x8]_x86()
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 08:28:19 +0000 (09:28 +0100)]
x86: cabac: change 'a' constraint to 'r' in get_cabac_inline()
Nothing requires this value in %eax.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 08:19:27 +0000 (09:19 +0100)]
x86: cabac: remove hardcoded esi in get_cabac_inline()
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 02:39:45 +0000 (03:39 +0100)]
x86: cabac: remove hardcoded edx in get_cabac_inline()
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 02:19:20 +0000 (03:19 +0100)]
x86: cabac: remove unused macro parameter
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 01:31:53 +0000 (02:31 +0100)]
x86: cabac: remove hardcoded ebx in inline asm
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 00:54:32 +0000 (01:54 +0100)]
x86: cabac: remove hardcoded struct offsets from inline asm
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 00:58:36 +0000 (01:58 +0100)]
cabac: remove inline asm under #if 0
A comment says it's not faster than the C code.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Mon, 20 Jun 2011 00:49:38 +0000 (01:49 +0100)]
cabac: remove BRANCHLESS_CABAC_DECODER switch
The code does not compile without this set.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Mans Rullgard [Sun, 19 Jun 2011 23:59:35 +0000 (00:59 +0100)]
cabac: remove #if 0 cascade under never-set #ifdef ARCH_X86_DISABLED
Signed-off-by: Mans Rullgard <mans@mansr.com>
Reinhard Tartler [Mon, 20 Jun 2011 20:40:31 +0000 (22:40 +0200)]
document libswscale bump
Ronald Bultje [Mon, 20 Jun 2011 15:07:56 +0000 (15:07 +0000)]
error_resilience: skip last-MV predictor step if MVs are not available.
Fixes crashes when playing broken MPEG2-TS streams.
Ronald Bultje [Mon, 20 Jun 2011 15:07:55 +0000 (15:07 +0000)]
error_resilience: actually add counter when adding a MV predictor.
Without, the predictor isn't actually used.
Michael Niedermayer [Sat, 18 Jun 2011 00:44:17 +0000 (02:44 +0200)]
qdm2: Fix alignment of local array.
Fixes ticket270
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Thu, 21 Apr 2011 08:31:48 +0000 (10:31 +0200)]
ocv: replace FF_INTERNAL_MEM_TYPE_MAX_VALUE with SIZE_MAX
Fix compilatin after removal of FF_INTERNAL_MEM_TYPE_MAX_VALUE.
Signed-off-by: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Mon, 16 May 2011 21:44:35 +0000 (23:44 +0200)]
drawtext: add braces around initialisers for option defaults
Fix warnings of the type:
vf_drawtext.c:NNN: warning: missing braces around initializer
vf_drawtext.c:NNN: warning: (near initialization for ‘drawtext_options[X].default_val’)
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Mon, 16 May 2011 21:48:00 +0000 (23:48 +0200)]
drawtext: reindent after the previous commit
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Stefano Sabatini [Mon, 16 May 2011 18:11:50 +0000 (20:11 +0200)]
drawtext: fix strftime() text expansion
The feature was dropped after the filter was partially rewritten and
recommitted.
Signed-off-by: Anton Khirnov <anton@khirnov.net>
Reinhard Tartler [Mon, 20 Jun 2011 10:20:54 +0000 (12:20 +0200)]
Bump major version for
b8e893399ff8755721dc117695ec5ff183c1e07b
On architectures such as x86 (both 32 bit and 64bit), the stack element
size is fixed, which maintains alignment. Here, this change does not
break anything. However, we also support also other architectures where
this property is not maintained and therefore, applications will crash
horribly.
This change effectively forces all applications to be recompiled against
libswscale.
Ronald S. Bultje [Mon, 20 Jun 2011 14:24:33 +0000 (10:24 -0400)]
h264-mt: fix deadlock in packets with multiple slices (e.g. MP4).
Yusuke Nakamura [Sat, 4 Jun 2011 16:28:43 +0000 (01:28 +0900)]
mov: Fix empty edit detection.
Mans Rullgard [Sun, 19 Jun 2011 18:31:56 +0000 (19:31 +0100)]
Drop git- prefix from version labels
Signed-off-by: Mans Rullgard <mans@mansr.com>
Reinhard Tartler [Sun, 19 Jun 2011 17:51:20 +0000 (19:51 +0200)]
Use the contents of RELEASE as version fallback instead of 'UNKNOWN'
This makes binaries produced by source tarballs identify themselves with
the version number of the corresponding release series, unless overriden
by a 'VERSION' file.
Mans Rullgard [Sun, 19 Jun 2011 17:03:21 +0000 (18:03 +0100)]
path64/open64: filter out unsupported flags
These flags are accepted without error but produce an annoying
warning. Filtering them out makes the build less noisy.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Reinhard Tartler [Sun, 19 Jun 2011 17:45:36 +0000 (19:45 +0200)]
Add Release Notes
Alex Converse [Sat, 18 Jun 2011 22:45:45 +0000 (15:45 -0700)]
sipr: Drop unused DSPContext
Alex Converse [Sat, 18 Jun 2011 22:44:31 +0000 (15:44 -0700)]
sipr: include string.h for mem*()
Alex Converse [Sat, 18 Jun 2011 22:42:57 +0000 (15:42 -0700)]
sipr: Use memmove() to copy overlapped buffers.
Alex Converse [Sat, 18 Jun 2011 05:03:24 +0000 (22:03 -0700)]
h264: drop some ugly casts
Ronald S. Bultje [Sat, 18 Jun 2011 19:33:49 +0000 (15:33 -0400)]
lavf: prevent crash in av_open_input_file() if ap == NULL.
Needed for proper behaviour in our old API compatibility code.